Mongoose – Chapter 5 (Motor Install)

There are about a billion step by step Youtube videos and forum posts describing the steps that go into installing a BBSHD onto an existing donor bicycle.  They already do a great job so I won’t try to get too deep into the weeds on that.  I will say that installation on the Envoy was simpler than my previous builds on different bikes, which is only a good thing for you if you aren’t familiar with installing a BBSHD onto a bike.

Step 1: Prepare the bicycle

You are going to be pulling apart some things and replacing them, so you’ve got some prep work to do.  Myself personally, I decided to do the following which you should consider optional.

9 Speed Conversion

I converted the bike to a 9-speed from its factory 8-speed.  You already have to say goodbye to the stock shifters Mongoose gives you with the bike since they are integral with the brake levers, which themselves have to go.  So I decided to bump up my options a tad.  8, 9 and 10-speed systems will fit interchangeably on the same hub.  I also have other bikes that are 9-speed and I like to have common parts between them.  Plus, I already had a cluster in my parts pile that was 9-speed :-).  So 9-speed it is.

As noted in the Motor Choice chapter, if you decide to stay 8-speed, you still need to replace your cluster as part of your mid drive upgrade.  A Shimano CS-HG31 8-speed cassette in 11-34T is about US$17.50.  You can bump that up to an HG-50 with nickel-plated steel cogs (not a bad idea) for about US$22.  At that point you will need to buy yourself some 8-speed shifters since, as noted above, the stockers can’t stay.  You’ll be able to keep the derailleur that comes with the bike (which frankly is not something that appealed to me… the one on my bike worked poorly and was clearly a basic, commodity component.

So here was my prep given this 9-speed conversion:

  1. Remove the handlebar grips without destroying them.  I used a jeweler’s screwdriver to lift up the edge of the grip so I could get a WD40 red straw nozzle in there and give it a spritz.  A couple of those and I was able to work the grip loose without having to cut it off.
  2. Release the brake cables.  This entails releasing the cable from the brake caliper so the lever flops free.  Then you can easily remove the cable from the lever.
  3. Release the shifter cables.  Release the cable from the derailleur and then pull the cable out of the housing so it is free.
  4. Remove the shifter/brake lever from the handlebars.  Just loosen the screw and slide them off.
  5. Remove the chain.  The Mongoose chain does not have a master link, so you will need to use a chainbreaker.  This chain will not be re-used.
  6. Remove the front derailleur from the bicycle.  Not used on a mid drive, so it goes into the bin.
  7. Remove the front derailleur cable housing.  Since the cable is routed thru the frame, this will give you an unused entry and exit hole in your frame which you’ll want to seal up with something.  I used a bit of rolled up mastic tape.
  8. Remove the rear wheel and replace the rear cluster.  You’ll need a chain whip and a cassette removal tool for this.
  9. Remove and replace the rear derailleur (optional.  See above.)
  10. Remove both pedals from the crankarms (or better yet don’t install them in the first place when you take the bike out of the box).
  11. Remove the crankarms.  You will need a proper sized socket to get the bolt out, then a crank puller tool plus a big wrench to finish the job.  Repeat for the other side.
  12. Remove the bottom bracket.  You’ll need a standard bottom bracket tool such as a Park BBT-32, and again a big wrench to turn it.

Prep work complete.  The bike now looks like the picture below, minus the big cargo net.  You’re ready to install the motor now.

IMG_20190928_102436
Fig. 2:  At this point your bike will look about like this:  No pedals, nothing on the bars and loose wires hanging around.

Step 2: Attach The Motor

It will fit right into the empty hole that used to house your bottom bracket.  At this point the motor is hanging straight down in the bottom bracket thanks to gravity.  You will want to dig into your installation kit’s parts bag and retrieve the two M6 bolts and the fixing plate that will clamp your motor to the bottom bracket.

One side of the fixing plate is ridged. These ridges are there to bite into the bottom bracket.  they go a long way (most of the way) towards holding the motor up and in place.  Before you slide that plate onto the motor’s axle, you need some spacers between it and the motor to support the plate, but not so many that you get in the way of the plate clamping to the bottom bracket.  Usually I use fixed-size steel spacers that I get from McMaster-Carr (you can buy them in various lengths in 1mm increments), but in this case I just used three M6 washers and got a perfect fit.

IMG_20191008_174717
Fig. 2a: The fixing plate, with those all-important ridges that must directly contact your bottom bracket.  In background: a couple of outer trim rings, and the  M6 bolts you’ll use to attach that plate to the motor.

Using just this plate clamped gently to the bottom bracket, its possible to get a provisional fit on the motor without fully tightening it.  Here’s where, essentially, the only hiccup in the installation occurs:  The rear shifter cable exits the frame just ahead of the bottom bracket, and travels under it.  If you rotate the motor up as high as it can go – which is the most desirable position – it will pinch the cable and seize it up.  So you have to back off a bit.  The picture below shows what my cable looks like after a successful positioning of the motor that allows the cable to move freely.

mongoose01.jpg
Fig. 3:  The arrows point to the shifter cable exit point and the allowable bend.  Also visible is a wad of mastic tape I used as a makeshift bumper to help solidify the motor’s position during installation

To ensure you’ve got the motor positioned right, use your replacement shifter cable and just manually run it down into the housing.  If you can get it thru easily so it pops out the other end, you are good to go.  Once you get the motor positioned up so that it a) gives max ground clearance and b) does not interfere with shifting, its time to tighten it down fully.  You will use two clamping lock rings over the axle and bottom bracket.  Typically, you use the inner and outer lock rings as shown in the photo below.

IMG_20191008_174043.jpg
Fig. 4: Note the thick black Sharpie registration marks running along the top of the two outer rings and the bottom bracket.  If I see those marks not lining up, it means something is slipping and needs some quality time with the torque wrench.

However, doing it the ‘normal’ way requires two tools.  The inner lockring is physically covered by the outer lock ring, so you can’t see it, but if you look in your parts bag the differences are clear.  Here’s the same installation done with two inner lock rings stacked on top of one another.

The outer ring used in Figure 4 is essentially a trim ring.  Its there to look nice.  Using two inner rings means you can torque the bejesus out of each of them.  And speaking of torque, I use a 1/2″ torque wrench to put 100 ft lbs of torque on each of these two rings.  Yes, 100 each is well above the ‘official’ spec but it is proven to work.

A note on tools:  Cheapie Bafang wrenches made of thin steel will let you ‘get by’ with a basic tightening of these lock rings. They are widely available for roughly US$15.  I don’t recommend their use given the lack of ability to use a torque specification, and their small size.  Pass on these and your knuckles and palms will thank me, as will you when your motor does NOT loosen since you used proper tools.  To do the job right, you need to use something like the socket found here.  You want the ‘inner’ socket only unless you want to use the trim ring instead of the two inners.  Worth noting: This socket is also sold at Luna Cycle (they also make their own version that is intermittently in stock) and California Ebike.  Pair this up with the torque wrench of your choice.  Mine is this bad boy.

Step 3: the Chainring

The stock BBSHD chainring is solid steel and, well… its awful.  Not only is it the ugliest chain ring in the solar system, its design is known to drop chains and generally make owners’ lives miserable.  And it also weighs a ton.

IMG_20191008_174708
Fig. 5: the Disc of Death.  A medieval throwing weapon that coincidentally is also used as a Bafang BBSHD chainring.

You want to use one of the aftermarket alternatives out there.  Generally I buy Lekkie Bling Rings, where the 42T size (there are many available) with its 18.3mm offset is the size of choice (there are now Chinese clones of the Lekkie and they are so expensive its not worth taking the quality risk to save so few dollars) and works great to correct the chain line offset that comes from the BBSHD’s secondary housing sticking out as it does on the drive side.

However, for the Mongoose project I finally had a bike whose frame would allow me to use the other king of the BBSHD chainrings:  The Luna Eclipse.

As you can see in the gallery above, the Luna Eclipse ring has a serious inward offset.

24.8mm in fact, which brings it in more than 6mm close to the frame than the Lekkie can.  Add to this the unique, wicked narrow-wide tooth profile on the Luna that effectively means you will never, ever suffer a dropped chain.  Lastly, needless to say, it looks gorgeous.  But, the biggest deal is the enormous offset, which allows perfectly centered chainline… not easy to accomplish on a BBSHD build.

Step 4: Size and attach the Chain

I made this part difficult to illustrate because I put the bags on – so they don’t come off – before I took pictures of the chain and derailleur.  I’ll do my best to get through this without good pictures.

First and foremost, let me say that a great many builders get chain length completely wrong.  Unfortunately, a lot of them may have gotten their examples from bike manufacturers who skimped on chain length and did ‘good enough’ instead of ‘best’ to save a few bucks.  What do they get wrong?  They let the derailleur cage stretch forward; not understanding that it is the cage’s job to wrap excess chain… so let it do its job!  The only reason to stretch the cage is to use less chain (i.e. you are a bike manufacturer out to save a nickel or so).  By the way, stretching the cage also extends its springs, and which lasts longer… a device whose springs are at rest, or that are under stress?

Here’s just about the only published source I’ve seen that boils down how to do it right, although they don’t shout out to the high heavens what they are doing right that so many others get wrong.

I’ll lay out the process very briefly here:

  1. Shift down into your highest gear (your smallest cog, the furthest outboard from the frame)
  2. Run your new chain thru the drivetrain, so the ends are down at the bottom and halfway between the front chainring and the rear axle (this is simply the easiest location to work with).
  3. Pull the chain together so the rear derailleur cage is only slightly tensioned.  That means in the case of the Mongoose, where we have a long cage derailleur added into the mix, that the cage is pointed straight back behind the bicycle.  Believe it or not, thats how its supposed to work on the low gear.  The cage is wrapping as much chain as it can while still maintaining tension on the chain.Here are low and hi gear shots taken at two different chain lengths on a different bike.  On the top setup, I was a little too generous.
  4. Assemble your chain with your master link.

Forget about all the other tricks associated with determining chain length.  If you have a single front chainring, this is all you need to do.  Ensure there is sufficient tension on the little cog in back.  At this point you have added as much chain as you possibly can, so the big chainring will have to take care of itself.  If you find you cannot wrap enough chain to make your big rear cog work, then that usually means you need to step up from a short cage to a mid-length, or from a mid- to a long cage.

Step 5: Cable Hookups

Sidebar: Why I used a triangle bag.
For this build, I chose to use a battery that is NOT hard surface mounted on the down tube of the bike (for now I am using the very safe, waterproof, crashproof LunaCycle Wolf Pack but I am not using its excellent magnetic mount).

I have decided to use a triangle pack simply because it hides a multitude of sins.  In particular it saves me the trouble of routing cables neatly and tidily across the frame, which in the case of the Envoy is made more difficult to do cleanly since there are no cable guides to piggyback onto, thanks to its internal cable routing.  the use of a dedicated ebike frame bag means I can just run my wires thru one hole and out the other without worrying about what they look like inside the bag.

A triangle bag also helps me with my goal of easy portability:  Long term I have a very specific small battery in mind that will sit inside of a cloth MOLLE pouch.  it will be a snug fit inside the triangle bag (thus no need for a hard mount), and when I am going into the store, I just unzip the triangle, grab the pouch, haul the battery out and set it in my shopping cart.  That internal pouch means my fellow shoppers are not looking at a bare high voltage battery with red wires sticking out one end.

Getting a triangle bag to fit your frame is often a challenge.  getting one with ebike wiring cutouts doesn’t make things any easier.  Especially since the Envoy’s triangle is relatively large.  larger than pretty much all mass produced battery bags.  Except one.  The FalconEV triangle bag is one of the largest bags of its type and just happens to be a nearly perfect fit for this frame.  I’m using that bag here.

Hook up the Speed Sensor

I was pleasantly surprised to find I did not need to use an extension on the Envoy’s mid-tail frame.  However it was a close call.  The speed sensor must pass within 1/4″ of the magnet to be reliable.  To accomplish this I used a trick I have used with fat bikes.  I first wrapped the chainstay in a bit of silicone tape to make it grippy.  Then I used more silicone tape to affix a small rope/cable crimp sleeve onto my chosen spot on the chainstay.  the rounded exterior of the sleeve will mimic the chainstay when I stick the speed sensor to it, and finally a couple of zip ties in the prescribed loops on the sensor lock that little sucker down so it works perfectly and is going nowhere.

Connect the Battery Adapter

The motor comes with the ubiquitous red and black wire for its power feed, terminated by a pair of quasi-standard 40a Anderson powerpole connectors.  Actually, the Andersons are kind of dated, with an XT90 (and particularly the anti-spark XT90S) being a much more common, reusable and weather resistant connection to your 48v or 52v battery.  My kit from Luna Cycle came with another length of red/black wire, also terminated with a pair of Andersons on just one side and bare wires on the other.  The point of this ‘pigtail’ is to connect to whatever your battery wants as a connector.  Most likely that battery is going to have either another pair of Andersons attached to it, or a (hopefully female, for your sake) XT90S.  You use the Anderson pigtail to connect to another pigtail that uses whatever your battery wants.  Yes, this means you either need to a) solder or b) crimp together the two to make a proper connection.  If you have no skills in this regard, you can probably find a pre-made cable somewhere for an exorbitant sum.  Make sure it uses 8 or 10-gauge (thick!) wire.

(I Didn’t) Install the Gear Sensor

One way to trash your bike with a mid drive motor is to go hammer down on the throttle and then try to shift.  the bike will shift alright, with about 1000w of power behind it.  The sound that makes will tell you what a horrible thing you just did, and if you want to walk home some day with your snapped chain lying somewhere behind you on the ground, go right ahead and keep doing that.

There are many ways to keep this from happening.  Most of them old-school techinques developed before gear sensors came into common use.  Since I started riding before that time, and I already know you Don’t Do That by instinct, a gear sensor is to me, optional.  And since I am lazy I didn’t feel like wiring it in.  For others – especially for those new to mid drives, its probably best to wire in the sensor.

Here’s what you do:

  1. attach the gear sensor to the motor – its the yellow cable end.  I am assuming you have one of the more modern BBSHDs and you don’t have to mess around with any sort of Y connector.  So we’ll pretend that problem doesn’t exist.
  2. Once the sensor is attached you now know where it can go on your frame.  See where it will line up best with your rear shifter cable, because you are going to fit it inline into/onto the cable.  So… pick your spot.
  3. Disconnect your shifter cable from your derailleur and pull the shifter cable out of the housing.  If you were following my prep steps above, you have already done this as I didn’t include a ‘replace the cable’ step there for this reason.
  4. With the shifter cable out of the housing, cut the housing at the place where you intend to install the gear sensor.  Place a shifter cable ferrule over each cut end of the now-separated cable.
  5. Run the shifter cable down the housing until it exits the cable where you cut it.  Push it out some more and then work it into the gear sensor, and out thru the other side.  Once you have it thru the sensor, run it in to the back portion of the separated cable housing and all the way back down to the derailleur.
  6. Cinch everything up so the sensor and the cable housing are all nice and snug against one another, and the now-inline sensor is placed as unobtrusively as possible.
  7. Reattach the derailleur cable.  Job done.
  8. Optional:  the gear sensor is nothing more than a little wheel inside of a box that senses when the cable is moving when your shifter and derailleur are pulling on it.  As soon as the little wheel detects motion, it kills the motor and its torque for a split second so you can shift – even under full throttle – safely.  So… that little wheel can get crusted up with grit and dirt, and fail.  Early designs were infamous for this.  Modern ones seem to be free of the problem but be sure… spiral wrap that rascal in some silicone tape to seal out any possible crud ingestion so it will work forever without maintenance.

Here’s what I do instead:  While pedaling I …

  1. Stop pedaling and while not moving my feet
  2. Click my shift (always just one gear)
  3. Perform a single pedal rotation to make the shift happen (this is not enough to engage the motor with any worrisome level of torque) and stop pedaling.  the motor will engage and spin a little further, completing the shift as part of its brief power up/power down.
  4. Resume normal pedaling

That sounds like a lot but it takes about 2 seconds total.  Other folks have perfected the use of a brake lever as a clutch.  With some experience in the saddle, they can gently engage the brake lever just enough to engage the safety cutoff without engaging the brakes.  Then they shift and release the lever.  Old school stuff that works great.  You decide what is best for you.

Wiring Harness Cable

A nice thing about the BBSHD is it has a single bundled cable that is meant to run up from the motor to the front of the bike, under the handlebars, where it splits off to connect up to the throttle, display and both brake safety cutoffs.  As clean as you can get given the multiple connections.  Just plug the harness cable into the motor, run it up and into the top rear of the bag, and out the hole in the front side.  The remaining dangly bits will connect right up to what you stick on the handlebars shortly.

On the Handlebars:  Display, Brake Levers and Throttle

These are pretty straightforward.  Place them where convenient for you, connect the color-coded plugs and job done.

Thats it!  You’re done.  I can guarantee you this:  It takes more time to write this article than it does to install the motor.  I think all in I needed only a couple of hours for the motor part, not counting all the extra work I made for myself swapping out everything else on the bike.

Li-Ion Ebike Battery Charge Charts

Quite some time ago, I produced a series of charge status charts for a variety of common lithium-ion battery voltages.  They’ve become a fairly common link to help folks out on various Facebook groups who use these battery voltages in their ebikes.

I built them using Google Sheets, so they are not web pages, which I suppose has kept them from being widely linked in search engine results when people are looking at such things.

Here for the first time are direct links to the charts on a normal web page.

Remember, these charts show ballpark values that are as numerically correct as they can be.  Individual cell characteristics will cause variation in the numbers here.  Read the notes on the chart for a little more detail.  Bottom line: cells degrade differently, but imperfect charts like this are still good baseline references.  Use these and teach yourself how to read the voltage gauge on your display screen.

36 Volt (10S) Battery Charge Chart

The first link is to the lowest voltage:  36v.  Generally this is the lowest voltage you will find on an ebike.  Note that its called ’36 volt’ but really that is the ‘nominal’ value.  A 36v battery is actually fully charged when it is at 42.0 volts.

36v_Thumbnail.jpg
Click on the image above to be taken to the actual 36-volt battery charge chart.

48 Volt (13S) Battery Charge Chart

The next common size is 48v.  These batteries are fully charged at 54.6 volts.

48v.jpg
Click on the image above to be taken to the actual 48-volt battery charge chart.

52 Volt (14S) Battery Charge Chart

The next battery voltage is 52v and very common.  52v batteries will work on systems designed for 48v, and why is easier to understand when you become aware that a ’48v’ battery really tops out at over 54 volts.  A ’52v’ battery tops out at 58.8v, so it essentially lets you use a 48v system for a longer time at higher voltage levels that it is already designed to utilize.

52v.jpg
Take a guess what you are supposed to do to see the 52-volt battery charge chart.

60 Volt (16S) Battery Charge Chart

With a 100% charge voltage of 67.2 volts, when you have one of these you are more or less now using high voltage electricity

60v.jpg
Yup.  Click it.

72v (20S) Battery Charge Chart

72v.jpg

 

An Ultra Reliable Ebike Battery Charger…

… Some Assembly Required

I originally published this article in the Sondors Owners Forum.

Please take note:  What I am describing here is not for everyone.  You need to do this right.  If you have any qualms about doing this kind of basic electrical work, don’t try it.  Mean Well LED power supplies have been used by the DIY ebike community for years.  The concept is not new.  The weak link here is you and if you screw this up consequences could be profound   If you know your way around a crimper, or a soldering iron… great this will be easy.  Kid stuff.  If not, don’t pick this project as your first learning experience.

Why reinvent the wheel here?  What benefit could be gained?

Ebike battery chargers tend to be dodgy.  The interwebs are filled with stories of frequent flyers whose chargers keep dying.  Its either a dead fan that in turn lets the charger heat up and fry, letting the smoke out of the internals (never a good sign) or perhaps the most common:  the charger stops cutting off at its cutoff voltage and keeps on charging … with potentially catastrophic results.

So… what is better?  You can see that in a popular commercial battery charger:  The Grin Satiator.  Its so efficient it needs no fan to cool it (or to fail).  It is also largely weatherproof and highly reliable.  The only demerits it gets from users – which have largely gone away over time – are programming/firmware issues.

Grin-technologies-Cycle-Satiator-24-48volt-waterproof.jpg

Oh and its cost is US$300+  once you figure in a programming cable, along with a couple of adapters.  I bought one.  It works perfectly.  But with an AWD bike with 2 batteries that I ride every single day and charge both at home and at work, I found convenient charging means walking up and plugging in.  Not carrying chargers with me, unloading them, setting them up etc.  So 2 batteries x 2 locations = four chargers.  $300×4= not happening.  And I carry a charger with me in case I get stranded.  $300×5=crazy talk.

What to do?  Use the same core hardware that gives us the $300 charger but without the fancy user interface.  That costs around $40.  We won’t have a fancy display screen or onboard memory, but it will still be adjustable with a screwdriver.

IMG_20180903_083144.jpg
The dollar is for scale – and to remind me what money looks like since I can’t seem to stop handing it away.

I have worked with three different models that can serve my purposes.  Remember that volts x amps = watts and this will be important when figuring out what to set your charger for:

CLG-150-48A

  • Available regularly on Amazon for about US$55
  • Rated to 150 watts
  • Rated as adjustable from 40 to 56v but actually adjusts from 39v to 58.1v
  • Usable as an 83% to 100% charger for a 36v battery
  • 80% to 100% for 48v battery
  • 80% to 96% for a 52v battery
  • Minimum amperage selectable is about 1a
  • Lower wattage rating means it must be set to lower amperage on 52v batteries (2.5a max for a 52v battery)
  • Designed for LED lighting and ‘moving sign’ lighting applications
  • IP65 rated for indoor and outdoor use.
  • Usable at EU or USA voltages.
  • Mean Time Between Failure (MTBF): 303,700 hours.  Yes, really.
  • Spec Sheet Here

HLG-185H-54A

  • Available often on ebay for about US$40.  Normally on sale in the $65-$75 range.
  • Rated to 185 watts
  • Rated as adjustable from 49 to 58v.  Actually adjusts from 48.3v to 60.0v
  • Usable as an 80% to 100% charger for 48v and 52v batteries.  Not usable on 36v systems
  • Higher wattage rating means can be used for faster charges than 150w CLG.  Typically this is a good 3a charger for 52v batteries.
  • Adjustable to very low current (about 0.85 amps) for safest trickle charging, ever.
  • Seems to ramp power down more slowly and precisely than 150w CLG as it approaches target voltage (not a meaningful feature; just an observation).
  • Mean Time Between Failure (MTBF): 192,200 hours, or almost 22 years of continuous use.
  • Designed for LED lighting and street lighting applications
  • IP65 rated for indoor, outdoor and wet/hazardous locations
  • Usable at EU or USA voltages
  • Spec Sheet Here

HLG-320H-54A

  • This is the Big Daddy.  As in bigger and heavier and more power output.
  • Street Price around $90.  Often available on EBay for around $50, even down to as little as $25 each if someone is selling off a pair of them wired together as Zero e-motorcycle chargers.
  • Essentially same specs as the HLG-185H-54A but is instead rated for 320 watts
  • Current can only dial down to about 2.0 amps.  But the high wattage rating means it can be dialed UP to make it a 5 amp charger (only aftermarket battery plugs like XT60 or Andersons are able to safely handle this current level).
  • MTBF: 157,100 hours (just under 18 years of continuous use)
  • The 320H units I have bought on the aftermarket were originally wired together in pairs in series and used as onboard 115v Zero Motorcycle chargers
  • Spec Sheet Here

– – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – – –

HLG-480H-54A

  • New addition to the product line since this thread was originally created.
  • $125 street price.
  • 480 watt capacity
  • Minimum 4.4a rated current rate (max 8.9a) means not safe for anything but an aftermarket battery connector
  • Strictly a fast charger for a BIG aftermarket battery that can safely take 4-9 amps.
  • Not for novices – know what you are doing insofar as batteries, chargers and battery chemistry is concerned before you go this big.
  • MTBF 345,500 hours
  • Yes it is big.  Not the sort of charger you want to be carrying around.
  • Spec Sheet Here

HLG-600H-54A

  • New addition to the product line since this thread was originally created
  • $170 street price
  • 600 watt capacity
  • Minimum 5.6a rated current rate (max 11.2a) means not safe for anything but an aftermarket battery connector
  • Multiple power outputs enable charging multiple batteries at once, so may make some sense for multi-battery households.
  • Not for novices – know what you are doing insofar as batteries, chargers, charge capacities and battery chemistry are concerned.
  • MTBF 76,900 hours (a mere 8.7 years of continuous use)
  • If you thought the 480H was big wait until you see this one.
  • Spec Sheet Here

SIDEBAR
IP65 Enclosure – IP rated as “dust tight” and protected against water projected from a nozzle.  So these chargers are safe with:

  • Garden hose spray – Yes
  • Ocean waves – Maybe 
  • Bottom of fish tank – Hell no

Myself, my bikes have 52v batteries.  I do use a couple of CLGs at work for my charging station there but only because I hadn’t found the HLG-185’s yet.  The HLG-185’s are ideal chargers as they can charge at levels safe for the Sondors battery plugs (3a max) and can handle any voltage asked of them for a 48v or 52v system.  If you have an aftermarket battery that does not use the pin plug as does the Sondors batteries, then you almost certainly have an Anderson Powerpole, an XT60 or an XLR connector.  Those plugs can handle the higher amperage the 320 is capable of delivering.  I use a 320 as a travel-with charger under the theory that if I am stuck somewhere I want to grab as much charge as I can, as fast as I can.  But a 185 is perfectly capable of being a 3a charger and weighs probably half what the 320 does.

IMG_20180902_085021.jpg
This pic was supposed to show everything needed to make the charger, but I sort of overdid it.  No need for a whole bag of connectors, or heat shrink, and that grey cutter isn’t needed either.

So… enough details already.  Lets make a charger!  Here’s what we need:

  • A Mean Well power supply.  The process is identical for all three models.
  • A pigtail’d grounded electrical plug.  They are sold on Amazon typically as replacements for corded drills and similar power tools.  NOTE: I am using a USA standard plug, but these units are made to accept worldwide voltage/current so just go to your local hardware store and choose your local version of a pigtailed, grounded power cord if you live outside the USA.  Oh, and read the spec sheet to confirm what I just said applies in your country.
  • A digital watt meter to tell us what we are outputting to the bike.  For almost all of my chargers I use $15 inline watt meters.  This is optional but very desirable.
  • An interface from the charger to the battery.  I will use an XT60 as the direct connection, which is what a lot of aftermarket batteries use.  You can then plug just about any adapter into that for your Sondors or whatever else you have.  Note in the picture above, bottom center just to the left of the little adjustment screwdriver we will keep with the charger, that there is a pin plug adapter for use with Sondors batteries.  That one came from Luna Cycles.

A note on battery/watt meters.  Here’s the short version:  They suck.  Or more accurately they are oftentimes off by  a bit, and there is no way to calibrate them.  Its not uncommon to see a battery meter accurate to within 2%.  That sounds ok unless you are charging to 58.8v, which could be 59.98v with a 2% error and that is very, very bad.  So you want to take a multimeter or similar *known safe benchmark* (in a pinch the reading on your LCD screen will work once you have disconnected any charger from it) and use it to learn where your chosen meter is in terms of its accuracy.  I do this and then I take a labelmaker and make a label telling me how much a battery is + or – actual voltage.

So for example, if my target voltage for a 52v battery is an 80% charge of 55.4v, and my watt meter is reading 0.50v higher than it should be, then I create a label that says

+0.50v

  • UPDATE:
    This link directly below is one of many cheap Chinese watt meters.  the last two I have used, purchased across a span of 4 months, exhibited a new and consistent behavior:  Plug them in and they are WAY high, by like 1.2v.  But sit and watch the meter over a span of about 5-10 minutes and you will see it slowly auto-correct itself down to a steady reading.  this steady reading will still be off by a bit but not so bad… these last two meters were both off by only about 0.20v… so I recommend these as your best option – and recognize they have a calibration stage at startup that you need to wait out.
    https://www.amazon.com/dp/B01LVTST80/ref=cm_sw_em_r_mt_dp_U_coFVCbD1VDC1Q

So now, we have our parts in hand and its time to assemble them.  In order the steps are

Step 1
Attach the pigtail’d cord to the input side of the Mean Well unit.  For the USA plug and the Hanvex drill cord I have been using the wire sequence is green (cord wire)  to green (charger wire) for ground, black (cord)  to brown (charger) for AC+ and white (cord)  to blue (charger) for AC-.  Note that the wire colors are noted on the charger left side, but as ACL and ACN.  DO NOT SCREW THIS UP.   These are international standard designations and colors which as usual the U.S. does not follow.  If you want to check my work, start googling.  Myself, I use marine heat shrink butt end connectors to connect the wires.  I also use rather expensive electrician’s-grade crimping pliers.  There is a big difference between proper crimping pliers and … well, pliers.  Use the right tools for the job.  After I crimp, I heat shrink the connectors, add heat shrink around each individual wire and then do a heat shrink around that entire assembly.  How you do it is up to you (i.e. soldering or whatever).  Remember that this is mains power you are fooling with here so get this right.

Step 2
Attach a battery side plug.  In this case I am using a male XT60 which both works for my aftermarket batteries that have female XT60 charge plugs, and my Sondors bottle batteries where I can use my XT60-to-pin-plug adapter.  Same procedure as in Step 1 although a little simpler as there are only two wires.  Note that some of these charger units do not use red and black wires.  If you are not familiar with what the colors mean, the casing on the unit specifically tells you which wire is which (V+ and V-).

When you are done, you will have something looking like this:

IMG_20180902_093002.jpg
A completed 320w charger without the watt meter stuck on the end

Step 3
OPTIONAL – attach an inline watt meter to the output side of the Mean Well unit.  This is your power display.  I call this step optional because you could just calibrate your charger output once and not use a meter to monitor progress (easy enough to turn on the bike display during charging, which will hurt nothing).   Myself personally, even though meters are pesky insofar as getting them calibrated, I vastly prefer to have a real time progress monitor I need only glance at.
NOTE:  Source side of the meter gets connected to the charger.   Load side goes to the battery.

Step 4
OPTIONAL – make an extended output cord.  Essentially one big extension cord on the battery side. You’ll know real fast if you’d like to have one of those as whatever you made doesn’t reach.  You could just hardwire this to your output lead on the charger.  But then you are stuck with that length alone.  I prefer to make a cable as I have no problem using a couple of 12 AWG XT60 pigtail ends to make a dedicated extension.

Step 5
Connect an interface to your battery.  For a Sondors, this is a bottle battery connector.  For many batteries the generic standard is a male XT60 connector.  You can either buy a direct-connect bottle battery adapter (see link) or connect a male XT60 pigtail and then buy an XT60 Female-to-bottle adapter.  Doing it the latter way makes your charger able to connect to any battery (if you have another battery with say XLR connectors you can make an XT60-to-XLR adapter via a couple of pigtails).  You just swap in the adapter you need.  In this case I am picturing a Luna-sourced XT60 female to pin plug adapter.  A different source for the same thing is in the parts list below

Step 6
Go out and buy a little Phillips head screwdriver.  This tool will live with your charger forever so you should buy a new one unless you have an extra already. Its a must-have for the next step.  Also required if you plan on changing your settings (lets say you want to charge 80% one day and 100% the next).

If you have performed all of the above steps, you now have a parts pile that looks like this (well sort of, the meter and the charger have already been labelled with calibrations but just pretend we haven’t done that yet):

IMG_20180904_074621.jpg
Charger, adjusting screwdriver, extension cable and watt meter complete with sticky note showing how far off it is.

Step 7
Dial in your output voltage.  Once you have connected an AC plug, and a battery side connector AND connected the inline watt meter, you simply have to plug the new charger into the wall.  Amps will read zero and volts will read whatever the unit is currently set for.  See the little rubber whatsits that are capping the voltage (Vo ADJ) and amperage (Lo ADJ) adjustors?  Pull those off and stick the screwdriver into the Vo ADJ hole.  Twiddle it around gently until you feel it seat into the adjustor.  Now turn it first one way, then the other.  Watch the voltage readout on your meter.  One way goes up, the other down … and the directions are different on my 185’s and 150’s vs. my 320 so you figure out what direction does what yourself with your own unit.

Step 8
Calibrate your meter to reality.  Remember what I said above about meters.  You need to figure out how far off your meter is from your display.  As you can see if you look closely above, this meter is off by +0.50v.  Thats a fair bit.  The good news is when these types of meters are off, they are consistently off so you just need to know by how much (and if you can find a meter that is consistently accurate tell me.  I can’t find one at any price).  this is a pain but you only have to do it once.

Step 9
Dial in your output amperage.  OK… moment of truth time.  You are plugged into the wall.  Time to plug into your battery.  Maybe you should do this out in a field with a long extension cord.  Don’t do it in the baby’s nursery or in Grandma’s bedroom while she’s asleep.  Plug the battery in and now watch the meter.  The voltage switches to now show the battery state of charge.  The amperage comes to life and shows the current level (amps) being fed into the battery.

Once again, like you did with the voltage adjustment, use your screwdriver this time in the Lo ADJ socket and twiddle it until you see the safe amperage rate you safely want to safely run your charger safely at.  Did I forget to mention safety?  And volts x amps = watts ?  Pay attention and get this right.  If your meter is off – especially if it is reading lower than actual voltage – you will want to find out by what percentage it is off and adjust your indicated meter amperage less that percentage amount.

IMPORTANT SAFETY TIP:

  • Your charger does not switch its power feed on and then off like a light switch.  Instead, it will slowly ramp down its current delivery level (amperage) as the battery approaches your target voltage.  So that means if you plug in a battery that is fully charged or nearly fully charge, you will get a really tiny reading of current going into the battery – and this will give you a false idea of the amperage your charger is set for.  Because of this, when performing calibrations you must have a battery that is at least a couple of volts low.  At least.  If you are charging to 54v (100% charge on a 48v battery) then plug in a battery at no higher than, say, 50v state of charge.

If you have a Sondors battery and are using a pin plug, NO MATTER WHAT make sure this value does not exceed 3 amps.  The plug can’t safely take more.  Again, remember that volts x amps = watts.  So if your 185w HLG-185 is feeding the max of 3.45 amps, that means at 58.8v it will be sending 203 watts which exceeds its 185w rating and thats VERY bad.  Here again.  Use your brain and don’t screw up.  Best to leave a safety margin.  For example I have one of these set to a ‘full’ charge of 58.3v and 3.0 amps.  175 watts.

Step 10
Add a carrying case?  Your basic MOLLE water bottle bag will fit this all beautifully.  the slightly larger Condor bags available on Amazon will do so with a little more fudge room.  I got two green ones on sale for $5 and $8 respectively.  Sometimes they are more.  Happy hunting.
In the end what do you have?  A charger that you can expect to be reliable literally for years.  Not necessarily cheaper, but dependable.  If you buy this once you won’t have to buy it again in 6 months or a year… and thats the usual story out there in ebikeland for the more demanding users in the DIY world.

Parts (remember oftentimes you can get these chargers for a lot less on clearance on Fleabay).  Especially the HLG-185 which is commonly used in street lights):

Mean Well CLG-150-48A
http://a.co/4IaEpLU  ($55)

Mean Well HLG-185H-54A
http://a.co/dy3U4qI  ($68)

https://ebay.us/H3w74L ($56.50)
https://www.onlinecomponents.com/mean-well/hlg185h54a-43123124.html ($48)

Mean Well HLG-320H-54A (84.10)
https://www.onlinecomponents.com/mean-well/hlg320h54a-43123431.html ($82)

Hanvex 18awg 3-prong AC power cord, 6ft, pigtail’d
http://a.co/iNJAOMX ($8.99)

XT60 male and female pigtails (need 5 total if you are using an inline watt meter, extension cable and xt60 lead for battery)
http://a.co/fuzw4cM (8.99)

Inline watt meter
http://a.co/b0iE00l ($15.99)

Option: Female XT60 to male barrel plug adapter
www.progressiverc.com/female-xt60-to-male-barrel-plug.html ($4.99)
https://lunacycle.com/xt60-female-to-barrel-male-plug/  ($6.95)

Carrying case for the CLG-150 or the HLG-185H- any similar bottle carrier should do – ($8.95)
http://a.co/3ln2fd6

Carrying case big enough for the HLG-320H (really, this is the best case for all of them, especially if you get them on sale?
http://a.co/d/6bSx1lt

IMG_20180904_074404.jpg
Everything nice and neat in its own traveling package
IMG_20180904_074322.jpg
Thats a size 10 1/2 shoe there for scale.

UPDATE:

I had the opportunity to make another charger over the weekend for my daughter and son-in-law, who have matching Original ‘X’ models.  They also live in the EU and as such I needed the appropriate plug – the charger will auto-sense the voltage coming in and adjust accordingly.  So for those of you folks outside the U.S., here’s what one looks like.

My daughter’s locale uses a 2-prong grounded ‘Schuko’ type plug which I got from a UK seller on Fleabay.  One nice thing about using international parts is they conform to the same international specs.  So there is none of the translation necessary to pick which wire goes to where.  Just match the colors and you are done.

IMG_20181014_105356.jpg
Proof I am not colorblind (ps that dress was white and gold)

This time I took the time to take pics before and after during assembly.  The heat shrink and adhesive on the marine-grade splice connectors make for a very solid connection.  There is a trick to doing the best crimping:

  • do it on the very ends
  • don’t crimp so hard you tear deeply into the plastic covering the splice
  • ensure the pointy prong on your crimper faces AWAY from the other wires so if you do overcrimp and tear into the plastic, you won’t expose metal facing the other wires.
  • Use a halfway decent crimper.  I think I made this point in the original post but it bears repeating.  Use the wrong tool for the job and your results will suck.

There is also a trick to heating the adhesive connectors – First, use a nozzle on your heat gun that narrows the heat exhaust so you can better direct it to a small area.  Next, heat the ends that you actually need to shrink up and grip the wire.  Stay away from directly heating the metal center.  If you do that, any tearing of the plastic over the crimp tends to actually seal itself.  If you heat the center, those tears will break open further as the adhesive plastic shrinks from the heat.  Its actually pretty easy to do… you just have know to do it… and now you do.

imageproxy.jpg
If you live in the USA, no I did not use the wrong plug.

Heat shrink over top of those adhesive connectors and you have a stable, solid connection you need to look for to notice.

IMG_20181014_110354.jpg

Do it again for the XT60 ‘universal’ output connector.  Make sure that external heatshrink is plenty long.  In this case I made sure I had plenty of exposed wire on the end because I like the flexibility.  If I wanted to reinforce it and maintain that flexibility, self-adhesive silicone tape (sticks only to itself; spiral wrap it around the wire) is the perfect solution.  The Sondors-compatible bottle connector I chose for this charger had a male plug end on it, so I needed to make another connection using a short female-to-female XT60 extension.  It is important to get your genders right on a charger.  You do NOT want a male XT60 or male anything else exposed on the battery side as an arc between the terminals is much more likely on a male plug, and that can destroy your battery.

IMG_20181014_113508.jpg

Here’s the whole thing put together with a meter added to the end and the Sondors-compatible 5.5mmx2.1mm barrel connector attached.  The meter is showing it is configured for an 80% charge on a 52v battery.  After I took this pic I realized I needed to set it up for a 48v battery on an Original X and changed the voltage on the charger and the label on the meter.

IMG_20181014_140924.jpg

Mongoose – Chapter 8 (Build Sheet)

I’ll try and keep this as simple as I can and simply list parts in what passes for a table in basic WordPress which does not support tables.  Over time as this project is filled out the reasons why I chose what I did will be covered in the various articles

Mongoose Envoy Bike                  Amazon                749.99
Magura MT5 disk brake set            bike-discount.de      137.00
Thudbuster LT 27.2 XL                Amazon                119.99
ISH-203 203mm rear disk adapter      bike-discount.de        6.86
QM5 203mm front disk adapter         bike-discount.de        6.86
Tektro 203-17 downhill rotors (2)    ebay (hi-powercyles)   42.40
Continental Contact Plus City (2)    Amazon                 69.54
   26x2.20 tires
Sunlite thornproof 26x2.35-2.50 (2)  Amazon                 33.90
   Presta valved inner tubes
BBSHD motor kit                      Luna Cycle            699.95
   68-73mm standard motor
   Mounting hardware
   wiring harness
   speed sensor
   basic crankarms
   Luna 500C mini color display
   Universal thumb throttle
Second Bafang inner lock ring        Luna Cycle              5.95
Battery Solution
   Wolf V2 52v 12ah battery pack     Luna Cycle            549.95
   Potted, QD mount, 50a BMS and 
   Samsung 30Q cells
    -OR-
   52v 12.5ah battery pack, basic    Bicycle Motorworks    369.99
   pack construction, 50a BMS and
   Samsung 25R cells
Luna Eclipse chainring               Luna Cycle             99.95
   Anodized black face                  
   Anodized gunmetal chainring
Lekkie Buzz Bars (crankarms)         California-ebike       99.00
SRAM EX1 144Lnk mid-drive chain      Amazon                 28.99
 -OR-
KMC X9.93 (7 feet - more links)      Luna Cycle             57.75
Shimano HG400-9 12-36T cluster       JensonUSA              25.99
 -OR-
Shimano HG400-9 11-34T cluster       Amazon                 24.60
Shimano RD-M591 9spd derailleur      JensonUSA              37.99
MicroSHIFT TS70-9 shifter            Amazon                 22.88
ROCK BROS Wide Platform Pedals       Amazon                 21.99
Raised rear deck
   Moose Skateboard Deck             Amazon                 31.50
   aluminum unthreaded spacer        McMaster-Carr          10.88
      13mm OD, 25mm long, for 
      M5 screw (qty 8)
   countersunk M5 wshrs (qty 10)     McMaster-Carr           4.84
   stainless hex/flat head screw,    McMaster-Carr           9.35
      M5x55mm (qty 25)
Front Rack
   Axiom Streamliner Front Rack      Amazon                 47.99
   Delta AxelRodz skewers            Amazon                 13.60


Mongoose – Chapter 4 (Motor Choice)

I have chosen a mid-drive as the best tool for the job on this bike.  It will not only be hauling cargo, it has to be able to do it in a steep hilly area.  If I was building for flat ground, a maintenance-free direct drive hub around 1.5-2kw would be the answer.

I’m not doing that though.  So mid-drive it is.  Now, which one?  there are plenty on the market and I have owned and in some cases still own several different ones.

Originally when I was planning this article, I was going to describe all of the major players in the market, and why I chose the one I did.  Instead I’ll limit my scope to the top two choices to better stick to the subject.

Second Choice: Bafang BBS02

This should be the first choice for most people.

I do want to say that as far as I know, as Bafang’s largest USA dealer, only Luna Cycle sells a BBS02 I would want to buy.  Luna uses their buying power to spec more robust controller internals that keep the BBS02 from frying its controller under sustained load … that was something of a known drawback of these motors in their heyday.  Also, Luna’s pricing strategy means their BBS02 kits are among the cheapest, if not *the* cheapest, BBS02 on the USA market.

Bafang mid drive motor kits are at the bottom of the the difficulty curve in terms of installation.  With one of these, you can have a bike up and running in an afternoon.  You’ll find a zillion Youtube videos and blog entries telling you what to do and how to do it, as well a multitude of experienced users in online communities ready to help you through whatever specific, quirky question you may have.

bbs02

Sidebar:  Bafang is a mainland Chinese motor manufacturer that is essentially the 800 lb gorilla of ebike motor manufacturers.  If you think companies like Bosch are market leaders, their volume is a fraction of Bafang thanks to their massive installed base in the Far East.  Bafang motors, while not perfect, are typically overbuilt, rugged, dependable and not very exciting.  Workhorses.  Chinese products have a reputation for ‘optimistic’ spec sheets and shaky quality.  Bafang is pretty much the opposite of that.  They underreport their motors’ capability – It seems one reason for that strategy could be so they can sell the same motor with different power ratings, with a higher price point for the ‘bigger’ one (It is alleged a certain hub motor they sell is identical across 250w, 350w and 500w ratings and price points).  Again they are not perfect by any stretch, but for the beginner, working with a USA dealer of their products (do NOT be tempted to buy cheaper from an overseas vendor), its hard to go wrong.

The ’02 has been around for awhile, and what is sold as the ‘BBS02’ in the USA amounts to Bafang’s first effort at a kit offering.  It has long since been eclipsed in terms of power, and its aftermarket support for upgrades is not what you find for its successor, the BBSHD, but still the ’02 remains a rock solid product (IF you heed the caveats above).

A 750w or 1000w ’02 would have been my motor of choice for this project, except I already own two other BBSHDs.  As such I decided to keep one set of common parts across the fleet, so to speak.  I went with the more powerful next generation: the BBSHD.  More on that below.

Another Sidebar:  If you see these motors with the label ‘8FUN’, thats Bafang’s house label.  Usually those motors were manufactured for sale overseas in the Far East.  You can also find BBS02’s and BBSHDs with private labels on them.  Thats common.  Bafang will sell a private-labeled motor to anyone who will give them a big enough order.  Some vendors – like Luna Cycle, Bafang’s largest USA dealer – private-label and sell at a competitive price to builders.  Others just add their own label and crank up the price.

First Choice:  Bafang BBSHD

Bafang the BBS02 with the BBSHD (its called the BBS03 in some overseas markets).  The ‘HD is essentially more of everything you get in the ’02.  More robust.  More aftermarket support.  More power.  And more money.  But its still a bargain, with a bare motor running around US$450.  A complete kit providing you with everything you could ask for, plus programming upgrades (Bafang motors have a robust capability to customize their behavior, ranging from total power output to precise tailoring of each of the 9 pedal assist levels). should run you about $750 before you get to the battery.

bbshd

Myself, I paid less than that simply because I already have two other bikes that use this motor, and I have quite a few spare parts on hand I can just pick up off the parts pile and use for zero added cost.

 

Mongoose – Chapter 3 (Motor Types)

I am making a particular choice with regard to the motor I am putting on the bike that is the subject of this series.  I’m familiar with the various types – their strengths and weaknesses – but for the sake of the reader who may not be, I’m going to do a quick-and-dirty on each mainstream type.  If I’m not covering that motor type (looking at you, friction motors)  there’s a good reason that further research on your part will reveal.

There are three types of motors that make it into the mainstream of the ebike world

  1. direct drive hubs
  2. internally geared hubs
  3. mid drives

Direct Drive Hub Motors

mxus.jpg

DD hubs work with no moving parts.  The axle of the bike is also the axle of the motor itself, which is just a brushless DC motor with magnets running around the outer case of the motor (the rotor), copper windings wrapped around the interior stator etc.  Apply electricity and the stator repels (or attracts) the rotor, which results in the rotor spinning, and there is your powered motion.  Relative to the other motor options, direct drive motors produce much less torque.  In order to get around that, you use a really big motor and a really big, powerful battery.  Once you get into the 3kw-5kw and greater range, with a 60v or higher voltage battery, then you are talking about a bike that is effectively a monster that can run around town at 50 mph or more.   Direct drive motors are effectively maintenance free.

The drawback to the above is that big motor with its big metal magnets and large amounts of copper wiring is … big.  Heavy.  And so is the 2XL battery you need to get solid performance out of that motor.  For a serious bike that can carry cargo, passengers etc. up a hill at speed you are probably talking about a 125 lb bike, with a lot of that weight on – in – the back wheel.  And the battery is likely going to be on your rack as well, not your cargo, reducing your carry capacity.  At lower power levels (particularly those that are less illegal than the above noted 3-5kw) you aren’t looking at that kind of weight penalty.  But you get lower performance as a result.  A direct drive motor that is not in the hi-power league will need a long run-up to get to cruising speed.  And if you want to climb a hill… well its the worst choice for that job of the motors you can choose from unless, again, you go really big.

If you want to research the ins and outs of this type of motor further, you need to also look into ‘torque arms’, why they are needed with this kind of motor, and when.

Geared Hub Motors

Geared hub motors do a good job of providing more torque than DD hubs at similar power levels.  They do this by installing a planetary gear reduction inside the motor which connects the stator to the outer casing.  the motor spins nice and fast as it likes to, so the axle in turn spins the planetary gear and this finally turns the outside casing and the wheel at a slower speed, so you and the bike can go down the road at a comfortable rate of acceleration.

img_20180704_054016
I opened this motor up for its semi-annual re-grease.  I also needed to deal with that rust you see.

Geared hubs also tend to be lighter than their direct drive cousins, which helps with range and acceleration.  At sufficient power levels – lower ones than what a direct drive motor needs to do the same job – a geared hub gives some pretty good torque.  80 Nm in the case of the above pictured motor.  Mate that to a 35 amp controller and a commonly available, medium-voltage 48v battery and you have a really peppy ebike.

Whats the down side?  Those nylon gears will take a lot of abuse (really a lot), but they won’t last forever.  Especially if you subject the motor to regular extended hill climbs, or you are subjecting it to a lot of stress… like a full cargo load.  If you try to solve this problem with steel gears you find out why motor manufacturers use nylon:  noise… and metal shavings.  You will also have to open those motors up every few thousand miles and re-grease them, as the grease perishes over time.  Lastly, geared hubs really do not exist at the higher power levels (most I have seen are the MAC motors peaking at around 1500w with a special controller to get them up that high).  Any higher than that and the gears really can’t handle the power.

Final thoughts on hub drives

Both geared and direct hub motors power your bicycle directly through your hub axle.  They are the hub in fact.  What this means is, your bicycle powertrain is entirely irrelevant to what the motor does.  The power to the ground is transmitted directly from your axle.  In fact, if you want to have some fun with your hub bike, you can remove the chain.  Then ride down the street with your pedal assist turned on, and pedal the bike.  It will work just great.  Of course you aren’t getting any exercise and likely you don’t want to ride like this, but it illustrates the fact that the traditional bicycle powertrain is not needed.

Your pedals and chain now exist almost solely to provide you, the rider, with exercise.  You put as much effort into pedaling as you want, and that can be just mild pressure or pushing hard the way you would riding an analog bike.   It is now the motor thru the axle that is doing the real transportation work.  This fact is not lost on ebike manufacturers, and this greatly reduced duty cycle means hub-based ebikes tend to have cranks, chainrings and rear clusters that would not survive long if given the hard life an old-school analog bicycle receives.

Geared hub motors need torque arms just like direct drive hub motors, albeit not so much at the lowest power levels (i.e. 250w and 350w).

Lastly, both types of hub drives have this significant benefit:  they don’t require any special knowledge or care to use.  You can jump on the bike as a complete newbie and start riding.  So long as you aren’t riding some kind of hot rod that is hard to control, you already know everything you need to ride happily down the road and not cause any issues.  That makes hub motors of one sort or another preferred for most casual riders who are not challenged by their terrain.  That isn’t true of your more powerful mid drives, but I’ll get to that below.

Mid Drive Motors

‘Mid drives’ are known as such because the motor sits at the middle of the bicycle.  Typically replacing the bottom bracket.  As much as hub drives dominate the lower-cost, DIY and upgrade ebike markets, mid drives dominate the big-name commercial-manufacture market.  In particular, and most telling as to the benefits of the mid drive, E-MTB’s are exclusively mid drives, and for good reason.

img_20190603_194127
Yes, this is an ebike (and yes, its mine).  Look closely at the chainring.  See the mid drive motor hiding behind it?  This motor peaks at around 440 watts of final output which in the USA is well under the legal limit.

Mid drives work on an entirely different principle than hub drives.  Hubs, as we noted above, power your bike thru the axle, and your drivetrain is just along for the ride.  In terms of assistance, the hub drive bike is a 1-speed, and this is part of the reason hub drives don’t do so well in hills or fast acceleration, unless you start getting into big power.

But a mid drive works just like you do:  It pours on the power thru the bicycle chain.  That means if you hit a hill, you can downshift into a lower gear, keep the chain spinning fast and get up the hill more easily.

Gee thats great.  And since much of the world has legal limitations to 250w of final drive power, you can’t really put enough power into the system to break anything.  An average person in good shape can put between 50 and 150 watts of power into their drivetrain during a ride.  Thats what an old-school analog bicycle drivetrain expects to put up with.  250w isn’t a whole lot more than that (a trained cyclist can pour on roughly 400w or so… and sprint briefly up to around 1500w… which is still not really enough to make a slice of toast).

About that 250w limit… First of all, here in the U.S. that number is typically “less than 750 watts” according to our national manufacturing/safety standard, and many of the individual states who have intersecting vehicle codes that define what is an ebike vs. a moped vs. a motor vehicle.  But beyond that, in the last year or so we’ve started to see rebellion from major E-MTB manufacturers against the almost-a-joke 250w EU limits, in the form of their completely dropping mention of the wattage output of their motors and instead referencing the Newton Meter (Nm = torque) output of the motor.

Really, torque output is what matters in terms of figuring out how much assist you are getting, and mid drives just pour on the torque.  A Bafang BBSHD, the de facto volume-sales king of the American DIY market, puts out 160 Nm continuous when it is fully utilized.  That is about double the momentary peak of what the big geared hub pictured above is capable of (and that hub is among the biggest of its genre).

So easy choice everyone needs a mid drive!  Well, not so fast.  With great power comes great responsibility repair bills if you don’t use your head.  That means build it right and learn how to ride it.

Remember the wattage output a normal human is capable of?  The level that a quality drivetrain is expected to be able to handle?  Well, the above referenced BBSHD in off-road mode, pouring out those 160 Nm, is feeding about 1500-1700 watts to the drivetrain.  Continuously.

You want to figure out how much wattage is going to your motor?  The formula is Volts * Amps = Watts.  So a 52v battery running at its nominal 52v rating, multiplied by a BBSHD running at 30 amps is… 52*30=1560 watts.  At a full charge that battery is 58.8v, so 58.8*30=1764.  Yes, really.

So, when we build a DIY mid drive bike, we first want to buy parts that are meant to take this kind of punishment.  They are out there on the market but frankly a lot of DIY builders, riders and even most ebike sellers are ignorant of this.  You want:

A good narrow/wide front chainring

Made of 7075 alloy most likely, but if you can get a steel ring, do it (Wolf Tooth is the only one I know of and they only come in 30T and 32T sizes).   This style of ring typically has wider, taller teeth that eliminate chain dropping issues.  In particular rings made specifically for mid drives are sold by Luna Cycle, who manufactures their own here in the USA, and Lekkie, a New Zealand company with a stellar reputation who sells thru ebike vendors everywhere.  The latter two names are focused primarily on the BBS02 and BBSHD markets although Luna does make rings that will work on other platforms.

An ebike-specific chain

The interwebs are filled with complainers crying about how their chain snapped.  When you ask how many of them re-used their $6 stock chain, or who just bought a ‘nicer’ bicycle chain, the numbers pretty much come close to 100% of chain failures (chain alignment will be dealt with below).  E-bike specific chains in various widths for various speeds are sold by KMC (my favorite is the X9E/E9 9-speed) in 136 link lengths.  SRAM’s EX1 ebike group has its own 144 link chain.  Lastly there are the Connex chains by Wipperman.  The common factor in why people don’t use them (besides not knowing they exist) is price.  These are US$35+ chains (you can buy smart and get them for less if you know where to shop).  But… They.  Don’t.  Break.

Lastly, you can find a specialty vendor and buy a specific length of chain all in one piece, so you don’t have to section two chains together, resulting in a weakness at the joining point.  This is by far the most expensive option.   I bought a 7-foot length of 9-speed chain from Luna Cycle and it ran me about $60.  But my Mongoose Envoy, with the long-cage Deore derailleur I added, needed 152 links to be set up right (I keep a 144-link SRAM EX1 as a hot spare and it will work fine in a pinch).

kmc x9e.jpeg
The KMC X9E.  Those mushroomed pin ends will be torn off if you use a chainbreaker on them, so if at all possible just remove links to make it shorter, don’t add any to lengthen.  re-attach with master links.

A steel cassette cluster

You have two choices for this, generally.  First is the SRAM EX1 cluster that is an 8-speed, has a range of 11-40, is made of tool steel and meant to be used with the EX1 shifter which will only shift one gear at a time.  The rub is the cluster alone runs about US$385.  Its worth every penny (I have one on my E-MTB, so it had better be), but that cost is insane.  How about spending US$15 or less instead?  Just buy any cheap Shimano rear cluster.  In particular the HG-200, the HG-400 or the HG-50.  All in 8 or 9-speed.  They use steel, not alloy, cogs, and most importantly the entire cluster is welded together into a single unit so the punishment dealt to the cassette body is distributed across its entire width.  These Shimano clusters are an excellent example of something that is awful for an analog bicycle and highly preferred on an ebike, where durability is vastly more important than light weight.

Shimano-CS-HG400-9-9-speed-Cassette-silver-12-36.jpg
This is the 12-36T 9-speed cluster I am using for my Envoy build.  Don’t look for any bolt heads to allow disassembly.  Only the 12T cog is separate from the 1-piece body.

A steel cassette body

Here again, what sucks for a bicycle is great for an ebike.  A steel cluster will last.  An alloy one won’t.  Take a look below.  On the left is an alloy DT Swiss cassette body with about 1600 miles on it.  It comes from a DT350 hub, which is very near the top of the line as bicycle brands go.  The cassette cluster I used was a welded steel Shimano, so those notches you see still tore into it despite the gentler damage the welded cluster does.  I almost exclusively used the 11T small gear on this bike and on the far right you can see that section is torn into further than the rest (the last cog is free floating so no help from the welded together body on that one… we’ll come back to this and discuss further below).

bodies
eek… and this happened with a 1-piece cluster, too.  With a nicer cluster the damage would have been much worse.

 

On the right side is a steel version of that DT350 cassette body.  Unlike the alloy version, it is much heavier – and I expect it to last forever.  Worth noting:  DT Swiss has now released a “Hybrid” version of the 350 hub specifically meant for ebikes.  It includes the steel cassette body out of the gate as just one of its durability improvements.

Get a ‘star ratchet’ rear hub

There aren’t many of them.  DT Swiss, Chris King and Hope are the only big names that sell freehubs with this sort of splined engagement instead of the traditional 3- or 4 pawls.  A splined engagement provides dramatically better contact with the hub from the cassette.  See that nearly-ruined cassette body above?  the stock 18-tooth star ratchet wheels inside went right back into the bike with the new steel cassette body… they were still perfect.  Since DT’s patent on their system ran out, other makers have begun to use it and you can now find star ratchet replacements and complete hub systems on Ali Express etc.  Myself, I still buy DT Swiss 350’s.  But you can save hundreds with the Chinese hubs.

Uf26dd6f62c374343bc8c305ec79489f7R

Learn how to ride it

I mentioned this briefly above.  With a couple of narrow exceptions (don’t mash the throttle going up a long hill) you already know how to ride a bike that has a hub drive.  thing is, no matter how seasoned and smart you think you are, chances are excellent you are clueless on how to ride a mid drive.

Here’s the short version:  Keep the motor spinning.

Now the longer one:

Keep the motor spinning

Lug it and the torque that is pouring out of the motor will focus on tearing your chain apart, or taco’ing your chainring or rear cog, not to mention generating enormous heat (remember the nylon gears in a geared hub motor?  Guess what?  Mid drives use nylon gears inside too).  Even a BBSHD set to off-road power levels is not strong enough to tear up your cogs or chainrings.  But it can snap a chain that you are mistreating.

When coming up to a stop light, downshift.

Always.  Either that or stay in a gear that is in the middle of your cluster so that when you start up again, the motor can spin up quickly without any brutality being visited on the chain.

When coming up to a hill, downshift.

Always.  Can you guess why?  Thats right so you can keep the motor spinning.  And ‘coming up to a hill’ does not mean ‘already started up the hill’.  Anticipate and shift in advance of the climb.

When you want to go faster, upshift.

But wait until your motor is spinning fast before you do.

When you up or downshift, NEVER do so under power.

Shifting while pouring huge watts into your chain is an ugly thing.  You will recognize your mistake the instant the result hits your ears.  It won’t kill the chain outright, but as you hear that gear smash from one cog to another you will know your bike hates you very, very much.

You can invest in a gear sensor that will protect you automagically from this.  It installs inline on your shifter cable and, when it senses the tiniest amount of movement, it cuts power for an instant.  The result is a safe shift.  I have them on one of my three BBSHD-equipped bikes and it works great.

But for the two I don’t, I just stop pedaling/freeze my legs, click-shift and then do a single crankarm rotation to seat the new gear at low power.  Result is perfect shifting and only a minor blip in pedaling rhythm.  But that is a learned behavior.

Others have perfected the use of the brake levers as a clutch where they only slightly actuate the lever.  This triggers the safety cutoff which in turn allows a safe shift.

Keep chain alignment as straight as you can

Mid drive motors tend to work in a lot wider range than humans do.  So you can leave the motor in a gear that would be too low for your cadence and let it spin away like crazy… it actually likes it that way.  So, this piece of advice is partly about how you ride the bike (i.e. what gears you let it sit in) but also about how you build it.  You really only need three or four gears in the middle of your cluster on a mid-drive-powered bike.  You want them to be the ones that lets the motor spin fast.  You also want the cogs the bike is happiest to not be cockeyed, front to back (i.e. bad chain alignment).

On an analog bike you can get away with a lot, since you are only feeding back 150 watts to it.  Feed it 1500 and that sideways-skewed chain will become a saw and chew right through your front chainring and rear cog teeth.  Be smart when you build the bike, and learn in your first outing or two whether there are any problem gears you should stay away from.  There are all sorts of offset chainrings (and 1mm and 2mm shims) available for a BBS02 and BBSHD… they cost money, but spending that money now means not spending it later after you have walked home.

What happens if you don’t do some or all of these things above to install and use a DIY mid drive bike properly?

Well of course it means you go on the internet and blame the equipment.  Its not your fault you used the wrong components.  And its not your fault you didn’t know how to ride it.  Bikes should just work and do your bidding.  So its the mid drive’s fault.

… If you build with appropriate components, and ride it smart, even a high powered mid drive will essentially last forever.  Yeah sure you will wear out the chain and rear cluster in say three thousand miles, the smallest cog in half that, and the chainrings in 10.  But thats peanuts considering how many miles you put on the bike.  How much does an 11T cog cost (about $6)?

Wrapping it all up

Whew that was a lot of typing.  This article lays out in broad terms the characteristics of each motor type.  It doesn’t get into what kind of riding each is good for.  Lets finish with that:

Direct drive hubs

  1. Lower/legal power:  maintenance free cruiser bikes for paved streets where speed and acceleration are secondary to bulletproof reliability.  Want a bike for Mom?  Your kids (who aren’t future BMX pro riders) A direct drive 250w hub motor is a good candidate.
  2. High power: Sky is the limit in terms of power:  Light electric motorcycles only thinly disguised as bicycles.  Can range across a wide variety of cycling genres including cargo, dirt and pavement.  But big and heavy.  Single speed is usually a bad choice for offroad/singletrack riding but there are exceptions, in particular the famous B52 Stealth Bomber and similar.  They have so much power they use brute force to overcome the weight issue.  But you’d never pedal one of these things.

Geared Hubs

  1. The Swiss Army Knife of motors.  Not ideal for anything but good for almost everything.  Only runs into trouble under very heavy use, particularly in an area that is all hills.
  2. Their only drawback is maintenance if heavily used.  Semi-annual teardowns to re-grease a high mileage motor is advisable.
  3. Bad choice for singletrack/offroad.

Mid Drives

  1. Best choice for singletrack/offroad
  2. Got hills?  Mid drive.  Strong power delivery without significant weight added to the bike.
  3. Good for hard-use applications.
  4. Arguably the most efficient in terms of power consumption.
  5. Requires the most attention to build detail and demands the most attention and learning from the rider.
  6. Even with proper use and setup, more wear and tear on the drivetrain than any other option.
  7. A kid who gets his right hand caught inside the chain of a mid drive is going to be named ‘Lefty’ from that day forward.

Mongoose – Chapter 1 (Raised Deck)

The Envoy’s rear rack is about 27″ long and 6″ wide.  For a cargo hauler, I wanted it to be able to handle more.  The rear deck is very low over the back wheel, and thanks to my inseam, the seatpost is going to be set high.  Add all that up and you have a lot of available vertical room for cargo mounted on top of the rack.

But we can do better than 6″ wide.  Especially since Mongoose has piled on M5 bosses seemingly everywhere on the rear of the frame.  There are attachment points everywhere.  I just did a quick count and saw 32.  And thats not counting the 4 additional lower rack bolts you can piggyback onto.  Also not counted are the 12 brazed-in hook mounts for the included panniers (that I’m only temporarily using due to their frustrating 6″ depth).  So what does that come out to?  This bike has 50 (fifty!) discrete attachment points.

IMG_20190917_174736.jpg
Visible here:  4 of the 5 sets of bosses that pepper the top of the rack.

This thing is begging for home brewed solutions so lets come up with one.  I don’t want just a regular deck on top of the rack, I want it to be wider than six inches so its a bona fide cargo platform.

My buddy Houshmand finished up his custom cargo bike build a few weeks ahead of me, and one of the things his framebuilder did was haul a skateboard deck off of his shop floor and decide it looked like a decent rack deck.  I agreed and informed my friend I would be stealing this idea.  So I did.

I started looking for a bare longboard deck.  I found decks as large as 48″x9.5″ for a dancer style, but it was something that doesn’t quite have that cool skateboard look I was going for.  I ended up settling on a 33″x10″ double kick deck that, while it doesn’t fully extend across the entire rack, its overall concave surface provides a sort of natural cradle for the duffel bag that will be cargo netted down to it.

The shorter 33″ deck also isn’t long enough to use all five sets of M5 bosses, but the four it does use it fits to perfectly.  Part of the concept I wanted was an overhang to the rear to ensure my back side stays dry if riding in the rain, and the board is bolted down to its most rearward flat location as a result.  Really it would not have fit well further forward or backward unless I shifted to the front four bosses and moved it quite a bit forward.

For hardware, I went to my old-standby for weird fittings: McMaster-Carr.  I know from previous projects they sell metric spacers in a variety of sizes and increments, and I had no trouble finding them in a wider 13mm diameter size.  However, my usual choice of stainless steel resulted in a hefty price tag.  I found they offered the same sizes in alloy which is perfectly fine for this application.

My original choice of 30mm spacers turned out to be problematic as well.  That size isn’t 30mm x 13mm in diameter isn’t stocked, so delivery would take weeks.  However, 25mm was available for next day delivery, and the cost per unit was about half what the 30mm’s would cost me.  Sold.

[hardware picture goes here]

Next, I wanted to make these as easy to maintain as possible while keeping them out of the way.  Ordinarily I will use socket caps, but their squared-off and tall caps aren’t the best fit for the top of a hauling rack.  I wanted to keep the job simple so I did not want to mess with countersinking the deck itself.  Besides for durability I wanted metal to metal contact, not metal to wood.  In the end I decided on countersunk Phillips head bolts (50mm), with countersunk washers sitting on top of large-area stainless washers – the latter to distribute the clamping force across as much of the board as possible.  Under the board, between it and the alloy spacer, is another large-area washer to help further stabilize the connection.  Multiply this by 8, torque to a modest 4 Nm (the threads in a frame boss, which could be alloy themselves, are not the sort of thing you want to apply serious torque to) and you have something solid enough to sit on, that is completely rattle free.

IMG_20190923_165920.jpg
Those bolt holes look a lot more even than they really are.  Thats the frame’s doing not mine.

Drilling the board to get the result above was something of an adventure.  I discovered very quickly that the frame bosses were all, well… totally uneven.  You might be able to see a little of that in the picture above but I assure you that the photo above is deceiving and its worse than it looks in the picture.  Trying to graph out a precise template that I would then map perfectly onto the board for drilling was clearly a Pain in the ass I didn’t want to suffer thru.  So I tried to think of a way to put something pointy in the hole so I could see the marks on the board.  Stuff like finishing nails or similar just didn’t have the precision to sit perfectly centered in the hole.  And then I remembered grub screws.  “What the hell is that?” you ask?  Let me just show you.

61O5H7aVGJL._SL1100_
Behold, the M5x16mm grub screw made of the finest Chinesium. $7 for a bag of 16.

They were delivered to me next day by Amazon.  I chose the longest ones I could get my hands on so I would have plenty of thread engagement to get them to center up.  And they worked splendidly.  After I used painters tape to cover my likely drill areas on the deck, (helps keep the wood from splintering), all I did next was screw in one of these screws, pointy side up, into the outer four boss holes, I then gently centered the board atop them just right (I tried measuring frame overhang and in the end just eyeballed it) … and pressed down.  Instant markings on the board, and my drill holes are already started.

Next I drilled the four marked holes, flipped over the board and test fit the screws to the rack.  Perfect match.  Now for the four inners:  I screwed the board down with the four outers I just drilled.  This gave me perfect positioning on top of the four inner holes … which now had pointy grub screws in them.  This in turn pressed the board onto the pointy bits.  Once done, I again had a set of marks that were perfectly aligned with an installed board.  Using the new marks I drilled again and perfect fit again.  The rest was just inserting the rather fiddly combination of spacer and washer under the board.  By the way, 50mm was the longest I could find locally.  Order 55 or 60mm’s from McMaster when you get your own spacers.

IMG_20190923_165853
On the road this new upper rack is rattle free, and so solid you can grab the bike by it and drag it around without worry

The end result works well I think.  I want the load’s center of gravity moved back a bit, effectively giving me a longer tail than I really have, for the top load, at least.  Additionally I still have a fair bit of unobstructed rack up front to play with.  When I put together my pannier solution, I have the option of using those standoffs, as well as what is now the ‘lower’ rack surface, for anchor points.