TSDZ2 Full Suspension MTB build (motor cooling, chainline, etc)

[scianiac]

This is just a quick build log of my TSDZ2 build, mostly focusing on things that others might find useful when doing a similar build. Started with a 2013 Diamondback Recoil Pro 29er that I got on craigslist. Quick list of all the basic parts, upgrades, tweaks and so on:

Motor
-TSDZ2 with open source firmware
-CNC aluminum heatsink shim (more details below)
-All the other normal TSDZ2 heatsink mods
-Throttle for those occasional uses
-Separate temperature monitor with temp alarm
-Custom 3D printed torque bracket (more details below)
-Gear sensor
-Magnetic brake sensors

Battery
-Battery made from 7x 2S 5200 lipos, 14S split into 7S for charging.
-Size small frame means there is no space for a battery so backpack it is, using a cheap lightweight cycling pack with a bit Spectra braid stitched in for reinforcement.

Drivetrain
-Removed about 6mm off the bottom bracket right side to get perfect chainline (see below)
-24T narrow wide chainring with a 3D printed chain guide because why not
-New wider range rear cassette paired with a microshift 9 speed
-Went with some DIY molten speed wax chain lube to counteract the added ebike chain wear

MISC
-Budget dropper seatpost, this actually works just fine with the power cable routing
-Stock calipers but replaced both rotors with 203mm
-About 5 million zip ties to hold together the at this point ridiculous number of wires, cables, hoses, and other bits.

A bit more details on some of the above
Full suspension means no place to mount the stock motor torque brace. Easy enough I’ll just make a bracket that wraps around the seat tube and locks to the motor using the original mounting bolt. Considering the amount of torque this needs to withstand one would think that it needs to be made of metal. But 1000s of hours of 3D printing experience and a fascination with advanced material composites says otherwise. I whipped up a quick design to be printed from carbon fiber reinforced nylon and included wrapping guides for additional fiber reinforcement. With a little epoxy and some 100lb test kevlar braid it’s easy enough to add a theoretical 4000lbs of extra tensile strength to the part. Kevlar was used instead of UHMWPE or Carbon fiber because it’s easier to work with than carbon fiber and glues better than UHMWPE which has a terribly low surface energy.

Next quick modification was grinding down about 6mm of the bottom bracket to get the motor far enough over for a perfect chainline. This was done, very carefully and precisely with an angle grinder and flapwheel. All kidding aside if you are very gentle, use wax to prevent the wheel from getting clogged up do to the affinity of aluminum to the aluminum oxide abrasive and take lots of measurements the final result was well under 0.5mm of parallelism with the factory side. A side bonus of this is the Q-factor is now just about perfect down the centerline with the help of a straight right side crank.

For all the advantages of the TSDZ2 it has some issues as well and one of the biggest issues is it’s poor motor cooling. For some unknown reason the designers though it was OK to just have the motor in an entirely sealed case with no contact to the aluminum body at all. With the stock setup it’s very easy to overheat the motor and if you don’t install the open source firmware thermal throttling or a thermal probe to keep an eye on the temperature it’s very easy to damage it in the process. Many past attempts have been made to remedy this and the most effective and easiest was by najbyk as shown in the wiki which filled the approximately 1.5mm gap between the motor and shell with thermal pads. The only problem with this solution is even at that thickness thermal pads have very poor performance so gains are limited by their performance. So my simple solution to this problem was to just replace the pads with aluminum which is an order of magnitude more thermal conductivity than the pads. As a first test I hand cut shim a shim and after that performed wonderfully proceeded to CNC cut them for an exact fit. With the original test shim and the back thermal pads between motor and outer shell I was unable to get the temperature above 42C even with hard use. I expect the CNC version will perform even better since it is one piece instead of two sheets but the cold weather means it’s hard to tell real results from cold ambient assisted cooling. (actually did get a chance to ride in some warmer weather a few days ago, wasn't watching the temp too closely but never saw it above low 30s)

Due to how machining works making one of these shims was just as easy as making several so I’ve got a few more available if anyone is interested. I just threw them up in a hidden listing here if anybody wants one. If people are interested I might make some more. Edit 1/7/21 - due to popular demand I've made some more of these, available at the link above






Future upgrades
-Replace the motor temperature readout with a much smaller unit without an alarm, motor never even gets close to overheating anyway
-Installing more aggressive tires. I don't ride super aggressively but for me one of the best parts about an ebike is climbing otherwise totally unclimbable things, hills you can barely walk up. And there are still a few I've found that I just don't quite have the traction to get all the way up
-32T chainring and adapter plate, tweak that drive ratio a little more, again for climbing crazy hills
-Some custom built throttle that doesn't suck

So what do you guy's think?

 

[sparked]

That is a really great (and great looking) build, scaniac. You implemented some very nifty solutions. Thanks for the detailed write-up. If the wires are bothering you this wireless project might be up your alley for a future update.

 

[Vbruun]

How much power do you run through the motor?

 

[scianiac]

I pretty much have the current settings maxed out so I get about as much as you can from the 48V motor 52V battery combo. Peaks at about 750-800w but quickly tapers off as the RPM gets to reasonable levels. So around 500-600w continuous with aggressive shifting and 300-500w with more relaxed shifting.

 

[sparked]

Did any of the people who received your shim install it yet? I am very interested in the results.

 

[scianiac]

Haven't heard from anybody yet but I did ask if they would post their results in the TSDZ2 hardware temperature control thread, several of them were overseas and will take some time to get there. Hopefully we'll see some more results soon, maybe from people with warmer weather than I've got!

 

[Alfie]

Haven't heard from anybody yet but I did ask if they would post their results in the TSDZ2 hardware temperature control thread, several of them were overseas and will take some time to get there. Hopefully we'll see some more results soon, maybe from people with warmer weather than I've got!

I have received my shim. Thank you very much for sending it. I will be fitting a temp sensor before christmas so will fit it at the same time. It's cold and wet here in the UK at the moment so probably not the best time to test it.

 

[Stealth_Chopper]

Oh, that's a nice eBike !
Please critique the ride quality, and performance of your choice for the type of suspension.
( I think it's great !)
Your attention to thermal management with bespoke heat sink is a nice feature.
Mike

 

[scianiac]

So I'm going to try and summarize my current thoughts on the bike and how I use it in the hopes someone else will find it useful. I've always found that when researching a build, reading someone's experience that is as close to yours as possible can be really helpful. Someone building a commuter will be of little use to someone building a mountain bike.

First, some background, I ride the bike on single track in New England and this means many trails that I would consider pretty technical both up and downhill, trails without rocks or roots are pretty rare. I don’t ride super fast or do many jumps or drops. Mostly just even paced riding clobbering over tech and this is where I find most of the fun, riding technical terrain that is challenging.

The stock bike and it's components where pretty much the lowest end of "real" mountain bike components when it was new which after aging on the used market did make the bike very affordable. This also means that many of it's features would be considered outdated and more in the territory of a XC bike at this point. The short 100mm of suspension travel and narrow rims being the two that come to mind. That being said I don’t think this holds me back at all, the mountain bike industry seems to be full of companies that love to tell you what new tech you need to have when it’s mostly skill, skills you can learn on almost any bike. Except maybe disc brakes because barrelling down steep technical wet terrain with wet rim brakes is something I would rather not try again. All that being said my next bike probably will be more serious than this one in both components and drive unit, but that will come after I feel my skils have progressed to warrant it.

The other key element I’ve realized from riding the bike a bit more and thinking about what I would change or improve on a future bike is the ways in which one actually uses a DIY E-MTB. The way I see it there are 3 ways:

Torque power assist – You pedal normally but the torque sensor adds power, very enjoyable and rewarding for difficult and steep terrain. This is of course is your only option on most commercial E-MTBs.

Throttle with pedal assist – You use the throttle to cruise along on low inclines to make slight uphills feel like slight downhills (after all downhills are fun) and then add some pedal power to boost you over steeper sections.

Throttle only – If you have enough power (which the tsdz2 does not) to just power around in E-moto mode. The key distinction here and one that I think a lot of people overlook is even though it doesn’t take very much power to move you around at speeds faster than you can pedal this often feels rather lame compared to the same or lower speeds where you are putting in effort. The action of pedalling just has that effect on your brain, that you are making progress through your effort.

From these 3 use cases I don’t think I would ever want a low or even what some would consider high power drive without a torque sensor, something like the BBSHD. Without torque sensing, option 1 is hampered, the boosting parts of 2 are better done like 1 and you need some serious power to do 3. Option 2 works well I think because the sense of just gliding along a trail is more enjoyable than the rewarding effect you get from pedalling on an almost level incline. Related I think is pedalling up a steep hill while putting in some real effort is more rewarding than to just pedal a few strokes of relatively little effort just to keep yourself moving on level or slightly inclined terrain.

P.S. Due to popular demand I’m now selling TSDZ2 heatsink plates here.

 

[Zambam]

Next quick modification was grinding down about 6mm of the bottom bracket to get the motor far enough over for a perfect chainline. This was done, very carefully and precisely with an angle grinder and flapwheel. All kidding aside if you are very gentle, use wax to prevent the wheel from getting clogged up do to the affinity of aluminum to the aluminum oxide abrasive and take lots of measurements the final result was well under 0.5mm of parallelism with the factory side. A side bonus of this is the Q-factor is now just about perfect down the centerline with the help of a straight right side crank.

I think the 210 mm Q factor on my TSDZ2 may be responsible for the discomfort in my hips after accumulating 100 miles on the bike. Where did you find the straight right side crank? Link?

 

[dgc]

I think I have fitted a Bafang crank on my bike, the Tongsheng has I believe a 20mm offset on the RH crank. Fitting Bafang crank does not eliminate the off-set but reduces the Q factor by 20 mm. Check it out. Have a few bikes, but I like the one in the photo best at the moment which is a lightweight mullet with a TSDZ2 and a 36v 10ah battery which gives me approx 30 mile range.

 

[Zambam]

I think I have fitted a Bafang crank on my bike, the Tongsheng has I believe a 20mm offset on the RH crank. Fitting Bafang crank does not eliminate the off-set but reduces the Q factor by 20 mm. Check it out. Have a few bikes, but I like the one in the photo best at the moment which is a lightweight mullet with a TSDZ2 and a 36v 10ah battery which gives me approx 30 mile range.

Will check out the Bafang crank. Can you provide a link to your bike? I don't see any frame weld marks. Is it carbon fiber? What's the weight as shown? Nice looking bike!

 

[dgc]

No it is a Boardman alloy frame with the welds all smoothed down will find a link for you.

 

[scianiac]

I think mine might also be a bafang crank, it came as on option on the kit from eco-cycles. Also my removing of 5mm off the bottom bracket also shifted the drive over a little so that centers the Q factor better as well but that's not as easy a mod to do.

 

[Zambam]

I think mine might also be a bafang crank, it came as on option on the kit from eco-cycles.

I have a set of Bafang cranks ordered. What display / OSF are you using and how has the motor been running? What's the weight of your bike? I may look for a FS bike to move the TSDZ2 to if the total package can be kept @ around 42 lbs.

 

[dgc]

Zambam, I have two ebikes which are based on Boardman smooth alloy frames, one is the 2014 Team FS 650B 18" with 26" wheels, 2.5" tyres and a under-slung 36V15Ah battery the other is a 2011 Team FS 16" with 27.5"F 26"R wheel, mullet and the 36V10Ah battery this weighs in at 18kgs, the 650B is not a lot heavier. Boardman do quote "The mainframe tubes are joined with deliberately smoothed-down welds to produce a seamless carbon-like look". Both are scratch built from brand new old stock frames from Ebay. Otherwise go for a Giant Anthem frame, sold this to a friend as I wanted to go with the new builds.

 

[dgc]

What display / OSF are you using and how has the motor been running?
Standard firmware with VLCD5 and VLCD13 displays, like to have brakes sensors so did a graft on the lead for the

 

[dgc]

Hit the wrong button, to continue.
Standard firmware with VLCD5 and VLCD13 displays, like brake sensors so grafted them onto the VLCD13 lead. Setting up my laptop to reprogram the drives to try the hybrid option. Covered about 2000+ miles on each no problems so far, thought I did have a problem with one but use the spokes coming loose in the back wheel so did a rebuild with Sapim stroke spokes.

 

[dgc]

Do not use carbon frames as I put extra M5 thread inserts in where needed for the battery mounts, glad you like the lightweight mullet version, great to ride, I got it just right. Carbon frames a little bit too costly for me at the moment.

 

[scianiac]

Last time I checked my bike weighs 51lbs and the battery is quite small (although also not very energy dense either). Running an older version of the OSF. At this point this bike is mostly used to loan out to friends as my other trail bike is superior in every way.

That 51lbs sounds heavier then I remember but is probably accurate. I have a weird relationship with ebike weight, I often jump through all sorts of hoops to reduce weight where I can and still end up with bikes way heavier than I expected. My other trailbike is around 58lbs and it's the light one compared to the fat bike and road bike. Thing is it doesn't really matter all that much, the only thing I notice about the weight is how hard they are to lift into the back of the truck. When riding you notice a heavy or light bike a lot when you switch back and forth but once you get used to it you don't notice very it very much and often the advantages you got from that weight are well worth it. Power, reliability, range, practicality, etc all cost weight so the thing I try to optimize for is performance per weight, if I can make something lighter without sacrificing performance (and at a reasonable cost) I'll do it.

Now that is strongly related to what you want the bike to do which is what determines what performance means, performance as in how good the bike is at doing what it's designed to do. My Small Block Santa Cruz V10 is 7 lbs heavier than this bike but has about 8-10+ times the power, 2.1x time battery, twice the suspension travel, is much more reliable, much better control and responsiveness and is just more capable overall so yes it's heavier but the performance per weight is vastly higher.

 

[dgc]

I do prefer the Boardman 2011 Team FS ebike as it is light and easy to flick around when riding along the winding country / forest paths. I fitted the 27.5" front wheel and forks to play with slackening the head stock angle and it seemed to work, anyway suits me. I do feel the ride and weight difference when riding the Boardman 2014 Team FS 650B, not quite as easy to flick around, used more on the road and down canal tow paths.

 

[Zambam]

Last time I checked my bike weighs 51lbs and the battery is quite small (although also not very energy dense either). Running an older version of the OSF. At this point this bike is mostly used to loan out to friends as my other trail bike is superior in every way.

That 51lbs sounds heavier then I remember but is probably accurate. ..................

My no suspension CrMo bike with the TSDZ2 is around 42 lbs. It's easy to lift up the 3 steps to get it into the house. I wouldn't want to do that with my Arrow 9 @ 75 lbs. I'd like to keep total weight around 42 lbs for any other potential bikes to move the TSDZ2 to.

I have not done the cooling mod. My main concern was putting the motor back together into the new casing and have it working. The hottest I could get the motor is 50C (122F) on a 75 F (24 C) day, measured by thermocouple on stator lamination. There are no steep hills to climb, nor can I go much faster than 20 MPH safely on streets around here. I am around 118 lbs at my heaviest. Do you think heat will be an issue when summer comes? Should I do the cooling mod anyway?

 

[Zambam]

I do prefer the Boardman 2011 Team FS ebike as it is light and easy to flick around when riding along the winding country / forest paths. I fitted the 27.5" front wheel and forks to play with slackening the head stock angle and it seemed to work, anyway suits me. I do feel the ride and weight difference when riding the Boardman 2014 Team FS 650B, not quite as easy to flick around, used more on the road and down canal tow paths.

Boardman is sold in the UK only? Is the 2011Team FS considered low/ mid or high end? What did it cost when new? One issue with FS is having a good rear rack. What do you use to carry stuff or get groceries? Back pack? Nothing wrong with that but it's nice to have a "real" rear rack. With the articulated rear suspension, did you have to discard the anti-torque clamp on the chain stay? What did you use in its place? Maybe I should consider hardtail and not FS.

 

[dgc]

Both my Boardman' are described as entry level full suspension bikes that ride very well by any standards, they both cost £1000 when new at the time. But I do have old Specialized Epic Comp and Giant Anthem X3 frames in "stock", may have a go at the Specialized first as it will take a larger battery in the triangle, always like the Giant frames, have a Trance with a MXUS XF15C rear hub motor, problem having two teenage grandsons who like coming out with me. It has been modified since this picture with air forks and a better frame mounted housing for the controller thanks to a Creality Ender 3 Pro 3D printer, present from wife.
Shopping goes in a 25 litre backpack, shop everyday like to read a newspaper with coffee in the morning, I call it weight training.