Tongsheng assist levels

To answer the OP's question-- and Rick, wow, it's a very good question, this motor's power delivery is unusual:

Level 1: 36%
Level 2: 85%
Level 3: 100%
Level 4: 250%

To which I will add:

Level 4 with extreme pedal pressure: 400%

This is a Marin bike which Pedaluma created, an experimental build that is still in development, but is now fully assembled and test ridden four times. What we think we will add/change: Move 11th gear to where 1st gear is currently for now, swap the 42t chainring for 38t to reduce or resolve cross chaining issues, which will require some modification that Pedal will spec and ship, and either just go tubeless with these tires or get new tubeless tires and go from 2.3 to 2.1.

Looking into my crystal ball, there's even a potential future where we find a way to get to a nine speed, hopefully a mega-range, from 52 to 10. As we've gotten deeper into this, and I've scoped around to find posts from other folks who have done something similar, there is some alchemy involved to get optimal power delivery and resolve the chainline issues. People have done it, it just takes some experimentation.

The motor is unrestricted; essentially, it's 'limited' by gear ratios, not software. This bike is a 29er, which complicates the assessment of power delivery even more! So: The motor is a bit anemic in levels 1 and 2 to begin with, and that issue is compounded by having larger wheels. Today I did my first 12-mile circuit of Griffith Park, which involves about 1,200 to 1,400 feet of vertical, if memory serves. Mostly broken pavement with stretches of sand and dirt.

The initial climb, which is long and moderate to steep, did seem like quite a grind if I tried to stay in Level 3 and out of Turbo-- it's a small, 336 Wh battery, and I'd already put about 10 miles on it, so I was concerned about range. It seemed just as hard as doing it on my 40nm Motobecane with a Shimano E5000. But here's what was strange: I'd look at my speedometer, and it's like, "Oh, I'm going 8 to 10 Mph, usually I'm going 6 Mph."

I kept trying to change my cadence and gearing so I'd be ascending at 6 MPH, so I could compare it to the Moto, but for some reason, it just wanted to go closer 8 or 9 MPH and kind of beat the crap out of me. Okay, fine, well, whatever-- bike's not done yet, the mods will probably change a lot of that.

The descent was uneventful-- handling is great, shocks need adjusting to reduce vibration, I'm getting used to the hard tail thing. High speed for the run was 28 MPH, about the same as for this route with the Motobecane.

What was really different was the super steep uphill dirt segment. I thought, "Okay, this probably won't work because I can't trust 1st gear, and barely trust 2nd." I start in turbo, and yeah-- immediately apparent that it's got way more grunt than I thought (and way more than the Moto.) There's a tricky hairpin right before the steepest part, and I get through it-- very different on a 29er, kinda hairy-- and then I pedal REALLY hard in second or third to manage the grade...

And it was like Turbo got a nitrous shot-- the bike just surged up that hill, faster than the Rossignol EP8 or E7000 or whatever I tried last summer, never felt that much power from an ebike-- it was a welcome rush, but unexpected, a little tricky with the grades and hairpins. It really is well-over 85nm, I'd believe it's 95, but only in this situation. May be how the torque sensor is set up, or maybe it overheated and freaked out... it did act a little odd on the last mile or two after that, like sometimes the motor stopped making noise, but still seemed to be delivering power?! Gotta test it again tomorrow. It was also on one bar of battery, occasionally bouncing back to two, so maybe the power delivery changed when I asked so much of it and the electrons were running low.

I had trouble on the last hill before my house, too... the same thing, I couldn't go 6 MPH, it wanted to go 8 MPH, and would only do that with more power from me. I wound up downshifting to first or second under load (probably first) and dropping the chain. Easy to plunk it back on, and restarting on a steep hill in low gear in Level 3 was no problem.

I came home utterly exhausted, drenched in sweat, but really excited. It's going to take some tinkering to get optimal power delivery out of it for my use-case scenario, but it will be so cool when we're done with it...
 
Since you're experimenting why not try installing the speed sensor/magnet that Pedaluma deletes and see how it performs with the addition of that input as well as the torque sensor? Maybe it's because of how eco-cycles programmed the controllers on my 36v and 48v versions but I have none of those issues that you report. They do respond more to pressure on the pedals than cadence, as does my bike with a Yamaha mid drive. FWIW when I removed the speed sensor I didn't notice increased performance on mine (other than a brief initial placebo response). As you note chainline/crosschaining is a problem on the larger rear cogs. It can be corrected to some degree using an aftermarket 10mm dished chainring or the 34t chainring that attaches without using a spider but the ones I have wouldn't work on the "B" version. TSDZ2 do seem very efficient, I've put over 130 miles on a single charge 48v 15ah battery (in very hilly route riding) but I use the assist when I need it, not constantly.
 
Since you're experimenting why not try installing the speed sensor/magnet that Pedaluma deletes and see how it performs with the addition of that input as well as the torque sensor? Maybe it's because of how eco-cycles programmed the controllers on my 36v and 48v versions but I have none of those issues that you report. They do respond more to pressure on the pedals than cadence, as does my bike with a Yamaha mid drive. FWIW when I removed the speed sensor I didn't notice increased performance on mine (other than a brief initial placebo response). As you note chainline/crosschaining is a problem on the larger rear cogs. It can be corrected to some degree using an aftermarket 10mm dished chainring or the 34t chainring that attaches without using a spider but the ones I have wouldn't work on the "B" version. TSDZ2 do seem very efficient, I've put over 130 miles on a single charge 48v 15ah battery (in very hilly route riding) but I use the assist when I need it, not constantly.
Ooh... 'dished chainring?' Very interesting, not sure what you mean. Do you have a link? I'm not sure we need to go all the way down to 34t, I would like to keep some of the potential for 28ish MPH top speed on level ground. One thing that bugs me about the Motobecane is how quickly I could overpedal the gears going downhill, though I did hit 43 MPH on one run on that Class 1 bike just by getting a flying start downhill. I'm not going anywhere near those speeds on this bike until I get the front shock dialed in better.

I certainly wouldn't rule out installing the speed sensor, but my hunch is that the problem is the gearing, particularly the chain ring. Fixing that will also make the bike more rideable if I do run out of battery power. Great news about the range, wow! I also shut off the assistance when I don't need it, but that's probably less than you do, only on the rare flat segments. (Faster rolling tires, e.g. at least these tires tubeless and maybe new tires, 2.1s) would allow me to use the motor less. This battery is 48v 7ah, which translates to 336 Wh, interesting how the math works at 48v.

A lot of this is so subjective, too. I'm used to 40nm, 46 pounds, 27.5 / 2.1 tubeless tires and F/S, so it makes sense that the bike still feels kinda weird. I think if I get the gearing right, I'm more likely to hit the sweet spot where the torque sensor gets what it needs and the electric power delivery feels more natural... when I don't feel like it wants me to go 9 MPH uphill. This is SO much more comfortable for my back and hands than the Motobecane, and seems to handle just as well, though it's strengths and weaknesses are a bit different.
 
I love that last photo! ^^ It is still a guitar (bike) but now it is electric.

 
Last edited:
Ooh... 'dished chainring?' Very interesting, not sure what you mean. Do you have a link? I'm not sure we need to go all the way down to 34t, I would like to keep some of the potential for 28ish MPH top speed on level ground. One thing that bugs me about the Motobecane is how quickly I could overpedal the gears going downhill, though I did hit 43 MPH on one run on that Class 1 bike just by getting a flying start downhill. I'm not going anywhere near those speeds on this bike until I get the front shock dialed in better.

I certainly wouldn't rule out installing the speed sensor, but my hunch is that the problem is the gearing, particularly the chain ring. Fixing that will also make the bike more rideable if I do run out of battery power. Great news about the range, wow! I also shut off the assistance when I don't need it, but that's probably less than you do, only on the rare flat segments. (Faster rolling tires, e.g. at least these tires tubeless and maybe new tires, 2.1s) would allow me to use the motor less. This battery is 48v 7ah, which translates to 336 Wh, interesting how the math works at 48v.

A lot of this is so subjective, too. I'm used to 40nm, 46 pounds, 27.5 / 2.1 tubeless tires and F/S, so it makes sense that the bike still feels kinda weird. I think if I get the gearing right, I'm more likely to hit the sweet spot where the torque sensor gets what it needs and the electric power delivery feels more natural... when I don't feel like it wants me to go 9 MPH uphill. This is SO much more comfortable for my back and hands than the Motobecane, and seems to handle just as well, though it's strengths and weaknesses are a bit different.
I bought my aftermarket chainrings from eco-ebike, in Tennessee, if I remember correctly. Their web site has been down for some time. Here is another seller that I found by googling

the stock steel chainring is 5mm dished, you can get 42t and 52t 10mm dished chainrings. The 30t and 34t attach directly without the spider, I'm not sure how much it corrects the chainline though (better but still not optimal). With the 10mm dished chainring the chainline isn't optimal either, but it's better. If you are thinking of using a flat (not dished) 38t chainring I'd guess that your chainline/crosschaining will be worse than it is currently with the stock 42t, 5mm dished chainring.

fwiw
cheap 48v 15ah battery used on TSDZ2 750w motor
this included ~7500 feet of elevation gain
Mileage was actually more than this because after 120 miles or so the odometer on this display under-records mileage significantly, until then it is very accurate.
1695408270813.png
 
Last edited:
@EMXG, You know what you are talking about and have great insights. @Catalyzt' MTB has an extra-wide BB area. We knew that it would cross chain to first gear, so I fell back to the standard 42-t 5mm offset, dished ring. We will try two things. With gears 12,11,10...1 on the cassette we will use 11 as a spacer behind 1. It will be 12,10...1,11. Then he will take in the low limit screw, maintaining the same overall range. A smaller ring is less prone to cross chaining and a narrow/wide holds the chain better and closer. And you are absolutely right about standard rings not working on the 'B'. I will modify it as I did with the 46-t narrow wide on the Stinson 2, and that Short Haul 36-t in LA.
 
@EMXG, You know what you are talking about and have great insights. @Catalyzt' MTB has an extra-wide BB area. We knew that it would cross chain to first gear, so I fell back to the standard 42-t 5mm offset, dished ring. We will try two things. With gears 12,11,10...1 on the cassette we will use 11 as a spacer behind 1. It will be 12,10...1,11. Then he will take in the low limit screw, maintaining the same overall range. A smaller ring is less prone to cross chaining and a narrow/wide holds the chain better and closer. And you are absolutely right about standard rings not working on the 'B'. I will modify it as I did with the 46-t narrow wide on the Stinson 2, and that Short Haul 36-t in LA.
@EMGX, thanks so much for all that! Those are some awesome numbers... I am just now calculating/guestimating how much range I got from my first 90% charge on my 7Ah battery, tricky b/c I didn't install the speedometer and odometer a few days, but I remember where I rode and can map it. What it looks like so far is close to 30 miles, I have to really dig in to calculate the elevation, but I would be shocked if it was less than 3,000 feet.

I rode the last 4.2 miles today, and was just hitting the blinking single-bar of battery when I got to the house and shut down. It really is a powerful motor, I could feel it kicking in hard, and often when I needed it, but not always. (I was also worked from yesterday, so I tired easily.) If the gearing were just a little different, I think I'd be getting optimal power-- enough to really help, but not so much that I'm running out the battery-- and getting it at a cadence that feels comfortable.

@PedalUma, yes, we did suspect we'd have cross chaining, that was one of the issues that came up when we discussed the possibility of a nine-speed option, (and at other points as well) but we didn't know what the impact would be. I am totally down for both mods cluster and chain ring mods and will get into the weeds with you about the details elsewhere, including whether I should get a bike stand and try to do the work myself or take it to the LBS. I'm thinking that 38t narrow-wide will be close to the sweet spot, just wondering if there's a dished 38t narrow wide that can be modded to be 5mm. THAT would make the re-ordered 11-speed cluster super stable. Also: I am confident now that the bike dropped the chain in first, not second. Second, I hear friction, but the chain won't drop under load. In third, I hear a tiny bit of friction, too. Wish we could do another offset somehow on the inside of the cluster, would be simpler...

I may even bite the bullet and go straight to 2.1 tubeless tires. The 29s are jus hard to get rolling-- not so much initially, but from 3 to 10 MPH there's drag. I also have to learn to change my technique. The 29 inch wheels do have more momentum, so when I can remember to really open it up and pedal hard at the bottom of a hill before the next one starts, I wind up using level 4 and even level 3 a lot less on the climbs.
 
This is worthy of its own thread. Some people will open a thread on a generic fat folder. This is not that. The thing is you can't dish a 38 or it will rub. A large ring can be dished, pie panned. This bike is pushing limits into uncharted territory. I should have time to make that experimental narrow/wide 38 tomorrow. It will take about 8oz. off the drivetrain. Blocks under the grips of a flipped bike will allow for easy work without a stand.
 
Right, the crank and gear choices for my bike should be a separate thread. When we get the gearing sorted, I'll post my final thoughts on assist levels, which was the original topic. We will see if the power delivery is as odd as it feels at first, or if it's just an artifact-- for this rider-- of unfamiliar gearing and wheel size.

Vaguely On Topic: Range of Tongshen with 7Ah (331Wh) battery 48 volts at 90% 1st full charge = 27 miles with 3,428 feet of vertical.

We can estimate that at 100% charge, this Marin build would get a bit over 30 miles and 4,000 feet of vertical.

For the sake of comparison, my Motobecane's (250 Watt 421Wh 36 volts) range at 100% charge is estimated at just about 40 miles with 4,000 feet of vertical.

The Marin is about 41 pounds, the Moto is about 46. My conclusion is that the Tongshen is quite efficient, it's just kinda mean the way it's set up now!

I do believe EMGX could get 120+ miles with 7,500 feet of vertical from a battery that's probably around 700Wh because I don't think the relationship between mileage, vertical feet, and battery capacity is linear. (@EMGX, what kind of bike are you riding?) Steeper grades and higher power levels use juice at a disproportionate rate. The routes I'm taking have some stretches that are either disgustingly steep or crazy long-- several miles of over 10%, and short segments that are 20%, one of them on dirt. (It's hard to account for the battery reserve he has, however!)

Put differently, if I were puttering along at power level 1 or 2, on asphalt with different tires, and the grades were not as steep, I bet this power plant could do over 50 miles with 3,000 feet of vertical, even on 29 inch wheels.
 
Right, the crank and gear choices for my bike should be a separate thread. When we get the gearing sorted, I'll post my final thoughts on assist levels, which was the original topic. We will see if the power delivery is as odd as it feels at first, or if it's just an artifact-- for this rider-- of unfamiliar gearing and wheel size.

Vaguely On Topic: Range of Tongshen with 7Ah (331Wh) battery 48 volts at 90% 1st full charge = 27 miles with 3,428 feet of vertical.

We can estimate that at 100% charge, this Marin build would get a bit over 30 miles and 4,000 feet of vertical.

For the sake of comparison, my Motobecane's (250 Watt 421Wh 36 volts) range at 100% charge is estimated at just about 40 miles with 4,000 feet of vertical.

The Marin is about 41 pounds, the Moto is about 46. My conclusion is that the Tongshen is quite efficient, it's just kinda mean the way it's set up now!

I do believe EMGX could get 120+ miles with 7,500 feet of vertical from a battery that's probably around 700Wh because I don't think the relationship between mileage, vertical feet, and battery capacity is linear. (@EMGX, what kind of bike are you riding?) Steeper grades and higher power levels use juice at a disproportionate rate. The routes I'm taking have some stretches that are either disgustingly steep or crazy long-- several miles of over 10%, and short segments that are 20%, one of them on dirt. (It's hard to account for the battery reserve he has, however!)

Put differently, if I were puttering along at power level 1 or 2, on asphalt with different tires, and the grades were not as steep, I bet this power plant could do over 50 miles with 3,000 feet of vertical, even on 29 inch wheels.
I definitely got 130+ miles on that battery (total over 3 or 4 separate rides) and it still had some juice left @46.4v. The display voltage is accurate, I like the C3 display which doesn't seem to be available anymore. Range has to do with how you ride. I rode 73 miles on a single ride with my Yamaha PW-SE and 36v 400w battery and still had 30+% remaining because I ride with no assist a lot of the time and save the battery for hills, wind or when I just plain get tired and going up hills usually means coasting down without assist. I also don't ride very fast, usually 12-15mph, slower up hills and I always put in as much pedaling effort that I'm able. I've put my two TSDZ2 (36v 500w and 48v 750w) on and off 6 different bikes. The 130+ miles was when it was on a Schwinn Avenue hybrid 700c bike with 35c high pressure tires. I ride in a hilly, mountainous area.
 
How much did you have to pay that athletic looking guy to stand there and pretend to be you? :p
It is a selfie, if one of my friends had been taking the picture, they would probably have told me I looked like a pretentious idiot.

I actually have the physique of a crackhead, the jacket has soft armor to protect me during a fall. I've never needed it, but my other armored vest actually did a lot to protect my back when I flipped the Motobecane (though my neck still hurts, and that was about a year ago.)
 
pay that athletic looking guy
It is Hollywood. They are all supper models. Hey, check out the arborists' climber jackets. These hardcore jackets are for guys who spend all day up in trees with chainsaws. They breathe, allow for free movement, and offer protection. They also have pockets that can be accessed with gloves. They have butt crack rain flaps and extra-long sleeves that are perfect for cycling. These climbers are true athletes who cannot afford inferior equipment. I have had one for four years that still looks new.

1695488224038.jpeg
1695488261270.jpeg
1695488437049.jpeg
1695488476611.jpeg
 
The most recent 90% charge was 33.8 miles and I didn't hit reserve yet. I don't know what the vertical was, a bit less, but I'm sure well over 2,000 feet.

I'm now estimating 90% = 35 miles with at least 2,500 feet of vertical. That is climbing slow for sure, about 6 MPH uphill, pedaling hard, but not to exhaustion, just trying to stay out of level 4 as much as possible, definitely using it when needed. Better than I expected. Seems significantly more efficient than the Motobecane, but now I want to go back to the Moto and see if I'm just working harder on the Marin. I don't think so.
 
The other thing I'm beginning to suspect: The TSDZ2B seems VERY efficient. Been out twice in steep hills and I'm still at 5 bars.
I agree that the TSDZ2B is super efficient. My average battery usage is between 5-7 Watt-hours per km (8-11Wh per mile), and that's with me only putting in moderate effort. I have a nice compact 36V 16Ah battery, so I could easily be looking at a 100km ride on a single charge.

Like many other people, I found it took me some time to get attuned to the torque sensor way of riding, but now I love the fine-grained power increments you can get out of this thing, which I guess contributes to its efficiency - you only use exactly the power you need at each point.
 
I agree that the TSDZ2B is super efficient. My average battery usage is between 5-7 Watt-hours per km (8-11Wh per mile), and that's with me only putting in moderate effort. I have a nice compact 36V 16Ah battery, so I could easily be looking at a 100km ride on a single charge.

Like many other people, I found it took me some time to get attuned to the torque sensor way of riding, but now I love the fine-grained power increments you can get out of this thing, which I guess contributes to its efficiency - you only use exactly the power you need at each point.
I believe it-- I've now gotten over 37 miles on moderately hilly terrain without hitting the reserve on a 90% charge. So even here, I might get over 40 miles.

However, I did go back to my other bike since the Marin got a flat, and there is no question: The little 40nm Shimano E5000 on the Motobecane provides far more assistance than the TSDZ2B tooling around my 4-mile fitness route, which has three steep hills. I was really surprised, but that has got to be part of why the motor is more efficient: I'm supplying more power. The battery fade on the Tongshen is also noticeably worse and kicks in sooner.

The Tongshen has a far higher top speed on flat terrain-- like, 7 MPH faster-- but it does require significant effort to get there. The Tongshen also provides more boost on the steepest hills in situations where the E5000 simply would not have enough grunt.

My opinion may change after I make some changes to the gears, hopefully also the chain ring, and go tubeless. Right now, it's not like I would prefer one motor or the other, they are just very different.
 
I believe it-- I've now gotten over 37 miles on moderately hilly terrain without hitting the reserve on a 90% charge. So even here, I might get over 40 miles.

However, I did go back to my other bike since the Marin got a flat, and there is no question: The little 40nm Shimano E5000 on the Motobecane provides far more assistance than the TSDZ2B tooling around my 4-mile fitness route, which has three steep hills. I was really surprised, but that has got to be part of why the motor is more efficient: I'm supplying more power. The battery fade on the Tongshen is also noticeably worse and kicks in sooner.

The Tongshen has a far higher top speed on flat terrain-- like, 7 MPH faster-- but it does require significant effort to get there. The Tongshen also provides more boost on the steepest hills in situations where the E5000 simply would not have enough grunt.

My opinion may change after I make some changes to the gears, hopefully also the chain ring, and go tubeless. Right now, it's not like I would prefer one motor or the other, they are just very different.


if that's the case there's something wrong with your tsdz2 setup
 
if that's the case there's something wrong with your tsdz2 setup
I replaced the initially supplied Tongsheng display with a 500c, and that made all the difference for me.

The 500c can be tweaked to have either 3, 5 or 9 levels of assist, and I find 5 levels to be ideal for the somewhat hilly terrain I ride on. Fine-grained enough, but not so much that I would have to keep pressing buttons.
 
if that's the case there's something wrong with your tsdz2 setup

Maybe, I'm certainly open to that possibility, but it's also possible that the E5000 has spoiled me. It's an unusual motor to find on an eMTB. I should also note that over the past few years, it actually seems more powerful than it was before, which I can't really explain-- eBike motors do not 'break in,' but this one sort of did, and I wonder if it's delivering more than 40nm.

Pedaluma rode the Marin before delivery, and it seemed powerful enough to him even in it's current configuration. I trust his opinion, I just wonder if we have a different basis for comparison, and are riding on very different terrain.

--> Do you experience battery fade? I wonder if connecting the speed sensor would impact the way the system rations battery power.
 
I live on a mountain and all of my rides involve pulling long steep grades. If I lived in an area with flat terrain I wouldn't even bother with an ebike. My 48v tongsheng has very similar assist as my 70nm Yamaha pw-se which is quite powerful. It's hard to understand how it wouldn't provide much higher assist than a 40nm rated motor even if power comparisons aren't exact. I experienced minimal battery fade until the battery is nearly exhausted even with the cheap Amazon sourced batteries that I've used with my tonsheng(s). Your description just doesn't seem right. Too bad you aren't local to test ride my bike, although my 48v version has been sitting on a shelf for months and my 36v version is on a heavy beach cruiser that I haven't ridden in over a year - but even my lower power 36v on the heavy beach cruiser offered excellent assist on the same hilly terrain.
 
Back