TSDZ8 throttle scaling

[rcflyers]

I have been trying to understand the throttle on my TSDZ8
It is almost an on-off switch with very little fine control.
I put some tape on the throttle and did some measurements.
The range of the throttle lever between min and max is 22mm
For the first 7mm of travel nothing happens.
For the next 8.2mm of travel is your throttle range
The last 6.8mm of travel does nothing.
So the throttle only operates over 1/3 the range of the lever travel
See pictures below, the 2 marks with the X in between them is where the throttle operates.
Next I measured the voltage on the throttle.
I connected a 5.23 volt supply to the throttle.
At throttle off there was 0.82v on the blue wire
At the point at where the motor is running at minimum power there is 1.8v
At the point at where the motor is running at minimum power there is 3.6v
At full throttle there was 4.48 volts.
So, the motor controller is looking for a voltage between 1.8V and 3.6V to make the motor run.
My bike has a 9 speed rear set.
I set it to 7th gear.
The bike was on a test stand during the test.
At minimum throttle the slowest it would go is 12.8mph.
At mid throttle setting it was reading 20mph
At max throttle it was reading 31mph
For the throttle to be more usable either the motor controller needs reprogramming to start providing power when the throttle is outputting say 1 volt and give max power at say 4.2 volts
Or, we try to find a throttle that outputs say 1.5v at minimum and 4.6v at full throttle.
This would make the throttle actually usable!

 

[amberwolf]

There are a few alternatives:

--use an op-amp (or set of them) to re-scale the throttle's output voltage to what the controller expects.

--use an MCU (arduino, etc) to do the same thing; this would be "easier" in that you can code it to recompute the entire throttle curve including changing it from linear to some other curve, etc.

--replace the physical throttle with a potentiometer type (either one that has the pot actually in the throttle, or a cable-pull throttle handle that pulls a COT cable operated throttle body that uses a pot rather than a hall. The wiper lead goes to the controller's throttle input. Then use resistors on the two non-wiper leads, from VCC and ground, calculated as a voltage divider with the actual pot resistnace across those leads, so that the pot itself has the exact voltage at it's start and end points that you want the controller to operate from.

--check what sensor your throttle actually uses. Some of them have a wider range of VCC, and most of the types that are used in throttles work by dividing the VCC in half for the "center" point of the output, when there is no magnetic field present. If the throttle presently outputs a total range of voltage that is lower than what is needed to operate the controller, you could raise the VCC on the throttle (separate supply not from the controller) to whatever point that brings the output up to where you need it, or as close as possible.

 

[E-HP]

For the throttle to be more usable either the motor controller needs reprogramming to start providing power when the throttle is outputting say 1 volt and give max power at say 4.2 volts
Or, we try to find a throttle that outputs say 1.5v at minimum and 4.6v at full throttle.
This would make the throttle actually usable!

If you’re running the stock firmware, the throttle is total crap. On the updated firmware, it’s almost ok. I haven’t tried the OSF yet.
If you’re planning to stay on the stock firmware, you can adjust your throttle min and max points to remove the dead zones. You can locate the necessary pots/resistors outside of the thumb throttle housing since it has no room.

Guide to Hall Sensor Throttle operation, testing, and modification. - Electricbike.com Ebike Forum

***** 8-16-2018 Welcome to the Hall Sensor Throttle Thread! If there is something you'd like to add, correct, needs better explanation, or have a question about... feel free to Private Message me. Better yet, open a new thread in the "Trouble Shooting...

electricbike.com

 

[TommyCat]

I just want to say how refreshing it is to have a post with such detail on the particulars! Job well done.

The only thing missing would be your preferred way to resolve the dead band issues.

As E-HP has mentioned, the resistor solution might be your easiest bet.

As seen in the GUIDE... make sure you read it thru for all the details and explanations on the how and why this works. As well as important safety considerations, if this looks like a path you would like to follow.


Regards,
T.C.

 

[rcflyers]

I just want to say how refreshing it is to have a post with such detail on the particulars! Job well done.

The only thing missing would be your preferred way to resolve the dead band issues.

As E-HP has mentioned, the resistor solution might be your easiest bet.

As seen in the GUIDE... make sure you read it thru for all the details and explanations on the how and why this works. As well as important safety considerations, if this looks like a path you would like to follow.


Regards,
T.C.

Thank you T.C.
Soldering in a switch and a couple of pots will be no problem.
Are these pots suitable please?

https://www.amazon.co.uk/Innfeeltech-Multiturn-Potentiometer-Adjustable-Variable/dp/B0D3X9FMB1?dib=eyJ2IjoiMSJ9.fiYNUW4z5HHxt4ME_zJn8G_e2d4mySCSeZKiL75RlkvuBwm9jBnpJV52eXxptzaIIoVoCtQDDZn01BH6rZoIz760vdYAEdCkyoXIhVnPGOLyGg181M233Vv-Drc5FTusM_pvclwjYoADZSq5eNbrSj_jVCDdO8sVKs6VKwXgCBEVa5cHpiJtu_7hUxeIxHhipIPxblu9lwLXhmqd5Td86FOEc9AGKpUSi0sHMMczS2FLZ_stVW9CT1GNtomED0c00OVJnpgmw2spGMn1xcCWrvYxpCYb8j53F7O_EuqM57o.NENXRy8Tu_z6ytVEgJwGgQC3hrhcilBDm12hg2yExc4&dib_tag=se&keywords=multiturn%2Bpotentiometer&qid=1752779509&refinements=p_85%3A20930949031&rnid=20930948031&rps=1&sr=8-8&th=1

 

[TommyCat]

Soldering in a switch and a couple of pots will be no problem.

Sounds great.
I just repurposed my unused light switch on my throttle handle for my safety switch. (Daytime rider...)
Just for a failsafe... but I do a lot of messing with the throttle.

Are these pots suitable please?

I think those would work well, although I didn't see a turn count.

For reference, this is an assortment kit I use with a turn count of 25. Which is quite a few turns... but give maximum resolution of desired voltage.

Multi-turn Trim Pot Assortment.

 

[E-HP]

Thank you T.C.
Soldering in a switch and a couple of pots will be no problem.
Are these pots suitable please?

https://www.amazon.co.uk/Innfeeltech-Multiturn-Potentiometer-Adjustable-Variable/dp/B0D3X9FMB1?dib=eyJ2IjoiMSJ9.fiYNUW4z5HHxt4ME_zJn8G_e2d4mySCSeZKiL75RlkvuBwm9jBnpJV52eXxptzaIIoVoCtQDDZn01BH6rZoIz760vdYAEdCkyoXIhVnPGOLyGg181M233Vv-Drc5FTusM_pvclwjYoADZSq5eNbrSj_jVCDdO8sVKs6VKwXgCBEVa5cHpiJtu_7hUxeIxHhipIPxblu9lwLXhmqd5Td86FOEc9AGKpUSi0sHMMczS2FLZ_stVW9CT1GNtomED0c00OVJnpgmw2spGMn1xcCWrvYxpCYb8j53F7O_EuqM57o.NENXRy8Tu_z6ytVEgJwGgQC3hrhcilBDm12hg2yExc4&dib_tag=se&keywords=multiturn%2Bpotentiometer&qid=1752779509&refinements=p_85%3A20930949031&rnid=20930948031&rps=1&sr=8-8&th=1

Don't set your hopes too high. The stock firmware is crap, not only for throttle, but for pedal assist as well. The non-OSF firmware update is a huge improvement by comparison, including more power (I get the same power in level 3 as with level 5 on the stock firmware) and less fatigue. It's hard to believe they let these motors leave the factory with that crap loaded on them.
I'm holding off on the OSF for now, since it looks like the power is lower, but the pedaling quality is improved. The pedaling quality and response I get now is great, so I don't feel I'm missing much, but keeping an eye on OSF developments.

 

[rcflyers]

Update:
I soldered in the 2 pots between the throttle and controller.
I adjusted trim pot 1 until the motor started to turn, I then backed off about 1/2 a turn to give it the deadband at rest.
I then adjusted trim pot 2 to use the full range of the throttle up to maximum.
Took it out for a test ride and it worked much better.
However, I stopped and turned the display off and on again but was met with an EO4 - throttle error.
The only way I could get rid of this was to turn pot 1 anticlockwise again.
This got rid of the error but also now means that the starting point of the throttle operation is no different than before i.e. a large initial deadband.
I assume that the controller wants to see a lower initial voltage that the pot adjustment was giving it.
So, a big improvement from 3/4 to full throttle but sadly no improvement at 0 to 1/4 throttle.
Pics show pots quickly soldered onto a breadboard just for testing purposes.
Thanks.

 

[rcflyers]

Don't set your hopes too high. The stock firmware is crap, not only for throttle, but for pedal assist as well. The non-OSF firmware update is a huge improvement by comparison, including more power (I get the same power in level 3 as with level 5 on the stock firmware) and less fatigue. It's hard to believe they let these motors leave the factory with that crap loaded on them.
I'm holding off on the OSF for now, since it looks like the power is lower, but the pedaling quality is improved. The pedaling quality and response I get now is great, so I don't feel I'm missing much, but keeping an eye on OSF developments.

I guess there is no easy way to tell which firmware version I have unless I buy the cables and download it?
Did you update the firmware yourself?
Was there much difference in the throttle operation?

 

[E-HP]

I guess there is no easy way to tell which firmware version I have unless I buy the cables and download it?
Did you update the firmware yourself?
Was there much difference in the throttle operation?

I got the hardware originally for OSF, but only loaded what they call the Facebook firmware (from the TSDZ8 Facebook group). I couldn’t get the OSF to work (yet). The throttle response and PAS levels are much improved, and slightly more power per pedal input, so more pleasant. Still, if you’re used to a decent ebike throttle with fine adjustment, it’s not there yet. I’ve found it has limited use anyway since the motor’s power band is narrow so the throttle response that’s there peters out quickly, requiring an up shift.
I rationalized the hardware purchase since a friend of mine built his first conversion with the TSDZ8, so I’ll send him the hardware so he can update his. Same with my neighbor who is planning a TSDZ8 build, but held off after trying my bike with stock firmware. He hasn’t tried it with the update yet.

 

[TommyCat]

I soldered in the 2 pots between the throttle and controller.

Nice.
The wiring looks correct.
I would encourage you to put your meter on it to directly measure the signal output voltages.
This way you can fine tune the “Home” position starting voltage to about 1.6 vdc to start with. Perhaps closer later if desired and possible.
And WOT ending voltage to about 3.7vdc. You may have to tweak after each adjustment…

If still having problems, try to find the exact voltage issue appears and set accordingly for as good as you can get.

That starting voltage seems quite high, but not much you can do about that…

Interested in seeing your testing voltages! Good luck.

 

[rcflyers]

Nice.
The wiring looks correct.
I would encourage you to put your meter on it to directly measure the signal output voltages.
This way you can fine tune the “Home” position starting voltage to about 1.6 vdc to start with. Perhaps closer later if desired and possible.
And WOT ending voltage to about 3.7vdc. You may have to tweak after each adjustment…

If still having problems, try to find the exact voltage issue appears and set accordingly for as good as you can get.

That starting voltage seems quite high, but not much you can do about that…

Interested in seeing your testing voltages! Good luck.

Did some more testing today with the bike on a bike stand and in 9th gear.
If the throttle voltage is any higher than 1.05 volts when you power up the display it shows the EO-4 throttle error.
Shame as the ideal voltage for minimum deadband is 1.5 volts.
At approx 3.4 volts the motor is producing maximum power/speed.
Interestingly the motor power cuts off at 37mph, it will not go any faster.
If I unplug the throttle from the back of the VLCD5 display and then power up the display it doesn't see the 1.5 volts high throttle voltage and boots up fine. If you then plug the throttle back in it still works fine.
I think the only way around this would to have a N/C momentary switch which would be wired in series with the throttle 5v red wire and keep it pressed while the display/controller boots up and goes through it's checks so it doesn't "see" the throttle at all.
Pot resistances measured with the battery disconnected.
Pot 1 pins 1-2 46 Ohms
Pot 2 pins 1-2 44 Ohms
Pot 2 pins 2-3 8.7k Ohms
I am considering changing to a twist type throttle as I find the thumb lever throttle is hard to keep in the same position when riding over bumpy ground.

Update: I have just ordered a momentary switch on Amazon so I will fit that and test again when it arrives tomorrow.

 

[TommyCat]

Hey, thanks for the great feedback! I’ll update the guide with your particular controller’s input reactions when you’re done.

I think the only way around this would to have a N/C momentary switch which would be wired in series with the throttle 5v red wire and keep it pressed while the display/controller boots up and goes through it's checks so it doesn't "see" the throttle at all.

I perhaps would look for a timed delay on switch and use that in series… amazing these days what is availible. (Or not… :-/)

I am considering changing to a twist type throttle as I find the thumb lever throttle is hard to keep in the same position when riding over bumpy ground.

Had the same problem with a “full” twist throttle myself. So would recommend a half-twist model if asked.
Amazing how much better control you have when half your hand is on steady ground so to speak.

Then you could get one with a voltage meter… and a key switch… or a momentary push button….

 

[rcflyers]

Hey, thanks for the great feedback! I’ll update the guide with your particular controller’s input reactions when you’re done.


I perhaps would look for a timed delay on switch and use that in series… amazing these days what is availible. (Or not… :-/)



Had the same problem with a “full” twist throttle myself. So would recommend a half-twist model if asked.
Amazing how much better control you have when half your hand is on steady ground so to speak.

Then you could get one with a voltage meter… and a key switch… or a momentary push button….

Thanks,
Time delay switch like this?

DC 5V/12V/24V Delay Timer Relay, Self-locking Delay On Off Time Switch Super 555 Timer (DC12V) : Amazon.co.uk: DIY & Tools

Great prices on your favourite Home brands, and free delivery on eligible orders.

www.amazon.co.uk

I worry about the vibration from off road riding and having to waterproof it so the push button switch seems a much simpler and more reliable option.

 

[TommyCat]

Time delay switch like this?

In concept that would work I think. But the 80 mA draw to me would be a deal breaker, being a bit high.

I worry about the vibration from off road riding and having to waterproof it so the push button switch seems a much simpler and more reliable option.

 

[n00b]

GAh! This thing's being a pain in the butt. Why is Varstrom all kinds of different than regular throttle standards?'
Grr. So..a VR inline on the hot of the throttle would fix it maybe?
I could tune it to max and then Hot glue it.
Wide open on the throttle is definitely too much, unless there's something I could do inside the motor.
I'll scratch some stuff or whatever, but this ain't gonna get it.
The thing achieves peak rpm at about 5/8 of the throttle I'm using, then bogs and lopes after that.
What kinda crap is that?! Grah!