TSDZ2 OSF for all displays, VLCD5-VLCD6-XH18, LCD3, 860C-850C-SW102.

[raylo32]

Yeah, me, too. But why not have both if you can for a few lines of code? I have another build coming that I guess will be more like an e-motorcycle... a CYC Pro Gen 2 going onto an old but in good shape Stumpjumper. I don't think there is any OSF for that and I am not sure of all the features in the OEM firmware. But I'd be highly surprised if it has cruise control. In any case it should be fun... and NOT be heat limited. Unfortunately all this is getting me out of shape for riding my pedal bikes that is (was?) my primary hobby.

What you said is the crux of this...I am sure at some point I will "try" this feature. ..

Tried, because I just don't need it. I want a bike, not a small motorcycle.

 

[raylo32]

Long day for me... The bike rides great with the VLCD5 version of OSF, but I hate the menus and interface and not having a real time display of motor temperature. Takes a press of the light button to get a brief glimpse. So I decided to install the OSF upgrade kit (850C display, brakes, controller programmed with OSF 1.1) I bought from Electrify a couple years ago that I hadn't messed with yet. Install was straight forward... swap the controller which was essentially what I did already to install the temp sensor in my OEM controller. Then swap the brakes, display and handlebar switch. Before I did this I updated the new controller and 850C display to OSF 20.1C.

Simple, right? Well, no. It didn't work. The 850 booted up, correctly showing the 20.1 firmware but after 45 seconds of "waiting" it throws a "brake error" and fails to activate the motor. I get this whether the brakes are connected or not so it is highly unlikely to be anything to do with brakes or cables which are all brand new. A search here indicates this is a communication problem between the controller and display, which to me basically says I have a bad controller. I also tried rolling the firmwares back to the V1.1 they came with, same result.

So, I swapped my OEM controller and VLCD5 back in (also on OSF 20.1C) and I am back in business. Lots of wasted time, but on a positive note I can now R&R an TSDZ2 motor and swap the controller in less than an hour. Maybe one more hour to also do the brakes and dress up all the cables.

I don't have the ability to diagnose this busted controller but I may try another one some day. Probably from a different provider. We'll see...

 

[raylo32]

....

 

[raylo32]

....

 

[freexxx]

Hi , sometimes I have problem with error 0x2 0x8006090 on 860C display.
The error is randomly displayed when I am in the settings menu and enter the submenu. The error only appears occasionally. Then all that remains is to turn off/on the switch on battery. Shutdown via LCD remote does not work. No error is displayed while driving.

 

[mbrusa]

This is an error that occurs randomly when navigating the configuration menus.
It does not affect the operation of the engine, it is only annoying the need to disconnect the battery.
I still don't understand the cause.
I am trying a new version, I have only added three new items to the menus and I have found that increasing the number increases the likelihood of this happening.

 

[freexxx]

This is an error that occurs randomly when navigating the configuration menus.
It does not affect the operation of the engine, it is only annoying the need to disconnect the battery.
I still don't understand the cause.
I am trying a new version, I have only added three new items to the menus and I have found that increasing the number increases the likelihood of this happening.

Exactly mbrusa.
it si only:
It does not affect the operation of the engine, it is only annoying the need to disconnect the battery.
For me no big problem.

 

[mff47025]

I recently installed this firmware on my bike, it works great and we love it.
My display is a SW102 that was bought programmed to work if stock firmware.
I did not reprogram the SW102. I just installed the firmware for the stock displays. Everything works great, especially now that I calibrated the torque sensor.
There are a few things to know as far as what could cause problems
1) I was unable to reprogram the power meter on the display. I think this must be programmed in the display so the bars still work, just based on the original programming which I never liked. This is actually not a big deal, since I can just turn on the lights and see the actual battery voltage. Huge advantage. I tried setting the display for VLCD5, VLCD6 and XH18. They all worked exactly the same in my case
2) Programming from the display is really tricky for a number of reasons.
A) I don't think this display has an E03, at least I never saw it when I cycled through the programming options. I do think I saw E02 twice. For me it didn't matter, I just wanted to get to the torque settings
B) The display is only 2 digits, so you sometimes have to add a 0 or another number to the end. For me I interpreted the 16 resting torque reading as 160. Seems about right. This is also true if you display current power. Though if you display current, batter voltage, or cadence you get the right number.

None of these issues do I care about. I am able to tweak and fine tune how my motor runs now and I didn't have to cut my display. I'm quite happy.
Thanks for all the hard work.

 

[raylo32]

I took my tsdz2 bike out for a nice hilly 30 mile loop yesterday, first real ride with the OSF installed. The bike ran great, I stayed mostly in eco mode, temp topped out at 62.4, I and returned with about half of the battery capacity remaining. Amazing.

But I have an issue with the wheel diameter/circumference and speed/ODO readings. Wheels are 26" with 1.75" tires. For OSF Java configurator I used the same wheel setting as for the Cateye computer on the bike, 2023mm as specified in the chart below. I also took along my Garmin Edge. Unfortunately the OSF ODO reported a full 13% higher miles than the Garmin and Cateye which were within a couple of % of each other. Based on the 13% delta I'd have to set the wheel diameter in the OSF to something totally wrong to make the ODO report the correct distance. Anyone have a solution?

https://www.cateye.com/data/resources/Tire_size_chart_ENG_151106.pdf

 

[w0utje]

Mine charts says 2114mm for 26*1.75

 

[raylo32]

The charts vary a few %, but my error is 13%. In any case putting a bigger number in will make the error larger, giving credit for more distance traveled per revolution.

Mine charts says 2114mm for 26*1.75

 

[raylo32]

So here's the math for the speed and ODO errors I am seeing with my TSDZ2 with VLCD5 OSF:

The setting I entered into the java config and flashed is 2023 mm for my 26x1.75" wheel and tire. From the Cateye chart, not perfectly measured or calibrated. But close.

So in one mile (measured by Garmin GPS) the wheel makes approximately 796 revolutions. But the VLCD5 on OSF is reporting 1.13 miles which means it is giving credit for a wheel circumference of about 2286mm, which equates to a diameter of about 28.6"... which is about right for a 700Cx47 or a 29er. And just to confirm there is only 1 magnet so it isn't counting extra pulses. (I actually measured the 13% delta on a 10 mile ride segment, the calcs above use 1 mile for clarity.)

Is OSF not taking or recognizing my setting and somehow locked onto that wrong one? No one else seeing this issue?

 

[Elinx]

..

Is OSF not taking or recognizing my setting and somehow locked onto that wrong one?

OSF ODO does only work accurate, if you enable also "Odometer compensation" (basic settings).
Otherwise the measurements of soc%, Voltage and other values are counting too for the distance.
The disadvantage of enabling this, is that you see for a certain time no speed reading.
Till compensated, the speed reading will be 00.0 km/h.

 

[raylo32]

Thanks, that fixed the issue. Suggestion to the developers: perhaps gray out the wheel circumference field in basic settings unless "Odometer Comp" is checked since it seems to be irrelevant until that is done.

OSF ODO does only work accurate, if you enable also "Odometer compensation" (basic settings).
Otherwise the measurements of soc%, Voltage and other values are counting too for the distance.
The disadvantage of enabling this, is that you see for a certain time no speed reading.
Till compensated, the speed reading will be 00.0 km/h.

 

[ilu]

Thanks, that fixed the issue. Suggestion to the developers: perhaps gray out the wheel circumference field in basic settings unless "Odometer Comp" is checked since it seems to be irrelevant until that is done.

It is used for accurate speed reading as well. I'd think it's more important for most users.

 

[raylo32]

Yes, but is it really giving accurate speed readings if the odo cal is not selected? I am not a coder and so I haven't dug into the code but the math is simple, if the speed is accurate and the time is accurate then should not the distance also be accurate? IOW if the speed is accurate and uses the input circumference why do we need to select "odo cal" to make the distance compute correctly? Who knows how they have it set up and I guess it really isn't important since mine is working fine now that I have odo cal selected.

Thanks, that fixed the issue. Suggestion to the developers: perhaps gray out the wheel circumference field in basic settings unless "Odometer Comp" is checked since it seems to be irrelevant until that is done.

It is used for accurate speed reading as well. I'd think it's more important for most users.

 

[Elinx]

Yes, but is it really giving accurate speed readings if the odo cal is not selected? .....

I did not take the value from a table, but measured the wheel circumference very precisely and entered it.
I also always measure distance and speed using GPS while cycling and I can confirm that the measured distance from the display differs only very slightly from the GPS distance. Speed is a bit more difficult to compare, but is fairly similar if you can drive a constant speed for some time.

If ODO compensation isn't enabled, the speed reading should be accurate.
Only the ODO distance will be considerably more depending on how often you want to read other values.
This these values show as speed reading too. A full battery give a (short) speedreading of 99,9 km/h

 

[ilu]

Yes, but is it really giving accurate speed readings if the odo cal is not selected? I am not a coder and so I haven't dug into the code but the math is simple, if the speed is accurate and the time is accurate then should not the distance also be accurate? IOW if the speed is accurate and uses the input circumference why do we need to select "odo cal" to make the distance compute correctly? Who knows how they have it set up and I guess it really isn't important since mine is working fine now that I have odo cal selected.

If you disable all the options for showing any additional data in the speed field, the "odometer compensation" option makes no difference. All these options are quite clearly explained in the manuals, which I think are included in the java configurator folder.

 

[raylo32]

Oh, I see. I did read the manual on this subject but not thoroughly enough. I thought the ODO reading went up only when making parameter changes, not just simply calling up the data whilst riding. Thanks, again. In the final analysis the ODO reading really isn't important these days since most of us are using GPS based devices for navigation and following cue sheets.

If ODO compensation isn't enabled, the speed reading should be accurate.
Only the ODO distance will be considerably more depending on how often you want to read other values.
This these values show as speed reading too. A full battery give a (short) speedreading of 99,9 km/h

 

[Elinx]

... I did read the manual on this subject but not thoroughly enough. ......

mbrusa spends a lot of time on these manuals. He deserves it, that you read them as intended.
Mbrusa made some clever handling solutions for the stock displays, but it is neccesary to understand why he made the choices he did, due to the limited options of these displays.

 

[hetm4n]

gentlemen, more and more motors are coming with new v2 controllers. my friend recently bought on pspower. bought 4 for the whole family from different sources, for 3 engines only 1 was on the old controller.

 

[Elinx]

....more motors are coming with new v2 controllers. ....bought on pspower. bought 4 for the whole family from different sources, for 3 engines only 1 was on the old controller.

It isn't complete clear for me. Do I understand well?
1 motor from PSW
4 motors from different stores, only 1 with the old controller?
Please, can you be more specific, which stores this is about?
The VeloFast type, with that (white) chainwheel, on the picture looks different from PSW. **

How do you know it is the v2 controller?
Have you tried to flash or measured with a multi-meter between swim and gnd pin.

I had the impression that PSW still sold the old controller for the international market and that the v2 type was currently specifically intended for EU, sold by EnerProf.

Edit:
** Found VeloFast here, but it looks an old type from 2015. (VeloFast.de doesn't exist anymore)

 

[casainho]

gentlemen, more and more motors are coming with new v2 controllers. my friend recently bought on pspower. bought 4 for the whole family from different sources, for 3 engines only 1 was on the old controller.

That are very good news, as the motor controller V2 is much better.
Our OpenSource firmware is not fully developed yet but we already have the motor running... We have all the tools needed to develop the firmware, we just need more developers working on it and here is our current firmware repository on Github:

https://github.com/OpenSourceEBike/TSDZ2_motor_controller_v2

 

[hetm4n]

Edit:
** Found VeloFast here, but it looks an old type from 2015. (VeloFast.de doesn't exist anymore)

I opened this Velofast, in fact an old controller, it had a different programming plug one motor for sure enerprof was purchased, my friend made a mistake, the motor bought with ps power was on the old controller. velofast from another store.

 

[Elinx]

....Velofast,....old controller, ....one motor..... enerprof ... ps power ... old controller.

So problably only the motor from Enerprof would have the v2 controller inside. All others the old v1.