This is really smart. I’ll have to do some comparisons between the two.
For anyone else with a similar issue, you can also take a screen shot of the motor foc settings from 5.3 and either compare that to 6.0 or manually enter those into 6.0.
6.0 uses a new observer compared to 5.3 so I wonder if that’s giving unreliable motor detections on some motors. I’ve had some issues with the big 46 pole direct drive hubs but I don’t know if it’s the same problem yet.
Haha to be honest, as much as I love the vesc, the observer is unfortunately totally wrecked on the 6.0b. Actually there is a big difference (please don’t be offended) between 56 and 59 (even if far from correct like with the 5.3 firmware). At least here on a “small” Bafang g310 (16 poles inrunner).
I have now to find why the PAS isn’t working. I have a simple one (three wires - one way) and nothing comes out right now with the modified code. May try first with a quadratic one I have laying.
Btw. Are you sure about the pinout of the DIO and CLK in hw? They seems to be GPIOA 15 and GPIOB 3 but on vesc 6 board they should be GPIOA 13 and GPIOA 14… I was a bit surprised.
Gruß,
H.
No offense taken haha. This is Vedder’s firmware that I created a hardware config for. Those issues are unrelated to the hardware config files. Still interesting to hear because I’ve had some issues with different vescs running 6 beta on non-esk8 motors, but haven’t been able to track down the exact cause.
PAS was added to the vesc project by Afzal. He can be found on the endless sphere forums or on the vesc discord. If everything is hooked up right he might be able to provide some answers. From the pull request it sounds like Vedder used the uart port for pas input if it’s not defined in the hwconfig.
For the 75100v201 yes, I used the st-link on it several times. I don’t remember if I ever used the st-link on the v202 once the new firmware was loaded.
How have you set up PAS? I’ve been trying to figure out how to PAS with vesc
Thanks mate 
I tried to write him directly on discord but no answer 
I will try on the ES. Something isn’t right I think. May be something totally stupid but wiring is 100% right and I didn’t touch anything in the soft/hw. Tbh. That’s something with what I really struggle with the vesc. I didn’t find any way to debug port activity. For example you get ADC voltage view and if there is an issue, you see it quickly in vesc tool… but for PAS there is nothing, except the good old voltmeter 
Thanks again, the 75100 ain’t that bad, get a bit warm tho… Where did you source your MOSFETs upgrade? I searched everywhere without succes.
Gruß,
H.
The mosfets the old 75100 uses are pretty good spec wise… I think the problem is mostly the TO-220 package and the lack of thermal mass. I don’t think a mosfet upgrade will help much unless you could somehow double them.
The 75100 aluminum pcb performs soooo much better thermally that it’s not even in the same category. 216 seconds vs 31 seconds at 80A. My theory is that because it uses SMD mosfets and the heatsink is directly on the other side of the PCB it can transfer heat much more efficiently. Cooling wise it performs about the same as a Stormcore which uses a very similar cooling design.
I did a comparison a few posts up
thanks mate,
yeah I’ve seen that the aluminum version was way stronger and had a better thermal behaviour. I thought you were the one having swap mosfets for the hsp0076a: http://www.icmkw.com/file-download-397530-left.html
they have 1.8mOhm and get 380A (the aluminium version has 1.2mOhm I think) if I recall right which is a great improvement compared to original IMHO. that said… not with original casing which is a complete joke. I want to go copper + bike frame cooling. but yes indeed in terms of Phase current per dollar… the aluminum version is hands down better 
I will try to find a way to get my hands on some, they seems pretty difficult to source
and the reason I want to220… is because fets die sometimes, and to220 to replace is pretty convenient compared to smd ones
LCSC has them in stock, but 1.8 vs 2.5mohms may not make a big difference.
I’m still thinking adding a second set of mosfets would be the best bang for the buck cooling wise. Would just need is own gate resistor added before the current gate resistor. This assumes the gate drivers have enough current for the job… but seems worth the experiment
TI-driving-parallel-mosfets.pdf (655.7 KB)
Thanks for the link mate
Yeah it’s not much more… still 30% win
The issue with dual fets is the 90 degree positioning of the second rail… this suck a lot and there is not much other way to assemble them… we should make a petitiion for " mosfet prinverted pining" production 
I finally got an answer from Afzal. could help me to start Debugging the PAS drama ( I can use printfs now… that really sweet)… but except to see that I plugged everything correctly and the port do get 0-1 when turning the sensor… well… nothing else works. 
and thanks for the pdf, thats pretty cool, 75-200 may come earlier as originally planned X-))
I was using these MOSFETs on my Flipsky 75100 and they are great, I can recommend them a lot! Much better than stock MOSFETs on the regular Flipsky 75100, but from spec not as good as the Aluminium PCB, but due to the TO220 design which can easily be mounted directly to heatsinks you can reach much better temps.
ah… it was you Thanks for sharing this. And yeah, I will use the aluminum frame and copper link. this should work just fine. aaaaaaaand now that PAS is working. I can soon test it correctly :))))
Hey mate:
got that thing working
I used the m365 dash from 1zuna + a xiaomi m365 dash to control everything (it’s really sweet)
PAS works now like it should be here in germany:
6kmh limit with throttle only + 25kmh with cadence pas
OR
6kmh limit with throttle + 25kmh with PAS active at 0% power & control from power with throttle.
and three level of power/speed whatever to be selected from display.
thats pretty awesome now just have to wait for my mosfets to pull a bit of amps
Good to hear!
The dash sounds like it works pretty well.
If you wanted to take it a step further, I’d also suggest looking into the metr module. The VESC app has some really limited profile support, but the metr app you can create profiles that can change every setting in the VESC.
Basically it means you can swipe on your phone for a “street legal” mode, then have another profile where it’s like a “power mode” when no one is looking
Haha, I currently don’t have a nrf connected right now X-))
I will make this later. I needed a fully legal setup to test first. and germans are pretty weird with this.
you can do the speed settings actually too with the vesc apk. simply connect, change max speed and hop… you are full… when you respawn the system or change it on the dash, it takes the lead and write legal settings again :)))
the trick was to be able to have a nice simple “one button” display (battery, power level,speed) change the max speed under PAS status…
I have the white pcb version and the blue pcb version. Can i run them both for a dual hub setup? Assuming both have same firmwsre version
Yes, no problems running a canbus link between different vesc models. They just need to be on the same firmware version (ie 5.3)
i use flipsky 75100V202 and uptated the firmware 6.0b56. i succesfully upgrade the firmware, but after i connect it with vesctool 4.0t9, there is prompt that say couldnt deserialize motor configuration. what is the problem and how i can fix it ?
i wanted to upgrade it because i want to test the silent hfi method
thanks
Hi, i’m running the official 75300 V2 firmware on the FSESC with Alu PCB (2pcs), however i seem to get less performance than expected @ 120 phase amps and 38 battery amps, the battery & vesc inputs don’t drop that much, and for some reason these 1200W motors top at 1300W a piece while my Duty cycle doesn’t ever go above 70%, i’ve checked everything in the vesc tool but i noticed the vref etc is different in your custom firmware, maybe that also will fix some of the other stuff like less power than expected (off FOC parameters)?
Also noticed that BLDC mode just wont work at all, only FOC mode works
Do you think the custom firmware could fix the issues i’ve got?
I think i had more acceleration out of the 2 JP 28A (28 battery amps) controllers that came stock in my scooter
The custom firmware certainly can’t hurt. It’s the same 5.3 firmware as the main branch, just configured properly for that hardware unlike Flipsky’s 75_300_R2 firmware used as a work around.
Are the connected using CANBUS?
Those silver box ebike controllers I don’t believe limit phase current, or if they do it’s very high. That’s going to give some very punchy acceleration from 0-5mph. Where the VESC will really shine is in the mid and top range performance since you can up the battery current to 100A on each controller if your battery could support it. 120A phase current on two motors should feel very powerful though.
If you are running the default firmware I’d make sure phase filters are off before running the motor config, then make sure not to reset defaults. Use the 2kg outrunner setting to detect the hub motors in the scooter. It seems counter-intuitive but it seems to yield better detection results than the hub settings.
If using the custom firmware the phase filters are disabled so that setting no longer matters. It should get over 70% duty, and if not then it’s most likely bad motor detection results or phase filters being active, but it could be a lack of power from the lowish battery amps… but that’s doubtful. You could confirm with some logging to see if you are hitting the battery amp limit at 70% duty. Also make sure you aren’t using any erpm or duty cycle limits.
Let us know how it goes with the custom firmware and the 2kg outrunner detection!
Thank you for the info, i’ve indeed used hub motor detection and the minimum read temp is around 31°C and i’ve read somewhere else that that is an issue with official firmware
The FSESC’s are connected through can, in my logs i currently get 185-187 phase A peaks (both together), so i guess it is a combination of the firmware and wrong detection, i’ll first flash your firmware and try the different detection! thanks for the info aswell!