esk8.news esk8 calc donate now

How to update firmware on the Flipsky 75100 FOC ESC

Im using on costum ebike, my motor is 10" 1200w from escooter. I will share some logs, where to upload them? New user can’t upload

2 Likes

Looks like the same style of 5" motor I am using.
So you run the ESC without enclosure?

@xsynatic

let your magic happen.

I unfortunately have no user related magic.

He can share it with someone via PM, so the other can share it for him.

1 Like

@Xgunax if you like you can pm me.

You might have to PM him, new users cant start PM dialogues either.

2 Likes

I loaded 5.03 beta and gave field weakening a shot. Wow.

The geared hub motor is actually an inrunner and seemed to take very well to FW compared to some disappointing bench tests on a standard esk8 outrunner. I did two runs back to back on the same stretch. There is a flat stretch long enough to get to top speed, then a fairly steep hill climb.

My metr was mounted in a bad location so I didn’t get very good logs, but I’ll try again later and post some.

65 battery amps (two 13s4p batteries in parallel) and 70 motor amps on 15awg wire LOL

Run 1: 0A FW (base speed). 22mph top speed on the flat, 20mph up the hill.
Run 2: 30A FW. 27.5mph top speed on the flat, 22mph up the hill
Run 3: 40A FW. 29mph top speed on the flat

Uphill performance barely improved and this makes sense now that I think about it. Field weakening current takes from the maximum motor current pool so there isn’t much additional power available as most is being used to overdrive the motor.

The power requirements are enormous above base speed. It’s pretty much pegged at 50A cruising at 28mph, vs when it’s below base speed the amps vary depending on the load. That also meant that the controller hit 70C pretty quickly. Feature wise it would be nice if field weakening only worked when a button was being held. That would allow it to work as a sort of a temporary afterburner so you didn’t drain your battery and cook your windings.

4 Likes

This would be really great!

2 Likes

Recorded a few logs testing field weakening and continuous amps before overheating/thermal throttling.

ESC was set to 65A battery & 70A motor. Mounted inside a plastic box with zero air flow. Couldn’t get this motor to take more than 50-60A (500w geared hub with 15awg phase wires).

Top speed test. This is a flat stretch, then uphill/downhill, turn around, steeper uphill/downhill and flat.

0A field weakening. 22mph top speed on full 13s battery.

40A field weakening. 30mph top speed!

Thermal throttle test. Mostly flat, some hills.

0A field weakening. Full throttle for about 5 miles, drawing about 600-700w on average at 20mph. Ran cool staying under 60C (started hot from previous runs). Sensible 33wh/mi

40A field weakening. Backed off the throttle to keep the amps steady at 35-45A, drawing about 1100-1200w on average at 25-28mph. Temperature slowly increased to 83C and started to thermal throttle after about 10 minutes. Without any airflow it looks like the max continuous motor amps is about 35A. Hilarious 50wh/mi

1 Like

The only thing he hacked was making a legal version that isn’t stolen property. It should perform the same as the version that Flipsky stole from Vedder. Just legally.

2 Likes

Even though I despise the locked down FW approach wondering if Flipsky actually had a non nefarious, technical reason as to why they would only enable FOC vs BLDC.

I quickly tried Hybrid mode and BLDC mode but had some issues. First, I get this strange condition where USB while using VESC tool. The USB port seems to freeze up once the motor draws any current. I have to disconnect and reconnect USB to regain communication. Problem is intermittent and happened on stock FW as well as the Hacked FW. I have two 75100 controllers and it’s happened to both.

I’m pretty competent with these sort of thing as well, Im positive it’s not created by a wiring error. Been an industrial electrician/electronics tech for +25 yrs plus Ive been active with electric RC since before brushless motors and Lipo were even introduced. Motors,motor drivers/escs and batteries are stuff I live for. Lol. VESC is admittedly different than anything Ive done before but only with the app side. The rest of it is straight forward.

2 Likes

Yeah, we all know this. The word " hacked" is just being used for lack of a better description for it. Jaykup did figure out a way to get it working though so technically it is a hack.

1 Like

Mostly true. I had to figure out what configuration settings they used and you can see my comments in the hw_75_100.h file. Things like resistor values for the input voltage divider, shunt config and resistance, beta values for the NTC sensor by measuring resistance at different temperatures, pin mapping for communication ports, etc.

If I got any of that wrong it won’t perform correctly, and I also don’t know if they made any changes to the core firmware that changed how certain features worked.

But yes, they did steal and are profiting off the hard work of an open source project without doing the bare minimum of releasing their changes to the source code.

2 Likes

It’s sad how it’s LITERALLY FREE for Flipsky to follow the law and yet they choose to steal instead.

6 Likes

Was thinking,. I did redo the wiring to eliminate all possible high resistance connections between the battery and motor in-between the FW change. I originally had a few extra XT 60s to create a series connection between my two 7s batteries. I eliminated those and soldered together the stock 4mm barrel connectors between the motor and ESC. No sign of heating with any of these connectors but…

That could account for the performance increase between stock and the hacked. Everything else was equal in FW settingg.

1 Like

It’s possible the USB cord you are using is too small of a wire gauge, or they are using a poor buck converter. I can test to see if I run into the same issue.

It looks like the latest beta VESC tool on the forum link doesn’t deserialize with the beta 76 build linked at the top. Can anyone build the latest?

Does it make sense to pull request the config files so the binaries get made in the upcoming release?

1 Like

I built the latest 5.03.78 and added it to the main post. Tested and confirmed it works with the 2022-01-12 tool.

I’ll work on a pull request in the next few days.

1 Like

I couldn’t get BLDC sensored or BLDC hybrid mode to spin the motor properly but that may be a configuration issue rather than a controller limitation. BLDC sensorless worked well. All FOC modes worked including HFI/VSS.

I couldn’t reproduce the USB issue. I had a stable USB connection using FOC and BLDC under load.

50A test over USB:

1 Like

I guess I was too slow… Vedder just ended the 5.03 beta and released firmware 5.3 and VESC tool 3.01.

I compiled the 5.3 firmware and that should work with the latest VESC tool for Windows and Android.

I also renamed the hardware back to just 75_100 so it will update properly in the future.

1 Like