FLIPSKY new 20s 100A tiny controller (vesc based)

[ypl]

I am the original author of 75100 hardware.
Due to my personal reasons, the schematic has not been completed. At present, there are only PCB files with chaotic component annotation and source code.There are also some typical features in the hardware design. Some resistors and capacitors are saved in some places, which is not a very perfect thing.

This is the 75100 which I have been working on for the last year March.

During this period, I made several 75100 by hand and sold them on the second-hand trading platform in order to test the stability of the scheme.

The hardware definition of 75100 is changed based on the hardware definition of 100250. The ADC channel, shunt setting and partial voltage setting are reconfigured, and phase filter, redundant temperature sensor and other peripherals are removed.
The maximum current of the firmware is defined as 120a and the maximum voltage is 120V. This is determined by the range of the shunt and voltage dividing resistance, not MOSFETs.

I have made two versions, one using stm32f405rgt6 and the other using stm32f407igh6.
At the same time, I also made a 75200, but the production process is more difficult.Because the copper column is difficult to accurately weld on the aluminum substrate,This is almost impossible to mass produce.

I modified the overcurrent protection part of the firmware, because in VESC, the overcurrent protection does not directly depend on the sampling value of a shunt, but the motor current calculated from three shunts. If the MOSFET fails, the current of a shunt will be very large, but the program will not return an error, Therefore, I modified the program so that when the Slow ABS Current Limit is False, the overcurrent protection is directly based on the value sampled on the shunt, and when the Slow ABS Current Limit is True, it is based on the original overcurrent protection.

I downloaded the firmware people got from flipsky fs75100 from the Internet and downloaded it to my hardware. Based on this, I can conclude that the firmware was obtained from a 75100 I had previously sold, and its version is 3.00.Later, I learned that the staff of flipsky bought two 75100 from me, but I didn't know who the buyer was at that time.

The reason for publishing this article is that I saw a product very similar to the 75100 I was using on the Internet. I searched on the Internet and found that the address and telephone number of the consignee I bought were consistent with those on the flipsky website.

Recently, I downloaded the beta version of VESC code and recompiled the newer 75100 firmware.

https://github.com/1611048264/vesc

Now you can experience the latest features of VESC on 75100, such as MTPA and field weakening.
https://endless-sphere.com/forums/viewtopic.php?f=2&t=115007&p=1700673&hilit=75100#p1700673

 

[ypl]

It's beyond strange, it's actually ILLEGAL and it's theft.

You sound like fun at a party.

You mean the party that is you having all this cool VESC stuff that's 100% made possible by folks adhering to copyright laws? The kind of fun allowing YOU to build cool electric vehicles? It's all because of the laws, buddy. This helps YOU if they follow the law, not some foreign abstract concept that has no real world benefits.

You should really be advocating for yourself.

Flipsky wouldn't even be building this at all --- if other people hadn't followed the law before them. And, you wouldn't be able to buy it.

When you realize that the person they're stealing from is actually the community and you personally, it starts to become clear why the copyleft license is actually so good. But crime is crime, and in this case, Flipsky is the criminal.

I've written code and released it and designs and released them, I don't ride vesc, I ride my own code. All my stuff is released under BSD 3 clause. So no I don't rely on the vesc party.

Meanwhile there's good reason to stop people modifying the code onboard like... It lets people trivially up the current to 2, 3, 10x the design intent and a pile of warranty and potentially safety claims ensues.

Like everyone in this thread is already trying to do.

I'm the author of 75100. See here for details,https://endless-sphere.com/forums/viewtopic.php?f=2&t=115007&p=1700673#p1700673

 

[mxlemming]

You sound like fun at a party.

You mean the party that is you having all this cool VESC stuff that's 100% made possible by folks adhering to copyright laws? The kind of fun allowing YOU to build cool electric vehicles? It's all because of the laws, buddy. This helps YOU if they follow the law, not some foreign abstract concept that has no real world benefits.

You should really be advocating for yourself.

Flipsky wouldn't even be building this at all --- if other people hadn't followed the law before them. And, you wouldn't be able to buy it.

When you realize that the person they're stealing from is actually the community and you personally, it starts to become clear why the copyleft license is actually so good. But crime is crime, and in this case, Flipsky is the criminal.

I've written code and released it and designs and released them, I don't ride vesc, I ride my own code. All my stuff is released under BSD 3 clause. So no I don't rely on the vesc party.

Meanwhile there's good reason to stop people modifying the code onboard like... It lets people trivially up the current to 2, 3, 10x the design intent and a pile of warranty and potentially safety claims ensues.

Like everyone in this thread is already trying to do.

I'm the author of 75100. See here for details,https://endless-sphere.com/forums/viewtopic.php?f=2&t=115007&p=1700673#p1700673

Hi
Lovely looking boards and your commentary on the overcurrent is very interesting. It's well known among those of us building VESC derivatives that the protection is frequently lacking and occasionally terminal.

Between your board and the flipsky one there are a few important differences to note... Firstly the flipsky one has 3x0.5mohm resistors stacked up (!?!) I just found when I opened mine. Your firmware files define 1 of them.

Are you aware of the extent to which they've copied your board? It looks similar, similar component position and choice but the detailed layout seems quite different. This isn't manufacture from your Gerber files, they look like they've been recreated from scratch/another source.

I only got mine a few days ago... I'll be writing about it soon. Leant my oscilloscope to someone for a few days unfortunately.

I still find it utterly perplexing that flipsky can create this complex hardware through semi reverse engineering but seem to be unable to modify the 10 lines of code required to create the header files and thus have used your image.

https://github.com/1611048264/vesc/blob/6128d0619073b8a3187ddeca7728a318bbef2ee5/mc_interface.c#L1829

This is your modification to the overcurrent?

I thoroughly approve. This is very similar to my implementation for my own code, and has resulted in precisely zero dead MOSFETs since I wrote those few lines.

 

[district9prawn]

Are you aware of the extent to which they've copied your board? It looks similar, similar component position and choice but the detailed layout seems quite different. This isn't manufacture from your Gerber files, they look like they've been recreated from scratch/another source.

I only got mine a few days ago... I'll be writing about it soon. Leant my oscilloscope to someone for a few days unfortunately.

I still find it utterly perplexing that flipsky can create this complex hardware through semi reverse engineering but seem to be unable to modify the 10 lines of code required to create the header files and thus have used your image.

This seems to be a pattern of behaviour with Flipsky. Very similar (but not identical) hardware emerges some time after they presumably get their hands on a design. Then they are not very forthcoming with firmware.

A couple of years ago I asked them if they planned to sell something higher power similar to the 75/300. Their reply was that they were but were having "firmware issues". I found this response odd given that it is fairly trivial to set the defines in the code.

 

[Ianhill]

This is very shady behaviour from flipsky to say the least, the most concerning part for me is how they took ypl work, theres no playing stupid with that one on licencing agreements its flatout theft from another member.

I for one wont be buying from flipsky again even if they have a rip off big power controller its fail rate will be high with the consumer always doing the majority of the testing, i get its not easy to start up an electronic business for hobbyists but the cut throat nature in which they operate has allowed them to survive stealing from the developer will force open source to disappear and stealing from the board creators will force a makers to keep cards close to they chest leaving forums like this out of the loop.

Shame on flipsky. Now ypl has come to light its put the spotlight on them more than ever but people with low money will still buy them most of us on this forum proberly been suckered with some they shit.

 

[ypl]

You mean the party that is you having all this cool VESC stuff that's 100% made possible by folks adhering to copyright laws? The kind of fun allowing YOU to build cool electric vehicles? It's all because of the laws, buddy. This helps YOU if they follow the law, not some foreign abstract concept that has no real world benefits.

You should really be advocating for yourself.

Flipsky wouldn't even be building this at all --- if other people hadn't followed the law before them. And, you wouldn't be able to buy it.

When you realize that the person they're stealing from is actually the community and you personally, it starts to become clear why the copyleft license is actually so good. But crime is crime, and in this case, Flipsky is the criminal.

I've written code and released it and designs and released them, I don't ride vesc, I ride my own code. All my stuff is released under BSD 3 clause. So no I don't rely on the vesc party.

Meanwhile there's good reason to stop people modifying the code onboard like... It lets people trivially up the current to 2, 3, 10x the design intent and a pile of warranty and potentially safety claims ensues.

Like everyone in this thread is already trying to do.

I'm the author of 75100. See here for details,https://endless-sphere.com/forums/viewtopic.php?f=2&t=115007&p=1700673#p1700673

Hi
Lovely looking boards and your commentary on the overcurrent is very interesting. It's well known among those of us building VESC derivatives that the protection is frequently lacking and occasionally terminal.

Between your board and the flipsky one there are a few important differences to note... Firstly the flipsky one has 3x0.5mohm resistors stacked up (!?!) I just found when I opened mine. Your firmware files define 1 of them.

Are you aware of the extent to which they've copied your board? It looks similar, similar component position and choice but the detailed layout seems quite different. This isn't manufacture from your Gerber files, they look like they've been recreated from scratch/another source.

I only got mine a few days ago... I'll be writing about it soon. Leant my oscilloscope to someone for a few days unfortunately.

I still find it utterly perplexing that flipsky can create this complex hardware through semi reverse engineering but seem to be unable to modify the 10 lines of code required to create the header files and thus have used your image.

https://github.com/1611048264/vesc/blob/6128d0619073b8a3187ddeca7728a318bbef2ee5/mc_interface.c#L1829

This is your modification to the overcurrent?

I thoroughly approve. This is very similar to my implementation for my own code, and has resulted in precisely zero dead MOSFETs since I wrote those few lines.

Before publishing this article, I haven't provided PCB files and code to others, but I sold a hardware to others before, so they should reverse my hardware. See here for details
https://endless-sphere.com/forums/viewtopic.php?f=2&t=115007

 

[mxlemming]

This confuses me more and more.

Looks like they took ypls design/ decided it was nice and cheaply mass producible and so recreated the entire thing (this is not a PCB scan and duplication...) swapped FETs, dcdcs, moved all the connectors, rerouted it...

Then after doing all that, they can't edit 20 lines of code and type make into a Linux terminal, so they suck out the firmware.

The mind boggles. It really does. Reverse engineering the pinout is more effort than making headers and compiling your own firmware.

The other thing that baffles me is why they didn't just ask ypl for the header files since he's "legally obliged" to provide them under the GPL licence having sold the device...

 

[ypl]

Are you aware of the extent to which they've copied your board? It looks similar, similar component position and choice but the detailed layout seems quite different. This isn't manufacture from your Gerber files, they look like they've been recreated from scratch/another source.

I only got mine a few days ago... I'll be writing about it soon. Leant my oscilloscope to someone for a few days unfortunately.

I still find it utterly perplexing that flipsky can create this complex hardware through semi reverse engineering but seem to be unable to modify the 10 lines of code required to create the header files and thus have used your image.

This seems to be a pattern of behaviour with Flipsky. Very similar (but not identical) hardware emerges some time after they presumably get their hands on a design. Then they are not very forthcoming with firmware.

A couple of years ago I asked them if they planned to sell something higher power similar to the 75/300. Their reply was that they were but were having "firmware issues". I found this response odd given that it is fairly trivial to set the defines in the code.

In fact, their employees do not have the ability to modify and compile firmware, so they use the firmware I modified, and other products use the open source firmware on the network

 

[ypl]

This confuses me more and more.

Looks like they took ypls design/ decided it was nice and cheaply mass producible and so recreated the entire thing (this is not a PCB scan and duplication...) swapped FETs, dcdcs, moved all the connectors, rerouted it...

Then after doing all that, they can't edit 20 lines of code and type make into a Linux terminal, so they suck out the firmware.

The mind boggles. It really does. Reverse engineering the pinout is more effort than making headers and compiling your own firmware.

The other thing that baffles me is why they didn't just ask ypl for the header files since he's "legally obliged" to provide them under the GPL licence having sold the device...

Because they know that such behavior is equivalent to trying to copy what I made for mass production. Naturally, they won't ask me for these

 

[mxlemming]

This confuses me more and more.

Looks like they took ypls design/ decided it was nice and cheaply mass producible and so recreated the entire thing (this is not a PCB scan and duplication...) swapped FETs, dcdcs, moved all the connectors, rerouted it...

Then after doing all that, they can't edit 20 lines of code and type make into a Linux terminal, so they suck out the firmware.

The mind boggles. It really does. Reverse engineering the pinout is more effort than making headers and compiling your own firmware.

The other thing that baffles me is why they didn't just ask ypl for the header files since he's "legally obliged" to provide them under the GPL licence having sold the device...

Because they know that such behavior is equivalent to trying to copy what I made for mass production. Naturally, they won't ask me for these

This is literally the point of GPL, that they can get the source code from you since you supplied the binary. The reason GPL exists is to ensure this...

GPL isn't there to protect individuals and the small guy.

How can they not have the ability to edit the firmware? The learning curve is like... 3 hours. Laying out a PCB and making working hardware is weeks of work, editing the vesc headers is like 1 hour if you have linux installed already and anything from 0.5-3000h if you have to install linux.

 

[ypl]

This confuses me more and more.

Looks like they took ypls design/ decided it was nice and cheaply mass producible and so recreated the entire thing (this is not a PCB scan and duplication...) swapped FETs, dcdcs, moved all the connectors, rerouted it...

Then after doing all that, they can't edit 20 lines of code and type make into a Linux terminal, so they suck out the firmware.

The mind boggles. It really does. Reverse engineering the pinout is more effort than making headers and compiling your own firmware.

The other thing that baffles me is why they didn't just ask ypl for the header files since he's "legally obliged" to provide them under the GPL licence having sold the device...

Because they know that such behavior is equivalent to trying to copy what I made for mass production. Naturally, they won't ask me for these

This is literally the point of GPL, that they can get the source code from you since you supplied the binary. The reason GPL exists is to ensure this...

GPL isn't there to protect individuals and the small guy.

How can they not have the ability to edit the firmware? The learning curve is like... 3 hours. Laying out a PCB and making working hardware is weeks of work, editing the vesc headers is like 1 hour if you have linux installed already and anything from 0.5-3000h if you have to install linux.

This is also my negligence. When I sold this ESC, I thought that the other party was just an electric skateboarding player and just wanted to use this ESC, so I didn't actively provide the modified code. Later, he didn't ask me for the code, but I can guarantee to provide the source code when he needs it.

 

[ypl]

This confuses me more and more.

Looks like they took ypls design/ decided it was nice and cheaply mass producible and so recreated the entire thing (this is not a PCB scan and duplication...) swapped FETs, dcdcs, moved all the connectors, rerouted it...

Then after doing all that, they can't edit 20 lines of code and type make into a Linux terminal, so they suck out the firmware.

The mind boggles. It really does. Reverse engineering the pinout is more effort than making headers and compiling your own firmware.

The other thing that baffles me is why they didn't just ask ypl for the header files since he's "legally obliged" to provide them under the GPL licence having sold the device...

Because they know that such behavior is equivalent to trying to copy what I made for mass production. Naturally, they won't ask me for these

This is literally the point of GPL, that they can get the source code from you since you supplied the binary. The reason GPL exists is to ensure this...

GPL isn't there to protect individuals and the small guy.

How can they not have the ability to edit the firmware? The learning curve is like... 3 hours. Laying out a PCB and making working hardware is weeks of work, editing the vesc headers is like 1 hour if you have linux installed already and anything from 0.5-3000h if you have to install linux.

The main difficulty lies in the configuration of this linux environment, which has cost me a lot of time.

 

[atkforever]

Btw Flipsky isn't selling it anymore.

Either on official website or aliexpress. It's not out of stock, just removed.

 

[badgineer]

Hi

I am the original author of 75100 hardware.

I still find it utterly perplexing that flipsky can create this complex hardware through semi reverse engineering but seem to be unable to modify the 10 lines of code required to create the header files and thus have used your image.

My respect for you ypl having made a controller worth cloning .
But what a plot twist!
And I thought that flipsky had some super secret fancy sauce on the software side.
And they can't even get a header file done.

Btw Flipsky isn't selling it anymore.

Either on official website or aliexpress. It's not out of stock, just removed.

And in yet another plot twist:

https://flipsky.net/collections/electronic-products/products/flipsky-75100-foc-75v-100a-single-esc-base-on-vesc-for-electric-skateboard-scooter-ebike-speed-controller

" Firmware: latest (firmware update supported) "

Br,

 

[mechhead]

The 75/100 is now back up on their website

 

[b264]

This confuses me more and more.

Looks like they took ypls design/ decided it was nice and cheaply mass producible and so recreated the entire thing (this is not a PCB scan and duplication...) swapped FETs, dcdcs, moved all the connectors, rerouted it...

Then after doing all that, they can't edit 20 lines of code and type make into a Linux terminal, so they suck out the firmware.

The mind boggles. It really does. Reverse engineering the pinout is more effort than making headers and compiling your own firmware.

The other thing that baffles me is why they didn't just ask ypl for the header files since he's "legally obliged" to provide them under the GPL licence having sold the device...

Because they know that such behavior is equivalent to trying to copy what I made for mass production. Naturally, they won't ask me for these

This is literally the point of GPL, that they can get the source code from you since you supplied the binary. The reason GPL exists is to ensure this...

GPL isn't there to protect individuals and the small guy.

How can they not have the ability to edit the firmware? The learning curve is like... 3 hours. Laying out a PCB and making working hardware is weeks of work, editing the vesc headers is like 1 hour if you have linux installed already and anything from 0.5-3000h if you have to install linux.

GPL was created explicitly to protect individuals and the small guy.

 

[mxlemming]

This confuses me more and more.

Looks like they took ypls design/ decided it was nice and cheaply mass producible and so recreated the entire thing (this is not a PCB scan and duplication...) swapped FETs, dcdcs, moved all the connectors, rerouted it...

Then after doing all that, they can't edit 20 lines of code and type make into a Linux terminal, so they suck out the firmware.

The mind boggles. It really does. Reverse engineering the pinout is more effort than making headers and compiling your own firmware.

The other thing that baffles me is why they didn't just ask ypl for the header files since he's "legally obliged" to provide them under the GPL licence having sold the device...

Because they know that such behavior is equivalent to trying to copy what I made for mass production. Naturally, they won't ask me for these

This is literally the point of GPL, that they can get the source code from you since you supplied the binary. The reason GPL exists is to ensure this...

GPL isn't there to protect individuals and the small guy.

How can they not have the ability to edit the firmware? The learning curve is like... 3 hours. Laying out a PCB and making working hardware is weeks of work, editing the vesc headers is like 1 hour if you have linux installed already and anything from 0.5-3000h if you have to install linux.

GPL was created explicitly to protect individuals and the small guy.

Edit... I'm electing to remove my last (this) comment because I've recently had direct contact with Benjamin, and concluded that since he's a thoroughly nice guy just working his ass off, I do not want to mix up my distaste for GPL and the followers and hangers on with his project.

 

[john61ct]

Sorry who is Benjamin in this context?

 

[mxlemming]

Sorry who is Benjamin in this context?

Benjamin Vedder, the author of VESC project. Writes 95% of the code.

VESC = Vedder ESC

 

[b264]

Because they know that such behavior is equivalent to trying to copy what I made for mass production. Naturally, they won't ask me for these

This is literally the point of GPL, that they can get the source code from you since you supplied the binary. The reason GPL exists is to ensure this...

GPL isn't there to protect individuals and the small guy.

How can they not have the ability to edit the firmware? The learning curve is like... 3 hours. Laying out a PCB and making working hardware is weeks of work, editing the vesc headers is like 1 hour if you have linux installed already and anything from 0.5-3000h if you have to install linux.

GPL was created explicitly to protect individuals and the small guy.

Edit... I'm electing to remove my last (this) comment because I've recently had direct contact with Benjamin, and concluded that since he's a thoroughly nice guy just working his ass off, I do not want to mix up my distaste for GPL and the followers and hangers on with his project.

That's good, because our feelings (mine, yours, and everyone's) about the GPL license are 100% irrelevant. That's the license the author chose --- and it has specific legal terms that must be followed.

How we feel about it, is irrelevant in that context.

If you want to use VESC, you follow the license terms. Otherwise use different code and follow its license terms. Otherwise, write your own code.

 

[john61ct]

really tiresome

"must be followed"

You and what army?

Laws may as well not exist if they are not enforced, as in with guns.

I mean, I agree with you as far as the morality goes, but in practice scofflaws abound and 99.99% of the time get away with their scoffing, even prosper.

 

[mxlemming]

really tiresome

"must be followed"

You and what army?

Laws may as well not exist if they are not enforced, as in with guns.

I mean, I agree with you as far as the morality goes, but in practice scofflaws abound and 99.99% of the time get away with their scoffing, even prosper.

He's actually really glad they don't enforce laws well, otherwise he'd get 500m down the road before the police arrested him for riding his unlicenced 3kW electric tea tray on the sidewalk.

 

[tostino]

really tiresome

"must be followed"

You and what army?

I get the sentiment, and I personally release most of my code under MIT because it fits better with my philosophy than GPL. But the disregard for other people's work when discussing open source projects really gets to me. There is a reason people choose the licenses they do, and I get that enforcement by yelling on a forum isn't exactly effective, but people raising their concerns are quite valid IMO.

Anyways, i'd much rather give the original creator my business if at all possible, so i'll be reaching out to them rather than dealing with Flipsky.

And I also echo the sentiment that some laws are inherently more just than others. I don't think it's asking much for a company to release whatever source changes when they are profiting off of someone else's work, and I think enforcement should be greatly increased in that area as it will do nothing but spur more innovation. I also don't give a shit about people riding "illegal" vehicles, so take from that what you will.

 

[john61ct]

Again, I 100% agree that if at all possible we should not give money to pirates and IP scofflaws.

But once that's been stated, in a given thread it should IMO not be reminded more than every couple weeks

and if others are knowingly breaking the law, let it be not our business

and yes especially restrictions on our DIY vehicles

 

[Archer321]

So all legalities aside, is the 75100 a good controller or not?
I'm looking for a super efficient controller with regen to use with my sensorless 36v 1500w motor. It's going on a small kart.