High Power VESC

[arvidb]

That motor connector looks nice. Do you like it? What brand/type is it?

[TheFallen]

The connector is an EC5 trinity connector and I bought it from here: http://www.ebay.co.uk/itm/331244976106.
It is 3x 5mm bullets in a blue plastic housing, theoretically rated for 120Amps but those may be Chinese Amps... I'm not thrilled with it, the wire must be fed through the housing, soldered into the bullets and then the bullets pushed into the housing until they click. So invariably you get too much solder around the bullet so that it doesn't fit anymore, or you push it too far and it comes out the back or... Yeh, not a great fan of that style.

For battery connections I've been using these EC5 connectors bought from the same shop: http://www.ebay.co.uk/itm/331267837473. As the bullet is already mounted to the housing there's no assembly, just add solder and heatshrink making sure not to melt the connector too much. I much prefer this but you can't get 3 pole versions of this style.

Oh and a 100W soldering iron is superb for soldering 12AWG or more to connectors/PCBs. Well worth the £20 an okay one costs.

[arvidb]

Ah, an EC5! I use them too, on my RC car. I think the worst problem with them (which they share with pretty much all RC type connectors) is that they have no strain relief for the cables; all strain is taken up by the solder joint. (And I know what you mean with "too much solder around the bullet" )

I'd like a version with a folding housing which clicks together around the connectors, and which also covers maybe 15 mm of cable, to get that strain relief.

I didn't know they come in a 3-pole version. Interesting!

[Thor]

Hi, just found your project of implementening what I call a VESC+ xD

First up the DRVs are very sensitive from end to end. Benjamin himself had issues with the DRV dying because of a inappropriate layout. I am not exactly suprised that it did not like your wiring^^. Just forget this chip. I realy do not like it.

I am designing a VESC+, too. You find my progress in this thread:
viewtopic.php?f=16&t=568

At the moment the status is mostly "finished". The controler is tested with 160V and 50A continous and was run in a car on a test track. There is a older video on youtube showing a endurance run at lower power:
https://www.youtube.com/watch?v=04ekDW2AEq0

My design is open source and can be found on github. Maybe it will help you figure your design out.
https://github.com/Thor368/HSWR_V2

Thor

[marcos]

I second the DRV opinion, and wanted to say watch out the shunt resistor setup. Its a very small signal with a big loop area and a long travel up to the control board. It will pick up noise.
For 10kw I'd go with isolated gate drivers to save money in blown stuff.

[TheFallen]

My design is open source and can be found on github. Maybe it will help you figure your design out.
https://github.com/Thor368/HSWR_V2

You say that but I can't see a PDF schematic in there at all.

I've not really touched this for ages but do really need to get back to it. Been playing around with an MGM 250A ESC which while lovely has awful support from MGM.

I want to try and get this power stage working with a VESC logic stage just to verify that it indeed works before I try and design another logic PCB.

[Thor]

Well, it is open source. That does not mean that I use github as my official release channel. If you want to use my design you will have to use my EDA toolchain, too.

Thor

[Marto]

Hello TheFallen, did you get any progress with your tests?

I am not expert at all, so at the risk of asking a completelly stupid question, have you considered scale factor and the offset for measuring the current in ACS sensors?

For example a reference code in Arduino would be:

RawValue = analogRead(analogIn);
Voltage = (RawValue / 1023.0) * 5000; // Gets you mV
Amps = ((Voltage - ACSoffset) / mVperAmp);

where mVperAmp=10 and ACSoffset=2500 for ACS758 bi-directional 200A sensor.

[TheFallen]

I did try with one of my odd 2 PCB VESCs back in June but immediately got the DRV fault so haven't been able to progress that. Once again I do need to get back on this.

I did consider the scale factor and offset but only the scale factor needs adjusting. The offset is calculated automatically by the VESC as I note in March.

It may be worth reading though the entire thread as you may find it of some interest.

[torg]

Very interested, subscribed for your updates