APD controllers?

AC5ME

10 mW
Joined
Oct 18, 2020
Messages
30
Hi all, any experience or opinions on these?


Thanks!
Jim
 
There are a few posts about them; this is from the archived old version since the new forum search can't find them (says "no results found").
 
Thanks! I don’t suppose you have any suggestions on a 28S 300A class controller?

Jim
 
Unfortunately I don't have enough direct experience to give you recommendations, especially since you probably need a high-RPM-capable type if this is for aircraft projects. All of the higher voltage / higher current controllers I've seen posted about either have been described as having poor support, poor quality, poorly written firmware, poorly written and/or translated setup software (most of the Chinese offerings), or like the Sevcon require a handful of rocket scientists ;) and some expensive hardware to configure because they aren't meant for DIYers, but only for OEM use where they configure one on a prototype EV then mass produce the EV using the settings determined in prototyping.

(if you were local to me and wanted to spend the money on finding their setup software and hardware, you could test out a Sevcon of some variant I think came off a Zero motorcycle that I can't use because I don't have the software and hardware required to set it up for any motor I actually have; presumably it is setup for one of the 75-x motors they used (but I have no idea which one, and there's no way I know of to find out other than trying motors on it till it works correctly...and I don't know what options, voltages, etc are setup in it either).

I think there are VESC variants, including some designed here on the forum, that could do the job. Not sure if any are commercially available, but since they're open-source, they could be DIYed if you're up for that.

If any of the automotive EV inverters have the voltage/current requirements you are after (most probably do) you could zombify it by cutting it's brain out and replacing it with either a VESC control brain or the Lebowski brain. There are some websites/pages around the web about doing the Lebowski transplant process including one I started (but haven't finished yet) here on ES. (the one I started should have enough info to complete the process, and it links to my sources of info where the process was already successfully completed by others before I started, but not detailed very well, which is why I documented mine).
 
I appreciate that, thanks. It's a bit beyond me to do brain surgery though! I've recently had my MGM-Compro controller succumb to whatever demon possesses these things, so the search continues...
 
What happened to it, under what circumstances? (it might be repairable, if you're up for that...if not, you could post it for sale as broken to at least recoup some costs if there are users out there that do fix them).
 
Everything was going great and it was happily putting out 33kW, until it didn't. Whatever feature that is supposed to prevent over-current didn't apparently.

I'm not sure how much value is left, but I'm happy to let it go to someone that can use it.

tempImageixB82S.jpg
 
YOWCH. I think that is probably going to require PCB repair as well as probably replacing all the FETs and quite likely the driver circuits, at least on the phase(s) that failed.


What usually causes catastrophic failures like this when there is no overheating is voltage spikes that exceed the Vds of the FETs. Sometimes this is a one-time instant failure from a big overshoot, sometimes it's cumulative damage that eventually results in POOF.

If the controller got really hot during normal then it's probably overcurrent, with FETs that couldnt' share current properly so one or more got more than it could handle for long enough to heat so much it couldn't take it and POOF.


In either case, it's probably the drive routine that shapes the current waveform that caused the problem (or settings in the setup program not being optimal for the usage / motor, thus causing the drive routine to send the wrong current waveform), if the controller was being used well within it's specs and wasn't overheating. If the hardware has limiting (temperature sensing, current sensing, etc) that rolls back power or shuts down in such situation, it should do so, but if it isnt' hardware doing the rollback or shutdown, it could take too much time for the software to respond and do this, or potentially even not respond if it's not correctly written to interrupt itself for such an event.
 
Last edited:
I've just discovered this:

Any thoughts?
 
Here is the thread for Hesc or 3shul:
 
Back
Top