Paul & Sabrina's cheap DIY 144v motor controller

[jackbauer]

I run a 400 A NH fuse. I got 3 of them complete with holders very cheaply from an electrical shop as they were old stock. I was a bit worried that they didn't have an obvious dc rating. However during an energy show earlier in the year I had about 5 guys from schneider electric looking at the car and they got me a datasheet on the fuse and its good to break 2ka dc.

This is a very strange problem with that truck. I think its safe to say the problem is using a bipolar transistor in place of a field effect. The battery problem is less clear. Bad smell is hydrogen sulphide if i'm remembering my chemistry. not good at all.

I'd look at getting some igbts with decent switching speed.

[snakub]

Wow a lot of questions where to start? I was not driving these modules directly with the revolt board the revolt board turned on a mosfet then the mosfet turned on the the base of the modules. The resistor to the base was 2.2 ohm with a pull down resistor of 39 ohms went from the base to the emitter to help with switching times. No stupid me there was no fuse in line with the batteries although there should have been. The power stage has four of these modules. There are separate resistors for each module. One of the modules shorted but the other three are still good.

The fact that the hardware overcurrent didn't stop anything was probably because the overcurrent circuit pulls the gate

Shouldn't the overcurrent still work because I was using a mosfet to drive the modules? I will write down all the component you said to check and get back to you this evening.

[MPaulHolmes]

If the darlington was failed shorted, turning the mosfet off wouldn't have an effect on anything.

By the way, a friend of mine sent me a link to Tesseract's comments about driving darlingtons. Here's what he said:
“What's the switching frequency? IIRC, you are using fairly old Darlington modules. Depending on the vintage, those were typically rated for 1.5kHz operation. Some VFDs of that time offered a "quiet mode" - 8khz, which is the default frequency today - but required a 25% or so derating.

Realize that Darlingtons require a proportional base drive circuit to achieve fast switching while minimizing Vce[sat]. Count on an overall beta of 100 (so a 500A controller will require 5A of base drive... and this is continuous, not impulse, current...).

The driver circuits used these days for IGBTs and MOSFETs will hard saturate a darlington which will lead to incredibly long turn-off times. This can be mitigated somewhat with a Baker Clamp, but then your driver power is going to shoot through the roof. Proportional drive really is best for Darlingtons (and, to be perfectly honest, is such a ***** to get working just right you might question whether you really got such a good deal on the Darlingtons in the first place ).

Here’s the link
Blackpanther_st's home brew DC motor controller. - Page 4 - DIY Electric Car Forums

[apowers]

I use a 400 amp fuse and a 200 amp breaker both from KTA.

[Weisheimer]

Is your Darlington power section modeled after BlackPanther's?

If so, what are you using for your 5 volt supply?

It would need to source 20 constant amps and also be isolated from the
other voltage sources and ground.

BlackPanther used a PSU from a desktop computer, driven via a 12v inverter.
Just be aware that they are not always isolated though.

In relative terms, we're talking about a 100 watt DC-DC converter here.

When you drive FETs and IGBTs, you hit them with a pulse of current, so
a similarly rated single IGBT module can be driven using the 6 watt
DC-DC converter that is included on the Revolt control board.

You've learned your lesson on fusing. Check with CTR-surplus.com
They will probably have a semiconductor rated 300 amp fuse for $10 or less.
As JackBauer noted, most of them are capable of breaking tens of thousands
of amps of DC current due to shorting. They are just what we want!

As a practical note, we should fuse every single wire that runs from power.
Even if we are only powering a small indicator or such.
It is cheap fire insurance!

[snakub]

Yes I have almost exactly what blackpantherst does. He used his control section at 12khertz it seems I couldn't really hear any noise on his test drive maybe I am just not getting it but 3us is 3 millionths of a second that seems pretty fast to me when you are only switching at 15 thousand times a second. I am using an 8 amp hour twelve volt battery and a battery charger powered off of the traction pack to power B not Bx which would take a lot more drive current I measured the current going to B and it is only about 2 amps so all in all my power section took 8 amps. But can I put the fuse in as one of the series connection to my batteries or would it be better to have it connect to the positive going out?

If the darlington was failed shorted, turning the mosfet off wouldn't have an effect on anything.

I wasn't really talking about if the mosfet shorted I was just trying to say that if I have the current sensor hooked up to motor negative wouldn't that still limit my current by resetting the PWM at appropriate times by turning off the mosfet? What did tesseract mean by hard saturate?

[MPaulHolmes]

I was saying that for some reason the darlington seemed to fail shorted. At that point, the mosfet could do whatever it wanted, and the gate of the mosfet could be kept low by the overcurrent circuit, but it wouldn't have any affect on the darlington being a dead short. The question is why would the darlington fail shorted before the current got too high, because as long as the darlington was working, it would have been disabled by the overcurrent circuit if the current got too high.

[snakub]

I still think the issue is in the control board here are some oddities that I found while measuring components while they were on the board. I couldn't test D3 but it was in the right way with the black line on the diode meeting up with the line on the PCB I tested it with the multimeter by putting it on ohms and taking the positive lead from the meter and putting it to the positive of the diode and negative to negative but I never could get anything on the meter other than OL maybe it won't measure in the circuit. R12 was a 10K. R9 was a 1.98K. R15 was a 220 ohm even though it looks like orange orange brown to me. VR1 measured 450 ohms even though I can remember measuring it before putting it on the board as 500 ohms maybe its just by being on the board. R14 measured 700 ohms. U8 is not backwards. D8 and D9 are not backwards but I couldn't test them properly either when I measured them with the multimeter positive to positive negative to negative I got 7.3k both ways.R18 measured 2.2k but I looked at it and the color code was definitely Orange black red 100 percent sure. Maybe I have to take these parts off to get a proper measurement? Does anyone have a board lying around maybe you could do some testing like I did and tell me what measurements you got?

[Dwffy]

Once again I come seeking advice from the great DC guru's of Revolt design.
OK fellows I have finally got my EV licensed and registered to travel by way of the
macadimised pathways of Canada. Today was cold so I turned on my heated, which is a 110volt 1kw ceramic heater built into the core of the old radiator for the heater. It only gave off moderate heat but was sufficient to defrost the windshield. However, when I turned it off, it kept on working until I finally stopped and shut the 120volt dc off. I open the 12v bosch relay which I used to switch the heater on and off because it says it can handle 30 amps. The contacts were welded together so I now conclude that there must be a way that I can turn my heater on and off without burning up relays, but I am at my wits end. Can anyone out there offer any advice please. Ps I also have it wired so that the relay can not engage unless the fan is operating.

[jackbauer]

It can handle 30 amps. At 12v! You'd need to use a kilovac or other ev rated hvdc contactor.