EV help

Thread Starter

scratch

Joined Aug 6, 2010
38
Please can anybody help me. I have converted a small car to electric [EV]. I am using a very simple homemade pwm speed controller, My problem is that i have no current sensing cut back so the FET keeps burning, when ever over current occures. Please please some body help. thank you.
 

marshallf3

Joined Jul 26, 2010
2,358
Going to need some more specifics, such as the motor drive circuitry, what FETs you're using and some info on the motor(s) involved.
 

Thread Starter

scratch

Joined Aug 6, 2010
38
Please can anybody help me. I have converted a small car to electric [EV]. I am using a very simple homemade pwm speed controller, My problem is that i have no current sensing cut back so the FET keeps burning, when ever over current occures.
motor is 48v series wournd motor of about 8hp from an old forklift.
FET's are RFP40n10 (24 of them in parallel).
Please please some body help. thank you. Marshallf3.

try send circuit but onable to.
 

Wendy

Joined Mar 24, 2008
23,415
Most circuits are design were only one transistor can be on at a time, if both are on then you have shoot through, where the power supply discharges through both transistors as a short, blowing both transistors (sometimes violently).

Pulse Width Modulation
 

Ron H

Joined Apr 14, 2005
7,063
Shoot-through only occurs on push-pull drivers. Perhaps he is not using an H-bridge or any other sort of push-pull driver.
Or maybe he is.:D
 

marshallf3

Joined Jul 26, 2010
2,358
Shoot-through only occurs on push-pull drivers. Perhaps he is not using an H-bridge or any other sort of push-pull driver.
Or maybe he is.:D
It's all academic here though, all we have to go on is that he has a battery, a motor and a "black box" between the two that contains MOSFETs. If you can draw us out some sort of schematic, even freehand, it's going to draw you plenty of help.
 

Ron H

Joined Apr 14, 2005
7,063
It's all academic here though, all we have to go on is that he has a battery, a motor and a "black box" between the two that contains MOSFETs. If you can draw us out some sort of schematic, even freehand, it's going to draw you plenty of help.
I know. I just thought that shoot-through could be a bit of a red herring if it isn't relevant. He could spend hours reading about something that doesn't apply to his situation. Not that that's necessarily a bad thing.:D
I agree that a schematic is required.
 

Potato Pudding

Joined Jun 11, 2010
688
Look at simple current capacity for the motor at stall speed.

An 8HP motor means this would have to be a go kart or a real car geared down to a top speed of 20 mph.

Does the car start to move and work with the motor running at the correct speed or does the Fet just blow up?

Also with PWM does he have plenty of protection against back EMF spikes? He also needs that protection just to shut the power off. Without fast rugged diodes his FET will blow up.

I agree that we can so far assume that he only has a motor connected on top of a big N Channel and if he has an H-bridge it is probably a mechanical switch from an old electric forklift like the motor. Lower Tech can be a comfort zone. He mentions one FET and if that is all he has then I don't think it will be shoot through.
 
Last edited:

Thread Starter

scratch

Joined Aug 6, 2010
38
Please can anybody help me. I have converted a small car to electric [EV]. I am using a very simple homemade pwm speed controller, My problem is that i have no current sensing cut back so the FET keeps burning, when ever over current occures.
motor is 48v series wournd motor of about 8hp from an old forklift.
FET's are RFP40n10 (24 of them in parallel).
 

Attachments

jpanhalt

Joined Jan 18, 2008
11,087
I agree with Ron H. Why use an H-bridge on a series wound DC motor? There has to be more to the circuit than the OP has presented to date. My suspicion is a simple mosfet with no protection. The load may even be connected to the source.

John
 

marshallf3

Joined Jul 26, 2010
2,358
Try adding another large current diode, cathode from the negative motor lead, anode to the ground. Many power FETs have these internally but some don't. This also has the benefit of giving you some regenerative braking, when you "let off the gas" the motor will turn into a generator and put some charge back into the battery.

You might also want to test the diode you already have in there, if it wasn't of a high enough rating to begin with it could have given up the ghost.
 

Potato Pudding

Joined Jun 11, 2010
688
I still think stall current is a big problem combined with not enough diode protection.

24 Mosfets in Parallel with 300nF per gate and I seriously doubt those are being driven up to fully on fast enough especially since as a logic drive they can have a max of 4 Volts on their Gate. Overvoltage on them is a possible problem but getting enough current is too.

Connect an npn transistors collector to the 48 Volts and Emitter to the Fets gate with a 220 Ohm resistor between the gate emitter and ground. On that transistor base you connect a 4.7 volt zener to ground and the signal from your PWM.

Being so many in parallel is also going to cause load imbalance problems.
Put a tiny amount of equal resistance or a two loop choke under each transistors source. That should balance the loads.
 

Ron H

Joined Apr 14, 2005
7,063
Connect an npn transistors collector to the 48 Volts and Emitter to the Fets gate with a 220 Ohm resistor between the gate emitter and ground. On that transistor base you connect a 4.7 volt zener to ground and the signal from your PWM.
Speaking of needing a schematic...:rolleyes:
 

Thread Starter

scratch

Joined Aug 6, 2010
38
Car runs well, then (wen i think all is well) out of the clear blue sky FET's smoks.
car is a dihatsu charade manual transmision.
 

marshallf3

Joined Jul 26, 2010
2,358
In addition to what I mentioned Potato has a good point, variations in Rdss_on is going to cause the drive current to favor certain FETs over the others. Being PWM driven only a few turns of wire between the source and GND of each FET will certainly help balance out the loads.

It was a shock to me but I found out that Radio Shack still carries magnet wire up to #20 and in short lengths it should handle the current well. Anything larger, such as some #16 I needed for a precision shunt, is available from a certain seller on eBay I'll post if anyone's interested.
 

jpanhalt

Joined Jan 18, 2008
11,087
When you post the schematic, be sure to include detail on the gate driver. I had a driver (LT1158) for as 2+ HP motor at 12V (lots of amps). It would work fine, then blow a mosfet. The problem was easily solved.

John
 
Top