Motor Speed Controller - MOSFET Killer

Status
Not open for further replies.

Thread Starter

oddball

Joined May 29, 2011
2
I'm restoring a '72 Cutlass and redoing the electrical system. To limit wires going through the firewall, I was hoping to switch to a speed controller instead of the normal resistor block and relay for the A/C blower motor (permanent magnet, brushed). I've managed to fry about 10 expensive FETs at this point, figure it's time to ask for help. Overall it's a silly project (two wires through firewall instead of four), but I think it's really interesting, keeps me out of trouble, and sets the stage for a likely fuel pump speed controller in the near future.

Note that I'm not very good at electronics, so please use small words. I can solder OK, but needs to be through-hole. I'm currently working on breadboards (other than the high current side, which is soldered), although I think the timing circuits are in good enough shape to build a board. That, of course, means perf board for me. I have three scopes at my disposal, but the youngest is about 25 years old. Each has its own interesting 'quirks', so I switch between them often. That's all to say that I can scope the board, but I'm never certain if noisy/unstable signals are the board or the scope or my cheap probes.
I'm currently testing from a car battery, but will need to live with spikes of 16V or more. I expect to add zeners or a voltage regulator to help with that as the design progresses.

This will be mounted in the engine bay, most likely in a sealed box, so heat, vibration and the variances of an auto electrical system are important, and want to minimize radio interference. Also don't want the motor to sing. Switching frequencies from 2khz to 20khz cause it to make a lot of noise. The speed is selected by a four position switch which I intended to use with resistor sets to drive a PWM signal. Would like for the controller to support 100% duty cycle so I don't have to separately wire a relay.
The blower fan motor pulls 12A at full speed.

Current design uses a 555 and comparator for the PWM then feeds a HIP 4082. I'm using IRF1324 for the power MOSFETs (Rds of 1.2mΩ, Id 195A). The bootstrap diode is a STPS2L60RL, a 60A schottky, fast recovery.




(full size at my flickr feed and attached below)

Current issue is the low side fet heats up quite a lot on anything other than 100% and quickly dies closed. I'm pretty sure the high side isn't switching as it's only getting 12V at the gate. A prior revision - using a TL494 to generate the PWM, but otherwise the same - had the high side hitting 24V at the gate, but I can't seem to recreate that. The AHB pin hits 20V, so I don't know why the gate isn't seeing that. There is a 'swooop' on AHO, where it bounces below ground when turning off and hits just above 12V when turning on. I had also tried a run-of-the-mill diode, but same effect.
Obviously if the high fet isn't switching, then the internal diode on the low fet is taking the brunt and causing the heat.
I also ran this design but using a 6A diode instead of the high fet, but same results.

I'm also trying to recreate my prior layout which, to the best of my recollection, ran great at these slow switching speeds, made noise in the middle, then killed fets when I hit >30khz. Motor was really quiet at that speed, though. I thought the motor was making grinding noises at this speed, but it was actually a resonance through my workbench. The TL494 also doesn't do 100% duty, so that was a problem. I've just used it successfully in the past for a PWM adaptation project.

So, at this point I'm at a loss. Please let me know where my errors are and what I can do to correct! Many, many thanks!
 

Attachments

beenthere

Joined Apr 20, 2004
15,819
Our Terms of Service do not let us discuss your topic -
6. Restricted topics. The following topics are regularly raised however are considered "off-topic" at all times and will results in Your thread being closed without question:

  • Any kind of over-unity devices and systems
  • Automotive modifications
  • Devices designed to electrocute or shock another person
  • LEDs to mains
  • Phone jammers
  • Rail guns and high-energy projectile devices
  • Transformer-less power supplies
 
Status
Not open for further replies.
Top