Paul & Sabrina's cheap DIY 144v motor controller

[amino]

I was just thinking, if the mosfet Q1 can be damaged by a spike from the contactor coil, wouldn't a small static discharge to the contactor control wire do the same, as it leads directly to Q1?

Amin

[MPaulHolmes]

I think so. I think that any static discharge that's noticeable represents a very high voltage, but a teensie weensie current. And Q1 is only rated for like 30 volts or something. I forgot. But the new fancy one you are using is rated for 100v. It's the Cadillac of Q1's.

[MPaulHolmes]

It's basically 99.999% done! I'm just debugging the control board a bit. One error was, I set the undervoltage trip point for the 15v supply to 15 volts. hahaha. Fortunately I had a resistor that was just right for setting the undervoltage trip point to 12.5 volts. The 15v output from the DC-DC converter drops to 12.5 volts when the 12v input drops to 6.7 volts. ya! So, if the battery powering the control board has a SERIOUS case of the sags, everything will still work fine, until it drops to 6.7v that is. Then the control board will gently shut everything down. Or if the 15v supply quits working it's also no biggie. And if the 5v supply quits it's also OK.

One minor catastrophe was I forgot to put the screws through the M- bar for mounting the current sensor. So then I slathered jb weld on the bar and squashed the sensor down onto it, and then I had a sudden flash of memory! That stuff is like 20% Iron Powder!!! That's not good! It's going to jack up my magnetic field. So I wiped it off before it dried. Man I love the smell of that stuff. I'm actually tempted to eat it. So then I got some 2 part clear epoxy that dries in 5 minutes and that sensor will never come off ever. Never ever. I centered it using the nylon screws with the heads cut off into the holes:


Heart attack averted!

I also added a few polypropylene caps and the TVS diodes. I jb welded the poly caps to the electrolytic caps just a bit so they won't vibrate loose:


Here's a picture of it basically done. I still need to solder the pwm, +5v, ground and 15v wires to the control board. I just wanted to make sure the control board worked as I hoped, since there were quite a few changes:


It's about 3.25" tall from the top of the baseplate to the top of the precharge resistor. I forgot to add a spot for a diode for the relay coil, so I'll have to add that on the underside of the board for now.

I tested the PWM after the mosfet driver, so the board can put out a 5v PWM signal of basically any current that is needed. I need about 60 mA from mine.

[MPaulHolmes]

And it works! There were a few times I put the wrong resistor value in, so I had to switch them out, but the mosfets are switching! It's officially tested at 0 volts. haha. But I could still see the mosfets switching from around 0 to 0.5v. PHEW! It all works! I can't believe it. Now to test it with my homemade inductors and that heating wire stuff to simulate a motor.

[jackbauer]

Heuston , we have a problem! Went out this morning about 8am. Temp was -2C outside but about +2C under the bonnet thanks to the 12v battery heater. Key on , contactor closes , no throttle response. Recycled the key and power comes up. Reversed out of the drive , got out , closed gates , into 2nd pull away a few yards then the lcd freezes at 133 motor amps and the contactor opens. No power. No amount of recycling the key would bring the controller back to life or engage the main contactor. Pushing 1.2 tons of German automobile 20 yards on a slippy road is no fun.

Hooked up the laptop , opened the traction disconnect (just in case) and key on. The green update indictor on rtd flashed a few times , the dialog box filled up with garbage and the precharging and high pedal lockout alarms stayed on. Controller would not respond to commands. Recycling the key produced the same results. The few flashes of the activity indicator and the garbage serial data got me thinking possible clock problem. Dragged out the extension lead and heatgun and warmed up the micro and crystal. Key on , normal operation restored. Went into town to get a few bits (3 mile round trip) without any problems. I have the log file saved on the laptop so i'll post it later. Looking very like a micro or crystal issue.

[adamj12b]

Originally Posted by jackbauer

Heuston , we have a problem! Went out this morning about 8am. Temp was -2C outside but about +2C under the bonnet thanks to the 12v battery heater. Key on , contactor closes , no throttle response. Recycled the key and power comes up. Reversed out of the drive , got out , closed gates , into 2nd pull away a few yards then the lcd freezes at 133 motor amps and the contactor opens. No power. No amount of recycling the key would bring the controller back to life or engage the main contactor. Pushing 1.2 tons of German automobile 20 yards on a slippy road is no fun.

Hooked up the laptop , opened the traction disconnect (just in case) and key on. The green update indictor on rtd flashed a few times , the dialog box filled up with garbage and the precharging and high pedal lockout alarms stayed on. Controller would not respond to commands. Recycling the key produced the same results. The few flashes of the activity indicator and the garbage serial data got me thinking possible clock problem. Dragged out the extension lead and heatgun and warmed up the micro and crystal. Key on , normal operation restored. Went into town to get a few bits (3 mile round trip) without any problems. I have the log file saved on the laptop so i'll post it later. Looking very like a micro or crystal issue.

Very interesting.

The atmaga168 is rated for -40C to 125C so my guess is its not that.

The crystal in the BOM for rev2C is PN: FOXLF160-20 which is good for -20C to 70C, but I know you used parts that were more available to you so this might not be the same one. My guess is this is the culprit.

On a side note, does anybody know what happened to the wiki?

-Adam

[jackbauer]

Forgot to mention that the green power led was active on the control board but not the orange led. With the mods i did to the board this would indicate the 5v line was on.

[evimarn]

Originally Posted by jackbauer

Heuston , we have a problem! Went out this morning about 8am. Temp was -2C outside but about +2C under the bonnet thanks to the 12v battery heater. Key on , contactor closes , no throttle response. Recycled the key and power comes up. Reversed out of the drive , got out , closed gates , into 2nd pull away a few yards then the lcd freezes at 133 motor amps and the contactor opens. No power. No amount of recycling the key would bring the controller back to life or engage the main contactor. Pushing 1.2 tons of German automobile 20 yards on a slippy road is no fun.

Hooked up the laptop , opened the traction disconnect (just in case) and key on. The green update indictor on rtd flashed a few times , the dialog box filled up with garbage and the precharging and high pedal lockout alarms stayed on. Controller would not respond to commands. Recycling the key produced the same results. The few flashes of the activity indicator and the garbage serial data got me thinking possible clock problem. Dragged out the extension lead and heatgun and warmed up the micro and crystal. Key on , normal operation restored. Went into town to get a few bits (3 mile round trip) without any problems. I have the log file saved on the laptop so i'll post it later. Looking very like a micro or crystal issue.

Back in the old times we used to get similar problems on electronic organs and synths which used to respond well to a slight increase in temperature and at times moisture control, these units used to sit in very cold and damp places with no heating or climatic control, so a small wattage bulb ( the filament type ) used to be left powered on in the cpu area enclosing all by a small plastic container.Being a small wattage lamp no excessive heat was liberated only enough to keep cold and moisture in control and voila it did the trick, untill another place or a revised version of cpu or crystal was obtained.

[dave koller]

Jack:

You checked to see if moisture droplets were around xtal, cps, leads, and so on?
Before using heatgun can you get to the caps on the xtal? If so, then grab one and warm it with your finger heat and see if the xtal kicks in. I have had bad xtal bypass cause the 168 to shut down in cold. If you could get a scope loop near the xtal to see if it is shut down or an rf probe. I does sound like the 168 is not getting clock by all you have said.

It would be nice to see the culprit! If the data stream stopped the darn thing has to be not getting clock - unless you have a bad 168!

I will be lurking....

[dave koller]

This is a PS...

I have cleaned and dried and sprayed a clear-coat over all the board to stop the moisture bridges from causing a problem in a sawmill (outdoor) -30 F LOL..