Overheating power regulator?

[Istas]

I'm not sure if it's supposed to get this way, but the power regulator on the back of the MPGuino is getting very hot, too hot to touch. And it shut off without provocation, then booted up a few moments later like it had just been plugged in.

Is it supposed to be getting that hot? I'm going to try putting a couple of mosfet heatsinks on there that I have sitting around. I'm wondering if I'll need to put a resistor in the power line leading into the MPGuino.

[dcb]

Hmm, that could be annoying. You are using a standard prebuilt without any mods? Is it in a case?

Definately have a look at the heat thread http://ecomodder.com/forum/showthrea...-etc-8684.html


The only time I've seen one go into thermal shutdown is when I tried a fancy vfd display.

I would leave the display backlight off during the day for starters.

An external 8 volt regulator feeding the guino power is probably the most effective mod if that doesn't help.

But if you want to tap your dimmer wire, I can put some pics together to show how to hack the guino to use that (minor bit of soldering required).

[Istas]

Thanks for the tips, and the link to the relevant thread. I've applied two heatsinks to the voltage regulator and turned off the backlighting. I've also redone some of the wiring connections that were not that well done, so that will eliminate that uncertainty. I'll look into using an external volt regulator too.

[Istas]

I am using a standard pre-built without any mods, in a case.

I checked out that thread, thanks for bringing it to my attention.

I didn't know about the backlight before, as I was still wiring it up. I'll definitely leave it off during the day from now on. I've also added some heatsinks to the regulator; that should help even if there are no overheating problems. I also secured some shoddy wiring that I'd attempted. (I didn't have a lot of luck with the default included connector with the little pins; instead, I scrounged a floppy drive power connector from an unused PSU)

I've also been having a weird problem where it reboots when I beep the horn. At first this was at least intuitive, if not expected, as I had the power coming from the horn fuse in the fuse box, but since then I've run a dedicated wire right from the positive lead on the battery to the MPGuino (through its own 3-amp fuse, I know less than that could probably damage the MPGuino but it was the least amp blade fuse they had) and through an on/off toggle switch to easily shut off power to it. And it still has rebooted when beeping the horn. Someone suggested a capacitor to smooth out the power supply some.

I'll look into the regulator and capacitor(s) when I get home. (I'm on vacation right now, and I have a 2500 mile drive home starting tomorrow at 7 AM)

EDIT: Sorry for the somewhat repeat of information there, the page didn't show my post even after rebooting the computer and I thought it had not posted correctly.

[dcb]

FYI, you might be able to salvage components from an old electronics device that you are planning on getting rid of (broken radio, answering machine, phone, ???).

It might be an interesting experiment to put a (not too) small speaker in line with the power lead. As an inductor in series it should help smooth things out when the horn goes off and provide 4-8 ohms of resistance on the power dissipation side of things.

Plastic case? Any metal involved?

[Istas]

I meant to say not in a case, sorry. Right now it's just sitting on a ledge in front of the instrument cluster, above my steering wheel. It's sitting on a completely plastic surface, leaning against the plastic cover to the instrument cluster. The wires are secured with duct tape currently.

That's a good idea on the salvaging, and also on the speaker. I'll see if there's anything I have that I can use for those purposes.

[tim3058]

I'm having the exact same issue with my MPGuino, and it's gotten worse over the last 2 months. It's pre-built, as-is from dcb. At first it would just randomly reset to 0.0 maybe once or every other tank, then last two tanks its been more often. This tank its done it constantly below half tank, and I started noticing the small black box (regulator?) on the back left-hand side (viewed from rear) was getting real hot. The other day the display finally went screwy and I unplugged it and haven't used it since. Note also the weather here just finally warmed up this last two weeks, which may be related. I followed the wiki wiring directions (an inj wire, the VSS wire, a ground under dash, and I have a +12V off the batt terminal with a 3A fuse in it like Istas has. The MPGuino is mounted to a face-plate on the column, the rear is completely exposed so trapping heat shouldn't be an issue

Have either of you fixed the overheating problem by adding a regulator to the +12V feed?

[Istas]

I haven't tried that yet, only added heatsinks and stopped using the backlight in the day, and only using the lowest brightness setting at night.

I might try that this summer, to help document more aero modding.

[Istas]

Oh, and, I never did completely solve the problem, it has still randomly reset itself, if rarely, without provocation from horn-beeping (which I also didn't solve yet). Still, it has been much improved since I added the heatsinks and stopped using the higher backlight brightnesses.

[dcb]

I posted some thoughts on it here
http://ecomodder.com/forum/showthrea...-etc-8684.html

The horn is likely not heat related, a capacitor across the horn terminals (the buzzy part) might help there.

I did make a couple revisions to the pcb recently (v1.3) which might help reliability namely:
the LED is powered via an 80 ohm resistor to 12v and an npn transistor. The resistor (orangish) does get hot so it is taking a lot of the load off the regulator.

This also allowed me to set it up so that the capacitors do not have to power the backlight in a brownout, so they should keep the cpu running for longer in a flakey power situation.

It also makes it easier to reloacate the resistor if you are using an enclosure, and move a large source of heat out of the enclosure, and/or to tie it into an existing dashboard dimmer circuit.

The other changes were mostly just making pads available for the unused cpu pins and establishing internal connections for avcc and agnd.