MPGuino rebooting and freezing also... Power issue?
 

I've browsed the many fine posts relating to the MPGuino locking up but still haven't found an answer yet.

Here's a brief history of the problem:

A mechanically-minded friend helped me install it in my 1990 Toyota Corolla 5 speed. (We also installed a tach and a vacuum gauge.)

We tapped the injector wire, ran a wire from the combination meter's VSS port, tapped an always-on 12v+ wire from either the stereo or the clock memory (can't remember which) and added a bolt to the firewall and ground both the MPGuino and the tach to it. All of these wires run to an RJ11 box so that I can easily plug and unplug the MPGuino if necessary.

We fired it up and configured the values (injectors for my 1990 Corolla seem to be in the ballpark of 322620341 usec/gal if you have the same injectors) and it worked great.

BUT... when doing EOC or turning the engine off/on, it would often reboot and lose my MPG info (especially when turning it off for some reason).

So, I tried running a wire directly from the battery (but kept the initial ground) and now its even worse! It reboots constantly while the car is on, totally loses all memory, and finally freezes and I have to leave it unplugged for a while. Also, the VSS doesn't even seem to register on the unit anymore. Curiously, as long as the car is OFF, the unit will run indefinitely. I tried running it with both the VSS and/or the INJ cables unplugged to see if they were the culprit, but it still misbehaves with just the power input.

I'm not sure what the problem is. Do I need to install some kind of car-based UPS to clean up the 12V power? I know a car's electrical system can very greatly in its output.

Thanks a lot and I hope my post makes sense. :o


PS: I'm not too good under the hood but I love FE had SG envy and so the MPGuino was a dream come true for me. :p A big THANKS to all the effort that went into producing it!

Do you have the pre-assembled version from dcb, or the kit version, that you assemble and solder yourself?

Are you sure that each wire is connected to the right spot? Like, are you sure the VSS and +12V lines aren't switched, or anything like that?

Thanks for getting back with me. :-)

Its the pre-assembled version from dcb, version 0.82

I thought I checked the wiring many times, but that's a good idea. However, if the unit +12V line was plugged into the VSS or the INJ line, would it have power with the car off, because with the car off, it seems to work perfectly.

If they're plugged into the wrong spots, the unit can behave very weird. I would definitely at least verify that they're in the right spot. I had a kit version, and the instructions for the RJ11 wires didn't work out for me, they were in the wrong spots. I ended up having to follow the leads on the board itself, mainly using this image to figure out which one went where. I think it had something to do with using a weird standard RJ11 cable, so the colors were backwards.

Since it's the pre-assembled version, we can probably rule out anything being shorted out, but it wouldn't hurt to take a look at the unit and make sure that none of the metal leads are touching anything else on another part of the board.

Thanks for the great suggestions! I'll definitely check that out tomorrow morning. I think I'll bring along a multimeter too and check the signals from the various wires to try and verify that they are what I think they are. Do you know what I should expect on the multimeter from the VSS and INJ wires if everything is working okay?

The injector signal consists of pulses, as seen in this thread, so it would be hard to read with a multimeter, since the pulses happen so quickly.

I assume the VSS signal is similar, and the pulses occur more frequently the faster the vehicle is going, but I'm not positive. So again, it would be difficult to read on a multimeter.

If you see the voltage varying on one of the wires, then it's probably the VSS or injector signal, since they're the only ones that should change. Both the +12V and GND signals should remain constant.

Thanks a bunch for the input. I'm going to review the whole setup tomorrow and do a sanity check.

Initial findings:

With the engine ON my 12V+ connection reads ~14.57V+ and the INJ signal is about the same. The VSS signal is reading in the millivolt range which I find kind of strange. Also, the MPGuino displays 0MPH. The only signal it registers is the INJ signal, and then the whole thing crashes.

I'm wondering if it could be a ground issue? Maybe it doesn't get along well with the tach? (The two have a common ground.) I might try running a new ground wire directly to the battery and see if that helps...

Also, I looked at the unit itself for possible signs of damage or bridged connections and couldn't find anything.

I made the floppy and RJ11 connectors myself using the supplied materials but just in case I'm going to pull a floppy connector off an old power supply and try that. *Sigh* :-)

Update:

Hooked the 12V+ and ground wires directly to the floppy connector and plugged in the MPGuino. Intentionally left the INJ and VSS wires off. It worked fine until I turned the car on. Then it reboot and froze.

Is it possible that this little guy just isn't liking the 14.57V my alternator is putting out?

I doubt that the 14.5V is the problem, there'd be a lot more reports of it not working if that were the case. The voltage regulator is designed to take anything over 5V, and output an even 5V for the entire board, and output the rest of the energy into heat.

A new ground couldn't hurt. Also, do you know if the VSS signal from the ECU is the same one that went to the speedometer? If it splits anywhere, then I would try tapping it right from the ECU.

On my old Corolla, the VSS signal is actually generated right there in the combination meter from the speedometer cable. I think I'm as close as I can get. Plus, the speed portion of the MPGuino was working perfectly when it was working.

I ran a new ground straight to the negative battery terminal and still had the same exact problems.

I definitely agree with you about the 14.5V not being the problem. I'm thinking that maybe the voltage regulator has blown/gone bad and I need to replace it. Bummer.

I tested the voltage regulator with a voltmeter. With the car off, it was reading +11.9V on the input side and +4.95V on the output side. With the car on (and the unit frozen I might add), input V was +13.5V and output a steady +4.95V.

I know output is supposed to be +5.1V. Is 4.95 a teensy weensy bit too low?

Also, I just performed the tests outlined here:

Mpguino tests - EcoModder

and here are the results:

A-B Inf
A-C Inf (Interesting. Not grounded?)
A-D 49.K (supposed to be INF. short?)

B-A 13.43M (yes, M)
B-C Inf
B-D 14M (That can't be good)

C-A Inf
C-B Inf
C-D Inf

D-A 48.9K (why so low?)
D-B Inf (Weird. B-D is 14M. Hmmm...)
D-C Inf

1-2 5M
1-3 5M
2-1 17.16M
2-3 10.54K
3-1 Keeps going up. Now around 18M.
3-2 10.54K

So I feel like these tests indicate that there's some kind of problem on the board. A visual inspection made my me doesn't reveal anything obvious. Perhaps one of the components is blown.

(sorry, been on vacation for a few days)

You can send this unit back for a reflash (via S.A.S.E) if you like, no guarantees where a donation and gift are concerned (would have to charge a lot more for that) but we will reflash it and test it and take a look. I don't know exactly why it konks out on ignition, even with the battery tap? Maybe a decent sized rcl circuit for the power taps might help too, sound like some sort of unusuall spike if a reflash doesn't fix it.

(No worries. I'm glad you took a vacation. :-))

Sent it off today. Thanks! :)

One thing I was going to try but didn't get a chance was plug it in while the car was already on. If it didn't freeze then there was probably some sort of spike or brownout issue. But I forgot to try that. :-o

I don't have a DC generator or I would have tried to test how it reacted to various voltage states.

So when it gets back hopefully it will work like a champ. Otherwise I'll look into creating an rlc circuit (though I'd probably need a few hints on which cap, resistor, inductor to purchase.)

Thanks for all your help!

Sounds like you may have a connection problem at the battery I would clean the post real good. It does not take much to drop the power to the other connections when the starter is engaged about 1000 amps or more will rush to the starter and then come down to the cranking amps any resistance at the battery will cause a voltage drop. I know that the battery should be checked also at cranking for the voltage drop it should not drop below 10.5 volts, this may give you a clue to the problem hope it helps.
Hugh

That's a really good idea Hugh. Thanks! I'll test the voltage while cranking and see what it looks like. Plus I'll clean the posts. Then when my MPGuino gets back we'll see what happens. :)

I checked the connection to the RJ11 box with a voltmeter and with the car off voltage was 12.65, nice and strong. While cranking, dropped down to about 10.6V one time and 11V the second time. Then the voltage shoots up to around 14.5V with the car on. So it seems like I've got a good strong connection. But I'm still going to clean the terminals again anyway.

Voltage of course fluctuates slightly with the car on and even with the car off when I turn on/off different devices that draw current.

Mine has been having similar issues. It'll lock up and restart itself after several minutes or it'll be working fine and randomly restart itself. 1993 Geo Metro, tapped to the injector wire by the injector and into the VSS signal at the ECU. Oddly enough, sometimes the vehicle speed will go crazy and will jump around going from 50 to 89mph when I'm going 55. I'm not sure if its radio frequency interference from houses or what, but that is probably a separate issue. However, I just put a fairly good size aluminum heatsink on the back of the power regulator today so that should keep it much cooler, I'm hoping that solves the restarting problem. I'll let you know what happens.

I'm interested to know if the heatsink helps you out any. Are you using an enclosure with your mpguino? When I get mine back I'd like to find out how to put it in an enclosure with a temperature-controlled fan in the back, perhaps in conjunction with a heatsink.

I'm not sure about the VSS signal oddity either. You could take a look at this thread:

http://ecomodder.com/forum/showthrea...ates-4236.html

Hello,

I've got the same issues. I'm using a spiffie kit in a 93 Toyota Previa. V.75
I already took mine apart and resoldered everything. It gets worse and worse. Crashes sooner and sooner in the same way described by Tygertec.

I even have mine hooked up to a 3.5 AH battery separate of the car.

One thing I noticed is that my injectors have 12V on each wire with the ignition on. car not running.

Recently the gph started to run like it was attache to the vss. Adding to itself.

Could this be a connection to ecu wires issue? I didn't solder the VSS. although I did just solder the inj.

Update
 

I'm back at it again

Cleaned the battery posts real good and cut more ventilation in my enclosure. So far it seems to be a lot more stable. It rebooted one time for sure, possibly more. I may cut more ventilation holes in the enclosure. If that doesn't do the trick I'm going to install a power filter to smooth out the dips/spikes.

The only problem now is that I think I shorted out the VSS in my gauge cluster while testing stuff with the multimeter and now I have no speed signal. Doh!:rolleyes: It was kind of irritating using the sensor off the combination meter anyway because it didn't put out a signal when the key was off. I'd like to get an aftermarket sensor that runs even with the car off. But I may create a new thread if I can't find any similar posts.

Thanks everyone! You guys are great! (Especially dcb!)

I am running into the same lockup issue described above. I put another diode from the cathode of the one that supplies +12V to a 9 v battery. This way, the regulator sees the higher of the two, and the 9 volt alkaline battery only supplies current occasionally and for short periods.

First try, I ran it in BIG INST mode driving to a friend's place. Upon restarting, the unit was stuck in BIG INST OOO, even when driving, and did not respond to the buttons. Fully interrupting power by disconnecting both supplies for 30 sec or so reinstated function. I believe the power is bone reliable, at least as far as dipping too low for the regulator.

I have v0.82, so maybe there is a bug in the code that shuts down after a timeout. Locking up has happened when just driving, or perhaps getting the display back to inst mpg when it has reset itself.

The car is my 49state 4door 1991 Metro. All connections are made about an inch from the connector on each of the 4 wires at the ECM. These are led out and spliced to a connector from an old power supply. After I messed up the first crimp, I just went to the junk box for a power supply. the 9 volt battery sits behind MPGuinio in front of the fuel indicator.

Do we perhaps need shielding?

Hmm, I am wondering about the commonality here. You both have old-ish cars with possibly old reed switched vss's. Can you try disconnecting the vss lead (lift the retaining tang at the plug or dyke the lead) and see if it runs more reliably? If you put it on the current screen it will show gallons to the 1/100th and the custom screen has an instant gph reading on it.

I was reviewing this thread and noticed that one of the posters brought this out:

"One thing I noticed is that my injectors have 12V on each wire with the ignition on. car not running."

He has a 93 Toyota Previa if I remember correctly. I believe my mechanic friend mentioned that there are 12V on each injector lead in my 90 Toyota Corolla. Perhaps that could be our commonality???

My unit hasn't been freezing much at all (Still haven't installed the rcl filter. Super busy!) Since my VSS signal isn't working I've been relying on the gph field, although I do still have the VSS lead connected to the MPGuino. Last night the unit froze with a bunch of random numbers in all of the fields and the weather was pretty cold, so I doubt that was a heat problem in that case.

I might look into scheduled eprom saves for the tank trip and adding a watchdog timer. That way it should restart if it hangs and pick up from the last saved trip. Have to find a balance between eprom update frequency and eprom life (like 1,000,000 writes), maybe update every 1/10 of a gallon or so.

dcb, i posted up some save-tank code to the wiki Code Hacks page. If you only want the Save-tank, and not the Archive feature, it should be easy enough to weed it out.

(Also, i used the last "edit" on the page, and somehow it appended my hack onto the end of "Singletree's More Precision". Can you tell me how to fix it, so that ours are separate, and there is a new blank "edit" below mine?)

n/m. i got it. i needed the "== Hack ==" separator.

OK, will look into it.

Maybe update the eprom every mile, if this thing is still in your car after a million miles then I'll be impressed :) And if you lose part of a mile 3 times in a tank, of 300 miles, you are still pretty accurate on the tank trip.

Driving around town (potholes galore) I note that the reset, usually the current ? screen (the one that is three left button pushes from BIG INST, my preferred usual setting), about 1 time in 5 I noticed the ver .82 screen, and the changes always came at a jarring pothole. I wonder if bouncing around on the dash is actuating the pushbutton switches. The speed and gph readings in the multi variable screens seem stable. With the hard alkaline backup of power, and behavior consistent with stray switch closures, this seems the most likely explanation.

Are most of us with this problem driving on rough roads with shot shocks?

i too have v.82 and mine acts weird like this. Freezing and having to unplug it to get it to function again. Mine is on a 97 honda civic. I tested all the connections, which are soldered and heat shrink tubed then taped. Wondering if its v.82 or something else.