Megasquirt dying/losing connection after running for ~20 seconds
#1
Megasquirt dying/losing connection after running for ~20 seconds
Hi guys, I know this isn't a Miata question but it is a Megasquirt question. I know you guys deal with that a lot.
this current project is EFI-ing a mate's TA22 celica.
The megasquirt is just controlling the fuel side of things. We still have a distributor.
When I start it, it runs like a champion for about 20 seconds, then it starts cutting in and out, the fuel relay starts ticking (meaning the fuel pump starts alternating between on and off) and the computer loses connection with the Megasquirt.
At first we thought it was the hall sensor not reading properly, we did every mod under the sun. Didn't solve the problem, switched to a VR. Exactly the same problem.
We replaced D19 and visually checked C22 (couldn't get a replacement) as per the troubleshooting guide on randomly-dying-megasquirt's. No difference.
HELP! PLEASE!! Supposed to be in QLD right now...in this car.
Basic set up:
- wired to take control of spark through single coil at a later date
- was running a hall sensor, converted to VR,
Update:
tried running it with the fuel pump hardwired. Exactly the same. Noticed something though: the middle LED (which I'm led to believe flashes each time the tach input gets a signal) went off being continuously illuminated (as expected with 36 triggers for every revolution) but when it starts ******* out it starts flashing around twice a second.
So internal timing problem? We changed the entire internal circuit from Hall to VR and (obviously) changed the external sensors and wires as well, only to have the same problem??
this current project is EFI-ing a mate's TA22 celica.
The megasquirt is just controlling the fuel side of things. We still have a distributor.
When I start it, it runs like a champion for about 20 seconds, then it starts cutting in and out, the fuel relay starts ticking (meaning the fuel pump starts alternating between on and off) and the computer loses connection with the Megasquirt.
At first we thought it was the hall sensor not reading properly, we did every mod under the sun. Didn't solve the problem, switched to a VR. Exactly the same problem.
We replaced D19 and visually checked C22 (couldn't get a replacement) as per the troubleshooting guide on randomly-dying-megasquirt's. No difference.
HELP! PLEASE!! Supposed to be in QLD right now...in this car.
Basic set up:
- wired to take control of spark through single coil at a later date
- was running a hall sensor, converted to VR,
Update:
tried running it with the fuel pump hardwired. Exactly the same. Noticed something though: the middle LED (which I'm led to believe flashes each time the tach input gets a signal) went off being continuously illuminated (as expected with 36 triggers for every revolution) but when it starts ******* out it starts flashing around twice a second.
So internal timing problem? We changed the entire internal circuit from Hall to VR and (obviously) changed the external sensors and wires as well, only to have the same problem??
Last edited by beelzeboss; 01-30-2014 at 07:10 AM.
#3
Hey mate, I just created an account. This is my car. Beezleboss has been helping me for the past 4 days including one 40+ hour straight stretch where we finished of the mechanical side and moved onto electrical.
1. The megasquirt is grounded to the battery negative.
2. I am using non-resistored spark plugs. I'll go and buy some others now.
thanks for the reply!
could you explain why the block is better? And whats wrong with non-resistored spark plugs in terms of megasquirts?
Update:
Changed the spark plugs to resistored set, and I changed the ground to the head of the engine. No change. Still starts dying after around 20 seconds, fuel pump starts clicking on and off etc.
Update 2:
It's a Megasquirt 2
Firmware is 3.3.1a (downloaded through the DIYautotune.com site)
I'm using Tunerstudio
Log is here: https://dl.dropboxusercontent.com/u/69963796/Log.msl
We have two tunes, one that we fiddled with quite a bit (CurrentTune1) and one that's pretty much as it's installed with the megasquirt (we did a fresh firmware and project install). Neither worked.
Fiddled with: https://dl.dropboxusercontent.com/u/...rrentTune1.msq
Mostly Virgin : https://dl.dropboxusercontent.com/u/...rrentTune2.msq
1. The megasquirt is grounded to the battery negative.
2. I am using non-resistored spark plugs. I'll go and buy some others now.
thanks for the reply!
could you explain why the block is better? And whats wrong with non-resistored spark plugs in terms of megasquirts?
Update:
Changed the spark plugs to resistored set, and I changed the ground to the head of the engine. No change. Still starts dying after around 20 seconds, fuel pump starts clicking on and off etc.
Update 2:
It's a Megasquirt 2
Firmware is 3.3.1a (downloaded through the DIYautotune.com site)
I'm using Tunerstudio
Log is here: https://dl.dropboxusercontent.com/u/69963796/Log.msl
We have two tunes, one that we fiddled with quite a bit (CurrentTune1) and one that's pretty much as it's installed with the megasquirt (we did a fresh firmware and project install). Neither worked.
Fiddled with: https://dl.dropboxusercontent.com/u/...rrentTune1.msq
Mostly Virgin : https://dl.dropboxusercontent.com/u/...rrentTune2.msq
Last edited by ashh; 01-31-2014 at 03:34 AM.
#5
Ok, so turns out I'm a complete newbbbbbb
*Puts on flame suit*
upon careful inspection of the board it became apparent that a diode was installed backwards.
Switched it around and now everything appears to be working brilliantly! (We haven't driven it yet). Fingers crossed
Thanks for all the help!
*Puts on flame suit*
upon careful inspection of the board it became apparent that a diode was installed backwards.
Switched it around and now everything appears to be working brilliantly! (We haven't driven it yet). Fingers crossed
Thanks for all the help!
Thread
Thread Starter
Forum
Replies
Last Post
Mikel
MEGAsquirt
4
09-28-2015 05:46 PM