HELP.....Sensor needles twitching

[BaldiGit]

HELP....

Got the car booked in to a rolling road to get it tuned on Monday and the IAT CLT and MAP sensor needles in TunerStudeo are twitching like crazy even with the engine not running. The weird thing is non of them react at the same time and if I pull a sensor connection off it's still happens, but more severe. Everything I try seems to make it worse.

So far I have rewired all 3 with twisted wires and joined the 5v and earths (Microsquirt pin 20) as close to the Microsquirt as possible. The ignition to the EDIS and injector wires are shielded. All earths are soldered to one terminal and securely connected.

I'm beginning to suspect the Microsquirt itself...

Help..!!

[LAV1000]

Grounded as on page 15 ?
http://www.msextra.com/doc/pdf/html/Mic ... e-3.3.html

[BaldiGit]

Yes.... Well the only thing I can see is the wideband lambda (LC-1) on that diagram has an earth back to pin 20. There are 2 earths on LC-1 and the blerb that comes with it states they should be soldered to the same tag and earthed to a good earth.

Anyway I disconnected the LC-1 and ran a wire to the Microsquirt and powered it up by itself, no other systems in the car running. And it still does it.......!!!!! Mamma Mea

[BaldiGit]

Tried wiring LC-1 through pin 20. Now it doesn't work...... The Lambda that is..

[LAV1000]

can you post a datalog?

Try this for instance;
Disconnect alternator and all other electrical systems.
Also the fuelpump and WB sensor/unit and sensors.
Make sure you only power the MS system.
If this works reconnect sensors one by one and check.
Next fuelpump, WB sensor/unit and other systems.
And ad system by system check each time.

Do you have exces to a scope?
Try measuring the 5V to sensors.

[BaldiGit]

I like your thinking....

The alternator has 2 wires as well as the big HD battery one; a constant live and ign light. I have pulled the fuses on those with no effect.

The car itself is very basic, the next technical system would be the windscreen wipers....!! I have removed the fuel pump relay and ran the Microsquirt as the only live system in the car
(without engine running) and it still does it.

Sorry I'm not sure what you mean by WB sensor! And no, don't have access to an oscilloscope.

I've only tried the CLT sensor removed at the moment. But even when not connected that gauge hiccups worse. I will try the others, but in frustration I've soldered the MAP (MPX4250AP) in to prove there where no bad contacts. but
I can always take the pin out at the MS.

Think I've knocked the LC-1 out of wack by switching the earths around. It only reads full scale when warmed up and calibrating as per the manual doesn't help. I think it resets to factory settings when a wire comes adrift, so I'll
have to load up the software... Hay-Ho

I've attached a log of it on without the engine running and the fuel relay removed. The MAP sensors going nuts. By the way I'm only using the MAP for reading baro, as I have an Alpha-N setup.

The other odd thing is; checking the earths with a DM, all sensors have about 0.2 ohm continuity with the battery as expected. But when the MS is turned on, that goes to 4.2 ohms.. Is that normal?

[LAV1000]

The other odd thing is; checking the earths with a DM, all sensors have about 0.2 ohm continuity with the battery as expected. But when the MS is turned on, that goes to 4.2 ohms.. Is that normal?

Not normal, looks like a bad connection.
Is there a way to connect the sensor wires directly to ground and bypass the ms unit.
leave the sensor grounds on the MS unit and get battery ground connected to the sensor grounds.
Only to test.

WB sensor: Wide Band sensor (AFR sensing)
These sensor needs to be heated electrical to get to a good working tempature.
This heating is done by pulsed 12Volt and this could cause interference.

[BaldiGit]

I thought it odd..

When you say bad connection, do you mean inside the Microsquirt? What I'll do is take pin 20 out and earth it and have another go.

If that doesn't work I was planning to remove all sensors and add one at a time as you suggested. Last time I removed the MAP the reading slowly climbed then started at the baseline
again. So the gauge was going round slowly.

The WB is powered up by the fuel relay. So as that has been out during most of these tests, it's probably not that. But I see what you mean, finding a system that pulses.


Thanks for the help, it's driving me crazy.

[LAV1000]

When you say bad connection, do you mean inside the Microsquirt? What I'll do is take pin 20 out and earth it and have another go.

Yep, I would first check the microsquirt connector, esspecialy the one that connects the "field" to the microsquirt unit.
Did you make wire loom yourself ?
Is it soldered or crimpconnected ?

[LAV1000]

Came acros an older post:
http://www.msextra.com/forums/viewtopic ... 01&t=57623
Lotus 7 type ?

Did you check the heavy duty cables and connections, lose, coroded.
Aspecially the ground one.
Because I noticed in both datalogs battery voltage seems a bit low.
A good battery shouldn't drop below 12V with only the MS system connected.

[BaldiGit]

Yes, that's me... The problem started with the IAT and has evolved. Nobody was reading that post so I started a new one.

I have unplugged/soldered the IAT, CLT, TPS and MAP sensors, so no sensors connected at all and run a log. The problem is still there and interestingly the water temp jumps from 60 to 80 degrees for a few seconds then drops back...!!

So then I connected the earths that where connected to pin 20 to a good earth and tried again with all sensors connected. Still the same.

The battery voltage had been bugging me too. I kept thinking it needed charging but the car would turn over fine...!! So today I checked the voltage at the battery and it's 12.06V (on a cool morning). Then I pulled the ampseal plug at the MS and checked pin 1 and that was also 12.06V. Hmm... But the log taken this morning states it as 11.5V

I believe you can compensate in the software somewhere, but would that make a difference?

[LAV1000]

The battery voltage had been bugging me too. I kept thinking it needed charging but the car would turn over fine...!! So today I checked the voltage at the battery and it's 12.06V (on a cool morning). Then I pulled the ampseal plug at the MS and checked pin 1 and that was also 12.06V. Hmm... But the log taken this morning states it as 11.5V

Looks fine to me.
Ms sytem uses this voltage to compensate for dwell and injector timing.
If car runs ok I would leave it for now.

Check settings for lag factor and settings.
Reflash ms unit ?

It seems worse when nothing is connected, signal drops deeper.
What's the condition of the cicuit bord, ever contacted water or long time in damp condition.
Bad soldering ?

I am runnig out of thoughts

[BaldiGit]

The frustrating thing is the car runs lovely. Albeit with a not much better than a basic map... I have it booked in with a rolling road tomorrow to finalise the map but thats all gone up in smoke.
The car you see is just completed and hasn't really been on the road much, nothing has had a chance to corrode. But I have checked all connections.

I've given up for today for my own sanity. As the more I do the worse things get, The alternator light has gone out now.... And still haven't checked out the LC-1. Doh.....!!

I will reflash the MS as the ghost seems to be in there. What did you mean by check settings for lag factor?

The MS is a sealed Microsquirt V3 so I can't look at the board. But it's inside the car behind the dash and has never seen any damp as it's been garaged.

I can't help feeling I've missed something, but the longer it goes on, I get the dreaded feeling there's a dodgy solder joint in the MS.

Thanks again for taking the time to reply.

[LAV1000]

What did you mean by check settings for lag factor?

I ment the avaraging lag factor.

[BaldiGit]

Mine is almost the same. Only I have the MAP set to baro and I see you don't have a MAP.

I'm a bit confused about the Multiply MAP option...!!!

[LAV1000]

I didn't use multiply map because I am running itb's and don't have a map sensor.
If you need an explanation on these settings click on the question mark next to it.

I don't think this causes the problem

How about that LC-1 unit did you disconnect it ?

[BaldiGit]

I did click on the ? and that's what confused me.

The LC-1 is wired through the fuel relay and I've been removing that so I'ts not pumping fuel into the engine each time. So that's been out of the picture for a while now....

Can't get to it now until Friday. But I'll reflash the firmware and see what that does. I wish I had a stim so I could prove if the problems in the car or the MS.

[BaldiGit]

I might try filtering out the spikes with a decoupling capacitor and see if that helps.

If it reduces it to a ripple I'd be happy. Any ideas what value capacitors would be best? I've been looking on the net, but the people who say they have done it, don't mension the value they've used...

[LAV1000]

Never done it, so can't help you on this.

[BaldiGit]

No problem, experimentation time....

I'll let you know if there is a definitive answer.