odd drop in rpm registererd and ecu (Megasquirt) reset
So having been up to Dale's we though we had Ironed out all the issues, any electrical noise, by putting a shielded cable all the way to the vr
crank sensor, adding a capacitor to both the 12v and 5v circuits.. however something odd is still happening.
I left fully thinking the issue is just unavoidable as at really light throttle, it will open up the butterflies but nothing registers at the ecu
(Megasquirt 1 extra) for a bit, which makes getting the fuelling right for that bit to be a problem, so very lightly going from idle to a tad more
makes it go lean and if you do so too slowly it could conk out. Dale tweaked the best he could within the limits of the throttle linkage. I left
happy after going for a spin..
However, it isn't just that, it after a good run ie, got properly hot etc hits points where it will show a loss in rpm signal, but I also
spotted that the LC2 wideband is being powered off and on as it takes a few seconds to register the afr again.. Ie the ecu is cutting off the fuel
pump relay ( which i use to run the LC2 from)
Dale did wonder if the resistance on the VR sensor was a but low when he was testing the shielded cable setup, so I will be replacing it, but thought
I would put this out to all.
here are a couple of vids which show what is happening.. It is bad enough that when sitting in traffic, or at the lights it can conk out, pulling away
slowly can cause it to bunny hop, even stall.. each time the rpms drop and the lc2 resets, so it does look like an ecu reset. It didn't happen
all day yesterday when it was on the rollers ! we only spotted some electrical interference when Dale was adding the tach output and we had a relay
coil in the mix, which led to the shielded cable... (the etb gauge actually ended up just needing a pull up resistor)
just to be clear, Dale didn't experience any of this and has done an awesome job, I'll be pointing him towards this post !
any way any thoughts on this odd behaviour?
does a loss of crank signal cause the ecu to reset?
it happens as per the video and slightly higher rpms and also at about 1200, so 30 in 4th 50 in 5th become troublesome.. the vids show what is
happening without me touching the throttle, so it should have settled.
Photo Archive
Building: Fisher Fury 4.3 V8 with Speeduino
posted on 14/5/15 at 05:25 AM
Nice to see MSDroid being used Is it a loss of RPM or an ECU reset? Capture a log file and look at the time column to see if it resets to zero
unexpectedly (although it will loop from 255, that's expected)
If you can use tuner studio you should be able to see if the ecu loses sync. Which will help you determine if its a sensor problem or a voltage
problem causing it to loose itself
Is your VR sensor really rigidly mounted? If the mounting is vibrating at certain engine speeds it's entirely likely for the MS to lose or gain
tooth counts which will cause problems.
Photo Archive
Building: Fisher Fury 4.3 V8 with Speeduino
posted on 14/5/15 at 07:51 PM
Yep, the ECU reboots a couple of times in that log file. It's the second column SecL, not the first that shows it (and the markers MSDroid
inserts when it happens)
cheers, Dale agrees - well and I can see it too, a second or so while the reading stops. very easy to use MSDROID
under advise from the boss man, before my trip on tuesday,( a mear 1800miles!) I am going to move the coils to earth on the block and also move the
MS earths to the block too..
then just drive fast all the time!
also going to do an oil change - might try and stop some of the leaks at the front.. the valley gasket is dry, the sump is pretty dry, the end
nearest the gearbox still manages to get a drip, but at the front I am sure the oil cooler pipes leak (new ones) and also around the oil pump..
I'll see how brave I am feeling and as I probably will have to take the pump off to change the filter, I'll try and seal it up some more.
MS now earthed to the block with 5 separate earths going there to the same point.
oil changed, filter changed, spotted one of the oil pump housing bolts was not tight - doh! still weeps a little now i've tightened it, but I
think it will do rather than risking taking the oil pump off etc.
it ran in the garage till very hot etc and I could not reproduce the ecu resets.. not the same as going for a drive, but fingers crossed!
On my setup a closed throttle registers at 58.62ADC so the bottom line on my table is at 59.
This guarantees that a closed throttle is on the bottom line.
The next line up is at 60ADC, so even the slightest application of throttle moves you up off the bottom line and moves up onto a seperate line that
can be tuned.
The next line is at 65.
The throttle axis looks like this...
220
180
150
120
110
100
90
80
70
65
60
59
I should mention that I am on Alpha N algorithm.
In percents it might look like this...
100
80
60
40
30
25
20
15
10
5
1
0
Load bins close together at low openings, wider apart at higher openings.
A throttle application of 1% will get you off the bottom line.
I suspect that your second line up is at 5% or similar. ( If you are on Alpha N, that is)
thanks for the post, however, I am not touching the tune, Dale got that bang on.. I know what you mean tho, get more resolution where you need it.
It is Alpha N.
going to get another log and capture the reset just to prove it is definitely a reset after switching the sensor still.
You don't say what PCB you are using or what set up your running but ecu resetting it usually a sign of a wiring problem sync loss is a
different story this is a problem with your trigger setup
I have only when cold, captured a tooth log, one with the alternator on and one without (just to see if the alt is putting out noise). So this
isn't when the car is suffering the reset.
anyone care to take a look and see what they think, it looks a bit odd to me, like there are drops, but no drop is shown. I need to read the manual
some more.
I was kinda just putting up with the reset, but a couple of month a go it started getting worse, to the point of becoming not drivable- why I still
don't know.
I essentially pulled every fuse that wasn't essential, ran without the wide band too and was still having the problem.
I also ran with just 4 coils, then the other 4, it would just about run on half cylinders, enough to show the reset problem.
I also pulled the plug on a coil at a time and in all situations I could get a reset.
So, I eventually got round to making a new bracket for the crank pos sensor in case it was moving / when hot flexing.
no change, but hey I had a go at welding so not all bad. didn't set fire to anything
more googling and found a msextra post where someone was having positive flyback from ls2 coils and was hearing a clicking on their alarm siren - yes
I hear that too.
so I tried a 100uF capacitor on the main power feed to the coils. 2 of as there are 2 banks of 4 coils.
this definitely helped, no siren clicking all the time but still resets with clicks occasionally.
so i thought perhaps the capacitor isn't big enough, so I now have 4700uF ones, seemed much better again.
I tried earthing the coils (main earth) at the ms earth point on the block - loads worse, so put back to the other side.
I then decided to take the feed for the coils directly (via 1 relay) from the battery and hey presto things seem to be good. I wonder if this was a
flaw in my design as if the manufacturer takes it straight from the battery and the battery is bigger it would absorb more. I now get a click as the
ignition goes on, almost like the coils firing once when they get turned on and the ms goes on, but then all seems good - well running on the drive
for a while anyway.
I still need to get the wiring tidied back up again so I can get it on the road again and see how it fares.
so questions!
1) any one see an issue with using a 4700uF capacitor?
2) I currently just badly have used connector blocks to get them in line and wrapped them up in wiring tape. These are just 105degree 35v maplins
electrolytic capacitors, are there any better (even possibly bigger) ones that are more suited to under the bonnet application?
quote:Originally posted by Chris_Xtreme
well, a progress report!
so I tried a 100uF capacitor on the main power feed to the coils. 2 of as there are 2 banks of 4 coils.
this definitely helped, no siren clicking all the time but still resets with clicks occasionally.
so i thought perhaps the capacitor isn't big enough, so I now have 4700uF ones, seemed much better again.
I would try it with a big (ie the 4700uF), and the 100uF, and another even smaller, say about 1uF, all in parallel. the idea being the small
ones act fast, the big ones more slowly.
quote:Originally posted by Chris_Xtreme
well, a progress report!
so I tried a 100uF capacitor on the main power feed to the coils. 2 of as there are 2 banks of 4 coils.
this definitely helped, no siren clicking all the time but still resets with clicks occasionally.
so i thought perhaps the capacitor isn't big enough, so I now have 4700uF ones, seemed much better again.
I would try it with a big (ie the 4700uF), and the 100uF, and another even smaller, say about 1uF, all in parallel. the idea being the small
ones act fast, the big ones more slowly.
interesting idea - I'll bear this in mind if I have more troubles.
ah yes, I was aware of those, not that I can find one to get!, but the value is far too small.
oddly, the MS2 doc has it down as optional to have a cap on the feed, the ms3 doc has had it removed and in the msextra thread I mentioned, James and
others declared it cannot be required, yet others and myself now have found they add benefit.
quote:Originally posted by coozer
Just been reading PPC. Will had a similar low speed problem on his V8 Capri and Dave walker found the TPS to be not reading correctly..
Have you checked/changed the tps??
thanks for the idea, I by chance had another and swapped it. no change.
I can see the TPS reading in the MS being the same for both and going up as expected, so I don't think it is that.
As much as I thought on the drive that the capacitors and move to taking the positive straight off the battery for the coils (and MS) it still is not
right.
I took it for spin round the block, thought it was all good, got round a corner and it was awefull. couldn't get it going again and had to push
it home.
thankfully I found I had missed a washer on the earth for the coils and it was a bad connection. put that back and all good.
except I noticed on the drive again that the MS would reset.
I think I am going to have to have a rewire from the start for my power supplies as I currently have everything behind 1 80 amp relay (controlled by
an alarm). Now the coils and MS come straight off the battery but via individual relays. the coil relay is controlled by the fuel pump relay and the
ms by the ignition switch.
I also have some wierd issue where the tachometer will bounce around, or just go up when turning on the head lights, or bouncing with the
indicators.
very odd.
I had tried with the alternator charge wire off , but the 2 wires still on and then with the charge on but the 2 wires (exciter and dash light)
off.
I need to try all 3 off - just in case. I have recently spotted that the dash light isn't working all the time (it is a little denso alt) so
possibly part or all of the regulator has gone.
it is ridiculous, which ever way I wire it, it makes no difference.
What is odd is that it has got worse without me changing anything.
I left yours a year and a half a go, without it happening at all on the rolling road.
then it was a mild occasional reset at idle / slow rpm.
it still is at idle / low rpms but it is much more violent.
what ever is doing it I can hear it via the alarm siren (yes I have run the car with as much as possible removed and it made no difference)
next step is to completely take the alternator out, especially now that I have spotted that the dash light doesn't work all the time(wiring is
fine, checked it with a direct power feed and it doesn't work some of the time)
so I'll keep you all posted once I have done that!
Is it a loss of RPM or an ECU reset? Capture a log file and look at the time column to see if it resets to zero
unexpectedly (although it will loop from 255, that's expected) 
