265 Hemi to efi

[Charlescrown]

Yes a dead coil. It all came back to me @ 1AM this morning. It was a Sigma 40 years ago where the owner said on rare occasion it bacfired and did a merry carryon wen you went to start it. I had the car for a day and tried umteen times and it ran fine every time and when the owner picked it up it did it. Hiss, bang wheeze. The Camira coil has a very low primary winding resistance (like the Sigma) of less than 1 ohm and the module has a current limiter that clips the primary current to 6 amps. The old coils it was easy to find the fault. They were filled with oil which leaked out over time and they gave off a burnt smell. The Camira coil being plastic can crack the insulation and smell burnt. If I was a betting man my money is on the coil. Their cheap I see Repco selling them for $24. My guess is the ignition system you fitted has cooked the coil. It should have less than a ohm on the primary and around 5,500-6000 ohms on the secondays but I have found coils test OK but are burnt inside. If the coil fixes the problem you need to look at the ignition on a scope to make sure the module current limiter is operating or the new coil will also fail. FINGERS CROSSED I'm right.

[Holden202T]

yeah map sensor looks good now!

the rpm spikes in the earlier log are worrying .... but i think to confirm that one you probably really need a scope on the trigger when its running...

[heff0018]

Yes a dead coil. It all came back to me @ 1AM this morning. It was a Sigma 40 years ago where the owner said on rare occasion it bacfired and did a merry carryon wen you went to start it. I had the car for a day and tried umteen times and it ran fine every time and when the owner picked it up it did it. Hiss, bang wheeze. The Camira coil has a very low primary winding resistance (like the Sigma) of less than 1 ohm and the module has a current limiter that clips the primary current to 6 amps. The old coils it was easy to find the fault. They were filled with oil which leaked out over time and they gave off a burnt smell. The Camira coil being plastic can crack the insulation and smell burnt. If I was a betting man my money is on the coil. Their cheap I see Repco selling them for $24. My guess is the ignition system you fitted has cooked the coil. It should have less than a ohm on the primary and around 5,500-6000 ohms on the secondays but I have found coils test OK but are burnt inside. If the coil fixes the problem you need to look at the ignition on a scope to make sure the module current limiter is operating or the new coil will also fail. FINGERS CROSSED I'm right.

Hopefully, but you need to consider this is a brand new coil that has never been hooked up to the CDI.

[heff0018]

yeah map sensor looks good now!

the rpm spikes in the earlier log are worrying .... but i think to confirm that one you probably really need a scope on the trigger when its running...

Thanks for the confirmation.

Good news is that the dyno tuner is back in the game after completing his Cummins powered 100 series Landcruiser conversion and said he will help me.

He lent me a new coil to try and said that if it doesn't fix it I can take him the module, dizzy and coil and he can set up on his bench and test with an oscilloscope.

One of the first things he asked me about was the state of the MAP Sensor when I described the earlier backfiring problem so that's interesting.

[heff0018]

Yes a dead coil. It all came back to me @ 1AM this morning. It was a Sigma 40 years ago where the owner said on rare occasion it bacfired and did a merry carryon wen you went to start it. I had the car for a day and tried umteen times and it ran fine every time and when the owner picked it up it did it. Hiss, bang wheeze. The Camira coil has a very low primary winding resistance (like the Sigma) of less than 1 ohm and the module has a current limiter that clips the primary current to 6 amps. The old coils it was easy to find the fault. They were filled with oil which leaked out over time and they gave off a burnt smell. The Camira coil being plastic can crack the insulation and smell burnt. If I was a betting man my money is on the coil. Their cheap I see Repco selling them for $24. My guess is the ignition system you fitted has cooked the coil. It should have less than a ohm on the primary and around 5,500-6000 ohms on the secondays but I have found coils test OK but are burnt inside. If the coil fixes the problem you need to look at the ignition on a scope to make sure the module current limiter is operating or the new coil will also fail. FINGERS CROSSED I'm right.

Hopefully, but you need to consider this is a brand new coil that has never been hooked up to the CDI. Thanks for the heads up and at $24 I will pick up a spare.

[Charlescrown]

That being the case I may well be barking up the wrong tree. As long as it was a new coil it should be fine. Perhaps making sure it's got the full battery voltage might be a check that needs doing. You could slip a paperclip in with the positive coil plug and measure the voltsge when trying to start it.
Can you disconnect the coil primary plug and do a log for say 30 seconds that way the spike should be gone and all the other data should look as it should.

You need to fix this hole in your igntion map.
It's got nothing to do with the problem but it must have been missed.

[heff0018]

That being the case I may well be barking up the wrong tree. As long as it was a new coil it should be fine. Perhaps making sure it's got the full battery voltage might be a check that needs doing. You could slip a paperclip in with the positive coil plug and measure the voltsge when trying to start it.
Can you disconnect the coil primary plug and do a log for say 30 seconds that way the spike should be gone and all the other data should look as it should.

Ok will do, thanks.

[heff0018]

This is a reply from Antus (Thanks) that I am saving in this thread so I can find it again.


So... this is documented elsewhere but I am tired and it'll be quicker to summarise.

The stock VN/VP bin is 16K. OSE12P is 32K, a VR '424 computer is 64k. The NVRAM is 128K. The ecm is 64k and cant see the whole 128k nvram chip. MAP_SW controls the last address pin so the ecm can see the high or low bank (half) of the chip.
The location of the data needs to align with the top of the address space for the CPU.

So, for 12P, on the nvram at the hardware level, that means 32K of blank, 32K of 12P (Low Bank), 32K of blank, 32K of 12P (High Bank). Two 64k banks of data, with 12P in the top 32k. The MAP_SW jumper lets you switch between the top and the bottom half of the chip in hardware.
So the stacked bin, is for when you put the chip in the programmer, and it writes all of the padding, and both copies of the tune. A good starting point for a new or to reset a previously bricked nvram. Then if you have the jumper on the MAP_SW your computer runs off the low Bank, remove it you get the high bank. Thats why sometimes if you mess it up, you can remove the map_sw jumper and it might come back to life on the second copy.

When you do it with the flashtool, you are going through the ecm, so it can only see half the nvram. Further, the computer needs to be working - not bricked. So you probably cant do it through the ose flash tool if you are bricked now. If you can, then you should be able to connect with tunerpro and you should not have this problem in the first place. The flash tool and ose plugin know how the ecm handles memory addressing and it hides it from you. So you dont need to worry about or think about that you have a 32k file on disk, and internally the ecm is using 32-64k. That just happens like magic. You should upload a 32K bin.

Also noteworthy, oseplugin, by default, only transfers the calibraiton. You can go in to its options and uncheck "calibration area only" if you want it to transfer the whole bin including calibration and operating system.
This means that to:

Read a complete bin and write to a memcal with a 32k eprom or eeprom you would read a bin with oseflashtool to get the whole file. Alternatively you'd deselect "calibration area only" in oseplugin and hit the down arrow then save the bin.

To recover a bricked nvram you would remove the chip carefully from the nvram base, being especially careful not to bend the corner pins when you remove it as they bend easily as the most of the pins let go and the corner pin is not free of the socket. Select the Dallas DS1245Y (or DS1230Y or AB variants if you have those) in the chip programmer software. Put it in the chip programmer with the alignment the software is telling you - note which end has the white dot marking pin 1. Write a stacked bin from the original zip file. Put it in the nvram base board, put it in the ecm. connect with tunerpro and test connection. carefully put the RW jumper on. Test connection again. Exit tunerpro (if you tested with tunerpro) or any other software talking to it. Use oseflashtool to write the calibration you last saved with your tune. This will upload the calibration only and keep the operating system you replaced in the programmer. Its much easier than trying too cut and slice bin files on a PC to get them stacked properly, but it is still possible to go that way if you know what you are doing.

[heff0018]

test RPM spike 06.02.25.xdl

(14.18 KiB) Downloaded 180 times

I have attached the log file with the power to the coil removed.

Also I tested the coil's primary and secondary windings and the resistance seems ok.

Its getting consistent spark to at least cylinder 1 but I didn't get a chance to see if the timing light was picking up the signal.

I didn't get a chance to swap the computer's over but will do so over the weekend.

I pulled the plugs and noticed something strange. Cylinder 1's plug was black and slightly fouled (but not wet) but the rest of the plugs were clean??

What a mystery.

Cheers

[Charlescrown]

That log confirms what I suspect and the voltage when the plug fires is being picked up and generating some strange oscillations probably back into the primary windings. Regarding the NVRAM info as I have said I haven't ever used one so at this stage would only get confused. I use 512 flash chips and ofset the write by 8000 and have never had a problem apart from the occasional write error caused by a bad connection with the Eprom reader. Regarding the ECU's thay will all 808's will work with 12P because they are all the same no matter what car they come out of except the VR/VS which the only difference is they have high speed coms and output the data on a different pin. The difference is in the Eprom. Anyway weekend coming up so you might have some time to look and think some more.