pcmhacking.net

Maybe try my copy?

vlad01 wrote:Maybe try my copy?

OSE $12P V112 APNX V6.BIN

Thanks vlad, I have just tried your copy with the same bad result

All the bins in the zip work fine. You have done the right things testing. I would assume there is some kind of hardware problem maybe with your memcal adapter. The factory bins are stacked in your implementation, so the bottom 16kb is just filler and is not used to run the car in your testing, so you are not testing all the address pins. The 12P is a 32kb bin so the bottom 16kb is used in car. You might be missing connectivity through the memcal of one of the address lines in the bottom 16kb, so when you read the chip its fine, but when its in the pcm the bottom 16kb might have a problem. its best to get a couple of rows of pins and some vero board and make up an adapter so you can plug the whole memcal with the eeprom in to the programmer, then you'll be able to test connectivity end to end through the memcal assembly you have made and any internal problem will show up at the programming/verifying stage.

antus wrote:All the bins in the zip work fine. You have done the right things testing. I would assume there is some kind of hardware problem maybe with your memcal adapter. The factory bins are stacked in your implementation, so the bottom 16kb is just filler and is not used to run the car in your testing, so you are not testing all the address pins. The 12P is a 32kb bin so the bottom 16kb is used in car. You might be missing connectivity through the memcal of one of the address lines in the bottom 16kb, so when you read the chip its fine, but when its in the pcm the bottom 16kb might have a problem. its best to get a couple of rows of pins and some vero board and make up an adapter so you can plug the whole memcal with the eeprom in to the programmer, then you'll be able to test connectivity end to end through the memcal assembly you have made and any internal problem will show up at the programming/verifying stage.

Hey Antus,

What you say makes absolute perfect sense. And explains why I read back the data OK using the EEPROM directly on a socket on my programmer.

I'll hook up the memcal adapter to my programmer as you suggest when I can find some pins tonight.

By the sounds of it, from what you explain, i would guess that pin A14 of the EEPROM is probably not connecting thus not reading the 2nd 16k worth of the BIN.

Will report back with my results once tested!

Hey Antus,

I haven't pinned the memcal adapter just yet to my programmer, however I thought id run some continuity tests on it, and it was OK. I.e. Probing all Pins of the EEPROM and probing the corresponding slot at the bottom of the adapter was OK.

HOWEVER, now here is where things get interesting and im sure is the problem.

on the underside of the ECU, in the location of the memcal adapter, I can see solder joints - So i tested each of these for continuity with the EEPROM attached, everything was OK, apart from 2 at the far left. These would be, on a 28c256 EEPROM, A14 and VCC.

These appear to be joined together!! as in, there is connectivity between VCC and A14 essentially.

Now on the original factory EPROM, a 27c128 - these pins would be VCC and VPP....probably not an issue with 14 address lines and none of them being at this pin.

Using the bigger 28c256, A14 is now joined to VCC - meaning A14 is always High.....probably causing my issue.

Here's a pic (sorry for bad quality)
Orientation is Looking at underside of ECU, with MEMCAL being at top, Harness sockets at bottom.

Labelled solder joints are EEPROM Pin labels....and circled is the short - joining A14 to VCC - note that this ECU has not been modified in anyway so i assume these were joined from the factory?

I assume that I should cut this track/join so that A14 is not connect to VCC?

Before I go ahead and do that, I'd like to ask you/the forum members reading this, is this usual? Have you seen this before?

Mine has that too and no problems with it's operation. I have the 305 ECU (808 with built in high speed comms)

From the trace on mine, there is only that cap connected to it. Maybe on yours the cap is shorted?

Hmm, the cap seems to be OK - no short, and reading about 0.6k Ohm

The next component in line, the resistor, (not in the pic, but I assume you can refer to your board and would be the same, 103 printed on the resistor) also seems to be OK, reading at around 10k Ohm

That cant be a used address pin, thats a factory trace not an accidental short and there is no hardware revision that doesnt work with 32kb bins.

I have seen on one of my memcals that heat when soldering the socket slightly melted the memcal internally, and it worked in the programmer but not on the ecm unless I pushed on the whole memcal and the sideways mechanical force made a connection to the pin and the ecm came alive until I took my hand off. I threw away the memcal and built it again with less heat.

looking at my memcal thats the pin1 end of the chip, and looking at this diagram its VPP and VCC so should be fine connected.

And digging a bit further with off the shelf eeproms, I find this variation isnt pin compatible with its earlier eprom counterparts. What chip are you running?
A bit of a 'dead bug' style patch on the memcal carrier might be able to make it compatible if this is the pinout your dealing with.

A bit of creative wiring under the socket on the chip to leave pin 1 ecm side and pin 27 on the chip side disconnected, and pin 1 on the chip to pin 27 on the memcal.