Reading/ Reflashing 29F chips - electrical gurus?

[chri0029]

Okay, so I've been doing some forum reading in the past month trying to figure a way around making someone really rich by buying software costing >$500+ that only works with 2 VINS for later OBDII's. Completely a rip-off in my opinion unless you're doing this comercially. The DIYer with only a few parameters to change should be able to do this somewhere in the $100-$200 range (without limiting VINS). I'm really surprised the DIY market hasn't driven this (maybe in the future with J2534 lobbying for emissions regulating for 'aged' component cals rather than stock ones based on new component wear). Different discussion for a different day.

It seems nobody has found a good way (YET) to reflash newer OBDII's through the J1962 ALDL without making certain software companies really rich. Hopefully that comes soon. That leads me to "plan B" for the meantime - reflashing directly the chip.

I bought a bone yard PCM for $30, and it turns out the stock chip in that PCM from my research is actually pretty common & powerful. It's a 2x22 pin AM29F400BB chip with 8/16bit flash memory. It even has a built-in algorithm for reading, erasing, and writing.

There is so much information out there about removable OBD1 & 1.5 chips & reflashes, but not much about post-2000 OBDII reflashes. The little information I do see involves reflashing the chip mostly through Burn2 or Willem EPROM programmer hardware, but with the chip removed from the motherboard. However, removing that soldered chip is not my cup of tea unless absolutely necessary. I don't think I could solder in a DIP socket into the board reliably for easy chip removal.

Has anyone done any reflashing of the chip while in the board or removed a soldered chip? I attached my AMD chip specs, including pin-out. It appears that only a few pins are used to read/erase/write (most likely only the CE#, OE#, VIL, WE#, and BYTE# pins) That's it!? Therefore, soldering or hot gluing wires directly to the board would seem the most logical rather than removing my entire chip. Then possibly hook the wires directly to a Willem programmer or another pass-through device to a computer to read/ erase/ rewrite the bin? Or even JTAG directly to a PC through the parallel or serial port? I have done this before on a simple computer application, but not ever a PCM. Has anydone done this successully? Any ideas/ advice in this area from the electrical gurus?

[chri0029]

Here are the chip specs including pin outs and read/erase/write commands. Probably similar to other GM 2004 V6 PCM's.

[Weatheralls Auto]

I am currently working on this, just waiting for another piece of hardware.

What Vehicle was that PCM out of? a CAN maybe?

That Flash is the newest I believe for GM, earlier ones are AM28F400BX

[chri0029]

It was out of a '04 Grand Am GT. 3.4L

I'm going to try my soldering skills this weekend with an IC from an old broken laptop. I've been reading online and actually getting interested in unsoldering the chip and installing a DIP in its place. The experienced soldering guys make it look so easy on YouTube. Chips are pretty cheap if you burn one.

I want to try this hardware direction in case your ALDL cable hacking doesn't work. My goal is to have something working (either your way or mine) in the next few months when this damn snow goes away and I actually want to get outside and tune.

I'm still interested if anyone has done this before (soldering/unsoldering the IC) or setting-up JTAG wires.

[robertisaar]

I've been reading online and actually getting interested in unsoldering the chip and installing a DIP in its place. The experienced soldering guys make it look so easy on YouTube. Chips are pretty cheap if you burn one.

DIP or ZIF? honestly, i would go ZIF since there is the possibility of junking a chip if the flash doesn't complete correctly.

[dimented24x7]

I have my current PCM set up like this. I used 30 gauge insulated copper wire to make the connections to the external ZIF socket. Its a real bitch of a soldering job, but it does work. You will also need to mount (or relocate) the PCM to inside the car, obviously, as it will no longer be weather tight. ESD can also be a problem, so you need to take that into consideration as well. Even a small discharge could burn out the entire PCM as the most sensitive components are now exposed.

[dimented24x7]

Getting the flash chips is also an issue. They seem to be hard to find in small quantities now. I managed to get 10 new AMD 29F400 chips for $5 a piece about a year ago, but it took a lot of looking around. You will want to have at least a few spare chips as the leads become damaged and corroded over time.

[chri0029]

Thanks, Dimented24x7. Good pic. At least I know somebody has done this before. I was actually thinking of a DIP socket (I believe) to solder in place of the IC still internal to the PCM. Then I can pry out the old IC, reprogram, then push in the DIP socket easily. It seems that a ZIF socket would be best mounted external to the PCM like you have (then you have all the moisture issues & static discharge potential you spoke of).

My tuning is fairly simple - just a few parameters and shift points (including the fan on/off which is waaaay too high for this vehicle). So I wouldn't be taking out the IC too much. But I agree that I would try to get a few spare chips in case unsuccessful programming frying the chip or corroded terminals.

I am still interested in how many other people have done this before. It's either this way or hacking through the ALDL cable or spend mondo $ to make someone else rich (I wouldn't be so opposed to this if $500 was one-time for unlimited VINs).

[SappySE107]

i love how you think $500 makes someone else rich.

[Weatheralls Auto]

I have to agree with 1 thing though, $500 plus vin licenses @ $80 to $160 is a bit much.

$500 is not making someone rich, $500 X ???? could be.

I am waiting for some hardware, but not just sitting idly by, have the 256 Algos now, just have to decipher them to add to my App.

[SappySE107]

You get 8 credits, and it takes 2 per VIN. 4 per model/year. I just bit the bullet and got the 70 credit V6 GM Cars setup...but I have a business. Otherwise, the initial 8 credits included at $500 would be sufficient for my cars.

You couldn't pay me to use DHP or Tiny Tuner. I had a DHP, and if I thought it was feasible to use it for all vehicles, you can bet I would have. HPT is so much easier to use, and its faster.

[dimented24x7]

i love how you think $500 makes someone else rich.

Those guys over at HPTuners seem to be making a pretty good living out of it. That, and from what I can tell, a lot of the information thats used to construct the definitions comes from info leaked by the manufacturers, so I dont think there is as much work as we would think in putting together the tuning packages.

[SappySE107]

From what you can tell? Unless you work for HPT, I think you are making this up.

[Weatheralls Auto]

2 per vin for 60*V6 etc, 4 for others like LS1 I believe for HPT.

So some will be $160, not saying one program is better than the other, just that it seems expensive.

This is why I am trying to make an alternative, may not be as easy or efficient as HPT or even DHP, but more for the DIY.

If I can make an Open Source alternative for $100 - $150, if it takes longer or is not as efficient, I think it will not only benefit the community but force the others to lower their prices? would you not agree?

[SappySE107]

3.6 costs more as well, but if you had one of those, you wouldn't complain about the cost of the tuner:P You certainly wouldn't be hacking it open to try to make something else work.

Until you have a scantool and the means to program without hacking a PCM, I do not think you are going to get HPT to lower their price. I would say we are lucky they made a setup that we can use. This market is niche. LSx are bank.