Mangus,

I'm trying to burn a chip after finaizing the bin using Ostrich. I have RBob's EBL so its a 16k bin. I'm using an ST M27C256B chip and Needham's EMP10 burner. I set the bin stacker utility at 16k bin/256kchip/32kblock. Then I input same bin file for each of the 8 positions. Comes up that bin has been stacked successfully. But then I go to burn it and burner software shows a
Program Verify Error. Its says the BUFF ADD is 6B9, Devce ADD is 6B9.
When I go to load the bin, it only shows that its loading 32k, not 256k.
Help! What am I doing wrong?

 

Dom,
I thought you said it was time to move on??? Just kidding. Anyway I've tried stacking a bin just like your trying to do but using a dataman programmer. I've had similar results. i have even got a chip to program correctly(successful burn) but failed to work in the ecm. Maybe if we could get a step by step and make it a sticky it would benefit many. My programmer guide was no help.

 

Very odd because I was able to successfully brun a run a chip using the
AT29C512. TPRT shows that bin was successfully stacked but then when I try to burn it, it comes up with error message. Is it necessary to have the .bsl file for this? RBob tried it and came up with same results.

 

I'm wondering if it has anything to do with being a B chip. I am currently using the A (M27C128) But have a mass of M27C256B that are available to me. I am not sure what the difference might be between M27C256A and M27C256B.

 

No idea. My burner shows either a M27C256 or a 27C256B both for ST which I assume is SGS Thomson. In either case, it comes up with same verify error message.

 

In the NS (Nat-semi) line the B chips can be programmed with a faster algorithm. Can use the non-B algo on them, but can also use the faster algo with the B chips.

When looking at BIN sizes and PROM sizes, one is in bytes, the other in bits. A 27C256 PROM will hold a max of 32KB BIN, such as $8D. That is: 256 / 8-bits-per-byte = 32.

A 16KB BIN file will need to be double stacked on a '256 PROM.

In the TP stacker tool I am not sure what the 'block' size is. In the stacker the BIN size is 16KB, chip is 32KB (correct for a 27C256), then the block size of? Ha, just figured it out!! I was playing with it to post this and have it.

Dom, use 16KB/32KB/16KB for BIN, Chip, & Block.

Now (this is where I was messing up), below that there are two blank entries as 'Bin Layout - Input Files.' Click on the browse button on each and give it the BIN to stack. In this case they will both be the same. Then click the Stack button.

I had thought it would use the currently open BIN, but it doesn't. Which makes sense when using a BIN switcher. Here we want to do a simple stacking. With a switcher the user will want to stack with a different BIN in each position.

RBob.

 

RBob,

As usual you hit the nail on the head. Just burned one. I tried the M27C256 first but came up with another Verify Error. But then I used the 27C256B with the Quick Pulse Alg. and BINGO! Thank you again, my friend. Now I'll be delivering my Xfire to its new owner. Took out the WB and Ostrich. Use those for the new ZR. Lovely day here, so I'll be taking the roof off for my last ride in the Xfire.

 

The "Block Size" parameter is used as another subdivision of the stack. It was put there because the Moates switchers had 32KB switch blocks. Think of a block as a chip within the chip. As an example, say you have 16KB bins, that you'd like stacked in 32KB switch blocks on a 256KB chip.

A 16KB bin would have to live at the upper portion of each 32KB block in order to work at a given position on the switch. Since we're talking about a 256KB chip, there would be 8 switch positions.

In a case where the ECM/device expects the bin to simply live at the top of the chip, the block size should match the chip size, so in your case, use 256KB block and chip size, and 16KB bin size.

I hope that helps!

EDIT: You could also choose 256KB chip size, 16KB bin size, and 16KB block size. That would cause the 16KB bin to be repeated without any "blank space" until the 256KB chip is full. The earlier suggestion would fill the first 240KB of the chip with zeroes, then add the 16KB bin to the last 16KB of the chip.

 

Mangus,

Just tried it as per your edit. Came up with same Program Verify Error. They only way I have burned it successfully is to do it as RBob described. I used a
16kb Bin/32kb chip/16kb block size.
Also, I'll note that some of the chips I bought don't appear to erase completely using my dataeraser UV. But even with chips that check out as erased, unless I do it like RBob stated, I end up with error.

 

Thanks Bob I have a slug of 256's that I can get from work. This will make tuning more pleasureable since I can only erase 4 chips at a time. Dom thanks for asking the question I kind of half given up on it do to frustration. Time to get after it! Thanks again gentlemen!

Tim

 

Yeah, sorry, 256Kb == 32KB. Was hoping you'd take away the idea rather than the exact numbers.

 

PS - you may consider going to flash chips and drop the UV erasor. UV erasing is so... 2001! Anyway, you might benefit from eliminating the time it takes to erase and bad erases from your life altogether.

 

I would have liked to use the AT29C512 but couldn't find any. Well it doesn't matter now. I got the chip done and the car is is someone else's garage now.
Thanks for the help.
Next its on to tuning a ZR-1. Are you game for that Mangus? RBob?

 

Sure, drop if off for a few month, err, weeks, uh, days and I'll see what I can do.

RBob.

 

"Sure, drop if off for a few month, err, weeks, uh, days and I'll see what I can do.

RBob"


Touche' Bob. Any chance EBL can run 16 injectors?

 

Not only run 16 injectors but keep the CCM Happy.

Speaking of, Hey mangus anything new on Scanner pro working with the vette and CCM? I have not had any luck

 

[hijack]Well, it seems to work great with 94/95 LT1s, so it should work fine with your Vette. It's all about the definition, though.

There's a new build up on the ScannerPro site.[/hijack]

 

Is there a way to put a bin file to chip ($8d) on a 27c512 eprom? I burn it its fine accept wont work in the ecu. I thought maybe this split/stack thing is what i need to do.

 

yes, you can burn a single bin to it if you get the offsets right.

or, open the stacker, set bin size to 32kb, set the chip size, set the block size to 32. select the bin files, i prefer to fill all positions. select an output file for the stacked bin & stack them.
now go to where you burn the chip & select the chip, load the stacked file into the buffer & burn.