Burning EPROMS

[dphirschler]

I seem to be having problems burning EEPROMs. I can’t even read them properly. Do they need to be erased first? Or is there some configuration mode I need to be in? Please help.

 

I have the MiniPRO EPROM burner, and "10PCS AM27C040 -150 27C040 EPROM 4M 27C040-150DC,B12" chips.

 

 

Darryl

[+Ksarul]

It is always a very good idea to erase the chip before you try to program it, as you never know which bits on it happen to be set incorrectly (and there will always be some). You will need an UV light to do that--the more powerful it is, the less time you need. The little cheap Chinese ones take about 30 minutes, while the industrial strength one I have takes about 10 minutes. . .but there is a significant difference in price between them.

[dphirschler]

I ordered some UV LEDs, but they are coming from China and will probably take another two weeks. I plan to solder four of them in series (with a resistor) and power them from an old USB cable. So, putting them out in the sunlight isn't good enough? I can't wait for the slow boat!

 

Darryl

[+Ksarul]

Sunlight usually takes a week or two. . .and I'm not sure you'll get good results with four LEDs either--it depends on their output level and the frequency they radiate at.

[dphirschler]

So, you tellin' me I need to buy one of those Chinese erasers?

 

 

Darryl

[+Ksarul]

It really depends on what those LEDs put out--they might be enough, they might need a longer time to erase, or they might not work at all. Most UV boxes have pretty powerful lights inside of them.

[SoundGammon]

I went to a electrical supply house and got a clear uv florescent lamp, about 4' long, wired it to one of those shop lights and set it on a bench about 2" above the bottom with blocks to slide the chips under. Worked great! Don't ever LOOK at it while it's on! I later put into a separate room and closed the door until done!

[Omega-TI]

If you don't want to spend $50.00 for one of these...

 

 

Check out this interesting HOMEBREW UV Eraser

[dphirschler]

Nice! I will check that out.

 

 

Darryl

[+hloberg]

I went to a electrical supply house and got a clear uv florescent lamp, about 4' long, wired it to one of those shop lights and set it on a bench about 2" above the bottom with blocks to slide the chips under. Worked great! Don't ever LOOK at it while it's on! I later put into a separate room and closed the door until done!

how long does it take to erase?

[+Lee Stewart]

how long does it take to erase?

 

Whatever it is, it would be quicker with the platform closer to the light. The link has 3 pics and shows it set up in a file box with the chip platform only about halfway from the bottom of the box to the light that is mounted in the lid. The energy from the light falls off with the square of the distance, dontcha know.

 

...lee

[sjt]

All eprom programming software should have a blank check.Select that first

Some eproms that seem faulty can be saved sometimes but filling the chip with FF if the software has that option as well.

Then erase it and try again.

[atrax27407]

Check eBay. There are usually EPROM erasers there that are cheap and effective.

[dphirschler]

OK, I managed to get it to do a blank check successfully. But it errors out when I try to program it. I think it writes one byte and fails. Then I have to erase it again(!). What options should I check (or uncheck) when programming it? Again, these are AM27C040-150DC.

 

Darryl

[UKRetrogamer]

The Minipro definitely burns 27C040. I burned a couple this weekend for the 512K cart.

 

Have you selected the device by manufacturer then part no. from the drop-down menu? Manufacturer: AMD, Device: AM27C040@DIP32.

 

If the burn fails after one byte, try the blank check again. It may NOT have written the byte. If it fails with something along the lines of "Wrote DE, Read FF", the chances are it hasn't even been able to change a single byte.

 

A properly blanked EPROM should contain ONLY FF values.

 

If it's failing because it's verifying the first byte as FF, it's probably not actually changed anything in the EPROM due to a configuration issue.

 

Using the Blank check after a single-byte write then verify error, may save you the time it takes to re-erase the EPROM.

 

Failing this, I've attached the datasheet. Check the programming software's VPP and VCC values before telling it to program the IC.

 

According to the "Device Programming" section on page 5, VPP should be set to 12.5V, VCC should be 6.25V. Pulse Delay ("Puls Delay") should read 100us.

 

Only when you're absolutely certain you've selected the correct device from the software's drop-down menu and checked the values against the attached datasheet, you can un-check/un-tick the "Check device ID" box as this checks to see if the chip you've selected can be identified exactly.

 

Incidentally, there's a TL866 minipro-specific thread on Atariage, here: http://atariage.com/forums/topic/230688-tl866cs-minipro-eprommer/

 

Edit: The values shown in the IC config information when selecting AM27C040 appear to differ from the datasheet, with the software setting VPP to 13.5V and VCC to 5v. Try forcing the values to those shown in the datasheet (as above). If programming is then successful, save out a Project as this will contain the corrected values. Load this when programming another AM27C040 but [obviously] replace the data before programming.

 

Note: My own 27C040 is a Ti variant. I cannot check these values will work on your device, so your milage may vary but in general; if unsure, check the datasheet.

AM27C040.pdf

Edited October 20, 2014 by UKRetrogamer

[atrax27407]

If you bought the chips from a Chinese vendor, they might be counterfeit. That is, the device ID bytes won't match the manufacturer's specs. I have had good results by unchecking the "Check device ID" block with the MiniPro. The chips aren't bad - just the ID bytes.

[dphirschler]

OK, more info here... I get "ERROR! Address:0x000000 Buffer Data: 0xAA Verify: 0xFF"

 

Darryl

[UKRetrogamer]

OK, more info here... I get "ERROR! Address:0x000000 Buffer Data: 0xAA Verify: 0xFF"

 

Darryl

Verify 0xFF would suggest it's not changing anything in the EPROM, so you shouldn't need to keep erasing it after this particular failure. Especially if it fails verifying the first byte as FF.

[UKRetrogamer]

OK, more info here... I get "ERROR! Address:0x000000 Buffer Data: 0xAA Verify: 0xFF"

This is with VPP set to 12.5V, VCC reading 6.25V and Pulse Delay showing 100us?

[dphirschler]

VPP Voltage: 13.50V

VCC Verify: 5.00V

VDD Write: 6.25V

Puls Delay: 100us

 

It's the values it got from its own internal database. Should I change it to your values?

 

BTW, it fails blank check after this error.

 

Darryl

Edited October 20, 2014 by dphirschler

[UKRetrogamer]

The values I quoted came from page 5 of the manufacturer's datasheet. I attached this to my earlier post.

 

Erase the EPROM, set the values as per my last post, blank-check the EPROM and see if you get any further.

 

Edit: Just a thought. If possible, use a USB2 port rather than a USB3 one. I've read of some users experiencing problems with the programmer plugged into USB3 ports.

Edited October 20, 2014 by UKRetrogamer

[dphirschler]

Still having trouble. Here is what I am doing. See if you can find a flaw in my process:

 

- Select IC "AM27040 @DIP32".

- "Read from chip" (shows all FF's. I get the same if I "Blank check").

- File > Open. I select "Gamecart.bin".

- File format: BINARY, Load mode: Normal, "Clear buffer with default".

- I see the Gamecart binary data.

- Options: check only in "Verify after" and SetRange: ALL.

- IC Config Info: VPP 12.50V, VCC 5.00V, VDD Write: 6.25V, Puls Delay: 100us.

- If I "Read ID", it gives me "8F 08"

 

- Then I select "Programming Chip" ("program chip"?)

- Check "CODE Memory" (what is this anyway?)

- then I click "Program"

 

Then it immediately gives me "ERROR! Address:0x000000 Buffer Data: 0xAA Verify: 0xFF". Then I have to erase the chip again.

 

 

Darryl

[atrax27407]

Select the IC and uncheck the "Check device ID" block. Don't Read from Chip". After you have set the IC Config Info, go directly to Programming Chip (I assume that you have previously erased the chip). That has worked very well for me. It should program and verify.

[UKRetrogamer]

Rather than loading the binary file, use the fill option to fill the memory with the value FF.

 

I know Blank EPROMS should contain the value FF but it's something I'd try, if this were my programmer/EPROM.

 

See if you get a failure again.

 

BTW. Code memory refers to the EPROMs programmable memory area. In an EPROM, you only have one writable [ Code ] area. In other ICs (such as AVRs), the IC is divided into multiple writable areas. Code (as with the EPROM), data and config (For device configuration - write inhibit, fuse settings, etc.).

Edited October 20, 2014 by UKRetrogamer

[UKRetrogamer]

If you want help figuring out if it's the chip or programmer which is potentially at fault, I'm sure someone here would test your IC in their programmer then return it.

 

I'm in the UK but if you'd be willing to send me the IC, I can blank it here (using the same EPROM eraser I always use), program it and send it back with a copy of the correct configuration for you to try again.

 

Though I think if you were to do this, you might prefer to send it to someone closer to Georgia. Perhaps on the same landmass?

 

As I said, my own 27C040 is Ti-branded. Not because it is going into a Ti cartridge, but because it was the first I found on eBay when I looked.

 

I'm finding the minipro to be quite reliable. I'd be tempted to think there was a problem with the EPROM rather than the programmer - especially after verifying the settings against AMDs own datasheet.

Edited October 20, 2014 by UKRetrogamer