Programming 27C256B problems

By Gregory

Expert (110)

Gregory's picture

03-05-2015, 20:30

For the people who burn their own eproms is their anyone who uses the 27C256B.
My problem is that I have some 27C256B 's lying around but my eprom programmer can't program this specific type (I think this version is a newer version of the older 27C256 with different VCC and programming pulses during the write process).

I have also some 27C512 and I can burn those just fine.
So if anyone uses the 27C256B can you tell me what programmer you use to burn the eproms.

So in the meanwhile I'll replace the 27C256 by a 27C512.
I know that I then have to start the programming at &h8000 but are their any other modifications needed?

Login or register to post comments

By RetroTechie

Paragon (1563)

RetroTechie's picture

03-05-2015, 23:38

What make/model eprom programmer are you using? And exactly what manufacturer 27C256B's are we talking here?

I'd be surprised if a properly designed programming algorithm for a 27C256(or -A) wouldn't work with a 27C256B. Doesn't the programmer have an option to manually specify the IC type, or force it to use a specific programming procedure?

As for the '512s: just prepare contents the exact same way like you would for a 27(C)256. Then binary copy 2x the 32 KB data into a 64 KB file, and write that into the 27C512. That way the highest address line becomes a "don't care", and all other signals retain the same function as for a 27C256.

By Gregory

Expert (110)

Gregory's picture

04-05-2015, 18:33

* The manufacturer of the 27C256B I have is ST microelectronics.
http://www.st.com/web/en/resource/technical/document/datasheet/CD00000515.pdf

The programmer I use is one that was published in elektuur march 1997.
ftp://54685ff3.cm-12-1b.dynamic.ziggo.nl/elektuur/1990/pdf/1997/pdf-401-38.pdf

The software that comes with this eprom programmer only allows me to select the type of eprom (27c64, 27c128, 27c256, 27c512), I can't change anything else. The problem I think is with the constant VCC of +5V, the datasheet of the eprom says that it should be 6,25V (+-0,25 V) during the programming of the eprom and the datasheet also says that the program pulses should be 100us and my programmer uses 1 ms.
The programming voltage I have is okay (12,55 Volt wich is within the specs of the datasheet =12,75 -0.25V).
I don't know if the duration of the program-pulses is a problem.

*Furthermore, a read (verify) of the eproms also requires 6,25V +- 0.25V and here I get also into troubles, because the eprom gives inconsistent results when reading the eprom.

By RetroTechie

Paragon (1563)

RetroTechie's picture

04-05-2015, 21:45

If I compare pinouts between the supported EPROMs, most differences appear with the use of pin 20 (/CE, or /PGM), and pin 22 (/OE, or Vpp). But 27(C)512 would be the most complex case for those pins' use, and that seems to be working.

But there is another difference: pin 1. A 27(C)512 has A15 on that pin, the other types have Vpp on that pin. So just to rule some things out, could you verify that:

  • When programming starts for a 27C256, pin 1 actually sees Vpp (12.5 or 12.75V, whatever) ? Might be difficult to say for sure if programming fails immediately, though. Perhaps you could attach a LED + zener (+resistor) directly to that pin, such that you'll see a short 'blip' if >10V or so goes on that pin, even if for a moment?
  • Both 27(C)64 and 27(C)128 can be programmed successfully?

If you don't see the 1st, or 27C64/27C128 programming fails too, then problem might be what programmer does with pin 1.

By Gregory

Expert (110)

Gregory's picture

04-05-2015, 22:11

Problem is that I don't have a 27c64 or a 27c128 at hand, so I can't verify if they program correctly.
I'll try to check the first item though, to see if pin 1 gets Vpp while programming(will be tomorrow).

By Gregory

Expert (110)

Gregory's picture

05-05-2015, 19:16

I just had some time to measure Vpp when the programming of the 27C256 starts and pin 1 actually sees 12,5 V.
What I noticed just know is that only the first bytes get written and then the programmers quits with the error messag. Tried it on several and same thing, only the first byte gets written.

By RetroTechie

Paragon (1563)

RetroTechie's picture

06-05-2015, 00:37

So some bytes DO get written? oO That means a blank check fails after you've tried a write attempt?

Could be a data integrity issue perhaps, a data or address line making intermittent contact, etc. I'd definitely get some more different types of EPROM (27(C)64, 27(C)128, 27C256 without -A or -B, 27512) to get a clearer picture of what's going on. Do you have a way to read these EPROMs outside the programmer? A few other things to try:

  • Replace electrolytic capacitors on the board (I see only two? 10 uF can be had in ceramic these days, Smile2 tantalum would also be a good choice here).
  • Where IC sockets are used: remove & re-socket IC's.
  • Try with a shorter parallel cable.
  • (100% !) Verify that reading EPROMs with known contents produces the correct data.
  • If reading works, but data not 100% same as what's in the EPROM: have a good look at how the read data is wrong.
  • Same if writing works to some (read: any!) extent.
  • If possible, reduce speed of the PC running the programmer software (some old Win9x PC?). I see there's some timing element on the board itself so that's good, but doesn't necessarily mean PC speed isn't an issue.
  • Figure out a way to (manually?) bump the EPROM's Vcc for the duration of a write attempt.

By Gregory

Expert (110)

Gregory's picture

06-05-2015, 18:54

I tried two things today:

- I increased Vcc to 6 V during writing but still the same error message.

- I decreased the width of the programming pulse. If I'm not mistaken the pulse width is given by R11 X C5. Wich is approximately 1 millisecond acording to the schematic. I changed R11 from 12 k to 1 k which should give a pulse around 100 uS (wich should be just fine acording to the datasheet) but the problem still remains.

With these changes I'm still able to write the 27C512 without any problems but the 27C256B is giving me a lot of problems.

One thing I noticed is that when erase the 27c256, so there's FF everywhere in the eprom, if I then write a Rom full of FF's to the 27c256, then I don't get an error, the program writes to all the adresses in the eprom. But on the other hand that isn't really writing is it, when de program writes a '0' to a certain adress then the eprom gets written...

I have a couple of 27c64 and 27c128 comming my way, so i'll give them a try and see what I get.

By raymond

Champion (441)

raymond's picture

04-07-2020, 09:02

Did you get this working? I have recently bought an Elektuur programmer, I can see the eprom is empty, however programming it is not successfull yet...