Legacy 27C256 eprom programmer

Tracy Allen

We have a need to update firmware on legacy systems that use the parallel 27C256, but our old eprom programmer doesn't have the settings specified for eproms now on the market. The new ones from ATMEL have a much faster programming cycle, for example, 100 microsecond pulses up to 10x, but the minimum setting on our old programmer is 2.5 millisecond pulses 7x. Also the Vpp and Vcc voltages are wrong. Does anyone have a recommendation for a cheap programmer? I'd hate to lay out $$$ for a universal programmer, when all we have to do is this one chip. The other stuff we need to program is covered by serial programming.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Phil Pilgrim (PhiPi)

Tracy,

I've got a windowed AMD AM27C256-250DC you can have, if you want it.

-Phil

Tracy Allen

Thanks Phil. We have the chips already in SOIC, mounted on adapters that permute the pins for use in the weird Tandy model 102 option ROM socket. We need to program about 20 of them. It didn't occur to me, but it probably should have, that the timing parameters have evolved so much.

Hang onto it! Those windowed chips can be made into excellent refrigerator magnets!

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com

Mike Green

What a great project for a Propeller! It's got enough I/O pins. You could use transparent latches that can do any level shifting needed. The Prop can even do the voltage generation for Vpp and Vcc using PWM and simple ADCs. I think Paul Baker posted something like that way back when (Prop driven boost voltage generator). If you use a Protoboard, you've even got an extra 32K bytes of EEPROM to store the data to be programmed.

Phil Pilgrim (PhiPi)

Migosh, those Tandy 100s/102s never die, do they? I've still got a Model 100 that I used for setting up and debugging fruit sizers more than twenty years ago. Later, it was tasked with driving a Sherline mill to cut out eyes for fishing lures. It's in the PhiPi museum now, but I'll bet it still runs — assuming I didn't leave batteries in it that've long since corroded.

So, Tracy, do you still remember how to program an 8085?

-Phil

pwillard

You know... I'be be interested in a DIY Eprom programmer for legacy Eproms as a project. At this stage I'm only good at the Voltage level shifter stuff and not the programming algorithms or I'd already be working on it. Anyone done this?

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔

---

There's nothing like a new idea and a warm soldering iron.

kjennejohn

Perhaps I can help. I have a Xeltek super 3000U, which is stated to program over 40,000 different parts. Tell me the part number for your device and I'll check to see if it's on the list. Or go to the Xeltek site and check for yourself. If it's a go, then send me the parts and the source and I'll burn them and send them back, no charge, except the cost of mailing them back. Of course these have to be insertable into a DIP-style ZIF socket.

Then again, if you will be doing these on a regular basis and want to keep your parts and programmer in house, let me suggest EETools(.com). Their ChipMax ($345) can burn 5000 devices. Runs under Windows, uses the parallel port. Check to see if your device is supported.

My email is kjennejohn@yahoo.com, if interested.
kenjj

Beau Schwabe

Tracy Allen,
·
Seems that the most difficult part is controlling the Vpp signal.· If you use a Propeller, the schematic below combined with 26 I/O's ( 15 Address lines ; 8 Output lines ; 3 Control lines - CE,OE, and Vpp Control)·should·be all of the discrete hardware·required to program a 27C256.· The rest of it would simply be·software in which case the Propeller would be quite capable.
·
·
·
Note: The schematic·indicates a 0V or 5V input signal to the "Vpp Control" line....··A 3.3V input would work just as well here.
······· According to the datasheet, the minimum input voltage·for a logic HIGH is 2V, so 3.3V from the Propeller should work without any level shifting.·
······· ( Don't forget current limiting resistors on all of the EPROM Outputs·... O0-O7 )
·


▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Beau Schwabe

IC Layout Engineer
Parallax, Inc.

Chris Savage

Phil Pilgrim (PhiPi) said...(trimmed)
So, Tracy, do you still remember how to program an 8085?

Phil,

··· Nice...You had me sitting here wondering if someone dropped a 6502, 6809, 680x0 or Z80 in front of me, would I still be able to program it?· Since the Z80 was the last thing I used I feel pretty comfortable saying I would.· Now what that has to do with the 8085 (for those that don't know) is that the Z80 used a superset of the 8085 instruction set.· I never did get used to the 8085 multiplexed data/address bus though...Take care.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Chris Savage
Parallax Tech Support

Phil Pilgrim (PhiPi)

Tracy,

Another option is this: If your PROM programmer can generate the correct programming voltages (Vpp = 13V, Vdd = 6.5V), perhaps you could insert an edge-triggered one-shot that shortens the /CE programming pulse to 100µS from the programmer's 2.5mS. Your programmer's seven reps will likely be enough to get a solid verify. I'm not sure whether /CE is also used for verification during programming. It almost certainly is during initial erasure test and final verification and is probably held low. In that case, your one-shot would have to be removed for those checks, since the chip will fail them if the circuit is present. Most programmers will let you perform all these steps independently, though.

-Phil

Tracy Allen

Thanks for all the suggestions.

This programmer is for my friend, Rick Hanson, who operates Club 100. He has an ongoing evening business in buying and selling Tandy 100, 102 and 200s, and in providing add-on ROMs. Yes, amazing, there are still a lot of them out there running printing presses and laundromats and the in the cruisers of small police departments in the desks of a certain breed of news reporters and in classrooms. I provide Rick with the OTP soic eproms mounted on a carrier that fits into the option ROM socket. Not only is it a proprietary deep socket, but in their great wisdom Rat Shack switched the pin connections so that it doesn't correspond to the standard 27C256. We use a bridge from the programmer to an SOIC clamp to program the chip, and the large plated through holes on the edge of the circuit board are cut through to expose 28 half-moon contacts that press down into the option socket.

Back in its heyday, I made a different design of option ROM adapter called ROMBO that I sold a lot of through the magazines. It used a special circuit board material called "bendflex" that has since been discontinued. And I also manufactured RAM expansions called the "extram" and the "XR4" and the RAMpak, and provided operating system extensions to used the additional banked memory. The T102 was also at the heart of my original OWL data loggers. At the time I could spin BC, DE and HL pretty fluently, a hazy hazy memory now.

I like the idea of a Prop based programmer. That makes a lot of sense. But I know I'm not going to have time to build it, and I'm sure Rick wants a plug and play that he can hook up to his PC. Maybe he'll have to shell out the $300 or $400, since he wants to continue with this.

The part is the ATMEL AT27C256R-70RU. The programming specs suggest CE pulses of 100 microseconds +/- 5 microseconds at up to 10 repeats, and Vcc=6.5 volts and Vpp=13.0 volts, both +/-0.25 volts. I'm not sure how bad it can be to stray from those parameters, but Rick sent me back a batch of chips that failed in programming, and I'm pretty sure it must have to do with his old programmer that was delivering 25 millisecond CE pulses with up to 3 repeats, with Vdd=6.0 volts and Vcc=12.5 volts. I'm going to get ahold of that programmer later on to see if maybe we can do a little hack to make it work.

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Tracy Allen
www.emesystems.com