EPROM programmer recommendations

[dgl]

Hi Mike,

Willem programmers have been mentioned, but willem dot org looks like it was abandoned and picked up by a squatter making money from searches.

Willem passed away a few years ago.

Willem programmers are inexpensive and more geared towards hobbyists who know what they're doing; but that doesn't mean the more expensive "professional" programmers are necessarily better. I found a couple instances where my old Stag programmer wasn't using the correct algorithms or voltages.

There are a couple main Willem variants.

The original versions didn't have a step-up regulator, so you needed to use an appropriate supply. I prefer this variant because it is more reliable for programming older EPROMs that require VPP > 12.5V. The last version from Willem, the 4.1, was the best for NMOS EPROMs (2716, 2732, ...) because it had an optional relay to address the voltage drop from devices that drew more than about 30mA from VCC.

The other variant is PCB3, which introduced the step-up regulator so the programmer could operate from 9-10VDC. A later modification was to add an option to power via a USB port. This variant has all of the problems associated with the original version plus issues with unstable VPP if an improper inductor was used with the step-up regulator. Most variants that I've seen used an inexpensive molded inductor that saturated at a low current.

Later PCB3 variants improved programming speed by changing the way the address was generated and data was read. Those changes have been reverse engineered and any Willem programmer can be modified to use the latest software from Sivava.

Willem was working on a USB version before he passed away that would have been superior to the versions sold by Mcumall. Unfortunately, he never got around to sending me an early version...


The Willem 4.1 seems to be hard to find these days. There was a Chinese company that appeared to have copied the design, but their boards seemed to have problems. I bought a bunch of boards from Willem and still have a few if you're the type who enjoys building your own. Willem's programmers were always more expensive than his competitors. As I recall, he was selling the assembled 4.1 for about $60; which was twice what the other variants were selling for at the time.

Regards,
Dennis

 

[mbbrutman]

Hi Dennis,

Thanks for the background on Willem and welcome to the forums!

I've had the luxury of time and have not made a decision on a purchase yet. The project that I was working on (XT-IDE for a PCjr) uses an EEPROM, and I was lucky enough not to completely screw the code up and brick the machine. I still expect to write some bad code, so I need a burner.

I also have a variety of ancient equipment ranging from 1979 vintage standup video games to mid 80s and 90s PC hardware - having the ability to read and write EPROMs with the correct tool would make the maintenance easier.

Of the two modern Willem derived designs, which do you consider to be better constructed and supported? I realize that for under $100 I'm not getting a lifetime support promise, but I would like to get a few years of use out of it. The MCUMall version appeals to me because it's Canadian. I've not looked at closely as the Sivava as somebody on CCTalk reported that it wasn't terribly compatible with existing Willem based software.

And of course the other option that ChuckG likes are the Genius programmers, which I'm also considering.


Regards,
Mike

 

[dabone]

What do you find bad about the GQ-4x Usb programmer? It's done everything I've asked of it. (Including lots of eproms for my commodores, cocos, ide-xt.) I do wish it could do gals.


Later,
dabone

 

[dgl]

Hi Mike,

Of the two modern Willem derived designs, which do you consider to be better constructed and supported? I realize that for under $100 I'm not getting a lifetime support promise, but I would like to get a few years of use out of it. The MCUMall version appeals to me because it's Canadian. I've not looked at closely as the Sivava as somebody on CCTalk reported that it wasn't terribly compatible with existing Willem based software.

I would only suggest designs using the original linear regulators. ATH3.1SS is a real DIY programmer; you make the board yourself. As I mentioned earlier, I think the Willem 4.1 with the optional relay is the best for NMOS EPROMs. I would change the power on all Willem variants to provide more VCC options. 5.0V and 5.6V are common options, some also provide 6.2V, and even fewer provide a 4.2V option. To program, you need 4.2, 4.5, 5.0, 5.5, 6.0, 6.25. You should blank check at < 4.5 volts, but high enough that the device can be expected to function; I use 4.45V. You should program verify at 6.0-6.25V. Some devices need to program at VCC=6-6.25V. That's one of the problems with Willem programmers; Users need to know this. I've modified most of my Willem type programmers to give 2 fixed VCC options and 1 variable 2.5-6.5V. The Willem 4.1 is the only Willem variant that can do this properly. It has a separate 5V regulator for interfacing to the parallel port and takes care of level shifting, albeit in a simple manner; and the inputs to buffers driving the parallel port should have pull-up resistors to avoid problems when running at VCC < 4.5V. Since the 4.1 has a relay and I have a variable VCC option, I can set the voltage to within 0.01V of what I want (relay contact has some resistance).

I don't have much confidence in either Mcumall or Sivava. I've tried to work with both of them to improve their designs and software (which is now maintained only by Sivava).

Before Mcumall developed their "true USB" programmer, they were selling a PCB3 variant called the "enhanced universal Willem". That design had so many design flaws that it wasn't suitable to program anything. It was designed by someone, or a group of people, who knew nothing about analog design. They actually used the output of an opamp to provide VCC to the devices being programmed. Anyone who has used opamps knows an LM324 maxes out at around 25mA. All NMOS devices will draw more than that and many of the larger capacity CMOS FLASH draw more. When I pointed out the deficiencies to them, they said that that product served a particular market and they would sell it as long as there was demand.

Mcumall designed their USB programmer, but I didn't consider it because Willem was designing his own and was addressing deficiencies in the Mcumall design.

I also contacted Sivava regarding the power problems with their design and deficiencies/bugs in the software, which they made a closed solution based on the open source solution Willem worked on. There was a language issue, but basically they weren't interested.

You can't always believe what you read on the net; it was likely a User issue. Sivava's programmers are the most compatible with the latest software because they're the only maintainer.

And of course the other option that ChuckG likes are the Genius programmers, which I'm also considering.

My needs are modest and I don't care to spend the money on a programmer that doesn't provide schematics.

Regards,
Dennis

 

[dgl]

Hi Dabone,

What do you find bad about the GQ-4x Usb programmer? It's done everything I've asked of it. (Including lots of eproms for my commodores, cocos, ide-xt.) I do wish it could do gals.

I don't have anything good or bad to say about Mcumall's USB programmers. I stopped looking at their products after their second true USB version; the changes were coming too quickly when meant they were releasing product too soon.

If they don't have an external power supply option and tell Users to use it for NMOS devices, that would be a negative. Most NMOS devices will draw about 100mA from VCC. That's the maximum that USB devices are allowed to draw without negotiating for up to 500mA. If the request is refused, the device is required to operate with 100mA. Since that isn't possible with most NMOS devices, an external supply would be required. Even Willem's initial USB programmer design was a USB decoration (not intelligent).

Many of the people I've encountered don't know much about what it takes to program an EPROM reliably. And they tend to make blanket statements like "always worked for me". There's a difference between programming an EPROM and programming one reliably.

Regards,
Dennis

 

[Chuck(G)]

The problem with EPROM programmers in general is that they're ancient technology. Flash has replaced UVEPROMs and many new designs are JTAG- or ISP-programmable. Back as recently as the mid-1990s, you could pick up very decent low-quantity programmers for not a lot of money. But most of those firms have either disappeared or no longer support those devices. I have a programmer from M2L; they offered lots of adapters and the software was easy to use. Around 2004, they simply disappeared from the face of the earth.

Most contemporary programmers don't have a lot of smarts, unless they employ a microprocessor to control things and can operate as standalone devices. So, at least for EPROMs, I don't care much about schematics--some latches or counters mostly, a pulse generator and a power supply (usually a bost supply) to provide Vpp and a bunch of "glue". Clean, well-documented, software in the form of source code is far more useful. I've been thinking about disassembling my M2L software (it's text-mode DOS, so it's not full of GUI glarf). Heck, I wonder if I could rework my old Sunshine gang programmer software to work on a more modern system than a PC XT (too many CPU timing loops).

I got the Genius to program GALs, of which I have several tubes. And I've found that I don't use GALs much at all--CPLDs are so much eaiser to work with.

And nothing you buy today new will do 2708s or 1702s.

 

[dgl]

Hi Chuck,

The problem with EPROM programmers in general is that they're ancient technology. Flash has replaced UVEPROMs and many new designs are JTAG- or ISP-programmable.

There's still a lot of equipment out there that uses EPROMs and properly programmed EPROMs were only spec'ed to hold their charge for 20 years. Whenever I have the opportunity, I copy the data from EPROMs in my equipment. I have several pieces of Tektronix equipment that have EPROMs and they're all older than 20 years. Charge loss will become a big problem for well built equipment that's still being used after 20+ years.

My BMW is over 20 years old and I intend to drive it for many more years. I replaced the original EPROM in the ECU with a tuned version a few years ago and will back up the data the next time I take the ECU apart; or maybe I'll ask the person I bought the tuned chip from if he'll give it to me...

FWIW, Willem programmers will also do EEPROM and NOR FLASH.

Regards,
Dennis

 

[RetroHacker_]

There's still a lot of equipment out there that uses EPROMs and properly programmed EPROMs were only spec'ed to hold their charge for 20 years.

Oh, yes - there most certainally is a large base of installed stuff that uses that sort of chip. It's just that the technology is obsolete from a "current generation" view, and therefore nobody is going to invest the time and effort into building quality equipment to program them. Hence why we're stuck with this Willem garbage.

EPROMs are also widely used in the amusement industy - in coin operated video games, slot machines, soda machines, etc. But brand new machines have gone to using modified PC's, loading data from a hard disk (or CF card). You'd be hard pressed to find an EPROM in the current crop of machines (except, I do believe pinball machines probably still use them).

Best bet, buy an older, quality device programmer. You'll have less headaches.

-Ian

 

[gslick]

Best bet, buy an older, quality device programmer. You'll have less headaches.

That is what I would recommend as well. Anyone interested in a working BP Microsystems BP-1140 for $50 + shipping from Seattle? The last software release was in October 2003 so anything newer than that is not supported. You need a system running Windows 95/98 or Windows 2000 with an LPT port to run the software. Programs EPROMs, FLASH, and some MCUs. Does not do PALs. I'd have to fire up the software to check the device list for any specific devices. I also have PLCC-28 and PLCC-32 adapters for it, maybe $25 each for those? Since I have a BP-1400 I don't need the BP-1140. I also have a same vintage CP-1128 which will do PALs and EPROMs up to 28-pins. I could list these on eBay but that's too much work.

-Glen

 

[Chuck(G)]

Oh, yes - there most certainally is a large base of installed stuff that uses that sort of chip. It's just that the technology is obsolete from a "current generation" view, and therefore nobody is going to invest the time and effort into building quality equipment to program them. Hence why we're stuck with this Willem garbage.

Thanks, Ian--you said it much better than I did.

I wonder (I haven't done any research) how practical it would be to develop a JTAG-programmable 28-pin Flash replacement for common (2764-27256) UVEPROMs?

 

[dgl]

Hi,

Hence why we're stuck with this Willem garbage.

When used by a knowledgeable person, Willem programmers do quite well... It's much easier than building a custom circuit to program; as I did for my first PROMs and EPROMs. The only "problem" is that you need to know what you're doing and not an insignificant number of people programming EPROMs are unaware of their ignorance...

Regards,
Dennis

 

[dgl]

Hi Glen,

Anyone interested in a working BP Microsystems BP-1140 for $50 + shipping from Seattle?

I sent you a PM.

Regards,
Dennis

 

[mbbrutman]

Hi,
When used by a knowledgeable person, Willem programmers do quite well... It's much easier than building a custom circuit to program; as I did for my first PROMs and EPROMs. The only "problem" is that you need to know what you're doing and not an insignificant number of people programming EPROMs are unaware of their ignorance...

Regards,
Dennis

That might be a fair criticism, but a lot of the audience for this kind of device is composed of people like myself - software types who can somewhat read a schematic, but are not EEs by trade. Hence my search for something I can use without getting a EE degree.

Engineers seem to be overly critical - based on what I'm reading in this thread they all stink. ;-0

 

[RetroHacker_]

Hi,
When used by a knowledgeable person, Willem programmers do quite well... It's much easier than building a custom circuit to program; as I did for my first PROMs and EPROMs. The only "problem" is that you need to know what you're doing and not an insignificant number of people programming EPROMs are unaware of their ignorance...

Oh, it's not that I don't know what I'm doing. I know EPROM technology very well. It's just that I don't want to have to mess with it. Why should I spend money on a programmer that has to be modified in order to function properly? A programmer that is, as shipped, unusable for reliable operation? That's a bad tool, right there. Sure, I've built a programmer from scratch - it's practically a rite of passage. But once I started using a well-designed programmer (Needham's), I didn't use my crummy little counter circuit anymore. Why? Because the Needham's does the job better, and I don't have to manually change the VPP voltage, or change timing loops in software to get it to program a different device.

Now, there may be a class of devices that the Willem is fine for. But it's not flexible, you have to manually change jumpers and fiddle with badly written (and stuck to Windows) software. If all you need it for is burining common 27C512's or something, then sure, it'll be great. But if you do what I do, and work on a large variety of old computers, coin operated video games, and all manner of other EPROM-using devices, you want something a little more versatile.

The Willem is, as shipped, a bad design. You said it yourself. And when you start using a tool regularly, you don't want to put up with an inferior tool. A mechanic does not use dollar store wrenches, a carpenter doesn't hammer nails by hitting them with a rock, and I don't want to program EPROMs with a Willem.

-Ian

 

[dgl]

Hi Ian,

Why should I spend money on a programmer that has to be modified in order to function properly? A programmer that is, as shipped, unusable for reliable operation? That's a bad tool, right there.

I downloaded a manual for a BP Microsystems programmer to see if it would give me an idea of whether it was, as you put it, functioning properly.

The manual indicated that blank checking was done at 4.7V. This voltage is incorrect. Some EPROMs are spec'ed for 5V +/-10% (4.5-5.5V) operation. To properly blank check such a device, VCC would have to be less than 4.5V. Many people overlook the fact that proper EPROM programming must begin with a device that's been erased properly. Manufacturers know this. I think I read that Intel programmed and erased 100% of their EPROMs before shipping so they could screen for stubborn bits.

Does this mean that the BP Microsystems programmers are a bad design? Probably not, but my statement of not placing blind trust in a piece of equipment just because it was made by a reputable company still applies. What I can say is that you can't depend on it to blank check correctly.

Do you know what voltages your Needham's blank checks and programs at? Or do you just assume it's doing it correctly? With a Willem, more information is available to you if you choose to use it.

Regards,
Dennis

 

[dgl]

Engineers seem to be overly critical - based on what I'm reading in this thread they all stink. ;-0

Must be quite some stench if it can be transmitted by electrons that didn't even touch them...

Good engineers tend to be very detail oriented... Not all engineers are very detail oriented, ergo not all engineers are good...

Regards,
Dennis

 

[Shadow Lord]

I like the price range on the Genius, but I have to admit that when I saw the web site I was immediately concerned about usability and ongoing support. My needs are fairly basic so I'm sure it would be ok.

MCU Mall (http://www.mcumall.com/comersus/store/comersus_dynamicIndex.asp) has some Willem derivatives that look to be nicely done that I'm also considering.

My project is adapting the XT-IDE BIOS for the PCjr. If I was really up the creek I could get by with just the bare XTIDE card and the flash utility, but I want something a little more elegant that I can use for other projects too. The ultimate toy would be a ROM simulator, but those start at $300 and go up from there.

I purchased a Genius based on Chuck(G)'s advice about 4 months ago. Heck, I almost did a double take of your post as it was almost identical to mine . Anyhow, this was my first programmer and I was able to get it going w/o to many issues under XP on a P4. Make sure to dl the latest SW for it. Also when buying on e-bay they all work out to be the smae price (more or less). I was able to program a new BIOS for my Adaptec card w/o too much trouble.

 

[carlsson]

I don't have much confidence in [..] Mcumall

Say what you want about the technical details, but I must give them tons of credit in one field. I ordered a GQ-4X on March 31, and it arrived at my post box eight days later including overseas shipping. Furthermore, without asking them, they had checked the gift box and declared a $10 product value for a $95 unit. You may have opinions on the good and bad practises to cheat local governments from getting bonus tax money, but as the customer I'm very happy not having to pay another ~$50 to my tax office just because I'm importing an item not possible to buy in my home country.

 

[atod]

Does anyone know what version of the Willem programmer this is? I purchased it a few years ago and am unable to find the associated software for it. Thanks!

 

[channelmaniac]

Man... those Willem units are all temperamental beasts. People either love them or absolutely hate them.

I have a TopMax by eeTools as my main programmer. It'll do 2716, 8516, 28xx EEPROMs, Motorola 68764/68766, 803x/4x/5x and 873x/4x/5x MCUs, a large # of bipolar PROMs, many 82S series chips - including 82S100/PLS100 PLAs, and a LOT more... All WITHOUT having to mess with adapters, dipswitches, etc.

Only downsides - requires a parallel port and Win XP is the latest OS supported.

It's a rock solid programmer and my main programmer. I picked up a NOS unit off of eBay for a spare since I use this thing so much.

For the real obscure stuff I have a Data I/O Series 22 with the 064A PROM module. It'll do all kinds of odd stuff.

The only things I can think of offhand I can't program are the 2804 and 1702A EPROMs (and the NS EPROMs of that 1702 era...)

The Series 22 will even handle the pesky TMS2716 tri-voltage EPROMs that most everything else can't.