Well, we all have differing levels of electronics knowledge ... and net worth... which tip the calculus one way or the other depending on the one's personal situation... As much as I do wish money could be no object, that's not realistic for me. Best, Sean On Tue, Sep 8, 2015 at 5:39 PM, Jay West <jwest at classiccmp.org> wrote:

Sean wrote... ------ my bench is all Racal, Tek, HP, Fluke, etc and the ------ You must be one of those high net worth folks you mentioned in your previous post, I'm jealous of that kind of bench gear! I'm not saying the CP-1128 isn't a great value. Of course it's a good solution! I am merely saying the Data I/O 29B currently and has been a good value (in response to your post saying it was "too expensive for 30+ year old gear"). But since it was brought up, here's some things to consider.... The Data I/O 29B is specifically sought out by arcade enthusiasts (and many vintage computer collectors that I know) because it supports an unusually wide array of vintage chips, most that we're likely to come across. Why is that important, if the CP-1128 does - as glen points out - program a subset of the needed (for this specific case) chips? Answer from past experience.... There may be a dozen chips that will work for the HP 1K loader roms. What I have found over the past many years is that the supply ebbs and flows. Good example - for about a year, MMI 6301's were unobtainium. All the suppliers I could find were all pushing the Harris 7611's, because they couldn't get the 6301's either. Then a year later, apparently a batch of fuji 7114's were found and those were available, because suppliers couldn't find any more 7611's (at the time). Now I'm finding the MMI 6301's again. Yes, these chips are all compatible once programmed, but some programmers will do one and not the other. This "ebb and flow" in supply may occur right at the least convenient time, like when your 2000/Access IOP firmware chips die (me, about two months ago). If your programmer supports the widest possible selection of chips, even post-programmed compatibles, you're in a better spot. If not, a post-programmed compatible chip may well be available, but your programmer doesn't support it. Not a great spot. My 29B will also do a lot more different types of devices than usual, thanks to the plethora of other adapters (not just unipak 2B) that came with mine. Another good example - the CP-1128 appears to require a usb attached computer to function. While I usually connect my 29B to a computer to program (via serial), there are a few occasions when I've been exceedingly glad that it can be fully operated from the built in control panel without having to have a computer to run it. Another example - Some manufacturers, over time, drop support for very early devices. That can't happen with the 29B ;) It sure can and does happen with current devices being manufactured. There's a downside of course... just try finding replacement parts for a 29b ;) Fortunately, schematics exist. Long story short, as I said before, there's a reason that programmer still holds a high value. For some people, those reasons are worth it. For others, they are not. Is the 29B "the best" programmer? Probably not, as "best" is both situational and subjective. I'm willing to allocate portions of my precious small bench gear budget for the above reasons. But several folks asked at the same time for advice on 21MX M/E/F proms and programmers, and that post was my advice. There are certainly other great programmers out there. Best, J

Glen wrote... ---- Just in case anyone is looking at BP Microsystems programmers: ---- I'm looking - heck, at that price, a good backup/insurance policy! Would love to see a full device list.... J

I must be mis-remembering how I designed my first EPROM programmer. I know I could do 2716s (single-rail, not the TI ones) and I thought I always applied Vpp and pulsed another pin at TTL levels to program them IIRC the 27128 was the last one to actually allow the 'dumb' 50ms programming algorithm. It may well work on larger chips, but I never tried it. That's why I designed my programmer to do 2716s, 2732s, 2764s and 27128s only. Not having another programmer I couldn't use a microprocessor or microcontroller (no way to load the firmware) so I had to do it all in TTL logic. Doing the 'intellegent' algorithm in TTL was a bit much ;-) -tony

The data sheet for the SGS-Thomson 2716 here : http://ee-classes.usc.edu/ee459/library/datasheets/2716.pdf says that you apply Vpp and keep it applied, set up address and data (with G/, pin 20, more commonly called output enable, high and E/P low) and then program each location by taking the E/P pin (Pin 18, more commonly called chip enable) high for 50ms. That is what I remember doing. You didn't have to pulse Vpp. -tony

Yep, looked it up. 2716 and 2732 used a pulsed Vpp( 2708 a well ). Beyond that they had a static Vpp and a PGM pin. Dwight

I am now totally confused. Do you believe the 2716 and 2732 require a Vpp pulse for each location or not? FWIW, I believe they do not (confirmed by the data sheets I have found on the web). You can apply Vpp to get the thing into program mode and then apply a TTL-level 50ms pulse to program a location. -tony

Anyone interested in doing 1702As should look at the schematic in the MCS4 user manual on bitsavers.org There is a schematic for a MP7-03 board. It can be easily controlled can easily be replaced by software. You mainly need the level converters. Dwight From: dkelvey at hotmail.com To: cctalk at classiccmp.org Subject: RE: 21MX proms (per request Date: Thu, 10 Sep 2015 07:03:53 -0700 I would be surprised if it did 1702s and 2708s. Even 2716s have a slightly different hardware configuration. 1702s require voltages like 50 volts ( I could be a little off ) and changing address and data lines over wide voltage ranges. 2708s require multiple rail voltages as well as programming voltages. 2716 are that last of the program by turning the programming voltage on and off. All the newer chips have a constant programming voltage and a separate program logic level pin. They can even be read while the programming voltage is high. I doubt any would make the special provisions for 1702s. 2708s are a stretch but if your not doing 2716s it won't be able to handle 2708 even if it had the extra rail voltages. I have a DataIO ( I think a 29B ) and I know it doesn't do 1702s. In any case, I don't think any of these would be fast enough for a 21MX boot code. I think these expect it to be faster than 100ns. Dwight

On 2015-Sep-08, at 2:39 PM, Jay West wrote: Alternatively, for the rare instance of need, burning fusible PROMs generally or often isn't very complex. When I needed to burn a replacement 74S387 (256*4) for a repair awhile ago, I hacked up a programmer on a breadboard, controlled by a program running on an RPi. The hardware was an LM317, 7805, 3 TTL ICs, 5 transistors, and a few R/C/D. Add 4 more transistors for 8-bits.

On 9/8/2015 9:38 PM, Glen Slick wrote: Anybody know if the CP-1128 supports any of the older EProms (2708, 2716 or 1702)? I couldn't find a device list online. JRJ