[time-nuts] HP 59309A Clock runs, sets via GPIB, but no GPIB output?
francesco.messineo at gmail.com
Mon Oct 10 04:10:38 EDT 2016
I will send you my dump, but you have to wait at least until tomorrow
evening. I'm not near my backups today.
I can confirm that 2 x 28C64 (I used ST parts, so if you can scope
your Atmels for glitch, I would be very much curious to see if they've
the same problem) worked in my 59309A, but for example the same
replacement failed in the 5328A HPIB option state machine (and worked
in the main 5328A state machine...) so it's probably a matter of luck
on the timings and the glitches.
I have already planned a CPLD replacement for these small ROMs, but
really it needs time
to get back the dumps and running each output into the program that
reconstruct the equation, then programming all the 16 equation into a
suitable CPLD, make the PCB, program it and so on...
On Mon, Oct 10, 2016 at 9:46 AM, Bob <bob at marinelli.org> wrote:
> Hi Francesco,
> That's wonderful, I hadn't dreamt that anyone has dumped out that ROM. If you can find your copy of your good 1818-2295A ROM it would let us avoid trying to fix the bit rot in the one I was able to read, especially as we don't know if the 1818-2295A exactly matches the tables in the user manual.
> Just this afternoon I ordered a couple Atmel AT28C64 EEPROMs and a 28 pin DIP header. When the parts arrive I'll try replacing U2 with them. The original part is 5v signals and the modern parts can ignore the +12 and -2 pins, and of course are erasable. As the Teensy++ easily reads the HP part, and will read the pin compatible new replacement, it should be possible to have the replacement match the bits in the manual. At least I'll not have a stuck LOAD pin :)
> Agreed about modern fast parts glitching but watching the logic analyzer, it seems the 59309A samples the ROM in a very relaxed manner using the TP2 slow clock, maybe we can get by without a CPLD.
> I'll report back after writing and testing EEPROMs*. Knowing that you had success is a great encouragement.
> Hi Paul and Don,
> Thank you both *very* much for checking the U2 part number on your clocks. That you both see 1818-2295A is terrific, it means that the ROM version in this clock matches its serial prefix.
> * The two main state machines have 8 and 53 states, and with logic analyzer watching the EEPROM it should be possible to follow along. It is a nice little computer with constant time instructions, 64 bits of RAM. Each instruction contains a conditional qualifier along with the next address, very 1960s.
>> On Oct 10, 2016, at 12:32 AM, Francesco Messineo <francesco.messineo at gmail.com> wrote:
>> Hi Dave,
>> right, once I find the dumps, I'll upload them.
>> Frank IZ8DWF
>> On Mon, Oct 10, 2016 at 9:24 AM, Artek Manuals <Manuals at artekmanuals.com> wrote:
>>> One of the best places to store ROM dumps for easy access by everyon is
>>> dit dit
>>> On 10/10/2016 3:20 AM, Francesco Messineo wrote:
>>>> I have a dump of the 1818-2295A somewhere, it should be archived in
>>>> one of my backups. I also made a replacement with a board having 2 x
>>>> 28C64 SO-28 eeproms and it worked in my 59309A as far as I could test
>>>> it. However these eeproms present many glitches on the outputs during
>>>> address toggling, so it's way better to use a suitable CPLD after
>>>> recovering the equations (I'm a bit stuck on this project due to lack
>>>> of time...).
>>>> If someone needs the dump, just let me know and I'll dig it out.
>>>> Frank IZ8DWF
>>> Manuals at ArtekManuals.com
> time-nuts mailing list -- time-nuts at febo.com
> To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/time-nuts
> and follow the instructions there.
More information about the time-nuts