HP 3456A inguard 8048 uP firmware?

[TimInCanada]

Hi Al,

Would anyone have the firmware from the 8048 uP of the inguard logic board?  It's IC A30U13, HP part # 1820-2539.

I picked up a 3456a showing error -3, inguard/outguard isolation failure.  The 5V regulator on the inguard power supply had failed.  It and all electrolytic caps are replaced now.

The inguard receiver logic was working properly.  Reset signals from the outguard processor were making it to the reset pin of U13.  U13 power is good.  Clock is also fine.  There was no response from U13 anywhere.  U13 gets an interrupt request from the receiver when a byte is available and is supposed to send out a READ signal on pin 8.  Even with pin 8 cut from the board there was no signal after the interrupt, so U13 is gone.

The 8048 can be replaced with the 8748 EPROM version, but I can't find the firmware anywhere.  The BAMA archive on the 3456A page does say "The checksum for a good 8048 binary ROM image is 1C936.", but they don't seem to have the file.  There have been a few discussions on different boards over the years but nobody seems to have posted this binary.

Some discussion mentioned there had only ever been one version of this firmware, so age of the meter doesn't matter.  It seems that the earliest models may have had the 8048 in a socket but most had it soldered in.

Any help finding this binary would be certainly appreciated.  (Otherwise, I might just have to buy another broken 3456A to work on.    )

Tim

[MadTux]

I think I had seen the 8048 binary somewhere in the wild once, but can't find it anywhere now 

Just bought a MCS48 adapter board for my willem programmer for $4, so I can keep my stack of 3456As working in case some of those inguard uCs will fail sometimes in the future.

I you can wait a few weeks until it arrives, I'll upload the binary here 

[TimInCanada]

Thanks Tux!

No worry about the wait.  I just ordered the MiniPro programmer featured on EEVblog #411, so it will be here anywhere between two weeks and two months from now.

I haven't been able to exercise the other ROMs in this meter yet, but am thinking of a pre-emptive strike in replacing them, too.

Tim

[TerraHertz]

Hmm. I have a couple of HP 3456As, and a device programmer that lists Intel 8748. But the question is, can the 8048 internal ROM be read out at all?

I'd have to dig out my old Intel data books and take a while to find out, and I'm busy atm.
Does anyone know?

If it can be read out (using 8748 programmer config) I don't mind desoldering the chip and doing it.

[pelule]

No sure, but have a look atthat link:
https://xdevs.com/doc/HP_Agilent_Keysight/3456A/fw/u7.bin

[JoeO]

Hmm. I have a couple of HP 3456As, and a device programmer that lists Intel 8748. But the question is, can the 8048 internal ROM be read out at all?

I'd have to dig out my old Intel data books and take a while to find out, and I'm busy atm.
Does anyone know?

If it can be read out (using 8748 programmer config) I don't mind desoldering the chip and doing it.

Link to 48 books by Intel
https://archive.org/search.php?query=intel%208048

[TimInCanada]

Wow, Archive.org is amazing.

The MCS-48 Users Manual says reading the 8048's program memory is a little different than the 8748 https://archive.org/stream/bitsavers_intel80481_45902725/1980_MCS-48_Users_Manual#page/n44/mode/1up

2.4.3 Reading Internal Prograin Memory
Just as the processor may be isolated from
internal program memory using EA, program
memory can be read independent of the
processor using the verification mode described
in the previous section, Programming/
Verification.
The processor is placed in the READ mode
by applying a high voltage (+23V for the
8748, +12V for the 8048/8049) to the EA pin
and +5V to the TO (8748 only) input pin.
RESET must be at OV when voltage is applied
to EA. The address of the location to
be read is then applied to the same lines
(TTL levels) of BUS and Port 2 which output
the address during single step (see below).
The address is latched by a "0" to "1"
transition on RESET and a high level on
RESET causes the contents of the program
memory location addressed to appear on
the eight lines of BUS. RESET must be
brought back to OV before leaving the
READ mode.

So EA gets 12V instead of 23V and T0 is not used.  Looks doable to get the firmware out.

pelule -- the U7 (and U5 and U8) firmware is for the outguard logic processor.  The inguard processor is a mask-programmed 8048, just to make repair more interesting.   

[MadTux]

Received the MCS-48 adapter today and dumped the AM-8048 firmware for you.

Used my Willem PCB5.0 and the MCS-48 adapters available from here:
http://www.sivava.com/b17-mcs48-dip40-to-dip32-adapter.html
Further programming instructions are in the attachment.

Also programmed an Intel-8748 and tested it in a 3456A to prove that the reading and programming procedure worked correctly.

[SaabFAN]

Be aware: The minipro-programmer cannot program the 8048. You need a Willem Programmer, or some other compatible device.
I made a programmer with a arduino Nano several months ago to repair my PM6654C counter in which the 8049 Co-Processor started to act up. Worked rather well.

[TimInCanada]

That is excellent.  Thank you MadTux.  Given the number of downloads already it looks like these files are helpful to others as well.

Thanks for the warning about the minipro programmer SaabFAN.  That is a good idea about using an Arduino to program, and making the programmer sounds like an interesting little project too.

Tim