I have three HP3458A's and decided it was time to back up the battery backed static RAM chips. I wrote a program to dump the RAM contents over the GPIB bus using the undocumented MREAD command as documented by Poul-Henning. I dumped both the 2kB CAL rams and the 32kW auxiliary data RAMs (hopefully that data is at the addresses mentioned by Poul-Henning. I used a NI GPIB-232CV-A RS232 to GPIB converter as the interface. All seems to have gone well and the CAL data looks reasonable.
As a check I dumped all the data from each machine three times and compared each dump... each CAL data set matched except for the third machine. On the first two, the units were powered on for a couple of hours while I tweaked with my code. I dumped the third machine from a cold start and one byte was different between each dump. I waited a couple of hours and tried again, this time the data dumps matched. It looks like that byte may be being updated while the machine warms up... If you dump your machine's data, it might be a good idea to let it stabilize for a couple of hours first.
The dumps of the 32kW of aux data have several words that change between each dump. It looks like that memory has some uses by the firmware other than storing user data, etc.
I think when the time comes to replace the battery backed memory chips I may try Cypress/Ramtron FRAM chips or Simtek STK16C88 AutoStore devices. The Simtek chips are available in 28 pin DIP packages that appear to be drop-in replacements for the DS1230 chips. The DS1220 chip would require a SO8 packaged part on a carrier board.
In message BLU170-W4732A073BE87463FF74529CEE10@phx.gbl, Mark Sims writes:
I have three HP3458A's and decided it was time to back up the
battery backed static RAM chips. [...]
I dumped the third
machine from a cold start and one byte was different between each
dump.
If you mail those files to me, I can try to see if that bit matters.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
Mark,
I am using the STMicroelectronics M48Z12 series in at least one of my
3458A's. I had trouble sourcing one of the two different memory modules
from a reputable seller when I replaced my NVRAM. I purchased a DS1220
previously from Jameco and it died shortly afterward. The datecode on the
Jameco part was several months old and may have been stored improperly.
The only thing I don't like about the STMicro NVRAM is that they use a
three digit datecode. One digit is for the year and two for the week. It
took me a while to track down a document to decipher their labeling system.
Todd
On Thu, Aug 7, 2014 at 12:20 AM, Mark Sims holrum@hotmail.com wrote:
I have three HP3458A's and decided it was time to back up the battery
backed static RAM chips. I wrote a program to dump the RAM contents over
the GPIB bus using the undocumented MREAD command as documented by
Poul-Henning. I dumped both the 2kB CAL rams and the 32kW auxiliary data
RAMs (hopefully that data is at the addresses mentioned by Poul-Henning.
I used a NI GPIB-232CV-A RS232 to GPIB converter as the interface. All
seems to have gone well and the CAL data looks reasonable.
As a check I dumped all the data from each machine three times and
compared each dump... each CAL data set matched except for the third
machine. On the first two, the units were powered on for a couple of
hours while I tweaked with my code. I dumped the third machine from a cold
start and one byte was different between each dump. I waited a couple of
hours and tried again, this time the data dumps matched. It looks like
that byte may be being updated while the machine warms up... If you dump
your machine's data, it might be a good idea to let it stabilize for a
couple of hours first.
The dumps of the 32kW of aux data have several words that change between
each dump. It looks like that memory has some uses by the firmware other
than storing user data, etc.
I think when the time comes to replace the battery backed memory chips I
may try Cypress/Ramtron FRAM chips or Simtek STK16C88 AutoStore devices.
The Simtek chips are available in 28 pin DIP packages that appear to be
drop-in replacements for the DS1230 chips. The DS1220 chip would require a
SO8 packaged part on a carrier board.
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
Poul,
I would love to know where the CALNUM value is stored.
Any idea?
Thanks.
Joe
-----Original Message-----
From: volt-nuts-bounces@febo.com [mailto:volt-nuts-bounces@febo.com] On
Behalf Of Poul-Henning Kamp
Sent: Thursday, August 07, 2014 12:36 AM
To: Discussion of precise voltage measurement; Mark Sims
Subject: Re: [volt-nuts] HP3458A calibration memory backup
In message BLU170-W4732A073BE87463FF74529CEE10@phx.gbl, Mark Sims writes:
I have three HP3458A's and decided it was time to back up the battery
backed static RAM chips. [...]
I dumped the third
machine from a cold start and one byte was different between each dump.
If you mail those files to me, I can try to see if that bit matters.
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
volt-nuts mailing list -- volt-nuts@febo.com To unsubscribe, go to
https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
In message 012401cfb249$f429ca90$dc7d5fb0$@att.net, "J. L. Trantham" writes:
I would love to know where the CALNUM value is stored.
This is a dump of the CALRAM, with device-addresse. If you use
MREAD, you need to multiply the address by two and add 0x60000:
--
Poul-Henning Kamp | UNIX since Zilog Zeus 3.20
phk@FreeBSD.ORG | TCP/IP since RFC 956
FreeBSD committer | BSD since 4.3-tahoe
Never attribute to malice what can adequately be explained by incompetence.
Hi Todd,
There is nothing that a human could endure that that part couldn't
take for a couple of years without harm.
The Chinese run a booming market in taking US scrap boards, removing
interesting parts, welding on new leads, adding plating, sandblasting
off old part labels, and putting on new. They will do both laser
engraving and stenciling... all this on scrap parts... which they
sell as new, or new old stock (depending on which gets more money).
I wonder if Jameco got your DS1220 from a Chinese "refurbisher" that
simply blasted off the old date code and put on a new?
I would send a note to Jameco telling them what happened, and I would
send a picture of the top of the part to Maxim telling them what
happened. Maxim can tell you pretty quickly if the part is counterfeit.
And, they will probably both send you a replacement either way.
-Chuck Harris
Todd Micallef wrote:
Mark,
I am using the STMicroelectronics M48Z12 series in at least one of my
3458A's. I had trouble sourcing one of the two different memory modules
from a reputable seller when I replaced my NVRAM. I purchased a DS1220
previously from Jameco and it died shortly afterward. The datecode on the
Jameco part was several months old and may have been stored improperly.
The only thing I don't like about the STMicro NVRAM is that they use a
three digit datecode. One digit is for the year and two for the week. It
took me a while to track down a document to decipher their labeling system.
Todd
My first thought was that I bought a counterfeit part. The STMicro part has worked well and I like their snaphat option too. I will have to look at making up a few adapter boards for a couple of Solartrons so I can dump those Varta batteries and use something that won't leak everywhere.
Todd
Sent from my iPad
On Aug 7, 2014, at 10:34, Chuck Harris cfharris@erols.com wrote:
Hi Todd,
There is nothing that a human could endure that that part couldn't
take for a couple of years without harm.
The Chinese run a booming market in taking US scrap boards, removing
interesting parts, welding on new leads, adding plating, sandblasting
off old part labels, and putting on new. They will do both laser
engraving and stenciling... all this on scrap parts... which they
sell as new, or new old stock (depending on which gets more money).
I wonder if Jameco got your DS1220 from a Chinese "refurbisher" that
simply blasted off the old date code and put on a new?
I would send a note to Jameco telling them what happened, and I would
send a picture of the top of the part to Maxim telling them what
happened. Maxim can tell you pretty quickly if the part is counterfeit.
And, they will probably both send you a replacement either way.
-Chuck Harris
Todd Micallef wrote:
Mark,
I am using the STMicroelectronics M48Z12 series in at least one of my
3458A's. I had trouble sourcing one of the two different memory modules
from a reputable seller when I replaced my NVRAM. I purchased a DS1220
previously from Jameco and it died shortly afterward. The datecode on the
Jameco part was several months old and may have been stored improperly.
The only thing I don't like about the STMicro NVRAM is that they use a
three digit datecode. One digit is for the year and two for the week. It
took me a while to track down a document to decipher their labeling system.
Todd
volt-nuts mailing list -- volt-nuts@febo.com
To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.