Re: Tricks and thoughts about the 3457A battery
Hello Martin,
?
The EPROM idea is interesting, I have explored it but I don't think it could work. The main issue is that the 8Kbyte RAM chip which holds the calibration data (TC5564APL-15) is also used to store the processor stack, firmware variables, user's memorized measurements etc.
?
From the schematic, we know the RAM is used as follows :
- Calibration RAM data is from 0040 hex to 01FF hex with a special hardware mechanism to prevent accidental overwrite of this area.
- General storage RAM (processor stack, miscellaneous variables, user measurement FIFO etc) from 0200 to 1FFF.
?
?.... In summary, replacing the RAM chip with an EPROM would not work. One (strong) reason is that the processor stack and RAM storage would not be writable.
?
Cheers,
Joel?
|
Re: Tricks and thoughts about the 3457A battery
Congrats!! Ing. Patricio A. GrecoTaller Aeronáutico de Reparación 1B-349Organización de Mantenimiento Aeronáutico de la Defensa OMAD-001Gral. Martín Rodríguez 2159San Miguel (1663)Buenos AiresT:?+5411-4455-2557F:?+5411-4032-0072
toggle quoted message
Show quoted text
On 30 Mar 2025, at 18:00, Joel Setton via groups.io <setton@...> wrote:
?
I did it ! I replaced the memory battery in my 3457A.
?
My used 3457A was built in the late '80s and still had its original lithium battery, with a 1989 date code on it. This battery obviously needed replacement.
Unfortunately, the battery is soldered to the processor PC board and provides permanent Vcc for a low-power CMOS SRAM chip which holds the multimeter's calibration constants. In other words, interrupting the SRAM supply would result in the loss of calibration data.
Thus, the battery must be replaced while the 3457A is powered up. This is a bit of a nail-biting experience, but it can be done. One useful trick I used is to prepare a cardboard shield which covers the CPU board and protects it from any solder drops and shorts, while allowing access to the battery. And of course the soldering iron must be floating.
?
Replacement batteries in the original size (lithium BR 2/3AA with solder pins) are still available for a few dollars, but I chose to install a slightly different battery, a Tadiran SL750 which has a 1/2 AA size, for which sockets are available. Using a socket means that in 10 years, the next battery replacement will be easier : with the multimeter powered up, just pop out the old battery and insert a new one. Another advantage is that 1/2 AA lithium batteries are very easy to find, there's a good chance that your local supermarket has them.
?
Now I'd like to share some thoughts and questions about this battery. As we all know, lithium batteries have a long, but finite life. Even though the current drain on the battery is less than 0.2 uA at room temperature, at some point in time the battery will die.
I read the 3457A user and service manuals, but I couldn't find a single mention of the battery. Nothing about battery life, no instructions on how to replace it, even nothing that says "contact your nearest HP sales office". To me, this doesn't make sense, especially when considering that this multimeter was introduced in 1986, at the time of the "old HP" which produced high-quality instruments, not expendable stuff. I could only think of a few possibilities :
?
- Maybe the engineers at that time believed that lithium batteries could last forever ;
- Or they thought the multimeter would be scrapped when the battery runs out, or before it does ;
- Or they decided that if a multimeter loses its calibration data because of a dead battery, it should be returned to HP for battery or processor board replacement followed by re-calibration. That's an expensive proposition.
?
None of these options make any sense to me in the context of the "old HP".
Further thoughts and comments will be most welcome !
?
Joel?
<cardboard_shield.jpg> <All done.jpg>
|
Re: Tricks and thoughts about the 3457A battery
Hi Joel,
Since you're able to read out the contents of the RAM, would it be possible to write it to an EPROM and swap it?
I never use the save and recall commands... could this work or will it refuse to start up? Advantage would be to never bother with batteries again.
Just a firther thought, as I have the same DMM but did not yet even look inside :-/
cheers
Martin
|
Re: PCB CAD used and/or sold by HP ?
Back in the 80s, Mentor Graphics' CAD tools ran on HP-UX workstations. Perhaps that is to what he was referring.
?
Some time ago I had a discussion with a friend over old times circuit board design.
He referred to where he worked where they used to do manual design using black tape and
transparent films. "The layouts were beautiful but difficult or even impossible to modify".?
?
Then "We got a CAD-system from HP" and it all changed.
?
I am unfamiliar with CAD-systems created by and/or sold by HP and what happened to that product.
Anyone that can shed some light?
?
Cheers
?
Ulf Kylenfall
SM6GXV
|
Re: PCB CAD used and/or sold by HP ?
On Sunday, 30 March 2025, Ulf Kylenfall via <ulf_r_k= [email protected]> wrote: > ? > Some time ago I had a discussion with a friend over old times circuit board design. > He referred to where he worked where they used to do manual design using black tape and > transparent films. "The layouts were beautiful but difficult or even impossible to modify".? > ? > Then "We got a CAD-system from HP" and it all changed. > ? > I am unfamiliar with CAD-systems created by and/or sold by HP and what happened to that product. > Anyone that can shed some light? ISTR using something similar in the early 80s. I suspect it might have been a Racal Redac CAD package using HP terminals for display. The terminals had a more widescreen aspect ratio than VT100s, and could draw lines well. I haven't spotted a decent reference to all that.
|
Re: PCB CAD used and/or sold by HP ?
"Ulf Kylenfall via groups.io" <ulf_r_k@...> writes:
Some time ago I had a discussion with a friend over old times circuit board design.
He referred to where he worked where they used to do manual design using black tape and
transparent films. "The layouts were beautiful but difficult or even impossible to modify".
Then "We got a CAD-system from HP" and it all changed.
I am unfamiliar with CAD-systems created by and/or sold by HP and what happened to that product.
Anyone that can shed some light? While working on emulating HP9000/300 i've read about "HP Engineering Graphics System", but don't know how much use it had.
|
Re: PCB CAD used and/or sold by HP ?
The system I remember from the early 1980’s was SCICARD from Scientific calculations. I think they were based somewhere near Rochester, NY.?
|
PCB CAD used and/or sold by HP ?
?
Some time ago I had a discussion with a friend over old times circuit board design.
He referred to where he worked where they used to do manual design using black tape and
transparent films. "The layouts were beautiful but difficult or even impossible to modify".?
?
Then "We got a CAD-system from HP" and it all changed.
?
I am unfamiliar with CAD-systems created by and/or sold by HP and what happened to that product.
Anyone that can shed some light?
?
Cheers
?
Ulf Kylenfall
SM6GXV
|
Re: Tricks and thoughts about the 3457A battery
Hello Joel,
Thanks for sharing your repair. I have a question related to the battery
holder.
Your original BR2/3A battery had 3 pins if I am not mistaken. One on one
side and two on the other side.
Did you find a socket that also has 3 pins and matches the holes on the
PCB but takes 1/2AA battery ?
Thanks,
Razvan
toggle quoted message
Show quoted text
On 30/03/2025 18:00, Joel Setton via groups.io wrote: I did it ! I replaced the memory battery in my 3457A.
My used 3457A was built in the late '80s and still had its original
lithium battery, with a 1989 date code on it. This battery obviously
needed replacement.
Unfortunately, the battery is soldered to the processor PC board and
provides permanent Vcc for a low-power CMOS SRAM chip which holds the
multimeter's calibration constants. In other words, interrupting the
SRAM supply would result in the loss of calibration data.
Thus, the battery must be replaced while the 3457A is powered up. This
is a bit of a nail-biting experience, but it can be done. One useful
trick I used is to prepare a cardboard shield which covers the CPU board
and protects it from any solder drops and shorts, while allowing access
to the battery. And of course the soldering iron must be floating.
Replacement batteries in the original size (lithium BR 2/3AA with solder
pins) are still available for a few dollars, but I chose to install a
slightly different battery, a Tadiran SL750 which has a 1/2 AA size, for
which sockets are available. Using a socket means that in 10 years, the
next battery replacement will be easier : with the multimeter powered
up, just pop out the old battery and insert a new one. Another advantage
is that 1/2 AA lithium batteries are very easy to find, there's a good
chance that your local supermarket has them.
Now I'd like to share some thoughts and questions about this battery. As
we all know, lithium batteries have a long, but finite life. Even though
the current drain on the battery is less than 0.2 uA at room
temperature, at some point in time the battery will die.
I read the 3457A user and service manuals, but I couldn't find a single
mention of the battery. Nothing about battery life, no instructions on
how to replace it, even nothing that says "contact your nearest HP sales
office". To me, this doesn't make sense, especially when considering
that this multimeter was introduced in 1986, at the time of the "old HP"
which produced high-quality instruments, not expendable stuff. I could
only think of a few possibilities :
- Maybe the engineers at that time believed that lithium batteries could
last forever ;
- Or they thought the multimeter would be scrapped when the battery runs
out, or before it does ;
- Or they decided that if a multimeter loses its calibration data
because of a dead battery, it should be returned to HP for battery or
processor board replacement followed by re-calibration. That's an
expensive proposition.
None of these options make any sense to me in the context of the "old HP".
Further thoughts and comments will be most welcome !
Joel
|
Tricks and thoughts about the 3457A battery
I did it ! I replaced the memory battery in my 3457A.
?
My used 3457A was built in the late '80s and still had its original lithium battery, with a 1989 date code on it. This battery obviously needed replacement.
Unfortunately, the battery is soldered to the processor PC board and provides permanent Vcc for a low-power CMOS SRAM chip which holds the multimeter's calibration constants. In other words, interrupting the SRAM supply would result in the loss of calibration data.
Thus, the battery must be replaced while the 3457A is powered up. This is a bit of a nail-biting experience, but it can be done. One useful trick I used is to prepare a cardboard shield which covers the CPU board and protects it from any solder drops and shorts, while allowing access to the battery. And of course the soldering iron must be floating.
?
Replacement batteries in the original size (lithium BR 2/3AA with solder pins) are still available for a few dollars, but I chose to install a slightly different battery, a Tadiran SL750 which has a 1/2 AA size, for which sockets are available. Using a socket means that in 10 years, the next battery replacement will be easier : with the multimeter powered up, just pop out the old battery and insert a new one. Another advantage is that 1/2 AA lithium batteries are very easy to find, there's a good chance that your local supermarket has them.
?
Now I'd like to share some thoughts and questions about this battery. As we all know, lithium batteries have a long, but finite life. Even though the current drain on the battery is less than 0.2 uA at room temperature, at some point in time the battery will die.
I read the 3457A user and service manuals, but I couldn't find a single mention of the battery. Nothing about battery life, no instructions on how to replace it, even nothing that says "contact your nearest HP sales office". To me, this doesn't make sense, especially when considering that this multimeter was introduced in 1986, at the time of the "old HP" which produced high-quality instruments, not expendable stuff. I could only think of a few possibilities :
?
- Maybe the engineers at that time believed that lithium batteries could last forever ;
- Or they thought the multimeter would be scrapped when the battery runs out, or before it does ;
- Or they decided that if a multimeter loses its calibration data because of a dead battery, it should be returned to HP for battery or processor board replacement followed by re-calibration. That's an expensive proposition.
?
None of these options make any sense to me in the context of the "old HP".
Further thoughts and comments will be most welcome !
?
Joel?
|
Re: Agilent E5071C Low power both ports, Vernier Leveler E5071-62083, Component IGC1-4213 replacement
Hej Sven,
Thanks for sharing your experience.??My only update is that I managed to obtain the needed components.? We have 3 units with this particular failure mechanism (out of roughly 10 total).? We have a special soldering department which hopefully can get these mounted effectively.? There is a warning on the package that they should not be warmed to above 260C to avoid damage.? It's mighty hard to do that and melt solder, but with a little luck we will have these units working again within the week.? From my perspective, this is one of the two most common failure mechanisms with these units.? These units were probably kept running in a lab for multiple years and under these conditions this part just seems to fail.
|
Took me a while to figure out why 7652A gave no results, a typo for the 7562A log voltmeter, the manual lists 1855-0064 as Siliconix DN643, which is also listed under the NSN number.
?
David
|
Looking for documentation on 4263A LCR meter
Hi all,
?
I'm currently on the search for the Operation Manual (p/n: 04263-90000) and the Service Manual (p/n: 04263-90031) to suit the 4263A LCR Meter.
I have the 4263A Users Guide (p/n: 04263-90001) from the files section here, and all the 4263B manuals, but the two for the 4263A mentioned above seem a bit elusive.
?
Does anyone have a copy of these two that they could send through or upload to the files section?
?
?
Thanks!
Jared
|
Re: Agilent E5071C Low power both ports, Vernier Leveler E5071-62083, Component IGC1-4213 replacement
"Sven Schnelle" <svens@...> writes: "Scott Vance via groups.io" <scott_l_vance@...> writes:
In the process of fixing 9 pcs E5071C, I have noted that at least
three of them have low output power on both channels
(-26 dBm on two, -31 dBm on the third). I confirmed that the faulty
component was the Vernier Leveler and then took this
apart to take a look. It seems like it's always the expensive Did you make any progress on this?
I just finished fixing the mainboard issue on my E5071C. Measured the
output power on the Ports when set to 0dBm and all have around -23dBm
level. So looks to me like my E5071C has the same issue. The only
thing that i find odd is that bridging two ports shows a straight
S12/S21 line at 0dB. I would expect it to be around -20dB. How is that
on your units? Seems to be some intermittent failure on my unit. I removed the Vernier leveling module this morning, to measure levels at the various stages. Surprisingly enough after inserting it again the output level of the E5071C is now at about -2.5dBm. Wasn't able to reproduce the issue by bending the PCB a bit. Lets see, i guess it will happen again. And it looks like the 1GG7-4243 was already replaced at some point in time.
|
1855-0064 isn't in the master reference list or the several cross reference lists I looked at. I found a hit on one of the NSN sites which isn't always reliable but this one shows the hallmark's of being correct.
This part number is not for a specific device. It describes a list of important specifications that are required. Structure: multiple JFET transistors Material: silicon Polarity: N-Channel Pd: 300 mW IDss(Id when Vgs=0V): 1mA per transistor, 8mA all transistors Case: 0.219in diameter, hermetically sealed?
Plugging those numbers in to I find 5 known part numbers that exceed those specs. 2N3921 2N3922 2SK146 IFN146 LSK389
Looking on Mouser it seems almost every modern dual JFET meets or exceeds those specs.
You will need to check the circuit to see what Vds and Vgs makes sense. The modern JFETs range form 10 to 60V for both.
On Sun, Mar 30, 2025, 5:12?AM John Warren via <johnwarren376= [email protected]> wrote: Looking for any info on this device. Dual JFET, N-channel, TO-71 package.? I have a 7652A and both were pulled from one of the boards.? I may pull one off another unit to curve trace but thought I ask if there's any spares out there first.??
|
Re: HP6209B component location layout
I think you will need to select R38, (~30k) section 5-45, it sets the meter zero by canceling the programming current that passes through the current sense resistor. R53,(3k) may need to be changed also.
Check here for a manual with schematic.
Hugh Gilbert
Hugh Gilbert
Correction to my previous post: meant to say MINIMUM OUTPUT VOLTAGE after adding the 47K resistor in series with R10 is about 160V, maximum output is still about 320V.
I am checking some voltages around the current meter circuit: the -3V shown on the schematic to be present at one side of R38/R53 measures only -2.29V. Same value at the cathode of CR8 in the bias supply where this voltage originates. Cathode of CR7 has +1.77V.
CR7 is a 1N4830 "stabistor" 2.4V @ 100mA. Guessing some sort of reference voltage or early zener diode. On the other hand CR8 is listed as 1N3828 and seems to be a regular silicon small signal diode. The way they are connected in series in the bias supply (for among others temperature stability) I would expect the -3V output to be the sum of the CR7 voltage, plus about 0.6-0.7V additional drop of CR8, so indeed about 3V. But in real life CR8 seems to have a drop of only 0.5V as measured, so do I have a bad CR7 that has gone low from the original 2.4V to only 1.77V?
The other bias supply voltages: +43V is at +37.5V, and the +16V is correct at +15.98V.
And at the reference regulator, the +6.2V (33) measures 6.43V, the +12.4V (37) measures 12.66V, and the -6.2V (31) measures -6.22V.
All voltage readings referenced to +S (same as positive/red output binding post with correct rear jumpers in place).
All other rear jumpers where also verified earlier to be in the standard configuration as shown on the schematic.
|
Re: Agilent E5071C Low power both ports, Vernier Leveler E5071-62083, Component IGC1-4213 replacement
"Scott Vance via groups.io" <scott_l_vance@...> writes: In the process of fixing 9 pcs E5071C, I have noted that at least
three of them have low output power on both channels
(-26 dBm on two, -31 dBm on the third). I confirmed that the faulty
component was the Vernier Leveler and then took this
apart to take a look. It seems like it's always the expensive
components that burn out, so I decided to remove IGC1-4213
and bridge the contacts on either side of the component. Sure enough,
the output power went up from -26 dBm to about -9
dBm. Some limited information about this part is available on the
Keysight. Looks like it's a 11 dB gain power
amplifier, so the numbers check out and it looks like this is the
failed component. Assuming that they drive it at 9
dBm gain, that would give me the right power at my port (0 dBm
nominally).
Replacements components appear to be hard to find and expensive
(around $500), which is a pretty good indication that
failures on this component are common place. The Keysight page
indicates that they sell parts (1GC1-8213) in lots of 10
pcs, so I've put in an inquiry.
My questions:
1) Is it possible to get a IGC1-4213 data sheet?
2) There are some similar component numbers (IGC1-4214, IGC1-8213).
How does one tell which components are compatible
and which are not? Data sheets seem to be unavailable and every time
I ask Keysight a question they respond that they're
happy to fix it for me, just send them a check for 6K USD.
3) Of the 9 units, there are two components that have failed and
knocked out about 6 of the units. In other words, two
failure mechanisms keep repeating themselves. Keysight has got to
know this. Why don't they have a service bulletin or
something? Why don't they offer a reasonable repair service for these
components which appear to have been designed to
fail? Just a rant I suppose, but I'm not too impressed. On the plus
side, it makes my job easier because once I figure
out the failure I just have to apply the same solution over and over
again. Did you make any progress on this? I just finished fixing the mainboard issue on my E5071C. Measured the output power on the Ports when set to 0dBm and all have around -23dBm level. So looks to me like my E5071C has the same issue. The only thing that i find odd is that bridging two ports shows a straight S12/S21 line at 0dB. I would expect it to be around -20dB. How is that on your units? Thanks Sven
|
Who has the fieldfox N9912A or N9923A inside SD card image?
Hello!
My internal sd card has died and filedfox cannot boot. That's why I can't install updates firmware from Keysight.com
|
Re: HP6209B component location layout
Correction to my previous post: meant to say MINIMUM OUTPUT VOLTAGE after adding the 47K resistor in series with R10 is about 160V, maximum output is still about 320V.
I am checking some voltages around the current meter circuit: the -3V shown on the schematic to be present at one side of R38/R53 measures only -2.29V. Same value at the cathode of CR8 in the bias supply where this voltage originates. Cathode of CR7 has +1.77V.
CR7 is a 1N4830 "stabistor" 2.4V @ 100mA. Guessing some sort of reference voltage or early zener diode. On the other hand CR8 is listed as 1N3828 and seems to be a regular silicon small signal diode. The way they are connected in series in the bias supply (for among others temperature stability) I would expect the -3V output to be the sum of the CR7 voltage, plus about 0.6-0.7V additional drop of CR8, so indeed about 3V. But in real life CR8 seems to have a drop of only 0.5V as measured, so do I have a bad CR7 that has gone low from the original 2.4V to only 1.77V?
The other bias supply voltages: +43V is at +37.5V, and the +16V is correct at +15.98V.
And at the reference regulator, the +6.2V (33) measures 6.43V, the +12.4V (37) measures 12.66V, and the -6.2V (31) measures -6.22V.
All voltage readings referenced to +S (same as positive/red output binding post with correct rear jumpers in place).
All other rear jumpers where also verified earlier to be in the standard configuration as shown on the schematic.
|
Looking for any info on this device. Dual JFET, N-channel, TO-71 package.? I have a 7652A and both were pulled from one of the boards.? I may pull one off another unit to curve trace but thought I ask if there's any spares out there first.??
|
Patricio A. Greco
Martin
Steve Byan
factory
Jared Cabot