HP 3457A
Posted: Sat Jul 26, 2025 1:36 pm
This little jewel popped up for parts only at a well known UK bomb-site trader, listed for £150.
I threw a semi-lowball, (£100), and they came back with £125 (+£16.95 p&p), so I pulled the trigger, as I could see that although it wouldn't power on, the cal stickers were intack, so there was a reasonable chance that there'd been no gorilla action.
It turned up in reasonable time, in almost reasonable packing (good box, nylon bands, thick paper wadding, but NONE on the bottom). Fortunately no damage to the box and no additional damage to the meter.
It was a bit grubby, with a RACAL asset label:
First port of call is the mains input fuse, which is an 80mA-T 6x32mm, and of course is blown. The voltage selectors are correct, but that isn't a guarantee someone didn't give it 240V on the 115V setting...
Now, I have some 80mA-T, that I got for the Fluke 8125A, but naturally they are 5x20mm...
The smallest value I can find is 250mA, which will have to do for fault finding.
I checked the mains input filter for insulation resistance, and it's fine. I checked the trafo, and it is fine also.
So, time to power it up (with a power meter in line) and see what happens.
I get GPIB address, but no beep, and the message "FAILED". And it's using >70W. Wait, what, how much?!? Off it goes.
At least at this point the digital board looks like it's working.
Thou shalt check power rails. A1 board earthy +5V fine, A2 floating -15V at about -13V, +5 and +15 both dead. Here's the analogue board with the AC piggy-backed:
And with the AC board removed:
The input hybrid removes pretty easily, though I didn't leave it off; only did it to eliminate it, and the reference board too.
Tried testing all the diodes in the psu, inconclusive results. Time to pull the v-regs, which fortunately in this are simple, jellybean parts; LM340AT-5, LM340AT15, MC7915CT. Marked the heatsinks with a white pen, as it would be "unfortunate" to put them back the wrong way around...
download/file.php?mode=view&id=2906
The +5 and +15 both test good, but the -15 is not happy:
Luckily, as a parts hoarder, a replacement is not a problem:
Much better. Now, does it work? Does it fuck! -15V rail is now good , but +5 and +15 are still dead.
Checked clamping zeners; good, pull all the diodes and retest; all good. Grrrrr.
Ok, testing the caps.
Yes, yes yes, I should have done this earlier, but there was no evidence of stress.
Fucking.
Little.
Bastard.
Resin sealed Mallory, on the i/p of +15V reg:
Damn thing measured 0.12Ω !
Annoyingly, I don't have a decent replacement. Have put a shitty 470uF 50V cap in, for now, and ordered some nice Panasonic 105˚C ones.
So, does it work now?
Yes, it lives! And uses less than 11W, considereably more plausible than >70W!
Time to clean the flux off the board and the labels off the case:
Pic limit, continued in next post.
I threw a semi-lowball, (£100), and they came back with £125 (+£16.95 p&p), so I pulled the trigger, as I could see that although it wouldn't power on, the cal stickers were intack, so there was a reasonable chance that there'd been no gorilla action.
It turned up in reasonable time, in almost reasonable packing (good box, nylon bands, thick paper wadding, but NONE on the bottom). Fortunately no damage to the box and no additional damage to the meter.
It was a bit grubby, with a RACAL asset label:
First port of call is the mains input fuse, which is an 80mA-T 6x32mm, and of course is blown. The voltage selectors are correct, but that isn't a guarantee someone didn't give it 240V on the 115V setting...
Now, I have some 80mA-T, that I got for the Fluke 8125A, but naturally they are 5x20mm...
The smallest value I can find is 250mA, which will have to do for fault finding.
I checked the mains input filter for insulation resistance, and it's fine. I checked the trafo, and it is fine also.
So, time to power it up (with a power meter in line) and see what happens.
I get GPIB address, but no beep, and the message "FAILED". And it's using >70W. Wait, what, how much?!? Off it goes.
At least at this point the digital board looks like it's working.
Thou shalt check power rails. A1 board earthy +5V fine, A2 floating -15V at about -13V, +5 and +15 both dead. Here's the analogue board with the AC piggy-backed:
And with the AC board removed:
The input hybrid removes pretty easily, though I didn't leave it off; only did it to eliminate it, and the reference board too.
Tried testing all the diodes in the psu, inconclusive results. Time to pull the v-regs, which fortunately in this are simple, jellybean parts; LM340AT-5, LM340AT15, MC7915CT. Marked the heatsinks with a white pen, as it would be "unfortunate" to put them back the wrong way around...
download/file.php?mode=view&id=2906
The +5 and +15 both test good, but the -15 is not happy:
Luckily, as a parts hoarder, a replacement is not a problem:
Much better. Now, does it work? Does it fuck! -15V rail is now good , but +5 and +15 are still dead.
Checked clamping zeners; good, pull all the diodes and retest; all good. Grrrrr.
Ok, testing the caps.
Yes, yes yes, I should have done this earlier, but there was no evidence of stress.
Fucking.
Little.
Bastard.
Resin sealed Mallory, on the i/p of +15V reg:
Damn thing measured 0.12Ω !
Annoyingly, I don't have a decent replacement. Have put a shitty 470uF 50V cap in, for now, and ordered some nice Panasonic 105˚C ones.
So, does it work now?
Yes, it lives! And uses less than 11W, considereably more plausible than >70W!
Time to clean the flux off the board and the labels off the case:
Pic limit, continued in next post.