[volt-nuts] HP 3458A repair.

J. L. Trantham jltran at att.net
Tue Sep 17 21:50:24 EDT 2013


Can you give us more information?  Serial number, Rev. number, CALNUM?  How
much to invest will be determined by age and other condition.

It doesn't sound like a simple CALRAM issue but changing the CALRAM is
relatively easy.  I removed all three DALLAS chips in mine and installed
sockets.  The CALRAM can be read with a chip programmer and the data written
to a new DALLAS chip.

I would also call Gary Bierman at the Loveland Cal Lab and have a long talk
with him.  He has a lot of insight into these meters and generally prefers
to do a component level repair rather than an assembly level repair.  The
charge sounds like their standard repair charge, no matter what the problem
is, and includes a 'fresh calibration' along with a warranty, a year I
think, but Gary will be able to answer that question.  

Also, once you get the meter calibrated by Agilent (and thus prove it is
functioning normally) it will be eligible for their 'repair agreement' which
is $178.68 per year.  I would consider buying a 5 year agreement after the

Good luck.


-----Original Message-----
From: volt-nuts-bounces at febo.com [mailto:volt-nuts-bounces at febo.com] On
Behalf Of John Phillips
Sent: Tuesday, September 17, 2013 3:36 PM
To: Discussion of precise voltage measurement
Subject: [volt-nuts] HP 3458A repair.

I have a 3458A that we sent to Agilent for calibration which it failed.
Before we sent it we calibrated it and it looked good to us. The infor. we
revived led us to believe that the cal memory may have caused the failure.
 We ask that it be sent buck to us and paid half the cal charges (about
$800) insted of the $2660.64 they wanted to repair it. We were just going to
repalce the ram in try again.
When we got the meter back it came with befor and afer data Like before 10
volts read 9.9999957 and after it read 10.00009 so they did something or the
meter drifted that much.
The problem is  0.1 volt and 1.0 volts failed at 8 and 10 MHz but passed at
4 MHz.
4MHZ 0.1 volt reads 0.097251 Lower Limit is 0.095930 PASSED 8MHZ 0.1 volt
reads 0.085712 Lower Limit is 0.0959
10MHZ 0.1 volt reads 0.75569 Lower Limit is 0.084900 FAILED

4MHZ 1 volt reads 0.97272 Lower Limit is 0.95930 PASSED 8MHZ 1 volt reads
0.86389 Lower Limit is 0.95920 FAILED 10MHZ 1 volt reads 0.73514 Lower Limit
is 0.84900 FAILED

The AC after readings are the same. I do not see how AC after could be that
identical even if they did not try to calibrate it. Did they just copy the
before data and call it after data?

My best guess is that if the 4 MHz is in and the higher frequencies are not
the meter requires some kind of mechanical adjustment to get the frequency
response   withing spec or the AC board needs to be repaid.

Are they charging a standard repair charge to do a calibration? I do not see
changing the memory to fix this.

Where would you go from here if this was your meter?

John Phillips
volt-nuts mailing list -- volt-nuts at febo.com To unsubscribe, go to
and follow the instructions there.

More information about the volt-nuts mailing list