Hi guys,

I've got a pair of Solartron 7150Plus 6.5-digit digital multimeters
sitting on my desk which have just come back from calibration. One unit
calibrated to standard perfectly (with the note "resistance and 0.2VAC
ranges readjusted to norm and retested"). The other one seems to have
some issues.

The calibration lab reported that "Calibration on AC V and I ranges
fails with message 'ERROR' on FP. Partial calibration performed, other
ranges fine."
DC voltage, current and resistance are reading fine and well within
acceptable norms and agree with the other meter (plus/minus a dozen or
so counts in 6-digit mode).

If I connect an AC source (my function generator set to 400Hz per
service manual spec), the FG display and calibrated meter return valid
results.

When I connect up the faulty meter, I can calibrate it fine for about an
hour after power up. After exiting CAL mode, the indicated voltage on
all AC ranges gradually drops to the point where after three hours it's
half a volt down on nominal. Putting the meter back into CAL mode and
attempting a recalibration results in the ERROR message appearing.

The service and ops manuals suggest this means the calibration source
isn't in range or there's a hardware fault. Obviously the cal source is
in range because the calibrated meter (and my Agilent U1251A) both say
it is... which leaves a fault on the DMM.

Has anyone seen anything like this before?

I'm thinking a component has gone temperature-sensitive -- and given
that this affects both AC voltage and current, it'd have to be in the
parts common to those sections. The A/D is obviously working fine
because the DC ranges are rock solid. I've yet to remove the cover and
test things, but my suspect list is as follows:

• True-RMS to DC converter - IC21, AD637K. (£65! I REALLY hope it
  isn't this thing)
• Opamp IC15 (AD528J / LH0062CH FET opamp, seemingly long obsolete)
• Analog switch IC18 (DG211 - a couple of quid each from Farnell)
• TR1 and TR2 (a pair of 3N163s - more expensive parts - about £20 each!)
• TR12 (WN1001, seems to handle gain setting around IC15 - can't even
  find a datasheet for this one so I really really REALLY hope it hasn't
  gone bad)
• A handful of passives (resistors and capacitors)

This is based on a look at the circuit diagrams in the service manual
and elimination of parts common to other circuit blocks which are
functioning correctly.

I'm thinking that a good starting point would be to cool each of these
parts in turn with freezer spray, then see if the meter is affected. If
it is, warm the part back up with a hair dryer and see if it drifts back
again.

However if someone has any other ideas or suggestions then I'd love to
hear them :)

Thanks,

Phil.
philpem@philpem.me.uk
http://www.philpem.me.uk/

Phil,
I suspect it to be the EAROM(s) IC 105/6 of this meter. Losing its
calibration constants.
Regards, Wim.

Wim de Jong
wim.de.jong.59@solcon.nl

-------- Origineel bericht --------
Onderwerp: [volt-nuts] Solartron 7150plus - Chasing AC gremlins
Datum: Fri, 07 Dec 2012 01:01:08 +0000
Van: Philip Pemberton philpem@philpem.me.uk
Antwoord-naar: Discussion of precise voltage measurement
volt-nuts@febo.com
Aan: volt-nuts@febo.com

Hi guys,

I've got a pair of Solartron 7150Plus 6.5-digit digital multimeters
sitting on my desk which have just come back from calibration. One unit
calibrated to standard perfectly (with the note "resistance and 0.2VAC
ranges readjusted to norm and retested"). The other one seems to have
some issues.

The calibration lab reported that "Calibration on AC V and I ranges
fails with message 'ERROR' on FP. Partial calibration performed, other
ranges fine."
DC voltage, current and resistance are reading fine and well within
acceptable norms and agree with the other meter (plus/minus a dozen or
so counts in 6-digit mode).

If I connect an AC source (my function generator set to 400Hz per
service manual spec), the FG display and calibrated meter return valid
results.

When I connect up the faulty meter, I can calibrate it fine for about an
hour after power up. After exiting CAL mode, the indicated voltage on
all AC ranges gradually drops to the point where after three hours it's
half a volt down on nominal. Putting the meter back into CAL mode and
attempting a recalibration results in the ERROR message appearing.

The service and ops manuals suggest this means the calibration source
isn't in range or there's a hardware fault. Obviously the cal source is
in range because the calibrated meter (and my Agilent U1251A) both say
it is... which leaves a fault on the DMM.

Has anyone seen anything like this before?

I'm thinking a component has gone temperature-sensitive -- and given
that this affects both AC voltage and current, it'd have to be in the
parts common to those sections. The A/D is obviously working fine
because the DC ranges are rock solid. I've yet to remove the cover and
test things, but my suspect list is as follows:

• True-RMS to DC converter - IC21, AD637K. (£65! I REALLY hope it
  isn't this thing)
• Opamp IC15 (AD528J / LH0062CH FET opamp, seemingly long obsolete)
• Analog switch IC18 (DG211 - a couple of quid each from Farnell)
• TR1 and TR2 (a pair of 3N163s - more expensive parts - about £20 each!)
• TR12 (WN1001, seems to handle gain setting around IC15 - can't even
  find a datasheet for this one so I really really REALLY hope it hasn't
  gone bad)
• A handful of passives (resistors and capacitors)

This is based on a look at the circuit diagrams in the service manual
and elimination of parts common to other circuit blocks which are
functioning correctly.

I'm thinking that a good starting point would be to cool each of these
parts in turn with freezer spray, then see if the meter is affected. If
it is, warm the part back up with a hair dryer and see if it drifts back
again.

However if someone has any other ideas or suggestions then I'd love to
hear them :)

Thanks,

Phil.
philpem@philpem.me.uk
http://www.philpem.me.uk/

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.

Wim

I don't think that is likely, as Phil clearly says that all the DC ranges are correct,  it is just the AC ranges that are a problem.  If the EAROMs were bad, you'd lose all the calibration constants.

Dave
-----Original Message-----
From: volt-nuts-bounces@febo.com [mailto:volt-nuts-bounces@febo.com] On Behalf Of Wim de Jong
Sent: 07 December 2012 08:04
To: volt-nuts@febo.com
Subject: [volt-nuts] Fwd: Solartron 7150plus - Chasing AC gremlins

Phil,
I suspect it to be the EAROM(s) IC 105/6 of this meter. Losing its calibration constants.
Regards, Wim.

Wim de Jong
wim.de.jong.59@solcon.nl

If it does turn out the be the AD637 (shudder), you can get the SOIC AD637B or AD637K for about GBP30 and use an SOIC to DIP adapter (e.g. Winslow). This should work even though it is SOIC16 versus DIP14, as the two NC pins are at one end of the SOIC.

HtH Dave