[volt-nuts] Solartron 7150plus - Chasing AC gremlins
David C. Partridge
david.partridge at perdrix.co.uk
Mon Dec 24 14:19:43 UTC 2012
Service manual (page 3.6 item 7) says to use 20VAC range, shorted inputs, adjust trim pot for zero volts
Re: gluing front panel - VERY tricky to remove without putting creases in - been there done that :(. Better to just *sparingly* inject contact adhesive in strategic areas, wait until mostly dry, and press together.
As the only part replaced was IC15, the calibration should not be upset, as its the resistors that determine the gain, not the op-amp.
From: volt-nuts-bounces at febo.com [mailto:volt-nuts-bounces at febo.com] On Behalf Of Philip Pemberton
Sent: 24 December 2012 14:07
To: volt-nuts at febo.com
Subject: Re: [volt-nuts] Solartron 7150plus - Chasing AC gremlins
Right then, time for a writeup :)
Long story short - the 7150Plus is just about fixed. Parts were sourced as follows:
* Three 3N163 "analog switch" FETs at £11 each from Littlediode on Ebay. These turned out not to be required...
* Analog Devices AD637 (SOIC16) RMS-to-DC converted with SOIC breakout board. About £30 of silicon and plastic... also not used (though may be needed later if the 637 in the meter doesn't stop drifting).
* Three LH0062H opamps. $45 US plus shipping for all three, or £27.81 in the Queen's currency - also obtained from ebay (<http://myworld.ebay.co.uk/techno*>).
To recap, the problem was that the AC ranges were drifting badly and calibration reported an error condition after warm-up. After applying freeze spray to the opamp/FET section of the AC stage, the whole AC stage failed and the meter reported an input overload.
Initial investigations suggested that IC15 (LH0062C, National
Semiconductor) was running open loop (output was clipped to within a few volts of the negative-15V rail). No reason for this could be readily ascertained and the feedback loop was producing erroneous results when probed with a DMM (due to the opamp feeding back into the circuit).
Initially I suspected IC18 (DG211, Siliconix) and replaced it. This had no effect on the failure symptoms.
This morning I desoldered the LH0062C and replaced it with an 8-pin IC socket (machined-pin type of course). Probing the opamp support circuitry with the opamp out of circuit suggested that the attenuator and feedback loop were now operating as per the service manual description.
For "Gain of 10" ranges, resistance between IC15 pin 6 and IC15 pin 2 was measured at 1M-ohm after a settling delay of several tens of seconds. The settling delay can be attributed to the charging of C29 (3u3); the resistance is from R72 (1Meg). TR12 (WN1001) appears to prevent the multimeter from measuring the R21a/R21b divider chain.
For "Gain of 1" (unity) ranges, IC18 SW3 (pins 9, 10, 11) shorts over
C24 and puts the opamp into a unity gain state. Resistance between IC15 pins 6 and 2 is approximately 30 Ohms in this mode.
The function of TR12 and the circuitry around it still elude me. I've never been much good with JFET circuits!
After replacing the opamp, the AC range was selected with a 2V range.
This sets the attenuator to 1:1 mode and IC15's gain to unity. The whole AC chain now acts as a buffer. Applying a 2V pk-pk (~0.7V RMS) 400Hz sine to the input with a signal generator and probing the opamp output at TP3 suggested that the opamp was now functioning to spec. Testing the AC ranges suggests that aside from some expected calibration issues, the meter is functioning normally.
I'm hoping that trimming the offset is as simple as grounding the input, selecting the lowest AC voltage range and adjusting for zero volts at TP3.
After that, I need to figure out how to calibrate the AC voltage and current ranges! I'm hoping the signal generator will work for the lowest ranges, but the 20V, 200V and 750V ranges (not to mention the AC current
range) may prove somewhat tricky...
Then finally I need to remove the front panel and re-glue it to the plastic base. Unfortunately the old glue (actually double-sided tape) has perished and the front panel is now looking a little sorry for itself...
philpem at philpem.me.uk
volt-nuts mailing list -- volt-nuts at febo.com To unsubscribe, go to https://www.febo.com/cgi-bin/mailman/listinfo/volt-nuts
and follow the instructions there.
More information about the volt-nuts