Guildline 9574 Multimeter Transistor Debacle

[Thunderbeam]

A while back, I picked up this really neat, somewhat obscure Guildline 9574 Multimeter with the full, original service and calibration manual for next to nothing. Recently, I noticed a bunch of erratic readings measuring sub 10 volt levels as well as absolute garbage on resistance. The source of my erratic readings came from 7 worn out filter capacitors (made in West Germany, even!), which was an easy enough repair… but then came the resistance.

Below, I’ve attached pictures from the relevant manual pages including the circuit description.

TR602 and TR603 are cooked, they’re both open circuit. The cause of that failure was from some whackiness on the power supply that put -36V on the -15 rail because a failed 79M20 (Don’t worry, I didn’t know about the 7820 and 7920s either until this thing, and I actually found some!)

TR602 and 603 are listed as WN1001 NPN transistors, but there is next to nothing I can find on any WN type transistor, let alone that. As weird as it sounds, it almost seems like it’s around a 2N3762 OR possibly a Toshiba RN1001 equivalent? I know the RNs are NPNs with built in bias resistors (4.7K/4.7K) and are much more specific than the more “generic” 2N3762… but without a working circuit to see what’s going on, I’m afraid I’m not experienced or knowledgeable enough to work this out.

If someone could help me work this out I sure would appreciate it. I don’t want this meter to be relegated to parts over something as silly as two transistors.

[coromonadalix]

check PN4118A for the  specs ... even this one is discontinued   used in solartron 7150

https://pdf1.alldatasheet.com/datasheet-pdf/view/84636/VISHAY/PN4118A.html

2N4117A    PN4117A   SST4117
2N4118A    PN4118A   SST4118
2N4119A    PN4119A   SST4119         


and this page may give you some similar ones  further down

[Thunderbeam]

It doesn’t work, I replaced the 2N4118A (TR601) because it was also shot, and I just so happened to have some from a previous repair I did. I put those in place of the WN1001s and it was giving me better readings, but it wouldn’t even measure my 10K standard without floating around between 9-11K the entire time. Prior to that it was dead on. I’m not sure what’s so special about those transistors but apparently Guildline found the most obscure one for the job.

[lowimpedance]

TR602 and 603 are listed as WN1001 NPN transistors

Just to be sure there's no reader confusion , these are N channel FET's not BJT's !

See the below thread about the Solartron 7150 , that coomonadalix referenced. Even though they are different DMM's there is some vague similarities that might give some hints.

https://www.eevblog.com/forum/repair/help-with-repair-of-solartron-7150/

I have repaired a few HP3456 DMM's that use similar FET switching with LM339 open collector output comparators and those can go faulty causing 'whacky' operation, so check IC604 for correct operation and IC601 too.

Edit : After looking up a photo of this multimeter it appears to be a rebadge Solartron 7060 systems multimeter !
Perhaps that model could broaden search results.

[Kleinstein]

The circuit is not very high impedance. So leakage current should not be that critical. At least for a first test one could use common JFET switches like  J113 or PN4393 or what ever THT part are still available.
I also don't see the speed being that relevant. The point is getting JFETs that can withstand a relatively high voltage.  If desparate, SK208 SMD parts would be candidates that at least have specs for 50 V.

+1 for checking the LM339 comparator.  There are quite some reports of those chips going bad for no good reason also in other instruments.

[Thunderbeam]

It’s kind of a shot in the dark, but after going through that thread as well, I’m thinking of trying a LS846 as a replacement, it’s a 60V part that seems like a darn good substitute for the mystery JFET. (See attachment)

I’m also wondering if MAYBE the LF365 was damaged from the over voltage condition. I didn’t think about it until last night, but theoretically, would that also cause the roughly 2k drift I experience if the op amp inputs were cooked?

846s are cheap, so I think I’ll pick a few up and give them a shot.

Thanks for all the help, I’m excited to document this more and keep more of these old instruments alive. Also, a quick shoutout to the team who designed and built that thing. It’s a gorgeous unit inside, and surprisingly straightforward once you figure out why they did what they did.