TM-500 Test Module Question


 

I recently completed one of Jared's fine test modules and after using it to discover bad and very bad filter caps in my TM-500 frames, and replacing them, I proceeded with re-testing and ran into a puzzling (for me) issue. When testing the MJF2955 PNP transistors in my TM502 for leakage at Step IV, the voltmeter readings only drifted down to about 18 mV, instead of below 10 mV as expected. I ordered new transistors and installed one, but obtained the same 18 mV minimum reading.

As the test unit circuitry is a little difficult to trace across multiple boards, I thought it easier to ask other constructors if they might have encountered this condition before looking for a fault in my test unit.

The low side of my oscilloscope BNC connector is grounded at the aluminum front panel and not connected to the selector switch.

Any thoughts?

Bruce, KG6OJI


Jared Cabot
 

I would first correct the BNC to eliminate that from the equation.
I used a scrap of PCB behind my BNC to help position the connectors during assembly (technically not needed so no worries to not have it, besides the BNC grounding thing) so I didn't catch the grounding issue some are seeing.
The fix is simple, just use an insulating washer or similar insulator of some type behind the BNC to insulate it from chassis, and install the small wire from the BNC shell to the provided pad on the Front Panel PCB.

You can see the details in these photos of my unit I uploaded to the photos section.

https://groups.io/g/TekScopes/album?id=263748


Let us know how you go!


 

Hi Jared,

Thanks for your kind assistance.

The shell of my BNC scope connector is firmly grounded, but only its center conductor is connected to the PC board. As there was no connection to the BNC connector while testing the transistors, I do not think the BNC connector influenced the results.

Testing of NPN transistors works as expected with the voltmeter reading dropping to below 10 mV. Thus it appears I may have a problem with the PNP side of the circuitry in the test module. I may have interchanged some resistors during assembly.

The C-E path of the PNP transistors appears completely open when tested at low voltage with an ohmmeter; however, I will check them with a megohmmeter at an appropriate voltage to be sure the transistors are not a fault before looking further into the test module.

I have a BNC connector with insulated shell that I can substitute for the grounded panel connector, allowing it's outer and inner conductor terminals to be connected to the Low and High meter binding posts respectively.

Regards,

Bruce, KG6OJI


Mac Perkins
 

I noticed something similar, some slots in some of my TM5XX frames drop down below 10 mV, others don't, with several in a late production TM506A that won't drop below 50mV.
I have not torn into that frame yet as it passes all other tests with reasonable beta on the transistors. I will be interested to see what transistors are in there.

I agree that connecting the BNC shell to ground makes the most sense, if it is floating and connected to the METER LOW terminal and a scope with a grounded input BNC (virtually all AC line powered units) is connected, the METER LOW terminal will be shorted to ground via the power connection safety ground.
The original construction note does not show a connection for the BNC shell.

I have tested about half of my frames and found several with issues, mostly pass transistors, but one slot with a phasing problem which I had suspected but was too lazy to test it.

Thanks again to Jared for making building the tester so painless and Larry for handling the group buy.

-Mac


Colin Herbert
 

Mac,

While I agree with you that the original Tek construction note doesn't show a connection for the BNC shell, it seems clear to me that while "normal" practice would be to ground the shell of the BNC, Tek might have had a good reason to ensure that the "Meter LO" terminal and the BNC shell were connected. The "LO" terminal is connected to ground in all switch positions except 2 and 5. These are only in operation when Beta of the pass transistors in the mainframe is being checked. It is my contention that Jared is correct in connecting BNC shell to the "LO" terminal and not to ground; after all, on pages 6 & 7 of his updated document, he does point out that the oscilloscope should be disconnected to prevent erroneous readings when checking Beta of the mainframe pass transistors, as it is in positions 2 and 5 of the switch.

I have to admit that I find the switching arrangements mind-boggling and that others with better observation, reasoning and knowledge might disagree with my contention. Of course, when Tek originated this tester, there were no cheap transistor checkers available and so simply bringing the terminals of the pass transistors out to front-panel connectors wasn't an option, but it could be these days. I have "bodged" such an arrangement in the past and have an idea to do a more permanent arrangement with another junk plug-in sometime in the future.

Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Mac Perkins
Sent: 05 May 2021 06:10
To: TekScopes@groups.io
Subject: Re: [TekScopes] TM-500 Test Module Question

I noticed something similar, some slots in some of my TM5XX frames drop down below 10 mV, others don't, with several in a late production TM506A that won't drop below 50mV.
I have not torn into that frame yet as it passes all other tests with reasonable beta on the transistors. I will be interested to see what transistors are in there.

I agree that connecting the BNC shell to ground makes the most sense, if it is floating and connected to the METER LOW terminal and a scope with a grounded input BNC (virtually all AC line powered units) is connected, the METER LOW terminal will be shorted to ground via the power connection safety ground.
The original construction note does not show a connection for the BNC shell.

I have tested about half of my frames and found several with issues, mostly pass transistors, but one slot with a phasing problem which I had suspected but was too lazy to test it.

Thanks again to Jared for making building the tester so painless and Larry for handling the group buy.

-Mac


Jared Cabot
 

Hmm, ok. Now I'm second guessing myself here. :D

When I designed my unit, I had interpreted from the original manual that the BNC was connected across the HI and LO terminals, like the multimeter when connected.
Now I'm not so sure....

It's not a big deal to configure one way or another, either isolate the BNC shell from chassis and install the wire to the PCB for connection to the LO terminal, or remove the wire and connect the BNC to the chassis.

I think I'll do some testing in each configuration and see what works best. I have a couple TM501 mainframes on my bench ready to be tested right now by chance.


Colin Herbert
 

Hi Jared,
I have just taken a look at the photos of the plug-in tester on Tekwiki and it looks rather like the BNC shell is connected to the ground simply by virtue of the fastening to the aluminium front panel. Take a look here:

https://w140.com/tekwiki/images/c/c2/Tek_067-1201-99_3.jpg

I think I should have taken a look at this earlier...

I recall also that Dave Casey of this Forum has an original which he says looks exactly like the one on Tekwiki. Perhaps he could be persuaded to check if his BNC shell is grounded at all of the switch positions?

Colin.

-----Original Message-----
From: TekScopes@groups.io [mailto:TekScopes@groups.io] On Behalf Of Jared Cabot via groups.io
Sent: 09 May 2021 16:35
To: TekScopes@groups.io
Subject: Re: [TekScopes] TM-500 Test Module Question

Hmm, ok. Now I'm second guessing myself here. :D

When I designed my unit, I had interpreted from the original manual that the BNC was connected across the HI and LO terminals, like the multimeter when connected.
Now I'm not so sure....

It's not a big deal to configure one way or another, either isolate the BNC shell from chassis and install the wire to the PCB for connection to the LO terminal, or remove the wire and connect the BNC to the chassis.

I think I'll do some testing in each configuration and see what works best. I have a couple TM501 mainframes on my bench ready to be tested right now by chance.


Jared Cabot
 

Yeah, I connected my BNC to ground and it all seemed to work properly, I think that is correct so I'll update the manual (yet again... :P ), and the gerbers too to reroute the pad for the BNC shield from LO terminal to GND.
Allowing the shield to fasten to the aluminium front panel and making no other connections would work fine and is the easiest mod for existing testers.


 

So if I have a non-conductive front panel, with the BNC ground wire installed, what should I do?

Thanks
David

-----Original Message-----
From: TekScopes@groups.io <TekScopes@groups.io> On Behalf Of Jared Cabot via groups.io
Sent: 10 May 2021 07:43
To: TekScopes@groups.io
Subject: Re: [TekScopes] TM-500 Test Module Question

Yeah, I connected my BNC to ground and it all seemed to work properly, I think that is correct so I'll update the manual (yet again... :P ), and the gerbers too to reroute the pad for the BNC shield from LO terminal to GND.
Allowing the shield to fasten to the aluminium front panel and making no other connections would work fine and is the easiest mod for existing testers.


Michael W. Lynch
 

On Mon, May 10, 2021 at 03:54 AM, David C. Partridge wrote:


So if I have a non-conductive front panel, with the BNC ground wire installed,
what should I do?

Thanks
David
David,

Great Question! I have the same issue.

--
Michael Lynch
Dardanelle, AR


Jared Cabot
 

With a non-conductive front panel, simply remove the wire linking the BNC shell to the Front Panel PCB and re-route it to one of the two lowest pins on the connector between the Main PCB and the Front Panel PCB. (The two closest to the bottom panel of the module)

Now you have the shell of the BNC connected to ground with a genuine bodge wire! :) (and my apologies for the bodge too :) )

I just performed this mod on my unit and it was trivial to do with the only dissassembly needed being the removal of the side panel for access.


Michael W. Lynch
 

On Mon, May 10, 2021 at 10:44 AM, Jared Cabot wrote:


With a non-conductive front panel, simply remove the wire linking the BNC
shell to the Front Panel PCB and re-route it to one of the two lowest pins on
the connector between the Main PCB and the Front Panel PCB. (The two closest
to the bottom panel of the module)

Now you have the shell of the BNC connected to ground with a genuine bodge
wire! :) (and my apologies for the bodge too :) )

I just performed this mod on my unit and it was trivial to do with the only
dissassembly needed being the removal of the side panel for access.
No Worries, Just wanted to make sure that I was understanding what needed to be done. I do not mind the "bodge" as you call it. It would not be the first time that I have done such a repair or modification.

--
Michael Lynch
Dardanelle, AR


Jared Cabot
 

Ok, I just updated the photo album with a description of the fix for future reference. This should make it clear how to perform the fix.

Let's call it a 'service update', just like you find with all good test equipment... :)


https://groups.io/g/TekScopes/album?id=263748