Tektronix 5A20N differential amplifier plug-in repair


magnustoelle
 

Good Day to the group,

I have just completed a simple repair of a type 5A20N differential amplifier plug-in for the 5000/5400 series of Tektronix oscilloscopes. There is not much to "write home about" this, but I thought I share my joy with you...

*Problem containment:* The input signal present at either the positive or negative input was not displayed properly. The trace was heavily distorted and shifted grossly over the whole screen when rotating the VOLTS/DIV deflection factor settings.
*Troubleshooting and repair:* Using my DIY-made extender, all supplies and the push-pull output stage composed of Q163/Q263 were measured with a DMM and found nominal. Injection of an external small signal into Q129A/B of the input stage resulted in a reasonable display. Both input protection fuses were found open and were replaced. A dual N-channel JFET Q120A/B (Tektronix part # 151-1049-00) in the input stage was found "toast".
As several other group member have already discussed, this is a hard-to-obtain transistor pair. So, I have used my 5CT1N curve tracer to match two standard "dirt cheap" Fairchild BF245A JFETs from my local stock. I have sanded the flat side of the package and super-glued the two units together. I was pleased to find that with this replacement the plug-in was fully restored.
*Conclusion:* A massive electrical over-load event has taken the N-channel JFET and both fuses at the input stage out.

A full calibration was completed after the repair.

I have uploaded a few pictures to https://groups.io/g/TekScopes/album?id=268328

Cheers,

Magnus


Dave Peterson
 

When I picked up a 5111A to drive the 5CT1N I got in a fried 5440, it was in a rack mount. Andy W. here was kind enough to send me a bench-top case to convert it.

I've since repaired the 5440, and also picked up a 5103N (wrapped around a 5L4N). I now have 3 5000 series scopes! How did that happen!?

In any case, why have I never thought of building a wood rack? I love it!

I'm in serious need of rebuilding my bench to accommodate the equipment I've collected. Rack mounting the 5000 scopes along with a hefty HP 141T is the solution I've needed. A bit of cabinetry is right up my alley. Just another thing to do, eh?

Another nice piece of work Magnus. Thanks for sharing.

Dave


Harvey White
 

I have an audio rack I constructed from a picture that the manufacturer very kindly provided.  It's two rack units with a bridge between them.  They're slightly angled and between them is a work surface.  The bridge section could hold a scope or so should I wish.

You can get bits and pieces of rack hardware without too much problem.  You can find black formica to coat the bridge sections and the inside of the rack sections.  The rack sections themselves are stained plywood.  You can also buy a press-in edge molding. The workbench itself is a green marbled kitchen top counter (without backstop).

Right now, it's got the audio editing setup on it, as well as one of the LCD TVs.

Harvey

On 9/21/2021 11:59 PM, Dave Peterson via groups.io wrote:
When I picked up a 5111A to drive the 5CT1N I got in a fried 5440, it was in a rack mount. Andy W. here was kind enough to send me a bench-top case to convert it.

I've since repaired the 5440, and also picked up a 5103N (wrapped around a 5L4N). I now have 3 5000 series scopes! How did that happen!?

In any case, why have I never thought of building a wood rack? I love it!

I'm in serious need of rebuilding my bench to accommodate the equipment I've collected. Rack mounting the 5000 scopes along with a hefty HP 141T is the solution I've needed. A bit of cabinetry is right up my alley. Just another thing to do, eh?

Another nice piece of work Magnus. Thanks for sharing.

Dave





Ron
 

Magnus,

Great work, thanks for sharing. Which parameters were used to match the JFETs? How close were the parameter readings to get an OK match?

Regards, Ron
-


magnustoelle
 

Hello Ron,

thank you for your kind words.

I have curve-traced the JFETs for their standard output characteristics, i.e. Id vs. Vds, and best similarity of the ohmic, say the linear region and best matching characteristics of all displayed curves.
The two JFETs I have picked as a replacement for the defect dual JFET were matching so closely that I could hardly see any trace shift on the CRT, probably a fraction of a millimeter when switching between them with my adapter. Some samples showed curves shifted by more than half of a screen division or so...

With the 5CT1N one can easily test for the DUT's behaviour under various conditions and check for other parameters, but I kept it easy for this repair.

I am also a big fan of the "test it in the application"-principle. So, with the two "matched" JFETs (glued together) inserted into the 5A20N, I have tested the positive and the negative input of the plugin for how well they are matching with various input signals. Considering the opposite polarity, of course. And they matched nicely across the whole VOLTS/DIV settings range.
For the fun of it, I have used my heat gun and heated the PCB/circuitry up to almost "too hot to touch" temperature. The 5A20N's data sheet specs say its operating range is from 0 to +50°C. And I was pleased to see only a minimum amount of trace drift when the temp. went from room temp. say 22°C up to about 50°C. The trace shifted barely at all, worst case by half a trace, or approx. 1/5th of a subdivision on the screen.

Touching or moving the side panel cover definitely had a much greater influence at the low VOLTS/DIV settings, so I have re-tested the unit without the adapter. I have found that's good enough by my standards and then carried on with the calibration :-)

Cheers,

Magnus

On 22/09/2021 05:12, Ron wrote:


Magnus,

Great work, thanks for sharing. Which parameters were used to
match the JFETs? How close were the parameter readings to get an OK
match?

Regards, Ron
-