How did I get it
I was searching for a curve tracer since a couple of years, but all offers I found were way to expensive or so far away so that the shipment costs of this heavy gear would immediately kill the deal.
I wanted to have a 577 since it is already transistorized, but still can go up to 1.6kV in collector voltage, which is necessary for some areas of device testing.
Just recently I found one on eBay – declared as “for parts” – and just gambled that I can fix it. It came with the necessary set of fixtures for transistors, diodes and power transistors and was offered as auction for a reasonable start price.
What did I get
To my surprise the unit was fully working – just some glitches in the switches made the beam disappear, which is a common problem with these devices.
Mine is dating from early 1973, but not all of it is in its original state. On the picture you can see the upper part of the instrument has a D11 storage unit – the top part of my unit has been borrowed from an 5111 oscilloscope, which you can identify because of the probing element, which is useless in a curve tracer.
However this is nice cause the tube is burn-mark-free and very precise. The storage function is a bit weak (even after adjustment), but generally speaking useable.
Parts of the storage board already have been overhauled since new, modern trimmers have been used to replace the original ones.
After a few tests the following todo-list was clear:
- Full cleaning of all switches and contacts
- Full calibration of the 177 and 577 units (which also is a good test if everything is working)
- I found some briddle contacts which have to be replaced
- I noticed a “wobble” or “flickering” in the step generator, which has to be fixed
Cleaning and calibration
I cleaned all switches with de-oxid (Kontakt-60) and a de-oxid-remover (Kontakt WL) and this made already a big difference. Actually I started with the 177 base fixture because switches and pots are easily accessible in this part of the device.
But of course and unfortunately all switches have to be cleaned, especially the big wafer switches which are deeply in the measurement path.
For calibration and adjustment I strictly followed the documentation in the service manual – here it is extremely important to read the settings description carefully and to do chapter by chapter. Some of the settings rely on previous settings and you will miss them.
There is an excellent video on YouTube from Zenwizard Studio which you should watch upfront.
I did the power supply alignment first (+30V,-30V) since this affects everything. Then I did the adjustment for the 177 (there is a separate service document) and then the full 577 alignment. According to Zenwizzard the 177 will otherwise defeat the 577 calibration.
Actually my device was completely off in almost all parameters – but everything could be realigned.
Briddle Connectors
The connectors used in the instrument are Tek-manufactured and they are not available any more. The TekScope-groups.io-pros call them harmonica connectors.
There are a couple of options to replace the stuff, in my case I only replaced the housings with DuPont housings. I left the metallic contacts as they were and just pushed them in the new housings. This is not 100% perfect, but way better than the broken housings before. Just find our replacement discussion here.
However the one I needed to replace were not DuPont compatible. They have a 4mm spacing and not 2.54mm. For now I secured the two faulty ones with some Kapton-tape.
Step Generator Issue
This became the hard part as you can see in the picture below – all four oscilloscope channels used!
The curves on the screen were not stable when using the step generator (e.g. with a FET or BJT) and you could see that sometimes an additional curve was displayed, which then disappeared again.
The step generator is synchronized to the mains frequency, since the collector supply is also based on the mains frequency – and both have to be in sync to provide a stable picture.
The three speeds (slow, norm, fast) you can select are created by extracting the high and/or low points of the sine wave of the collector supply voltage and selecting some of them to advance the generator one step.
To analyze the problem, I traced the step generator output at TP305, which is the output of the integrating op-amp U305A (amber trace). Blue is the reference waveform (so the collector supply), yellow is the U305B output and green is the 74121 monoflop output U270.
As you can see, with a bit of clever triggering I could identify that the last “step” was sometimes too short which then resulted in the whole step system getting out of sync – hence the flickering. But why is this pulse sometimes to small?
As a first step I checked all zero crossing detectors (Q210/Q2012 and Q260/Q262) and the reference waveform: No issues. Tip: you have to trigger on the reference waveform, otherwise always something prevents to flicker around.
The “end of steps” op-amp U305B an the following RS-flip-flop (U220D and U220B. U220C in the schematics is wrong) create small state machine, which allows the step generator to “finish” the last step although the U305B already signalled “level reached”
Above you can see some of the involved signal in one of the perfect step runs.
And below a faulty one:
When you look very carefully on the green line, you can see that there is a small glitch where it goes low and immediately up again. The going down is correct: this is the same duration as for all other pulses and determined by the 74121 mono flop and its external components (1.8ms). But the going up again is a fault.
I checked all inputs to see if this glitch is a result of an external event or signal, but everything is fine, even the power supply lines. So last resort is a somehow faulty 74121. Replacing this, fixed the issue asap.
That’s how it should look like!
What’s next?
Now I have to put everything together, clean a bit the outside and wait for my safety cover replacement I designed….but this is for another post! Until the I will enjoy these perfect trace of a BC547B
