Repair of Wavetek 132 signal generator

[angdis]

I got an old wavetek 132 function generator off Ebay that I am attempting to repair for entertainment and troubleshooting practice. I got one that was advertised as "no power up", so I can't have fun with it until I fix it. I just wanted to bounce some basic questions off of the very nice folks here to get ideas about how to proceed with the troubleshoot.



The first thing I checked was the fuse: OK. then power rails: Not OK! This unit has a power supply on the main board that provides +/-22V, +/-15V, +/-6V to the device. The +15V rail was only 7V. The other rails were off as well but not as much. Below is a screenshot of the power supply portion of the schematic.



The next thing I did was to remove the other two boards. One was a digital board that is used to produce noise, the other was an amplifier board that drives the output of the device. What remained was the mainboard and power supply. I checked the voltage rails again, and they were still the same (only 7V on 15V rail). Whatever problems there are, I think it is safe to say that at least ONE of them is on the mainboard and it is bringing down the rail.

At this point, I noticed that there were wire jumps on the mainboard that connect the +/-15 V rails of the power supply to the rest of the board (see TP2/TP3 in schematic). I cut these and measured the voltage rails again. This time the +/-22V and +/-15V rails were perfect (after minor adjustment of R105). I could not check the +/-6V rails because they're generated after the jumper.


So here's my questions:

• Do I need to provide a load to this power supply to really ensure that it is functional? The power supply I think is not a switch mode. There are a couple of opamps (IC6,IC7) that appear to control power transistors (Q17,Q24). Is there a name for this particular type of power supply? Unloaded, the power supply is producing the correct voltage. Is that "good enough" to proceed with troubleshooting beyond the power supply?

• I can foresee needing to isolate the different functional circuit "blocks". This means I need to cut power rail traces and restore them as each area is fixed/cleared. My concern is that if I just cut power rail traces on the PCB, I will be delivering signal from powered portions of the board into unpowered portions. Does this mean I need to cut signal traces AS WELL AS power rail traces? That will make it all more time consuming. The signal chain is really simple: an integrator that is fed into a comparator which feeds back to the integrator, toggling polarity to generate a triangle wave. The triangle wave is then fed to a diode shaper circuit that fakes a sine wave, after that there's an amplifier.

I should add that there are some electrolytic caps in the device. I am going to replace them, but I wanted to proceed as far as I did now because the power supply caps seem to be OK. Somebody replaced these in 1995 and wrote it on the PCB! I also notice that a few matched diodes (in a can "CA3039") where replaced with individual diodes. So someone has been "in there", but that is further downstream.

Many thanks for any advice you may have !

[SeanB]

Check R90 is not open circuit. If it is there will be low 15V rail at any load over around 5mA as Q19 thinks there is a short on the rail. Measure the current through jumper J, it should be pretty low, under 50mA is a good guess.

[JacquesBBB]

This reminds me my first steps in electronics last year.  I spend  two weeks at christmass on a Wavetek 164 to fix it (I did not know this forum at the time).

There were several problems, starting with the power supply, where one of  the TIP30   was down (I replaced it by a TIP126 which I had). 
I did not changed the caps as  although they were the original ones, they looked OK.
I had also to replace two transistors in the triangle part.

The very good thing was a fantastic
service manual which  provided a detailed description of all the procedures to search for faulty components, and at least to
isolate the regions of the problems. I hope you have the same for this model.

I have not seen such a nice service manual in the next devices I tried to fix, but this time I was helped by the forum.

[TimFox]

My first guess would be bad electrolytics on the +/- 15 V rails in the circuitry not shown in your excerpt, since the + 15  V rail is OK with the two wire jumpers disconnected.  Unfortunately, everything on the rails is probably in parallel, and it is hard to isolate the problem without brute force disconnection of parts or very specialized equipment.  Replacing old electrolytics is generally a good idea.
Note that the -15 V regulator is slaved to the +15 V rail to make a balanced supply, so anything bad connected to +15V will take down the -15V rail proportionately.

[angdis]

Thanks everybody:

I will apply a light load to the power supply while it is disconnected to confirm that it is operational (that the rails remain steady), but first checking R90 as SeanB suggested.

TimFox, yes, I can see that the neg rail is referenced from the positive rail, thanks!

The very good thing was a fantastic service manual which  provided a detailed description of all the procedures to search for faulty components, and at least to
isolate the regions of the problems. I hope you have the same for this model.

Yes, that was one reason I decided to attempt this repair. These old wavetek manuals are so concise and easy to follow. Also, the drafting on the schematics is remarkably clear and consistent despite being hand-drawn. The penmanship of the "Mr Gray" (the engineer whose name is on the schematics) is something you just don't see anymore.