Tektronix AFG2020 repair

[Mysion]

My work place recently threw out a totally dead AFG 2020, A tek function gen, beast of a machine. 


At first glace it looks to be an issue in the early stages of the PSU. Mains make it from the filter to the psu then nothing, not even the hard wired fan.
However by removing power cable I was able to deduce that it might be the CPU board and not the psu. IF a certain 3 wire cable is plunged in AND the cpu board is plunged in it won't do squat. Any other time the fan get power.

The cable in question the the small unplugged orange-red-brown cable right above the large PCB(cpu board). When not plunged in this wave form is present. Interestingly enough when the cable isn't plugged in nothing shows on the display. It's not booting either way it seems.




When plugged in this is what I see. It looks like the signal is being really loaded down, a short I presume. I measured ~100 ohms to ground and 14 mA of current. 



The short to ground made tracing the line impossible. The continuity tester keeps picking up ground pins.... I had a brief moment of happiness thinking I found it only to stupidly discover the pin was just ground... 

I'm running short of ideas of what to try next.  I might try to tear into the psu more to try and find what that pin does exactly. But I'm leaning more to the blow the snot out of it approach. 

I've got a constant current power supply and I'm thinking that if I run 50-100 ma though it by providing a supply it can't pull low I may be able to track down the hot chip.
Interestingly enough the 10 MHz out still works as long as the cable is unplugged. I could use it as a very large 10 MHz reference.

What do you repair guru's think so far? Any idea's? I could reallyyy use a signal gen so it'd be nice if it's not totally dead.
The biggest issue is that I can't find the service/manual for it. I emailed Tek but I don't have high hopes.

[tautech]

Try this link:
http://exodus.poly.edu/~kurt/manuals/manuals/Tektronix/TEK%20AWG%202021%20Service.pdf

[Mysion]

I did come across that manual initial but wrote it off due to some significant differences. That one has a floppy drive and some boards are shaped differently, that sorta of thing. I still can't find the use of that darn cable in it.

Well it's worth a shot comparing the two so I started plugging cables in and out again and HOLY CRAP!! HE BOOTS! 



Can't navigate any menus but I assume that's due to 50% of the unit being on the bench.

Some of the cables must have had a bad connection or some thing. I still have to reconnect tons of internal wiring but this might be it! Going to bed, update in a few hours!
Thank you so much tautech, you provided close enough advice at exactly the right time!

Hmm maybe the cable to the front panel is bad. Still can't use the buttons, would also explain why I couldn't turn it on before.

[tautech]

Some of the cables must have had a bad connection or some thing.

Yep, thats one of the specified checks in Siglent service manuals and it makes perfect sense....reseat all connectors.
Sometimes what looks daunting(pulls cover off....holy shit) often is the hidden easy stuff like connectors, sockets, switches etc.

[lowimpedance]

That is quite an interesting beastie nice score. I like the use of the same format as the early digital oscilloscopes from TEK. 
I wonder given its vintage if TEK had used the infamous SMD electrolytics in this ?. Probably worth a check and perhaps a few more photos of the component side  .

[Mysion]

Hmm I woke up and now while the fan turns on, that's it. The front panel lights are all wrong and not screen (not what the similar manual expects). According to the trouble shooting guide it's either cables or CPU board. I ordered a replacement cable but the constantly changing symptoms have me worried. So far the issue of the fan could be traced to the CPU board and the front panel as well. While not conclusive it seems to be a common factor.

Lowimpedance if you want I can take high res pictures of the boards. My back up plan if the whole thing is kaput is display the 3 boards on the wall sort of like a poster.

[Mysion]

Depending on which part of the front panel I press down on different lights come on. Once in a while the unit even boots, still no controls though. I get a similar effect if I press on it from the rear. It's looking like the board might have a bad solder joint or some thing. I've got a hot air gun and a toaster. Going to try and re-flow it with the gun. 

Hope it works.

[Mysion]

I'm going to hold off on re-flowing it. When I took it apart it looked like the frame was just a bit flimsy and when you press down on it buttons might get pushed down with it. Plus the lack of any visible damage. I wipe all the connectors between it and the CPU board with rubbing alcohol, and re-seated the chip.



Now I'm just waiting for it to dry.

[tautech]

Check those headers for cracked solder joints and don't foget what lowimpedance said about SMD caps, they have been a real problem in some Tek gear.

[free_electron]

Immediatley replace ALL electrolytic smd capacitors on the cpu and gen boards !! these machines are plagued by leaking caps that eat off half the traces ....

[Mysion]

Holy crap I just looked it up. I had no idea the caps were THAT bad.

I took a close look on the CPU and front panel board and it looks like no corrosion yet... I'm going to hold off on testing it until I replace all SMD (electrolytic )caps.

[tautech]

Holy crap I just looked it up. I had no idea the caps were THAT bad.

I took a close look on the CPU and front panel board and it looks like no corrosion yet... I'm going to hold off on testing it until I replace all SMD caps.

electrolytic's

[Mysion]

Haha sorry, bit of a typo. I meant electrolytic, just forgot to put it down.

[tautech]

Haha sorry, bit of a typo. I meant electrolytic, just forgot to put it down.

Just checking you were on your game. 

[Mysion]

Well snap replacing the SMD electrolytic plus the supper cap did not fix it. I did figure out how to get it to the screen in an earlier post though. I have to press the on button as soon as I plug it in. Not sure how that information helps but it's all I have. It looks like its freezing during boot up. Whats odd is that unless I turn it on immediately after power connection it won't even try to boot.

I have no idea where to attack this unit next.... I did check for corrosion, nothing visible.  I'm just going to sit on it tonight and maybe I'll have an idea in the morning.

[Mysion]

Upon closer inspection I did find evidence of corrosion. I checked with the meter and none of the tracks were severed but it's a 6 layer board. Who knows whats going on in the internal layers. I'm calling this unit DEAD. Short of a schematic it is a former AFG2020. I think I'll pick up a siglent SDG1025  instead.

[wjmoltz]

HI, I just joined the Blog and saw that you were trying to repair a Tektronix AFG2020.  I have the same generator with the same problem, but was able to repair it just today.  First I replaced three leaking SMT caps, 22uF 35V.  One was on the front control PCB, and two were on the CPU PCB.  Of course changing the bad capacitors did not solve the problem, because there was also corrosion damage caused by the leaking caps.  I finally located a severed track between Q421 and CR420, 421, and 422.  Repairing this trace restored the generator to working condition.  If you haven't already tossed your generator out, I hope this information helps. 

[tautech]

Welcome to the forum.

Thanks for sharing your findings, no doubt it will help others.