Tektronix TDS320 boots intermittently, with Front panel dead or alive etc

[morzh]

I got it from work; they wanted to get rid of it.

When turned on, it displayed the traces but no controls: all LEDs were lit and no reaction to any button/encoder, other than the power-up button.
Saw some oxidation on the flat cable connector solder pads on the Front panel board; touched up, no effect.
Eventually, after one of the start-ups, it worked. Then worked every time. Then, when left for an hour or so, started doing the same, this time not showing traces either. Then eventually, after many restarts, worked.

The voltages are all good, as per the table in the manual.
The troublehsooting chart points to main computer board, as the LEDs, at least some, or all, are lit.

The computer board has DS1230AB-120 Dallas chip, which, considering the scope is at least 30 years old, probably has its lithium source long dead. However it keeps the settings (mv/div, timing, the traces displayed), so maybe it is not dead yet.
Changing the computer module, as well as any other board, makes little sense due to the fact that the whole scope might sell for comparable money.

I am running continuous diagnostic until fail now. Not sure this will yield anything.
Anyone has experience with these?



PS. The test stopped, the message displayed was with all 5 tests "PASS". I switch it to run forever.

[morzh]

I found out that during hung state the TXRDY signal from the FP board to the Main board is in some indeterminate state, bouncing around 0.5V state with some sort of contention-like signal.

I proceeded back to the FP board, which I thought should be OK, as the troubleshooting diagram pointed to the Main, and found a few more oxidized points, including two electrolytic caps pads, vias and such.

I only have solder tweezers at work, so will have to wait to remove the capacitors.

[Swake]

Indeed corrosion.... might come from that cap, or not, it looks like it was stored in a humid environment before.

You do not really need tweezer irons to remove those caps. But it is destructive:
 

[morzh]

Indeed corrosion.... might come from that cap, or not, it looks like it was stored in a humid environment before.

You do not really need tweezer irons to remove those caps. But it is destructive:
 

Not sure about the environment; it was given to me within the company, and at that point it was not the outside storage items; it was sitting on someone's desk, and we needed more workspace.
I think it is the capacitors: the oxidation points are fairly close to them, and I think I saw a similar discussion about the same scope either here, or on some other forum, and same thing happened: the caps leaked and etched a trace.



As for the tweezers - no, I do not necessarily need them, but I need two soldering tips touching the pads at the same time. The reason being, as I might've said before, I was not impressed with the Tek's PCB routing skills (I am an EE, and one of the things  I do is PCB routing), especially the way their routing very thin traces into rectangular pads without support vias. Lifting one side off its pad almost inevitably will result in the other side starting pulling its pad off the board. And the aforementioned 5-mil trace into rectangular pad will result almost inevitably in that pad breaking off of that trace.  The cap is short, and so the angle will be large (much longer cap would result in smaller lifting angle and in lesser risk of the pad's detachment). The tweezers are low power, but this is where the bad routing, think traces, small pads and no vias will work in my favor. Or so I hope. (I could also try to use a Milwaukee gun, but that could stress the board).
(sigh) - I need a second soldering iron at home. 

PS. Repairing that scope is more of a sport thing. For my hobby of old radio restoration, even a 20MHz scope is overkill. I just like fixing things.

[morzh]

So, today I popped up at work (I am on vacation) and then I had time to use solder tweezers to remove the aluminum electrolytics that leaked.
I measured them with my Fluke, and it refused to tell me the value. It was kinda infinite. Despite them not being shorted.
So, I cleaned the board again, touched up some vias, put a bandage on the trace that got etched right by the via (invisible but the continuity was not there), though the trace is from the encoder and would not result in the issue I was seeing; then I replaced the electrolytics (33uF 10V) with X5R 22uF 25V caps: the value is sufficient and the size allowed me to solder them to the footprint.
C131 and C132.
The bandage is seen at the right upper corner of the CPU.

So, full of hope, I installed the board and...same issue. Wakes up in indeterminate mode.

The problem is, there is no schematic for the Front panel board. I found the sch for a similar scope, but it is not exactly the same. But the interface is the same, and so I decided to look at the last signal I never looked at, and which I should've done, considering the CPU wakes up in different modes: the RESET.
Which I did.
It was floating close to the Low level. And...it did not connect to any of the the connector's pins, especially to the pin 9, which is the RESET.
I traced the trace.  I could not see under the connector, and it was mu guess that the pin 9 should connect to the processor's pin 18 (RESET), which has RC to GND on it.

So, this is the via, after which I could see the beginning of the stup, that supposedly went to pin 9, and which was the last point that would show connection to that RESET pin of the CPU. The via is pointed to by the red arrow. And the Pin 9 of the connector (RESET) is pointed to by another arrow.

So I put a short between them on the bottom side.
I installed the board, and it immediately worked!

The photos of the new caps, the via and the pin that it was not connected to, and the jumper that I installed are in the photos.

[Swake]

Good job 

Scary how those fine traces 'disconnect' without it even be noticeable with the naked eye. Very fragile stuff. Might, or not, be due to that corrosion.
In the pictures there are still spots of 'blue/green', what I think is some remaining corrosion. I suggest using liberal amounts of IPA or similar to clean it out.

[morzh]

I tried. Lots of IPA, and probably half-box-full of cotton swabs.
Plus Exacto blades Every one I saw I tried to clean and then touch up.
Hopefully that blue - green is by itself and no longer in contact with the copper; plus there is no source of it anymore, with the lytics removed.
Yes it was due to that corrosion, as all the visible spotting is close to the leaked aluminum caps; what helped it was the caps are at the very top of the board, and gravity pulled it down. Of course thhere was some of that stuff going up a bit, (including that trace under the connector) which I attribute to the electrolyte wetting the board material and spreading due to it being "philic" rather than "phobic" and all that good Van der Waals stuff.
Anywho...I am 63, and so if this one lives past me, that's all I need for it to do 

I assembled it, checked all the buttons; all seem to work.
Hopefully, this concludes it.