Test Equipment: Oscillocopes

[MarkE]

Actually the parts list in the manual I have identifies all the tantalums and other electrolytics just as "electrolytic". The page numbers I'm citing are the pages of the .pdf file that I downloaded from the electrotanya link I provided earlier. It's a real pain that the manual itself doesn't have page numbers. Perhaps the easiest thing would be for you to download that manual too so we can refer to page numbers-- I obviously don't have the paper copy like you do. You have the advantage of seeing the good drawings of the boards, the ones in the .pdf file kind of suck. But it is nice to be able to refer to page numbers--- you could look in the .pdf file at the page numbers I cite, and then find the correct page in your paper copy.
The box referring to the tantalum decoupling caps is in the Power Supply troubleshooting flowchart for the low voltage (15, -8, 5 v) supplies. The schematic is the "power supply" tab , marked on the tab #12 in a diamond box. And the box referring to the tantalum decoupling caps is on the tab marked "+15V, +5V and -8V supplies Troubleshooting"
Good on the iron. I'd check the voltage at the testpoint first, (read the box on the troubleshooting chart about how they behave as the scope warms up), then I'd replace the caps if you have the right capacitance values. You can use higher voltage ratings of course, if there is room physically. You may even be able to use other capacitance values as long as they are close: the originals are 20 percent tolerance, really cheap tantalums.
I'd say, let the scope cool completely, then power it up, test the  +5V TP1558 with the DMM right away, then let it warm up to the point where the glitching gets bad and test the TP again to see if there is any difference. Then, regardless of what you see, shut down and replace the caps if you've got suitable replacements. Then repeat the process starting with a cold scope again.

Another thing you might like to try, since the transistors are mostly all in sockets, is to wiggle them or tap on them with a wooden or plastic tapper and see if the display responds in any way. If you find one that does cause a change in the display when you tap on it, shut down the scope, carefully pull the transistor and re-seat it, maybe the it wasn't making a good connection and reseating will fix that sometimes. Lather rinse repeat.

Within reason you can generally go larger with bulk power supply capacitors without a problem.  A 47uF or even 68uF capacitor can be used in place of a 33uF in almost all situations.  What you can't safely do is use a lower voltage rated capacitor.

[picowatt]

Just some food for thought.

Whenever troubleshooting any piece of electronic gear, it is always wise to measure the rails first thing, and as expeditiously as possible.  I often pre-attach my scope/voltmeter to supply test points using spring hooks and with the unit's power off and then power the unit on just long enough to get a stable reading.  Power off and then repeat for the next supply test point.  If a supply has gone overvoltage, or is not regulating current, this method may minimize damage.  With regard to the current topic, I agree with TK... measure the rails!  Although tantalums often fail just because that's what they like to do (particularly older gen tants), tantalums do not like overvoltage, or even more so, application of reverse polarity.  So check the rails first thing.     

I just had one of my SG505's in a TM500 rack develop a noise.  Tore it apart and probed the supplies first thing and noted one rail was noisey.  The rail used a 33uf 20V tantalum.  It was one of the old early tants, a big honkin' thing the size of a large peanut M&M with horizontal color stripes (in familiar M&M colors).  There was what looked like a tear dropped shaped solder blob stuck to the side of the tant.  Upon closer inspection, a very small hole had formed in the side of the tant's "candy coating" and the apparently molten at the time metal blob had oozed thru the almost microscopic hole and solidified.  A very light and unfamiliar metal was that blob.  Anyone who has ever worked on electronic equipment that uses tants will have many stories to tell regarding them.  Definitely a love hate thing.

I ordered new tants to recap all my SG505's, but in order to put this unit back in service, I temporarily replaced the tant with a 47uf 35V electrolytic (1/4 the size of the tant) and then tacked a .1uf ceramic to the back of the board in parallel with the electrolytic.  This parallel combination will work well in most applications that use a tant, and some of the newer electrolytics approach or exceed the HF performance of early gen tantalums.  A functioning electrolytic will always perform better than a non-functioning tantalum!

Also, regarding probes.  Yes indeed, expensive probes are nice and sometimes necessary, and I agree with everything MarkE has said about them.  However, when repairing a piece of equipment where one might have to get a bit creative just to get a probe hook attached to a component lead or other measurement point that is not easy to get to, I use the cheap probes.  Most of the time you are looking for a DC level, noise, or just the presence of a specific signal, and the cheap probes are fine for this.  If a probe hook gets bent or broken from somewhat abusive probing, well, it was just a cheap probe, no worries.

Now, if you are looking at a fast rise time digital signal, switching supply, something you cant load, a circuit that is screaming with an unwanted HF oscillation, or you need absolute measurement precision, that is the time to break out the expensive or active probes.  And since you did not use them as your "daily driver", they are always in great shape (and of course you will slow down and drive more carefully when using the expensive probes).

Anyway, like I said, just food for thought...

PW

[MarkE]

Just some food for thought.

Whenever troubleshooting any piece of electronic gear, it is always wise to measure the rails first thing, and as expeditiously as possible.  I often pre-attach my scope/voltmeter to supply test points using spring hooks and with the unit's power off and then power the unit on just long enough to get a stable reading.  Power off and then repeat for the next supply test point.  If a supply has gone overvoltage, or is not regulating current, this method may minimize damage.  With regard to the current topic, I agree with TK... measure the rails!  Although tantalums often fail just because that's what they like to do (particularly older gen tants), tantalums do not like overvoltage, or even more so, application of reverse polarity.  So check the rails first thing.     

I just had one of my SG505's in a TM500 rack develop a noise.  Tore it apart and probed the supplies first thing and noted one rail was noisey.  The rail used a 33uf 20V tantalum.  It was one of the old early tants, a big honkin' thing the size of a large peanut M&M with horizontal color stripes (in familiar M&M colors).  There was what looked like a tear dropped shaped solder blob stuck to the side of the tant.  Upon closer inspection, a very small hole had formed in the side of the tant's "candy coating" and the apparently molten at the time metal blob had oozed thru the almost microscopic hole and solidified.  A very light and unfamiliar metal was that blob.  Anyone who has ever worked on electronic equipment that uses tants will have many stories to tell regarding them.  Definitely a love hate thing.

I ordered new tants to recap all my SG505's, but in order to put this unit back in service, I temporarily replaced the tant with a 47uf 35V electrolytic (1/4 the size of the tant) and then tacked a .1uf ceramic to the back of the board in parallel with the electrolytic.  This parallel combination will work well in most applications that use a tant, and some of the newer electrolytics approach or exceed the HF performance of early gen tantalums.  A functioning electrolytic will always perform better than a non-functioning tantalum!

Also, regarding probes.  Yes indeed, expensive probes are nice and sometimes necessary, and I agree with everything MarkE has said about them.  However, when repairing a piece of equipment where one might have to get a bit creative just to get a probe hook attached to a component lead or other measurement point that is not easy to get to, I use the cheap probes.  Most of the time you are looking for a DC level, noise, or just the presence of a specific signal, and the cheap probes are fine for this.  If a probe hook gets bent or broken from somewhat abusive probing, well, it was just a cheap probe, no worries.

Now, if you are looking at a fast rise time digital signal, switching supply, something you cant load, a circuit that is screaming with an unwanted HF oscillation, or you need absolute measurement precision, that is the time to break out the expensive or active probes.  And since you did not use them as your "daily driver", they are always in great shape (and of course you will slow down and drive more carefully when using the expensive probes).

Anyway, like I said, just food for thought...

PW

One thing to be a bit careful about is that there are a lot of circuits that rely on the ESR of the capacitors for stability.  A really low ESR may break a regulator circuit causing it to oscillate.  Those old "candy wrapper" tantalums tended to have ESRs in the 2-10 Ohm range. 

I like the robustness of passive probes.  A colleague of mine set me up with some really sweet passive probes that I can clip almost anywhere and they perform very politely.  The oscilloscope trace below was snapped with one hooked to a 5V < 1ns risetime CMOS logic circuit.

[picowatt]

One thing to be a bit careful about is that there are a lot of circuits that rely on the ESR of the capacitors for stability.  A really low ESR may break a regulator circuit causing it to oscillate.  Those old "candy wrapper" tantalums tended to have ESRs in the 2-10 Ohm range. 

I like the robustness of passive probes.  A colleague of mine set me up with some really sweet passive probes that I can clip almost anywhere and they perform very politely.  The oscilloscope trace below was snapped with one hooked to a 5V < 1ns risetime CMOS logic circuit.

MarkE,

Yes, but the tants claim to fame when they first arrived was their low inductance/HF performance.  A "one stop shop" for decoupling applications.

What model probe?

PW

[MarkE]

They didn't quite live up to that claim.

My colleague custom made the probes he gave me.  They are quite handy.