Topics

Smoking 2465B (Rifa/30-ohm) + Recap = PS will not start

tekscopegroup@...
 

Yesterday was using the 2465B (SN+50K) and a loud POP and a puff of smoke came out the back. After inspection sure enough I spotted the remains of one of the RIFA caps. The 30-ohm resistor in series with it was also toast. I have been putting off the full recap of this scope for a while now, with new Mouser parts waiting in a bag. Since I am on 120VAC (more like 127-129V in my area) I figured I had ample time, although the Rifa caps looked quite crazed last time I saw them when I did the NVRAM/A5 work. So maybe now would be a good time to recap, not just fix the damage.

Full recap, only one cap at a time to avoid mixups or mistakes. The only ones I did not have was the two 2200pf 250VAC, one 0.01uF 250V and one 0.056uF 250V film caps, back then forgot to order them. The two 2200pf look a bit crazed, definitively will replace them asap. The 0.068 Rifas where replaced with 0.1uF X2 300V Kemet Safety caps that where suggested as a better substitute in one of the cap lists.

Also replaced the 4.7uF 10V Tantalum with a fresh 4.7/50V same composition type unit. The 3x 1uF 50V BP caps where also replaced, and the new ones are much smaller, tiny, in comparison to the originals so hope they still are ok to use (Mouser PN 647-UMP1H010MDD1TP in case anyone wants to check). Plus two new beefier 15-ohm 3W AC input resistors, and 2W for the 270K on the power supply. Even replaced the two Sprage main caps with 330uF 250V Radial Nichicons by extending their negative wire to reach the board. The originals still look pristine but since I already had the new ones on hand went ahead and swapped them as well.

BTW, all of the tall blue caps (180uF and 250uF) where more or less starting to leak from underneath, not much residue on the board, but definitively the smell when desoldering some of them was the biggest tell, and all of them had a slight oily residue between the leads. C1111 was the worst offender. All other caps overall looked quite good, have not tested any of them yet.

Before putting the PS back together double-triple checked that all caps where at the right locations, correct value and polarity orientation.

Upon scope power on the fan runs, all the front panel indicators flash only once in unison and then turn off, except for the green Trig/Sweep led staying on. The 3 GPIB Leds stay on continuously. Nothing else happens. Nothing on screen, no trace, nada.

I checked that all cables in/out of the power supply where plugged in correctly: the one going to the A5 is foolproof. The other two connectors on the power supply (the one with the multi-colored cables under the fuse and the other that only has white cables next to it) are properly seated and aligned with the pins on the board. The only other plug gors to J2474 has 3 small cables going towards the GBIB board and it is orientated correctly (printed letters/numbers up). Checked continuity of all the "stab" pin connectors as I understand they sometime can create a loose/intermittent contact if inserted the wrong way. All pins checked out fine.

Here are the the rail voltages at J119:

+15 -> +10.4V
-15 -> -11.7v
+5D -> +4.3V
+5A -> +3.3V
-5 -> -3.5V
-8 -> -5.3V
+10 -> +7.3
+42.4 -> +28.8V
+87 -> +59.6V

Needless to say all rails are way low. I even measured the voltage across the two new main caps on the J231A/B connector and I do read about 150V on each, total a bit over 300V, so I'm thinking that juice should in fact be getting into the PS.

Verified that the voltage selector switch was securely connected to the two bottom tabs, and the wires from fuse holder and line filter go to the remaining two other taps further up the board. Just checked continuity of the voltage selector switch at the tips of the connectors that go into the board and its almost zero ohms as expected (115V position). I would think that if this switch or the wires where not connected properly, the PS would assume 240V input, and with only 120 going in it might create a low voltage situation like the one I seem to be having. But no dice.

At this point I have the PS out again and on the bench, no way to safely poke into it while inside the scope. First thing again checked all the new caps and everything is confirmed correct. At this point was also looking at the PS troubleshooting chart in the SM, and the W1060 test jumper procedure. But it requires a test load for which at this point I don't have the correct specified resistors at hand. BTW just to know, what is the spec of the diode and voltage of the 3.3uF cap used for the load? Manual only provides Tek PNs.

I would think that with the exploding Rifa cap and burned resistor being well ahead of the actual power supply nothing else would have been damaged. Not even the main fuse blew, and I think it was just time for the cap to let go, and not because of an incomming huge spike or surge although could not be 100% sure. But I had my 2247A also turned on and some other equipment at the same time and on the same AC circuit, and everything else its fine. Nothing else around reacted to any abnormalities in the AC line either.

Any words of wisdom as always much appreciated.
Alex

Chuck Harris
 

Check the connection of the faston connectors that
connect the 110/220V selection switch. It looks like
you forgot to connect them back up when you connected
the supply back up.

-Chuck Harris

tekscopegroup@... wrote:

Yesterday was using the 2465B (SN+50K) and a loud POP and a puff of smoke came out the back. After inspection sure enough I spotted the remains of one of the RIFA caps. The 30-ohm resistor in series with it was also toast. I have been putting off the full recap of this scope for a while now, with new Mouser parts waiting in a bag. Since I am on 120VAC (more like 127-129V in my area) I figured I had ample time, although the Rifa caps looked quite crazed last time I saw them when I did the NVRAM/A5 work. So maybe now would be a good time to recap, not just fix the damage.

Full recap, only one cap at a time to avoid mixups or mistakes. The only ones I did not have was the two 2200pf 250VAC, one 0.01uF 250V and one 0.056uF 250V film caps, back then forgot to order them. The two 2200pf look a bit crazed, definitively will replace them asap. The 0.068 Rifas where replaced with 0.1uF X2 300V Kemet Safety caps that where suggested as a better substitute in one of the cap lists.

Also replaced the 4.7uF 10V Tantalum with a fresh 4.7/50V same composition type unit. The 3x 1uF 50V BP caps where also replaced, and the new ones are much smaller, tiny, in comparison to the originals so hope they still are ok to use (Mouser PN 647-UMP1H010MDD1TP in case anyone wants to check). Plus two new beefier 15-ohm 3W AC input resistors, and 2W for the 270K on the power supply. Even replaced the two Sprage main caps with 330uF 250V Radial Nichicons by extending their negative wire to reach the board. The originals still look pristine but since I already had the new ones on hand went ahead and swapped them as well.

BTW, all of the tall blue caps (180uF and 250uF) where more or less starting to leak from underneath, not much residue on the board, but definitively the smell when desoldering some of them was the biggest tell, and all of them had a slight oily residue between the leads. C1111 was the worst offender. All other caps overall looked quite good, have not tested any of them yet.

Before putting the PS back together double-triple checked that all caps where at the right locations, correct value and polarity orientation.

Upon scope power on the fan runs, all the front panel indicators flash only once in unison and then turn off, except for the green Trig/Sweep led staying on. The 3 GPIB Leds stay on continuously. Nothing else happens. Nothing on screen, no trace, nada.

I checked that all cables in/out of the power supply where plugged in correctly: the one going to the A5 is foolproof. The other two connectors on the power supply (the one with the multi-colored cables under the fuse and the other that only has white cables next to it) are properly seated and aligned with the pins on the board. The only other plug gors to J2474 has 3 small cables going towards the GBIB board and it is orientated correctly (printed letters/numbers up). Checked continuity of all the "stab" pin connectors as I understand they sometime can create a loose/intermittent contact if inserted the wrong way. All pins checked out fine.

Here are the the rail voltages at J119:

+15 -> +10.4V
-15 -> -11.7v
+5D -> +4.3V
+5A -> +3.3V
-5 -> -3.5V
-8 -> -5.3V
+10 -> +7.3
+42.4 -> +28.8V
+87 -> +59.6V

Needless to say all rails are way low. I even measured the voltage across the two new main caps on the J231A/B connector and I do read about 150V on each, total a bit over 300V, so I'm thinking that juice should in fact be getting into the PS.

Verified that the voltage selector switch was securely connected to the two bottom tabs, and the wires from fuse holder and line filter go to the remaining two other taps further up the board. Just checked continuity of the voltage selector switch at the tips of the connectors that go into the board and its almost zero ohms as expected (115V position). I would think that if this switch or the wires where not connected properly, the PS would assume 240V input, and with only 120 going in it might create a low voltage situation like the one I seem to be having. But no dice.

At this point I have the PS out again and on the bench, no way to safely poke into it while inside the scope. First thing again checked all the new caps and everything is confirmed correct. At this point was also looking at the PS troubleshooting chart in the SM, and the W1060 test jumper procedure. But it requires a test load for which at this point I don't have the correct specified resistors at hand. BTW just to know, what is the spec of the diode and voltage of the 3.3uF cap used for the load? Manual only provides Tek PNs.

I would think that with the exploding Rifa cap and burned resistor being well ahead of the actual power supply nothing else would have been damaged. Not even the main fuse blew, and I think it was just time for the cap to let go, and not because of an incomming huge spike or surge although could not be 100% sure. But I had my 2247A also turned on and some other equipment at the same time and on the same AC circuit, and everything else its fine. Nothing else around reacted to any abnormalities in the AC line either.

Any words of wisdom as always much appreciated.
Alex



Roy Thistle
 

On Thu, Oct 17, 2019 at 10:14 AM, <tekscopegroup@...> wrote:


So maybe now would be a good time to recap, not just fix the damage
Hi Alex, Hi All!
First a disclaimer: The following are not.. for sure... the "words of the wise"... or maybe not even the words of the competent.
Next, and suspiciously, I noticed that all the voltages in your table are about 2/3 of what they are supposed to be. (Could you have created some kind of capacitive divider, or something causing a line drop, in the "primary" of the SMPS)… will the SMPS even start at 80VAC?
Also, did you check the SMPS part number correction on Tekwiki?
Lastly, did you check the "new" caps before installing them? (And, have you checked the removed ones to see if they are bad?)
Best regards and wishes.
Roy

Roy Thistle
 

On Thu, Oct 17, 2019 at 11:50 AM, Chuck Harris wrote:


faston connectors
Hi All:
Speaking of Faston Connectors (TE Connectivity)… what's the favorite way people have of removing the female "plugs"?... you know the crimp-on ones with indents or tabs, that lock somehow... and that pull the male lug off the pcb, or right out of the speaker... when you yank on them, in frustration. (I use a small screw driver, thin metal blade [or appropriate width]… or sometimes I do just yank like hell, if I want to see it the thing will let go before it, or something breaks.
Best regards. Best wishes.
Roy

tekscopegroup@...
 

On Thu, Oct 17, 2019 at 11:50 AM, Chuck Harris wrote:
Check the connection of the faston connectors that
connect the 110/220V selection switch. It looks like
you forgot to connect them back up when you connected
the supply back up.

-Chuck Harris
Hi Chuck, with the switch connected I can measure zero ohms across R1013 a 68K resistor that is in parallel with the voltage selection switch, so it seems all is ok in that respect. I just measured again the voltages across C1021 and C1022 the main input caps at J231A and each measures 144V across each, and total 287V. Does that sound right? I am using an isolation transformer that also drops the AC down to 117V, so those cap voltages are now a bit lower than what I reported earlier when connected directly to the AC line.

Chuck Harris
 

Roy, Faston connectors are supposed to be rarely
mated... Kind of like panda bears.

There is no way to de activate the dimple that latches
them together.

You can remove them more safely by using something to pry
them apart carefully. Yanking usually only busts your
knuckles, or gashes your hand in some way.

-Chuck Harris

Roy Thistle wrote:

On Thu, Oct 17, 2019 at 11:50 AM, Chuck Harris wrote:


faston connectors
Hi All:
Speaking of Faston Connectors (TE Connectivity)… what's the favorite way people have of removing the female "plugs"?... you know the crimp-on ones with indents or tabs, that lock somehow... and that pull the male lug off the pcb, or right out of the speaker... when you yank on them, in frustration. (I use a small screw driver, thin metal blade [or appropriate width]… or sometimes I do just yank like hell, if I want to see it the thing will let go before it, or something breaks.
Best regards. Best wishes.
Roy



Chuck Harris
 

The caps each should have about 1.4 x 117V = 163V on them.

In the 110V position, the bridge rectifier, and the
capacitors are arranged to be a half-wave voltage doubler.
In the 220V position, they form a full wave rectifier.

Being half-wave, I would expect the voltage to be have
a lot more ripple, and be a little lower than it would be
as a full wave bridge... So, they are about right.

Is your supply loaded, or unloaded? Unloaded, all the
voltages will be low as the supply works feverishly trying
not to destroy itself.

-Chuck Harris


tekscopegroup@... wrote:

On Thu, Oct 17, 2019 at 11:50 AM, Chuck Harris wrote:
Check the connection of the faston connectors that
connect the 110/220V selection switch. It looks like
you forgot to connect them back up when you connected
the supply back up.

-Chuck Harris
Hi Chuck, with the switch connected I can measure zero ohms across R1013 a 68K resistor that is in parallel with the voltage selection switch, so it seems all is ok in that respect. I just measured again the voltages across C1021 and C1022 the main input caps at J231A and each measures 144V across each, and total 287V. Does that sound right? I am using an isolation transformer that also drops the AC down to 117V, so those cap voltages are now a bit lower than what I reported earlier when connected directly to the AC line.



tekscopegroup@...
 

On Thu, Oct 17, 2019 at 02:41 PM, Chuck Harris wrote:
Is your supply loaded, or unloaded? Unloaded, all the
voltages will be low as the supply works feverishly trying
not to destroy itself.

-Chuck Harris
Chuck I only powered the supply while it was connected to the scope, never unloaded.

Anyway, I found a problem and now the power supply is working again. But I am being greeted now by a 04 11 error message which indicated a faulty cal constant checksum. Maybe the FRAN did not like the lower power supply voltages or spikes etc. When I swapped out the old NVRAM I also programmed a fresh one and put it away just in case the FRAM would not work out. But so far the FRAM had been in the scope for almost a year without any problems until now. Guess I could just swap in that new programmed NVRAM to confirm the error goes away, and then try to reprogram the data back into the FRAM.

What happened with the power supply? Well as said I had double/triple checked all caps, but... against my OWN NOTES which I had taken prior to start removing and replacing caps one by one. Turns out that in the case of C1220 and C1240 I read 10uF 100V but wrote down 100uF 10V. Then when it came to replace those two I grabbed a bag with two extra and somewhat different looking 100uF 25V separate from the other 100uF 25V caps that I was correctly using for the other spots. So I figured I had ordered these two extra ones just to "upgrade" those pesky 10V units up to 25V.

Found out when I left the power on for a few minutes and heard another pop. Upon inspection C1220 had vented its top. It was obliviously pulling down the 87V rail. Oooops...

tekscopegroup@...
 

Happy to report that I just swapped out the FRAM for the old original 1990 NVRAM, which I had completely forgotten about, and the error code went away.
Will try to erase and reprogram the FRAM, otherwise will just order a new one, plus another Aries adapter.

tekscopegroup@...
 

Just finished reprogramming the FRAM with backup data and put it back into the scope, its once again working fine.
Time to put the cover back on.

Roy Thistle
 

On Thu, Oct 17, 2019 at 02:31 PM, Chuck Harris wrote:


Faston connectors are supposed to be rarely
mated
Hi Chuck:
Yes... I don't know about Pandas... but Fastons can bite back when you try to interrupt their mating!
Just the same, the Real McCoy, TE Faston is supposed to conform to UL-310... and so is speced as to the insertion and withdrawal force, for multiple insertions and withdrawals.The speced withdrawal force is hefty...about the same as the speced insertion force, which is about 20 pounds.
I've done the prying thing, when necessary... with mixed results...using a screwdriver, or pliers... or both...but, its sometimes hard to generate that kind of force, in tight fits, or without breaking something. And, I've never seen a sold tool to help do the withdrawal in the field. At least, TE doesn't seem to sell, or make one.
Does anyone know if their is an appropriate tool available?