Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

Right now I'm tearing my 1KW+ ( Yes REALLY!!!) PSU down to see if I can change the fan... this is my 2nd one of these, and the fan in it sounds like a effing jet engine (unlike the one in my first unit) 100% Duty Cycle so I'm thinking to add load-based control integration for its fan as well.

mnem
*recursion-ily*

Uggghhh... I think it REALLY IS running 100% duty cycle; comparing against the pics from this teardown, it looks like they omitted a thermal-control loop and are running the fan WOT/100% Duty.     Gonna be off the internet for a while then... need to dismantle 1KW PSU #1 to draw up a BOM so I can retrofit. 

mnem
See you all on the other side...

Profiling isn't racist when it's thermistors, right? 

grrr.  Figuring the BOM was easy; they only omitted the thermistor and pass transistor (2SD882); all the rest of the circuit is still populated.
Looking for one in the junk bins with a similar slope cuz I don't wanna wait until for one to be delivered. 

mnem
This will probably not end well.

[Specmaster]

You could also use freeze spray. But it tend to leave water all over the place.

To find shorts, I had some success using y 34401a in ohms mode.

Yeh, the advantage with the fluid method is that it is far quicker to locate without having to isolate one end of each part, find the dry spot, and that is where the heat is, jackpot.

I don't isolate anything. Just connect the lo terminal of the multimeter to ground and use the hi terminal to probe following the rail path. The point where the resistance is the lowest is usually really close to the short you are looking for.

True, that's one of the advantages of having meters with 6.5 or better digit resolution set in low ohms range. But not so clever if your working with lower resolution meters say 3.5 to 4.5. The liquid and current limited PSU approach is going be far quicker as well.

Sent from my POT-LX1 using Tapatalk

[tggzzz]

You could also use freeze spray. But it tend to leave water all over the place.

To find shorts, I had some success using y 34401a in ohms mode.

Yeh, the advantage with the fluid method is that it is far quicker to locate without having to isolate one end of each part, find the dry spot, and that is where the heat is, jackpot.

I don't isolate anything. Just connect the lo terminal of the multimeter to ground and use the hi terminal to probe following the rail path. The point where the resistance is the lowest is usually really close to the short you are looking for.

True, that's one of the advantages of having meters with 6.5 or better digit resolution set in low ohms range. But not so clever if your working with lower resolution meters say 3.5 to 4.5.

If you are using that resolution, you are depending on having a the same contact resistance every time you probe along the line. That's unlikely with low resistances.

Measuring the voltage or voltage drop doesn't have that problem.

[Specmaster]

You could also use freeze spray. But it tend to leave water all over the place.

To find shorts, I had some success using y 34401a in ohms mode.

Yeh, the advantage with the fluid method is that it is far quicker to locate without having to isolate one end of each part, find the dry spot, and that is where the heat is, jackpot.

I don't isolate anything. Just connect the lo terminal of the multimeter to ground and use the hi terminal to probe following the rail path. The point where the resistance is the lowest is usually really close to the short you are looking for.

True, that's one of the advantages of having meters with 6.5 or better digit resolution set in low ohms range. But not so clever if your working with lower resolution meters say 3.5 to 4.5.

If you are using that resolution, you are depending on having a the same contact resistance every time you probe along the line. That's unlikely with low resistances.

Measuring the voltage or voltage drop doesn't have that problem.

How does that work then? Surely if your probing along a supply rail measuring resistance / voltage or volt drop then the same contact resistance each time you probe as you mention is also just as affective in those modes as well?

Here Louis Rossmann uses alcohol and in this one uses petrol and see just how quick this proves the theory.

[BU508A]

Nope the tants disguised as silver mica ones if you look from above 

SCNR    but I saw these brown tantalums not at the first sight. Had to have a really close look. 

[bd139]

Nope the tants disguised as silver mica ones if you look from above 

SCNR    but I saw these brown tantalums not at the first sight. Had to have a really close look. 

Occasionally when it's late and you've hacked your way through an entire Tek plugin, one gets missed. I thought this was a silver mica but it wasn't:



Also it's always 2AM when you find these things, right in the middle of the damn plugins 



Limited forum partition over next day or two as have terrible man flu. Going back to bed

[BU508A]

Take care and get well soon, then.

[Specmaster]

Nope the tants disguised as silver mica ones if you look from above 

SCNR    but I saw these brown tantalums not at the first sight. Had to have a really close look. 

Occasionally when it's late and you've hacked your way through an entire Tek plugin, one gets missed. I thought this was a silver mica but it wasn't:



Also it's always 2AM when you find these things, right in the middle of the damn plugins 



Limited forum partition over next day or two as have terrible man flu. Going back to bed

Oh no, not the dreaded man flu, crikey, these are old photos, not on that bench that you still have 

[mnementh]

Both PCs are now reassembled and jet engines evicted; replaced with nice, ear-friendly thermally-controlled fans in both PSUs.

Take care of yourself, bd... lots of fluids, lots of soup and of course, copious amounts of TEA. 

mnem
*heeding Hypnos' inevitable summons*

[tautech]

Limited forum partition over next day or two as have terrible man flu. Going back to bed

Wife's special remedy.....she starts polishing a bullet with you name on it !
Sorta brings special meaning to: Get well soon ! 

[BU508A]

Wife's special remedy.....she starts polishing a bullet with you name on it !

Hmm, one should not fool around with women from Down Under ....   

[tautech]

Wife's special remedy.....she starts polishing a bullet with you name on it !

Hmm, one should not fool around with women from Down Under ....   

Managed to dodge that polished bullet for 35+ years ! 

[PA0PBZ]

How does that work then? Surely if your probing along a supply rail measuring resistance / voltage or volt drop then the same contact resistance each time you probe as you mention is also just as affective in those modes as well?

Contact resistance doesn't really matter when measuring voltages with a 10M input resistance meter, but it is a different story trying to measure mOhms.

[Specmaster]

Maybe not, but it does require a meter that is just a bit bit better then a lot of people have.

[beanflying]

Maybe not, but it does require a meter that is just a bit bit better then a lot of people have.

When I got the new 34461A I looked at selling my Quadtech 1880 milliohm meter few quick test measurements below 10 ohms and changed my mind quickly. Moral of the story is modern and more digits don't always get you more accuracy

[bd139]

Got my Fluke 8050A. Arrived with some serious problems. Decided to attack it rather than look after myself. Looks like the front end has been blown up. Varistors and the big 1k protection resistor have been replaced and the board has burned where the first set of varistors went up. Whoops. So it reads wrong by about ten percent on 20V and above. Popped 87V on the front end and input impedance registered as 8.5M which is low. Traced out front end and found Q2, a 2n3904 used as clamp was short. Replaced that and the other clamp transistor just to be sure, cleaned out the switches and it’s bang on cal  . Just needs a good clean now. Display is good too.

Going back to bed now though 

[Zucca]

Upgrade it! 



bd stay with us please:

[Specmaster]

Got my Fluke 8050A. Arrived with some serious problems. Decided to attack it rather than look after myself. Looks like the front end has been blown up. Varistors and the big 1k protection resistor have been replaced and the board has burned where the first set of varistors went up. Whoops. So it reads wrong by about ten percent on 20V and above. Popped 87V on the front end and input impedance registered as 8.5M which is low. Traced out front end and found Q2, a 2n3904 used as clamp was short. Replaced that and the other clamp transistor just to be sure, cleaned out the switches and it’s bang on cal  . Just needs a good clean now. Display is good too.

Going back to bed now though 

Good work, man flu never held you back on this one. Get better soon eh, hot toddy and plenty of night nurse.

[BU508A]

The 3460A is only 5 1/2 digit unless you got the H04 option, but the standard 5 1/2 version does accept the option PCB's for the 3461A AC/ohms converter.

Is it possible to add this H04 option afterwards? Are there any schematics around?
Or has somebody done this before and described it by any chance?

I have bought the scanned OSM for the 3460A here: https://www.artekmanuals.com

The scan quality is very good.
Example:



Thanks.

[Kosmic]

You could also use freeze spray. But it tend to leave water all over the place.

To find shorts, I had some success using y 34401a in ohms mode.

Yeh, the advantage with the fluid method is that it is far quicker to locate without having to isolate one end of each part, find the dry spot, and that is where the heat is, jackpot.

I don't isolate anything. Just connect the lo terminal of the multimeter to ground and use the hi terminal to probe following the rail path. The point where the resistance is the lowest is usually really close to the short you are looking for.

True, that's one of the advantages of having meters with 6.5 or better digit resolution set in low ohms range. But not so clever if your working with lower resolution meters say 3.5 to 4.5.

If you are using that resolution, you are depending on having a the same contact resistance every time you probe along the line. That's unlikely with low resistances.

Measuring the voltage or voltage drop doesn't have that problem.

Should be similar no ? the resistance mode of the multimeter is measuring the voltage drop right ? the multimeter is using lower voltage though.

But I agree providing external voltage is a better idea.

[mindcrime]

New multi-meter came in yesterday.  This one effectively replaces the old Radio Shack one that has gone to the great multimeter repository in the sky...

Anyway, this thing is a complete Wun Hung Low brand, but I'm relatively impressed with it given the price. It feels solidly built, came with two sets of leads, and the one set appear to be "gold plated" to some extent, and they pass the Dave "how quickly does the continuity buzz go off" test.  And the flip-stand thing seems relatively solid.  On the downside, you have to use a screwdriver to open the battery compartment, and it runs on one 9v battery, so not expecting great battery life.

[Kosmic]

New multi-meter came in yesterday.  This one effectively replaces the old Radio Shack one that has gone to the great multimeter repository in the sky...

Anyway, this thing is a complete Wun Hung Low brand, but I'm relatively impressed with it given the price. It feels solidly built, came with two sets of leads, and the one set appear to be "gold plated" to some extent, and they pass the Dave "how quickly does the continuity buzz go off" test.  And the flip-stand thing seems relatively solid.  On the downside, you have to use a screwdriver to open the battery compartment, and it runs on one 9v battery, so not expecting great battery life.

Nice big screen 

Normally those meters are not that bad with DC voltage. Got a cheap 15$ 4.5 digits here and it's pretty accurate.

Just don't try to probe main with it 

[med6753]

While you guys have been standing around the water cooler debating the merits of one short finding technique vs another I've been working.  And after hours of tracing and dealing with schematics that DON'T match the board (How many times have I had THAT happen) I have FINALLY found the source of the +15V short. 

I present to you bastard 15uf tant. Reads 0.20 ohms. 



It's going to take me at least an hour to restore all the pulled parts on the supply board then install it. Then apply power and see what happens. If I get power up that's great but I'm definitely going order replacement caps for the PSU and do a total recap.   

[Kosmic]

While you guys have been standing around the water cooler debating the merits of one short finding technique vs another I've been working.  And after hours of tracing and dealing with schematics that DON'T match the board (How many times have I had THAT happen) I have FINALLY found the source of the +15V short. 

I present to you bastard 15uf tant. Reads 0.20 ohms. 



It's going to take me at least an hour to restore all the pulled parts on the supply board then install it. Then apply power and see what happens. If I get power up that's great but I'm definitely going order replacement caps for the PSU and do a total recap.

Good job!

[Mr. Scram]

Well you knew it was going to be a tant, so I guess that one's on you.