Test Equipment Anonymous (TEA) group therapy thread

[capt bullshot]

Indeed we see only 40 and 20V on these boards. Voltage wise, we see one pin that carries 40V (first from the top). That's good on both boards. However the next 4 pints carry 20V on onre board... and same pins on the other boars all read 6+  or 7votls ! Here is my 7V again....

I am getting somewhere it looks like.

Will check these 2N3055 of course indeed, easy and well, power semi conductors like to fail after all...

So apparently you've got two indentical amplifiers. Most probably they're some kind of single supply Class AB amplifiers, similar to your commonplace audio amp. The electrolytics provide AC coupling to the output transformers in Scott-T configuration. To me, this looks quite plausible.
Having 7V instead of 20V at various pins connecting the amp board to its output transistors looks like a defect on the amp board. If one of the 2N3055 was dead, I'd expect near 40V or near 0V here. You might want to swap the amp boards and check if the fault swaps with the boards or stays where it is. This is a simple indicator pointing to either the transistors or the boards.
Once you confirmed the offender, you're lucky to have a working channel to compare the other one to. Might be as simple as a broken or bad solder joint on the board, a resistor that drifted, a borked transistor, maybe even a capacitor.
To me, having 7V instead of half supply voltage (20V) at the transistors points to a bias problem -> broken joint, drifted resistor or borked transistor.
Of course, I might be completely wrong with my guesses.

[BU508A]

Yes! 
Finally I got one for a reasonable price!

What it is you're asking?

A Rohde & Schwarz AM300 Dual Arbitrary Function Generator.   


(picture is from the R&S website)

Here are some spec's, for more details see attached datasheet:

Frequency
Sine: 10 μHz to 35 MHz
Triangle, ramp, square, exponential: 10 μHz to 500 kHz
Low-jitter square: 10 μHz to 50 MHz
      
Phase   
Setting range: –180° to +180°
Resolution: 0.01°

Spectral purity (sinewave)
Harmonic distortion         < 3V_pp            ≥ 3V_pp
20Hz ≤ f ≤ 1MHz         < –65dBc            < –60dBc
1MHz < f ≤ 5MHz         < –55dBc            < –55dBc
5MHz < f ≤ 35MHz       < –40dBc            < –35dBc

Nonharmonic distortion   
20Hz ≤ f ≤ 1MHz         < –60dBc            typ. (–70dBc)
1MHz < f ≤ 5MHz         < –45dBc            typ. (–55dBc)
5MHz < f ≤ 35MHz       < –40dBc            typ. (–50dBc)

More pictures and a teardown will follow. 

[Robert763]

No don't stay tuned.... at least not tonight. I am tired, time to go to bed.... will resume work on it tomorrow.

Did you see how lovely the wiring is ? 99.99% of the wires are white ? That makes it so easy to trace wires eh ?!   

I guess someone at Matra ordered by mistake a palette of white wires instead of just one spool, and the supplier refused to take the excess back. So Matra said oh boy what are we gonna do to use up all that white wire ! Hmmm.... what about making some of these 400Hz 3Ph PSU for that missile contract we just got the other day !!! 

Single colour wire is normal practice in aerospace. However on aircraft each and every wire has a unique identification number on it. This even chages when it passes through a connector. The first wire  coming from a unit for example would be 34-1-22A second 34-2-22A. If thes went through a connector they woud become 34-1-22B etc. The 34 i the ATA "chapter" (system) number, 34 is Navigation. 1 is wire number, 22 is wire gauge and A is segment letter. The exception is twisted pair/triple/quad. Each conductor is colour coded and the group is given a number. There are other exceptions. Flight test wiring is normally orange and on military aircraft armament wiring is red.
Inside equipment single colour is th norm too. In the UK often pink PTFE but with ident sleeves at each end.

[Robert763]

www.ebay.co.uk/itm/255529155186   

[med6753]

www.ebay.co.uk/itm/255529155186   

If you're going to make a suicide cord you might as well go top tier. 

[Peter_O]

www.ebay.co.uk/itm/255529155186   

Exactly what's needed to connect old devices with outdated power inlets.
Luckily you can change the clips to bare metal ones easyly, which would be slimmer to better reach into crammed inlets.   

[tggzzz]

www.ebay.co.uk/itm/255529155186   

If you're going to make a suicide cord you might as well go top tier. 

Seller is in Leytonstone. If I lived there, I'd buy it

[tggzzz]

www.ebay.co.uk/itm/255529155186   

Exactly what's needed to connect old devices with outdated power inlets.
Luckily you can change the clips to bare metal ones easyly, which would be slimmer to better reach into crammed inlets.   

For outdated power outlets (without shutters) you could change the plug to bare metal knitting needles, or, to avoid soldering, bolts

[med6753]

www.ebay.co.uk/itm/255529155186   

If you're going to make a suicide cord you might as well go top tier. 

Seller is in Leytonstone. If I lived there, I'd buy it

I take it that it's an "avoidance area" like for example Newark, NJ? 

[bd139]

www.ebay.co.uk/itm/255529155186   

If you're going to make a suicide cord you might as well go top tier. 

Seller is in Leytonstone. If I lived there, I'd buy it

I lived there once which was too many times.

I think Cerebus is out that way. Mains plug with nipple clamps is right up his street

[tggzzz]

www.ebay.co.uk/itm/255529155186   

If you're going to make a suicide cord you might as well go top tier. 

Seller is in Leytonstone. If I lived there, I'd buy it

I take it that it's an "avoidance area" like for example Newark, NJ? 

I used to drive through it occasionally, without stopping. I've no idea whether stopping then or now would be risky. Most big cities have some areas with enhanced risk, but I'd be surprised if any had as much risk as the worst in the US.

I certainly wouldn't want to live in Leytonstone since the only positive would be that you aren't too far from some nice places. Likewise Bracknell - except that has many fast roads out of there.

(My brief summary: Leyonstone is a suburb hastily built ~120 years ago for mechano-drones, whereas Bracknell was an overflow town hastily built  ~50 years ago for techno-drones)

[Ice-Tea]

Bricked my laptop downgrading the bios.

I'm not happy.

[Peter_O]

Bricked my laptop downgrading the bios.

I'm not happy.

Bricks never are a joy to switch on.   
(expression of condolences)

My full size Asus mainboard would provide fallback mechanisms. Don't know, if there are any on laptops.

[Ice-Tea]

You're supposed to be able to blind flash it but I haven't been able to. No official guidance, just some forum posts.

[xrunner]

You're supposed to be able to blind flash it but I haven't been able to. No official guidance, just some forum posts.

Bummer. I did that to a PC motherboard once trying to upgrade the BIOS. I wasn't paying attention and didn't notice they removed support for the CPU I was using in it (AMD A8-9600). The only way to "downgrade" back again would have been to buy a new CPU that was supported by the new BIOS, downgrade it, and then put back the old CPU.

But then if I had a new CPU why not just use that and move on? 

[Cerebus]

www.ebay.co.uk/itm/255529155186   

If you're going to make a suicide cord you might as well go top tier. 

Seller is in Leytonstone. If I lived there, I'd buy it

I lived there once which was too many times.

I think Cerebus is out that way. Mains plug with nipple clamps is right up his street

I have Forest Gate and Leyton as a buffer zone before you get to Leytonstone. Then you've got Leytonstone as a buffer zone between there and Walthamstow and Chigford (where traditionally you might encounter the zombie Norman Tebbut and his cohort of ghouls).

I think I'd go for something a little more violet and current limited than a 13A plug if I was going to dabble in that territory.

[Cerebus]

Bricked my laptop downgrading the bios.

I'm not happy.

No you wouldn't be. Is this something you can revive by directly flashing the BIOS into an SPI EPROM with a cheapo clip on programmer?

[srb1954]

www.ebay.co.uk/itm/255529155186   

Exactly what's needed to connect old devices with outdated power inlets.
Luckily you can change the clips to bare metal ones easyly, which would be slimmer to better reach into crammed inlets.   

For outdated power outlets (without shutters) you could change the plug to bare metal knitting needles, or, to avoid soldering, bolts

Nails would be better. With a large enough hammer you could even use them to defeat those pesky shutters.

[mansaxel]

www.ebay.co.uk/itm/255529155186   

Exactly what's needed to connect old devices with outdated power inlets.
Luckily you can change the clips to bare metal ones easyly, which would be slimmer to better reach into crammed inlets.   

I think the thing one wants, actually, is a safeblock, as in Bigclives video:

[mnementh]

Indeed we see only 40 and 20V on these boards. Voltage wise, we see one pin that carries 40V (first from the top). That's good on both boards. However the next 4 pints carry 20V on onre board... and same pins on the other boars all read 6+  or 7votls ! Here is my 7V again....

I am getting somewhere it looks like.

Will check these 2N3055 of course indeed, easy and well, power semi conductors like to fail after all...

So apparently you've got two indentical amplifiers. Most probably they're some kind of single supply Class AB amplifiers, similar to your commonplace audio amp. The electrolytics provide AC coupling to the output transformers in Scott-T configuration. To me, this looks quite plausible.
Having 7V instead of 20V at various pins connecting the amp board to its output transistors looks like a defect on the amp board. If one of the 2N3055 was dead, I'd expect near 40V or near 0V here. You might want to swap the amp boards and check if the fault swaps with the boards or stays where it is. This is a simple indicator pointing to either the transistors or the boards.
Once you confirmed the offender, you're lucky to have a working channel to compare the other one to. Might be as simple as a broken or bad solder joint on the board, a resistor that drifted, a borked transistor, maybe even a capacitor.
To me, having 7V instead of half supply voltage (20V) at the transistors points to a bias problem -> broken joint, drifted resistor or borked transistor.
Of course, I might be completely wrong with my guesses.

I think one of Vince's trannysistors is sulking because it wants a new dress. 

mnem

[mnementh]

   www.ebay.co.uk/itm/255529155186   

If you're going to make a suicide cord you might as well go top tier. 

Seller is in Leytonstone. If I lived there, I'd buy it

I lived there once which was too many times. I think Cerebus is out that way. Mains plug with nipple clamps is right up his street

Oh, and look... it has nice convenient banana plug ends so you can insert them in... places... and a GND wire..? Because, I suppose, you don't wanna accidentally electrocute yourself while trying to commit suicide...? 

mnem
OSHA-approved.

[Vince]

MATRA 400Hz 3 Phase PSU repair update

Today is another day, sun light and motivation are back. I am working on the thing.

White wires :

Nope their ends have no labeling whatsoever sadly. Just.. white wires, nothing buy white, a desert of white.... terminated on both ends with "lovely" transparent heat shrink tubing just in case I might want to freaking PROBE stuff !    I have murder envies, I really have...

Swapping amp boards :

I pulled them out and put them side by side to inspect them closely. They are 100% identical, same layout, same transistor types, same resistor values, capacitor values.... and on the solder side, there is a part number etched into the copper... of course they are identical. Maybe I could have looked at that first....

So, now happy that they were identical and therefore safe to swap.. I did swap them.
Result : on the back plane connections, the 4 pins that read 6V instead of 20V, DO follow the bad board, yeah ! So the board is indeed bad.
Then I measured the 3 phases on the output connector at the rear.
Phases 2 and 3 that used to read about 100V, now read 60V... but the good news is that Ph1 that used to be only 3 or 4 Volts, now jumped to 115V !!!!   
So I guess that means that all the surrounding components are good, capable of producing the required high voltage : the transformers, those blue caps in the PSU area, and the power transistors mounted on those big heat sinks. So that's excellent news. The problem seems limited to that bad board and nothing else.



Checking the amp boards :

So I pulled the amp boards and started the inspection.
No obvious signs of rework or bad joints, so I started to probe around.

On the bad board, I found a group of 3 close together resistors that all read 30% lower than they should. Suspicious... Pulled one leg on each to quickly make sure that they were.... good, as I suspected. They were indeed good. So it must be some other component, related to these 3 resistors, that makes them read low.  So I then looked at all the electrolytic caps. 6 of them. Pulled a leg on each, used the cheap chinese meter... they were all good but one that detected as a 60 ohm resistor rather than a capacitor... 22uF 40V. I replaced it. Does not look pretty because the pad holes were too small to host the replacement cap leads. So had to solder two bits of wires into the holes first, then solder the cap to those wires. OTHO I would have needed to do that anyway because replacing an old axial cap with a radial cap that's salvaged hence with tiny / short leads... doesn't work.

Then I checked all the diodes, 4 of them. Vintage looking packaging, but not old enough that they were Germanium : they were not, reading 600mV drop. So Silicon it is. All 4 of them checked good. HOWEVER.... on the GOOD board... one of these diodes was open circuit ! So that board was good "enough", but not 100% good it appears !
So I replaced that diode with... well a 1N4148 because when in doubt, that's what you put right ?!    At least it was Silicon and probably much better than having the old open diode I thought...


So I put the boards back into the PSU and fired it up.

That didn't change much on the outputs at the rear, back to what they were from the beginning : About 100V for Ph 2 and 3, and 3V only for Ph1.
However the 20V pins on the back plane now read higher / better : 10V instead 6+ V before. Still not quite 20V but well, it's going in the right direction at least...


Conclusion

So it looks like :

A) The problem is contained in that bad amplifier board and all the supporting power / beefy components scattered around the chassis, are not at fault. That's a big relief....

B) The bad cap and diode were not the root cause of the problem... but they definitely needed replacing so at least I am now closer to a working system for sure..


So I must keep digging, and I now now I need to dig into that bad board specifically, an dI now I can compare anything to the other, working board, which is a huge help.

So I will check the remaining components. I have checked the x6 electrolytic caps, the x4 diodes, and all the resistors... well the FIXED ones.
Now what's left to check on the board ?

- A couple multi-turn trimmers
- A couple small film caps, look like the expensive PS type ?
- x6 transistors


Stay tuned 

[mnementh]

Bricked my laptop downgrading the bios.  I'm not happy.

   CH341A programmer bundle for 24/25 series flash; $14 via Amazon Prime:   

https://www.amazon.com/Organizer-EEPROM-CH341A-Adapter-Programmer/dp/B07V2M5MVH/

mnem

  "There's no such thing as too many dongles..."

[mnementh]

No don't stay tuned.... at least not tonight. I am tired, time to go to bed.... will resume work on it tomorrow.

Did you see how lovely the wiring is ? 99.99% of the wires are white ? That makes it so easy to trace wires eh ?!   

I guess someone at Matra ordered by mistake a palette of white wires instead of just one spool, and the supplier refused to take the excess back. So Matra said oh boy what are we gonna do to use up all that white wire ! Hmmm.... what about making some of these 400Hz 3Ph PSU for that missile contract we just got the other day !!! 

Nothing wrong with that, no stupid manager or scrambled eggs general involved. These wire looms are made according to a carefully planned layout, as you can see, the person who had to solder the wires to the pins can identify the wires by their position. Just like a PC Board, all the traces got the same colour on a PCB, too, don't they?

Which is fabulous as long as you're authorized service personnel working on the thing at the time it was relevant; 40-ish years later, trying to resurrect the thing without documentation it is more likely to encourage homicidal rage. 

One of the things I did appreciate working as a large appliance tech: The push in the the industry to mandate certain minimum documentation including at least a basic schematic, and that it must be included inside each and every damned unit that went out the door. 

Nowadays with China-direct manufacturing dominating this market as pretty much all manufacturing, I imagine that is also a thing of the past... 

mnem
"Beware he who would deny you access to information; for in his heart, he dreams himself your master." - Commissioner Pravin Lal, Alpha Centauri United Forces

[mnementh]

...I think Cerebus is out that way. Mains plug with nipple clamps is right up his street

I have Forest Gate and Leyton as a buffer zone before you get to Leytonstone. Then you've got Leytonstone as a buffer zone between there and Walthamstow and Chigford (where traditionally you might encounter the zombie Norman Tebbut and his cohort of ghouls).

I think I'd go for something a little more violet and current limited than a 13A plug if I was going to dabble in that territory.

mnem
Can I watch and post video to antisocial media...?