Test Equipment Anonymous (TEA) group therapy thread

[squadchannel]

SHIT HEAVY Tektronix 576
38,000yen(35000yen + 3000yen shipment)

missing collector variac knob, bezel panel. few readout bulb dead. but working(not correctly?).
no crts double peaking, little burning the center and left bottom corner, not early high voltage transformer! lucky!

[TERRA Operative]

Oh, you won that one! Nice!

I had my eye on it but didn't bid as I already have a 577. I'm glad it went to someone in the TEA family.

[44kgk1lkf6u]

What does the label on the lower left corner mean?  I know that the definition of high voltage can span orders of magnitude depending on what field one works in.  But I have never seen 0 V being called high voltage.

[wkb]

Oh.. it followed me home.. 

HP 33120A AWG in absolute pristine condition, no speck of dust inside.

[wkb]

Bah.. balderdash of course: I took our 1995 Mini out for a spin to pick it up, 200km in total, beautiful sunny weather so why not 

[squadchannel]

What does the label on the lower left corner mean?  I know that the definition of high voltage can span orders of magnitude depending on what field one works in.  But I have never seen 0 V being called high voltage.

sunburned. Maybe 高電圧注意 2000V(HIGH VOLTAGE 2000V).  

[Cyclotron]

Oh what a beautiful specimen this is of a Tektronix 456M.  Picked it up for 70 USD.  Cover is in perfect condition with inside door and all. Even all the BNC dust covers were on it.

The case is great and since there's no cooling fan the inside is perfect.  I should have taken some pics while her clothes were off but I plan to go back to do a thorough look over and probably a calibration. 

My guess is that this unit sat on a shelf at some military base for a long while as it does not appear to be used much.

[n1ist]

The MIT flea has started up again.  Steal of the day was an Agilent 34410A for $25.  Bright display, passes self test, and matches my other uncalibrated meters at home.  Still need to fix my other one that has a malfunctioning button on the front panel...

[wkb]

That absolutely is a steal. No way you get one like that for that money in NL.

[Vince]

Wow, a miracle happened, the Easter week-end was an excellent one for me : I somehow managed to fix my very sick RF sig gen despite my total incompetence and zero experience in this field :     




I got lucky and it was all simple stuff in the end.. 
Luckily I have the service manual for this instrument, which made it all possible.

It had two problems : no output and issues powering up and staying up. Eventually it died for good and would not power up at all, which made it easier to diagnose.

So I first worked on the no power issue of course, then the no output issue.

Power Supply
There are no less than 7 rails in this thing, but really 5 main ones : +/- 5.2V +/- 18V and +26V.
Unfortunately these 5 rails are all inter-dependent on the regulation board, so all of them were down but no idea which one(s) were actually at fault.
All these rails have a shunt resistor to sense current, with a transistor across it to turn off the rail in case of an over current condition.
Luckily, these shunt resistors are dead easy to access : they are all located on a dedicated little interconnect board, underneath the instrument.

First thing that struck me was that one of them was clearly cooked : massive discolouration in the middle, and even of of its legs came off the board somehow !  So it must have gotten real hot in there, at least briefly.... however it measured still just fine, spot on 0.56R, so I soldered it back in place. Later once it was all said and done, turned out this rail was not the problem, believe it or not.  Maybe a tantalum cap on that rail shorted which briefly shorted the rail and smoked that resistor briefly, but the over current protection tripped, then all went back to normal because the cap probably then died open-circuit. I don't know...  This was the +18V rail.

Anyway. So I measured the voltage across them all, as a means to see if one of the rails were drawing a lot of current. Got lucky. One of these resistors was dropping 0.7V ! Hmm, interesting value isn't it... sounds like a Vbe drop, this current protection must have tripped on that rail ! 
It was the -18V rail. So then I measured the resistance of all power rail, and that rail was indeed shorted !  Like... 3 ohms or something, real bad.
So that's it. I then spent whole day pulling every board from every shielded module I could get to from the top of the instrument, one by one, each time checking the resistance of the rail.... with no luck ! On each board I pulled I gave the tantalum caps a good look and measured them all. 10 to 15 of them on each and every board ! It's like the plague, a cancer !! 
Good luck mass recapping this thing !    One would need much more money and time than sense   
So I had no choice but try to pin-point WHICH particular cap was at fault (assuming it was a cap problem to begin with of course... could be a dead IC for all I know...). After spending a lot of time pulling al the board I could see from the top, with no luck at all, I moved to the bottom of the instrument where there is a giant RF flat box containing two boards stacked over one another, separated by yet another shield. Here is the top-most board, the one that's visible when you open the RF box, that's the one with the frequency divider chips :



I unplugged it and.... the short disappeared ! Resistance of the rail went up to 200 ohms or so, much better !
PHEW ! At last I am getting somewhere, perseverance finally paid. Checked all tantalum caps... were all good.
Visual inspection revealed a burn mark by a small jumper link running real close to a copper rigid coax cable. The jumper is insulated but not its legs.
I pried the link slightly away from the coax cable and check with the DMM that the short was gone... it was.
Put the board back in, short gone, and instrument now powers up, and stays up, reliably !!!   

Trouble is : I looked at the schematic for this board and... it is not even CONNECTED to the faulty rail at all ! 

However it is connected to the +18V that had that burnt shunt resistor though, so maybe it was all related somehow.

Anyway, I was overjoyed to say the least.
I spent 15 Euros to recap, for good measure/peace of mind, at least the power supply regulation board. 5 electrolytic caps, and no less than 12 tantalum caps, insane. Then checked all rails, voltage and ripple is now just fine on all of them... victory.





No Output issue

With the instrument now reliably powering up, I could move on to investigating the no output issue, as well as the display/frequency meter displaying a solid zero at all times.
Of course that's were things get difficult for me... I am no Shariar... but he inspired me and it worked somehow ! 
I looked at the overall block diagram of the instrument, basic theory of operation and got cracking, hoping for the best but fully expecting the worst ! 

So I dug out my beloved Tek 2467B scope, as it was an excellent opportunity to enjoy its B/W and bright display at such speeds...
It's rated at 400MHz but was still very comfortable displaying and triggering on a 500MHz signal, with plenty of brightness to spare. You gotta love this scope...

The basic signal is generated by an RF cavity that can be adjusted by the knob on the front panel, from 250MHz to 500Mhz or so. Then it gets divided by a chain of digital ECL counter chips, then there is a large board full of filters, to turn the square signal into a sine wave again.

So I unplugged a few rigid coax cables here and there to try to trace the signal.
The cavity was outputting a nice clean stable sine wave of the appropriate frequency, so that was a good start. No idea if the amplitude was good enough to make it all work, but at least I knew "something" was coming out of it, which was a good start in my book.

Then that goes into an RF box that has 3 outputs.... nothing was coming out of any of those outputs !



So I pulled that box, opened it and... after a quick visual inspection, found a broken solder joint on the PCB, where a rigid coax cable brings the signal from the RF cavity.



I soldered it back, and now the instrument works just fine ! 

The levelled output... levels just fine. All ranges work fine, AM modulation works fine too. It's a winner !

Now I can put it up for sale and hope to get a bit more than 50 Euros for it hopefully !

Looks like I successfully passed my RF gear baptism of fire, I am thrilled !
Thanks Shahriar, your videos paid ! 

Anyway, I just happened to be very lucky on that one and I will put it up for sale immediately, before one of the 15,554 tantalum caps it contains, blows up and it's broken again !   

[Zoli]

This weekend thrifting results: Sanwa analog multi-meter for five bucks. Well maintained for his age(later), and according to my(limited) testing, well within specs.
I've checked for battery leaks before I've paid; the two pinches of leaks, which I've scraped down, are acceptable for me. After finishing all the checks, opened the manual, to find the original bill of sale...

[TERRA Operative]

The Advantest R6161 just got a RAM backup battery with date code 2408 to replace the existing with date code 9802, 27 years was a good run... 
And I didn't even get an error on boot, so success! 

I also gave the R6161 a cheeky calibration.
It's soooo close! :'(
(Output set to 10.00000V)

I'll have to use a better interconnect that's lower EMF to see if I can improve things. I should have some shielded Cat6A cable here somewhere that would do the trick (currently using Pomona low EMF banana cables that I added a shield to).

[Hydron]

Very jealous of the R6161, nearly got sent to Tokyo for work this week and I discovered them in my pre-trip TEA (Test Equipment Acquisiton in this case) research. Sadly the trip didn't happen and it's WAY too heavy to easily lug back to the UK anyway (plus I have no idea how to deal with .jp auction sites).

[TERRA Operative]

Well, if you do make it here, we can work something out. And meet up in Akihabara or something.

[Hydron]

Well, if you do make it here, we can work something out. And meet up in Akihabara or something.

Thanks, that would be great - I've actually been to Akihabara before on my last trip to Japan, but that was 25 years ago when I was in high school! Was pretty darn exciting as a young electronics hobbiest coming from NZ.

[rdenney]

I have a Tektronix CFC250 frequency counter, which was, I think, made by Kenwood and badged for the budget line by Tektronix. It's fine for what I do. But we don't need group therapy for buying things that are fine what we do. We need therapy for going well beyond what we need, again and again.

Hence: I have an incoming HP 5334B counter with the OCXO and 1.3 GHz options. Two-channel to 100MHz, plus the optional third channel to 1.3GHz. 1PPM accuracy at ten years with the OCXO; God knows how long it's been since this one was calibrated. I was going to buy one of the Chinese GPSDO frequency references, but I'm afraid the price will double and then some on its way here. I was looking for one located in the U.S. already, but can't find one. I'll check if someone has a frequency I reference I can borrow to check this thing out when it arrives.

Here's what I'm also buying to replace the RIFA caps:

For the line-ground filter: Vishay https://www.mouser.com/ProductDetail/Vishay-Roederstein/WYO222MCMBF0K?qs=rSCQ%2F9oRLB6sFYPdn%2FQFZQ%3D%3D
And for the line-neutral filter: Mex-Tenta MKP polyester-film caps, 0.1 uF, 275VAC, X2

And, of course, a pic:



Rick "can stack it on the HP3456a, which is already the size and shape of the deck of a Nimitz-class carrier" Denney

[Zoli]

Craigslist find for a hundred canadian bucks: doesn't count, two leaky drivers, opto version as can seen from the spare boards. But it has all the nick-c's, the original power cord, good leg, and the 18GHz plugin.
More pix on the road to full recovery.

[Анатолий]

Добрый день! Скажите пожалуйста , может кто знает , как можно clear
(удалить) лог ошибок в tds340 после ремонта?

[TERRA Operative]

Добрый день! Скажите пожалуйста , может кто знает , как можно clear
(удалить) лог ошибок в tds340 после ремонта?

This is an English language board, you'll need to post in English for anyone to understand you.

[qurvax]

I can translate this guy's message:

"Hello. Does anyone know how can I clear error log of tds340 after repair?"

[TERRA Operative]

Gave the R6161 a cal tweak the other day, now it's much closer to spot-on. 

Zero's make me happy.

[tggzzz]

I can translate this guy's message:

"Hello. Does anyone know how can I clear error log of tds340 after repair?"

So could he.

He has written 3 posts, all in Russian. Often such behaviour is the prelude to doing something that gets the poster banned.

[Анатолий]

Good day! Sorry! I was told that I need to write in English, so I am writing through a translator). I will repeat the question: can anyone tell me how to clear the log (error log) of the tds340 oscilloscope?
PS : I read the forum using the "translate" option, so the entire page is in Russian, and I also write in Russian, not suspecting that people will find it inconvenient to read) I apologize again)(

[TERRA Operative]

Try sending the following command via GPIB:

Code: [Select]

ERR CLEAR
This works for the TDS500/600/700 series scopes, so the TDS300 series might be the same.


To read the error log via GPIB (It can be helpful if the error message is long), send the following command for the first error:

Code: [Select]

ERRLOG:FIRST?

Then for the remaining errors, send the following command:
(Each time you send this command, the next error will be sent back as you go down the list).

Code: [Select]

ERRLOG:NEXT?

[Анатолий]

Thank you for such a quick and detailed answer! I'm at work now, but I'll try it at home. I tried to upload the 1644 dump via a floppy disk, but I couldn't. Maybe it will work, thank you!)