Test Equipment Anonymous (TEA) group therapy thread

[mnementh]

Mayhaps the difference is that in racing applications, the engines are torn down and rebuilt often enough it doesn't matter. Also, I've long been a true believer in tiny torque wrenches (and medium-sized ones, and big ones) and never-sieze.

Copper or nickel based ?

Ugh neither, ceramic of course!

No-one in Discord yet, I may go back to my freshly arrived Bluray box set of Game of Thrones...   

Well, back then it was copper. Works nicely on brake-part sliders too.

mnem
*poot!*

And now ?
And why ? 

Pretty much neither since I quit racing, tho I have had the same tube of copper never-sieze for easily 10 years. I still use it on spark plugs, lug nutzz, etc.

mnem
 

[shakalnokturn]

Worst case, if I can't figure out how to make it talk to my 3DP I can send it back and keep the rubber duckie. At least that doesn't need a FW update to work.

mnem
 

Meaning the duckie isn't an active (vibrating) model?
What kind of gift is that...!?

[mnementh]

mnem
 

[srjaynes49]

Yes that’s the delay line. I sold about 1,000, 485s in my years with Tek. They didn’t seem any more sensitive to power supply issues than the other 2nd generation 400 scopes including the 465 and 475. The 485 had a very different type CRT than the 465 & 475 in that it was smaller and if memory serves, it didn’t have a dome mesh lens to magnify the sweep. It was that lens that gave a good many 465s & 475s fat traces and halos.   485 didn’t suffer from that design feature.  I still have my personal 475, but sold my 485 and it’s lovely P6201 FET probes.

[med6753]

Those adapters are really cool...

Yes, they really work well at adapting the new capacitors to the board layout. I used them once before when I re-capped a 465B PSU.

[med6753]

The military 7603 PSU is fully reassembled and ready to be installed. Tomorrow's activity.

[BU508A]

Sorry, couldn't join discord today, because I was fast asleep on the couch after a busy day doing different types of dough for my xmas bakery.
Once a dough has been made, I put it on a plate, wrap it in a tea towel and then put it outside for cooling. A cat from the neighbourhood found one dough that attractive, that she has eaten 1/5 of it! 
Really, I do love cats but when it comes to the dough of my xmas bakery, I'm becoming REALLY insufferable and grumpy!   

Looking nice & tidy there, med.

mnem
 

Except the double up. 
 

Had no choice with that double. For some strange reason could not find an 18000uf/15V at a reasonable price. So I paralleled two 10000uf/16V.

I have to agree with tautech. Instead of using two 10mF/16V in parallel I'd go for a nice 22mF one.
Mouser has them in two different voltages / sizes: 16V and 25V from Nichicon for a reasonable price.

[mnementh]

Worst case, if I can't figure out how to make it talk to my 3DP I can send it back and keep the rubber duckie. At least that doesn't need a FW update to work.

mnem
 

Meaning the duckie isn't an active (vibrating) model?   What kind of gift is that...!?

Here you go. If nothing else, an amusing waste of 5 minutes. 

mnem
*Impairing productivity since nineteen-mumblety-mumble*

[Robert763]

Finally getting somewhere with the HP 5248M, bit tired tonight but will explain more tomorrow. 

Well I went through testing all the transistors in the 5MHz to 10MHz multiplier stage (A26) but couldn't find anything that was causing the low output voltage, I came to the conclusion that the crystal oven might be the cause.
To verify the board was OK before troubleshooting the oven, I disconnected the DIV output from the oven & used a Philips RF source in it's place (which was set it up using the oscilloscope for the correct rms level), well that made a big difference as I was now getting the correct output from the multiplier. 



Before & after at A26(6),


Oh dear not another faulty HP crystal oven I thought, the only accessible adjustments are the fine/course for the frequency & a filter for the main 5MHz output (BNC).
The manual only tells you to adjust the 5MHz filter if the output (BNC) on the back of the counter is below 2.5Vpp, which I had checked & it was well above. I had also checked the input to the A26 board was around 70mV or so I thought, turns out I should have checked the rms level not peak to peak  and this was a bit low.
After looking at the circuit I couldn't work out why the DIV output was low if the main output was fine. The 5MHz filter adjustment is in this part of circuit so I made a note of the original position before adjusting it, to my surprise the A26 board starting working as it was supposed to and I had well over the required 5Vpp at A26(6)  , I guess the original owner must have twiddled with it instead of fixing the separate oven power supply. 
 

The self check was finally useable, note the two LSD are still faulty (probably the last parts in the repair queue), still not getting anything through from either input, which I expected as the function control board still needs working on.

David

Walter Shawlee at Sphere has put a bunch of HP Nixie couner / drive modules on their discount pages - $9 each
https://www.sphere.bc.ca/test/stuffday.html

There is a note that he has some tubes too. but no price. You might get a better price if he knows you are using them for repair rather than making a clock.

[med6753]

Looking nice & tidy there, med.

mnem
 

Except the double up. 
 

Had no choice with that double. For some strange reason could not find an 18000uf/15V at a reasonable price. So I paralleled two 10000uf/16V.

I have to agree with tautech. Instead of using two 10mF/16V in parallel I'd go for a nice 22mF one.
Mouser has them in two different voltages / sizes: 16V and 25V from Nichicon for a reasonable price.

Yes, I could have gone with a single 22000uf but decided to try to stay as close to design spec as possible. And this technique of paralleled  capacitors is employed by Tek quite often in their equipment. Two other 9600uf capacitors in this PSU are actually wired in parallel when you study the schematic. And in vacuum tube equipment I often see multi-section 20uf and 40uf cans connected in parallel to achieve 40uf and 80uf respectively. 

[med6753]

Yes that’s the delay line. I sold about 1,000, 485s in my years with Tek. They didn’t seem any more sensitive to power supply issues than the other 2nd generation 400 scopes including the 465 and 475. The 485 had a very different type CRT than the 465 & 475 in that it was smaller and if memory serves, it didn’t have a dome mesh lens to magnify the sweep. It was that lens that gave a good many 465s & 475s fat traces and halos.   485 didn’t suffer from that design feature.  I still have my personal 475, but sold my 485 and it’s lovely P6201 FET probes.

What is this in reference to? 

I have all three scopes and I'm aware of their differences.

[mansaxel]

I think this is why we cherish our anchors.

-hp- sez BAM. And Marconi agrees.

Edit: The 428b measures current through a 120Ω resistor driven by the 6206b, and with the 974A inline. The ampère meters agree within 200µA or so, and I was some 2mA off when I'd set the voltage on the 6206b, using just the internal meter. The Marconi and the 3438 measure voltage across the setup. The FeelTech is not involved, and neither is the 465..

[mnementh]

You COULD involve the 465... from this side of a computer screen, that looks like a textbook example demonstrating common-mode noise.

mnem

[med6753]

As they say in North Jersey.....bada bing, bada boom.  It's back together and it survived my rough treatment. Powered up first try. 

I still have to adjust the +50V reference and I'll check all the other PSU voltages. It probably still has the intermittent trace disappearing act which I'll have to troubleshoot. But at least the PSU rebuild was successful. 


That spot on the screen is a reflection.
 

[bd139]

Nice job 

[mnementh]

Any valve-technology repair you walk away from is a good repair. 

How many digits (yours, not the scope's ) were lost or mutilated...? 

mnem
*agitating-ily*

[mansaxel]

You COULD involve the 465... from this side of a computer screen, that looks like a textbook example demonstrating common-mode noise.

mnem

Indeed I could, and it was turned on and working, because I wanted to test my TT counter in compare mode, which it managed to do; and I somehow had missed that it actually can do that. Looks like I've got to bring it to work for some 10MHz reference comparisons.  The 10MHz from my GPSDO also has been connected up a bit more steadily, and I wanted to test how the signal looks as it arrives. Apparently it is a 3V p-p square wave, that gets a bit rounded as it goes through the 12-15m RG58 under the raised floor. And, it needs termination. The unterminated signal looks like shit.

[med6753]

Any valve-technology repair you walk away from is a good repair. 

How many digits (yours, not the scope's ) were lost or mutilated...? 

mnem
*agitating-ily*

I count 9 so I'm good.  Oh...wait...

[med6753]

Correction, it's the -50V in this scope that's the reference. And it's dead nuts at -50.0V. No adjustment needed. All other PSU voltages are in spec.

And yes, after 10 minutes or so the trace went bye-bye. I had no illusions that the PSU re-cap would fix that. Suspect something in my least favorite area, the HV, is breaking down. If I leave the scope off for about a half hour it will come back. So I'll catch it while it's working and take some measurements and see what drops out. First place I'll start is the input to the HV Oscillator.


 

[mnementh]

Any valve-technology repair you walk away from is a good repair. 

How many digits (yours, not the scope's ) were lost or mutilated...? 

mnem
*agitating-ily*

I count 9 so I'm good.  Oh...wait...

Oh, so you've grown a extra then...? 

mnem
 

[med6753]

Any valve-technology repair you walk away from is a good repair. 

How many digits (yours, not the scope's ) were lost or mutilated...? 

mnem
*agitating-ily*

I count 9 so I'm good.  Oh...wait...

Oh, so you've grown a extra then...? 

mnem
 

I have the important ones which is why I have the girl, and you, well.....eat your heart out. 

[nixiefreqq]

Any valve-technology repair you walk away from is a good repair. 

How many digits (yours, not the scope's ) were lost or mutilated...? 

mnem
*agitating-ily*

I count 9 so I'm good.  Oh...wait...

Oh, so you've grown a extra then...? 

mnem
 

I have the important ones which is why I have the girl, and you, well.....eat your heart out. 

oh dear........smurf porn?

that's it, goin' back to utube.   (currently watching something called "it aint half hot mum".)

[bd139]

Omg don’t Google Smurf porn 

[factory]

The self check was finally useable, note the two LSD are still faulty (probably the last parts in the repair queue), still not getting anything through from either input, which I expected as the function control board still needs working on.

David

Walter Shawlee at Sphere has put a bunch of HP Nixie couner / drive modules on their discount pages - $9 each
https://www.sphere.bc.ca/test/stuffday.html

There is a note that he has some tubes too. but no price. You might get a better price if he knows you are using them for repair rather than making a clock.

Thanks for letting me about those as I hadn't looked for at the stuffday page for a while, I will have to enquire about a few of those*, as the parts mule I have is a parts mule because the previous owner stripped out all the easy to remove boards, I didn't ask what he did with them but I suspect he scrapped them.

*None are correct for the two faulty boards in the 5248M, but I do need a replacement for the 5243L as one of the display boards is cracked at the edge connector.

David

[tggzzz]

that's it, goin' back to utube.   (currently watching something called "it aint half hot mum".)

From the same writers, but much better and more fondly remembered is "Dad's Army". Thats still on terrestrial TV occasionally.