Test Equipment Anonymous (TEA) group therapy thread

[EggertEnjoyer123]

Time for another free scope (which I found today)
I fixed most of the issues already. The trace had a lot of noise when I got it, but that was fixed after replacing a capacitor which went high ESR. (Interestingly it measured 0.17 ohms for a 280uF capacitor, but after replacing it the problem went away). There was 80mV ripple on the 30V rail, when the manual said 20mV. The other issue was a broken focus and intensity control, but that was just because a few carbon composition resistors went open.

[TERRA Operative]

Interesting notice, it explains there is a 'B' and 'C' version. Mine is a 'B', 10-7 stability, and the 'C' has a better OCXO at 10-8 hmmm..... oh well 

Just hook it up to a GPSDO or Rubidium standard. 

Time for another free scope (which I found today)
I fixed most of the issues already. The trace had a lot of noise when I got it, but that was fixed after replacing a capacitor which went high ESR. (Interestingly it measured 0.17 ohms for a 280uF capacitor, but after replacing it the problem went away). There was 80mV ripple on the 30V rail, when the manual said 20mV. The other issue was a broken focus and intensity control, but that was just because a few carbon composition resistors went open.

Free? Nice!

Be sure those 510K resistors have a good voltage rating, I use Vishay 1/2W HVR series.

Also, check the service notes, there are a few upgrades for the 2235 that might apply to this one too, fixes and updates for the power supply, and the common overvolt issue for the CRT filament.
Also, check for Rifa madness!

I have a video on restoring and upgrading the 2235 which you might find handy.

https://youtu.be/lrXCVg6T-ek



For last night's fun, I replaced the VFD in the 53181A counter I'm poking at.
My brother called dibs on it and wanted the display replaced, so now it's all shiny again.
Once I get the 3GHz option built, this thing will be better than the day it was sold.
The display looks way better in real life, the dynamic range of even flagship phone cameras isn't great.....

[EggertEnjoyer123]

I soldered 5 normal 100k resistors in series to get the correct voltage and power ratings.

Also any tips on repairing ganged switches? I sprayed contact cleaner in there three times, and it unsticks the switches for a while, but then they get stuck again after the contact cleaner dries.

[TERRA Operative]

If they are the typical Tek ones, you can take them apart and usually you don't even need to desolder them.
I give them a good clean as the old grease gets really sticky (A q-tip is perfectly sized to clean inside) and then a good dose of deoxit on the contacts, some plastic safe grease on the plastic shaft, and a fiddle (with your choice of expletives muttered under your breath) to get them reassembled and they'll be as good as new.

Here's a Tek document on how it's done:
https://drive.google.com/file/d/1hT_iLZHV3wUyZeQgO4G8gT1L94NEQLXz/view?usp=sharing

[EggertEnjoyer123]

Looks like I need to desolder mine because there are components in the way, so I can't take out the plunger. I was able to remove one of the plungers and the contacts were covered in some kind of yellow oxide. The switch worked better after I cleaned the contacts and switch with IPA.

Also, there is another weird fault with the oscilloscope. There is an intermittent problem which causes a large amount of noise to show up on the oscilloscope. The noise is at 22 kHz. Holding a scope probe near the scope shows that the 22kHz signal increases in amplitude significantly when this problem happens. I think this is probably a bad ground connection somewhere.

I think the power supply oscillates at 22kHz (or one of its harmonics) but I'm not sure.

[TERRA Operative]

Good luck! I found a wire was never soldered for the pot at the far top right of the control panel (whatever it does, too many beers to remember right now). I guess whoever owned it before me never used that function so it was never found until I just happened to notice it by chance.


On the subject of test equipment acquisition, I got a new rear plastic bezel for the 53181A, fresh from Keysight. Cost me about 2500yen, cheaper than what people are asking for 3D printed crap on ebay!

I also got the bucking coils for my CT-4 current clamp adapter for my TCP202 and A6312 current clamps. Now I can offset up to 300A of DC bias. The setup should be good for all my railgun and gaussgun design needs. hahahaha.

[factory]

A while ago I started work on the HP 5090A I got from Robert.
This time I started with checking both the Zeners in the power supply, that had failed in the others, they are OK in this one & got reinstalled, the red plastic cased Plessey capacitor on the other hand was knackered.



Looking at the manual there are a bunch more electrolytic capacitors in the oscillator box, this is the larger of the two metal boxes inside, it's supposed to slide out of the side of the frame, I gave up on that method and ended up with this mess on the bench. 





It appears I'm not the first person to have been inside since this 5090A left HP in 1965, lots of the solder joints have been disturbed at some point. The red Plessey capacitors were removed one by one (noting the polarity) and were found to be dead or completely dead. The only capacitor that wasn't bad was the orange Plessey, this was left in place and the rest replaced with newer ones from my spares.







At this point I got very distracted with other TE and it didn't get reassembled until last weekend, I gave all the covers a good clean before refitting them, as they were pretty dirty.

I've done some testing today, the outputs are OK in free run mode, well OK apart from me swapping the 1Mc & 100kc leads when I re-soldered them.  The RF input doesn't seem to be working, at the moment, it supposed to work with 200kc at 1uV from a signal generator, or at least 10uV of radio signal (which hasn't been available on 200kc for several decades). I'm going to try & see if the RF amplifier is off frequency tomorrow, before worrying about the output leads I mixed up.



David

[Vince]

OK the Rochar has been "processed" so to speak....

Friday evening took the chassis out of the enclosure, removed all the knobs, and spent an hour cleaning the enclosure and front panel.
Looks cleaner for sure.

Then looked at how jam packed it was in there, and how unfriendly the mechanical construction is, and immediately gave up diagnosing anything.. it's out of my league. Not worth spending tens of hours and cursing and bleeding knuckles, knowing it's for parts only at the end of the day.



So the only work I did on it was get to the display tubes and do some swapping around, to figure out if the two apparently dead tubes, actually were dead or not. Turns out they are. So I get 6 good tubes....

Then spent my entire Saturday taking the thing apart to salvage 99% of the parts. It's such a nightmare to take apart. Certainly on par with the big 4 bay 7000 series Tek scope I dismantled last year. There are "boat anchors", and "boat anchors".... not all are equal. That Rochar is a very early unit, first generation (A.11xx series), and it shows in its mechanical construction which is not at all serviced friendly.
My other Rochar gear is only maybe 2 or 3 years younger but they improved on the serviceability and I don't mind working on those, thought still not exactly fun and straightforward.

Anyway, I bought if for parts and in parts it is now. 6 display tubes and bucket load of other parts. I certainly have my 45 Euros worth...
I kept the engraved face plate so I can practice painting the grooves to restore the coloured engraved plates on some of my old Tek scopes.




Worthy of note : all voltage rails had zero ripple on them, incredible... bar the -18V that had 3Vpp ripple. Filter cap clearly was shot.
Well, turns out the filter caps in this gear are not all the same model. There is a big one that looks like most you see in old gear, but most were of a kind I have never seen before. They look much higher quality.... look at that : they say "AVIATION3 on them, and the base of the cap is not rubber but metal, the terminals are they traverse the metal, are sealed with a milky looking bead of glass / ceramic ?! So maybe it's possible the electrolyte has not leaked out just yet, hence the caps still perform like new which would explain the ridiculously low ripple ??

Hmmm... interesting caps to me... I will definitely keep them around, if just as part of my vintage parts museum...

Maybe Robert knows about these caps ?

[Vince]

EDIT : CAPS. googled a bit. Apparently the manufacturer's name is " NOVEA".

Found an identical looking one for sale at 13 Euros :

https://radio-antigua.com/en/electrolytic-capacitors/2139-ELECTROLYTIC-CAPACITOR-NOVEA-300mF-25V-NOS--NEW-OLD-STOCK--1PC--CA330U6F270717.html

Except this one is a series " iTA " whereas mine are marked " TTO ". Or maybe it's just a plant code or whatever, but not a series / model designator.

[factory]

That doesn't look that bad to service, with the exception of access to the bottom of the capacitors, but a lot of HP TE is the same. I'm guessing you haven't got the right extender board, I have the same problem with some of the British TE that use card edge PCBs of differing pitch or strange connectors I haven't got.

Some suitable gloves might help if you having problems with sharp bits, they are compulsory where I work for that exact reason.

Those round end view readout bulbs are some of the most common types, of course they can be a bit pricey, but they were in the 1970s (compared with regular vacuum bulbs).

Got any pictures of the boards?

David

[Vince]

Yes the plugin boards are easy of course, assuming you can make an extender. The problem is everything else... it's not meant to be worked not, and quite frankly, not even really meant to be assembled easily either.... I feel for the worked that had to put these things together. I hope they were well paid... I know I would not last more than a day on such a job...

YMMV of course... you have a lto more experience than me working on old stuff, more patience and are more "productive" as well..... but for me, I know my limits... I don't enjoy this old a chassis design. Too many wires, too cramped...working on the controls at the back side of the front panel, good luck. Simply taking the thing apart too hours, so doing delicate diagnosing work... I don't fancy that even remotely.
At least on my Ferisol counter, the entire front panel detaches and comes off, so you can work in all comfort on the back of it.. brilliant design.
But the Ferisol also used all Hypertac FRB card connectors on the chassis... not so much the Rochar. So I guess the Ferisol had more money to work with than Rochar.
So I don't blame Rochar... they did well with the money they had to make this counter...

Problem with Rochar is the convoluted mechanical construction when you get to the details... and a millions tiny flat slotted screws that you can't get a reliable grip on, plus the nuts are mostly impossible to get a wrench onto, never mind operate it. And even if you do, but you don't... Rochar put half a pint of red Loctite on anything that vaguely resembles a nut. Not just screws, but any fuse holder, trimmer, pilot light switch, earthing point.. absolutely anything. Even taking it apart for parts , destructive manner with regards to the wiring, was still a day long chore. No, not my cup of tea... 
I am not into boat anchors to be honest.... I love my old hollow states Tek scopes which is how I came to boar anchorland, but the old Tek scopes I actually enjoy working on, much easier and pleasing.


Board pics ? No I don't have that sorry.... so let's make some just for you then ! Here they are ! 

Boards are very modern I find, for something this old : they use what looks like modern day FR4 material rather than the usual Rochar crappy opaque brown phenolic (?) material where the traces and pads lift and break just by looking at them the wrong way.

First pic are the 8 display tube boards. One of them is the exception to the rule : it DOES use the old brown PCB material.... so I guess some previous owner must have replaced this board in the past.

Second pic is a close-up of one of those display boards.

Third pic shows all the other boards

Fourth pic is the power supply board, and the OCXO board/ assembly. The insulating material is so dry, it's become very hard / dry / brittle, but has not yet turned to dust.

Fifth pic is one from my previous post, showing everything I salvaged. Somehow it failed to display the first time around.

[zepto]

I don't think I've seen discrete nixie drivers like that, very interesting. Not having card extender is disappointing, but at least the pcb connectors are exposed at the bottom. It looks solid mechanically, if a bit primitive. How much did it weigh?

[Vince]

The display boards are nothing unusual. specialized IC's didn't exist in the early days of Niksee tubes, so they used 10 discrete transistors instead.
Same for the 4 bit counter, was implemented with 8 discrete transistors.

Weight ? Couldn't find the info in the service manual somehow, but the sales brochure says it's about 14kg.

It's quite heavy for it's relatively small size.
I found out half of the weight is the enclosure alone !

[factory]

Solartron & Advance of the same age use discrete transistors for the displays too.

Also I know what you mean about bad access, someday I will be sorting out this Racal counter, with it's soldered in PCBs. 



Thanks for adding pictures, some people just don't believe me, when I mention the axial yellow (high-vis) capacitors were available in the 1960s.
The display boards aren't all the same, the higher speed ones are using smaller transistors, much like the HP of the same age.

I hope the parts help restore your Rochar voltmeters back to full health, for my second 5245L (one of the few left that were built in the Bedford HP factory), I've had to use whatever used parts I could get to build the new boards, the cost of new transistors & mica caps would have killed the project. 





David

[factory]

And the HP 5090A is finally working. 

The Marconi sig gen was just too drifty, to use to test the RF amplifier, even after many hours warming up it wasn't stable enough. I did see the signal meter moving if I got the frequency close to 200kHz.





After that I got the 198kHz to 200kHz converter board setup for testing, I've got the board & ferrite rod antenna screwed down to an offcut of wood and I had to make a slight modification the trimmer capacitor position. There won't be a Rev B. of this PCB, as the radio 4 transmitter on LW is due to EOL soon. I predict a lot of "untested" off air standards appearing for sale, at all the usual places.



After connecting the converter to the HP 5090A, I was surprized it was working straight away, I had loads of trouble last time I tried to get this to work. And the HP 5327B (with oven option) settled to exactly 1MHz, measuring the 5090A output.





Just need to get the 1kHz/1MHz output leads swapped back over & get it all back together again. Need to revisit the 5090B too.

Here are all the bad red plastic cased Plessey capacitors that were replaced.



David

[Vince]

Also I know what you mean about bad access, someday I will be sorting out this Racal counter, with it's soldered in PCBs. 

Good luck with that 
You don't lack motivation and perseverance but still, good luck...

Thanks for adding pictures, some people just don't believe me, when I mention the axial yellow (high-vis) capacitors were available in the 1960s.

Yes I sure did notice them right away, looked anachronistic at first, but then I checked the solder side, and no sign of rework. Plus that side of the boards has conformal coating anyway, so it would be hard to hide any rework.
So, original they are for sure.... I guess back then they must have been high spec super expensive caps, so they only used them when they had no choice. In the same vein, on that same board I notice what looks like fancy polystyrene axial caps as well ! No ? So clearly they needed high spec caps on that board.
So, I looked at the markings on that board, it read " Y 285 ". I looked it up in the service manual to see what was the purpose of that board, its schematic, why they might need fancy caps and.... guess what ? Y286 is nowhere to be found in the manual !
Board numbers in this instrument, first generation as I said, are all two digits only, bar two that are low 3 digits : 106 and 114.
From experience, board numbers on Rochar TE increases more and more with every new generation of instrument. For example my A13xx instruments are all in the high 4 digits bracket : Y9xxx.
So it's highly likely that this Y285 board belongs to a similar but slightly newer Rochar counter, for example this one for sale on Leboncoin at the moment : a A.1211

https://www.leboncoin.fr/equipements_industriels/2222216075.htm





Less digits but same function/mode switch, similar I/O at the back, same styling... this rust bucket is yours for the low low price of 300 Euros ! 

So I checked the manual and went through all the boards I took out of the instrument, and saw that there was another board that did not belong to this counter ! I have also a board marked Y286 ! A 285 and a 286...
All other boards are accounted for. These 2 alien boards appear to have taken the place of boards Y75 and Y81. These two boards basically handle the front end and the function/mode switch. So no wonder I couldn't get the counter to perform any of its functions !
So... no regrets, this counter could never have been fixed, because it was no even complete !
And god knows what damage these two alien boards might have done to the rest of the counter...

Thanks David, you allowed me to diagnose this counter post-mortem, what an achievement ! 

The display boards aren't all the same, the higher speed ones are using smaller transistors, much like the HP of the same age.

Yes sure, I mean they all use the same basic topology in their schematic. The operation is identical.

[zepto]

I have too many multimeters, dc sources, and electrometers. I picked up this counter a while back. I think I have two counters now. I really like the 'blue case' era of Takeda Riken equipment, and this counter has a large frequency range and presumably some degree of spectrum analysis capability. I haven't found any documentation for it and I don't currently have the capability to test it. It was definitely one of my more questionable acquisitions.

[Ice-Tea]

I have too many multimeters, dc sources, and electrometers.

No you don't.

[EggertEnjoyer123]

I received my replacement RF switch IC from Robert763.

After swapping out my wire bodge with it everything worked. (I did manage to mess up two of the pads though but with careful soldering I was able to get it back again).

[m k]

Seems that some wires are actually pretty good.

[EggertEnjoyer123]

The Tek 2236 broke after 10 minutes of being plugged in. Initially, the power supply kept on restarting. However, after connecting 42 volts to the test points, the NPN transistors blew up. One of them seems to have failed open, and the other seems to have failed short.

Probably should replace both power transistors and the MOSFET.

The power supply circuit is the weirdest one I've ever seen. I'd like to shoot the guy who designed it

[TERRA Operative]

https://w140.com/tekwiki/images/2/22/050-2242-03.pdf

[Robert763]

I received my replacement RF switch IC from Robert763.

After swapping out my wire bodge with it everything worked. (I did manage to mess up two of the pads though but with careful soldering I was able to get it back again).

Glad it worked. the little metallic / ceramic pakages are a bit of a pain to desolder. I used hot air and a couple of loops if solid copper wire under the legs where they meet the body to lift it off.

Robert.

[tautech]

An old but interesting CRO for sale on NZ Trademe:
Oscilloscope C.T 84. Vintage 1954
https://www.trademe.co.nz/a/marketplace/antiques-collectables/militaria/other/listing/4551543489?bof=Cru5zaWj

[EggertEnjoyer123]

I got the oscilloscope to power on after replacing the NPN transistors and finding a short on the 100V rail. The scope works when I feed in 37 volts into the test points, but higher voltages seem to cause it to not oscillate correctly. The transistors I'm using are the MJE15030.

Obviously it fails to power up when I plug it in. This power supply is horribly designed and I'm wondering how it got past a design review.

The other issue is that the neon bulbs do not light up anymore.