Test Equipment Anonymous (TEA) group therapy thread

[m k]

I have a small 'had to be bought' on my table.

What is your best method to get 40 years of gunk out of HP knob grooves?

Either sopay water or isopropyl alcohol and an old toothbrush.

I had to google it.

[McBryce]

Though sometimes, only a new toothbrush will do...

But then the wife gets all angry when she finds out what you've been using her new toothbrush for!

McBryce.

[Wolfgang]

I think in my case a new front panel was the solution to the gunk problem.
Mine is certainly older but now has a Keysight logo on the front.
From the standpoint of stability - the older the better, because drift gets less over time

[Robert763]

I have a small 'had to be bought' on my table.

What is your best method to get 40 years of gunk out of HP knob grooves?

1/ Ultrasonic cleaner with water, ammonia and a bit of dish washing soap.

2/ When household management are not around put them on the rods in the dishwasher.

3/ Foam cleaner and a electric toothbrush (use an old head or somebody elses)

[AVGresponding]

Though sometimes, only a new toothbrush will do...

But then the wife gets all angry when she finds out what you've been using her new toothbrush for!

McBryce.

You haven't even learned to hide the evidence yet? What happens when you need to use her nail products?

[TERRA Operative]

I have a small 'had to be bought' on my table.

What is your best method to get 40 years of gunk out of HP knob grooves?

1/ Ultrasonic cleaner with water, ammonia and a bit of dish washing soap.

2/ When household management are not around put them on the rods in the dishwasher.

3/ Foam cleaner and a electric toothbrush (use an old head or somebody elses)

I have committed offence number 2 on more than one occasion see, and I'll do it again you hear!

[Peter_O]

I have a small 'had to be bought' on my table.

What is your best method to get 40 years of gunk out of HP knob grooves?

Either sopay water or isopropyl alcohol and an old toothbrush.

I had to google it.

point taken.

But without the simple question we would not be able to discuss and relish complicated answers or the diversity of social side effects. 

Thx to all!

And to reveal the object:

It's something I definitely do not need, because I have a 3310A, a 3325B, a 8904A, a 2042X ,...
And it's something I definitely need and have been searching for, because it's HPs last fully analogue function generator.

It's a HP3312A from 1975, and it was in a really gunky shape and of unknown functional condition, but - for germany - cheap: 70 Euros incl. shipment.

After some fast first cleaning of case and front it dares to show up here, still with gunky knobs of course. 
Main and modulation generator are working generally. Offset calbration and some more are off, switches are shaky, etc., but overall I'm happy.



How do you fix these tiny button caps?


How would you mount the pastic strip? One of the fittings is missing, one does not clamp any more.

[DC1MC]

Shameless shilling for the last remaining AR488 GPIB adapter (8 already gone, 1 stays with me), this is number 9 ready for shipping. I sell them at cost, but the parts (buffers, GPIB connector) are from Mouser, so it may be a bit more then the Ali version.
This is the last one of the buffered version, I still have a countless number of PCB for the unbuffered (but physically smaller) version, but I guess no one in interested in these.
Just in case you're in DE or around, where the shipping makes sense, and need one, PM me.

 Cheers,
 DC1MC

[Vince]

Tektronix 7603 scope

Refresher : 4 months ago spot on, I bought that 7603 rack mount Tek scope.
Was displaying weird and wonderful things on the screen, found that the 5V rail for all those TTL chips, was kaput. Filter cap was dead.

Today, I finally replaced that cap. Was a big can, 18,000µF 15V , replaced it with two 10,000µF 25V.

My 50W Weller Magnastat soldering iron was totally inadequate for the job, it struggled big time, barely made it... but it managed. With patience and lots of flux to help, solder sucker, some wick, some fighting... the old can was extracted. One of the reasons this job waited for 4 months is that I wanted to get a more powerful iron to tackle this job, but I have no money burning a hole in my pocket right now, and just don't know what to buy anyway, so eventually I got fed up of having the scope in bits for so long, and gave it a try anyway. Luckily it went alright, no damage to the board... it's a beefy PSU board not a delicate 16 layer digital board eh....

Other reason I wanted to do that job now, is that there are many... many cables going to, from, and through that PSU, and memory was starting to fade as to how to plug it all back together properly...

A bit anxious when I fired up the scope... but I guess I did things right, no explosion, no smoke, scope still works, and the 5V rail is now perfect.

However we now had a new problem on our hands : the entire display (readout + the traces, in unison) was shifted south quite a bit... the bottom line of the readout was half hidden. Then after a few seconds the display started jumping up and down erratically... like there was a bad connection somewhere or something...
I powered it off, waited a few seconds and powered it up again... problem gone.

So I guess some more caps in the PSU need replacing ?!

Also, I noticed that the beam is hardly sharp (readout or the traces, all the same). Best I can make it using the brightness and focus controls is... average, like in... 5 bucks '70s junk hobby scope.. average. Usable for sure, no worries, but hardly nice never mind great.

Also, I notice that even though the inputs were ground coupled, the traces were very noisy... and the readout when looked at close up, is trembling a little, it's not rock stable.

So I guess some more caps in the PSU need fixing here as well.


So.....  looks like I will bit the bullet and replace every single of those can caps in the PSU. 7 of them. One down today, remain 6. Ordering this evening... want to do it all while I am in the mood and while I remember how / where to route / plug all the cables.

I am glad to report that no pMOS chips got maltreated during this cap job, and even though I checked the 5V rail with that nasty horrible Chauvin-Arnoux DMM I fixed the other day, that you warned me all about, ... it did OK, did not explode at me. I was a bit scared I must say, given that 5V is quite a dangerous voltage as we all know.. but I am so stupid I tried it anyway. I may not have as much luck when I get to check the 15V rail... maybe I will wear thick rubber gloves and a beefy welding mask.

[factory]

...snip...



How do you fix these tiny button caps?


How would you mount the pastic strip? One of the fittings is missing, one does not clamp any more.

I think someone on here 3D printed some button caps, alternately look for something else in poor condition to take them from, of course this can backfire & you have another item of TE you decide to keep. 

They used double sided tape on the older HP cases, to attach the aluminium trim strips.

David

[xrunner]

I think someone on here 3D printed some button caps, alternately look for something else in poor condition to take them from, of course this can backfire & you have another item of TE you decide to keep. 

I have in the past - they don't look too bad. You can get real anal if you want to and paint the plastic with acrylic paint to perfectly match color.

[factory]

I've not posted much the last few weeks, been busy last weekend moving stuff into storage again, this time some of the component cabinets I don't use often, as well as some TE.

Also been busy working on the HP blanking BCD decade counter IC problem, took me quite a few evenings & weekends trying & failing to get a replacement design to work, then by chance I found a block diagram in the 8443A manual, showing five JK MS flip flops, one AND gate & some NAND gates.



Even this didn't work at first, but I eventually got it working with NOR gates, a 74LS90, a 74HCT107 flip flop and a few extra gates for inverting inputs & outputs. Now got to get this mess sorted in the PCB layout programme on Kicad, hoping to get it to fit on a 2 layer board, but doubt it will be possible for the small horizontal board. Might try replacing the 74LS90 with four more JK flip flops for the vertical board as they are cheaper.







Once this board design is done I'll get the various boards made, probably just five of each at first, to make sure they actually work. 

P.S. Another block diagram from the 8443A manual, showing the buffer/store IC and a gated output that doesn't seem to be used in the 52/53xx IC based frequency counters, I believe this is pin 6, I might incorporate this into Rev.B of the board, as I will probably find something that uses it one day.



P.P.S. Part numbers have been omitted while these are at the design stage, will add them once the boards arrive, get built & have been verified as working.

David

[Specmaster]

I have a small 'had to be bought' on my table.

What is your best method to get 40 years of gunk out of HP knob grooves?

Either sopay water or isopropyl alcohol and an old toothbrush.

I had to google it.

point taken.

But without the simple question we would not be able to discuss and relish complicated answers or the diversity of social side effects. 

Thx to all!

And to reveal the object:

It's something I definitely do not need, because I have a 3310A, a 3325B, a 8904A, a 2042X ,...
And it's something I definitely need and have been searching for, because it's HPs last fully analogue function generator.

It's a HP3312A from 1975, and it was in a really gunky shape and of unknown functional condition, but - for germany - cheap: 70 Euros incl. shipment.

After some fast first cleaning of case and front it dares to show up here, still with gunky knobs of course. 
Main and modulation generator are working generally. Offset calbration and some more are off, switches are shaky, etc., but overall I'm happy.



How do you fix these tiny button caps?


How would you mount the pastic strip? One of the fittings is missing, one does not clamp any more.

Those knobs are not too bad at all, with my HP knobs like that, all I did was to soak them for a few minutes in hot (not boiling) soapy water and used a cheap toothbrush running along the grooves and the dirt almost fell out, and it looked like new after a few moments soaking.

That style of knob always reminds of the Armstrong Hi-Fi units of the 1960s, my parents owned one, a valve, or hollow tube jobbie if you prefer to call them that. I remember that I was not very popular for a while because I managed to put an entire Reader's Digest vinyl box set of Bert Kaemfert & Joe Last 33.33rpm records on top of it while switched on, and the lot became badly warped as a result. I remember having to place them between 2 sheets of glass to remove some of the warping, so they could be played. 

[xrunner]

I finished the Siglent 200 MHz probe rebuild. I trimmed back the shield and soldered it to the area that was only crimped (slightly poorly) before, so there won't be any loose goings-on there now.

All the parts are laid out for assembly in case any viewers out there want to do this repair. First I inserted the plastic end with the probe tip into the long metal case. Then the metal case slides onto the PCB. From the factory it would have been crimped with a tool I do not have. I tried to see if any of my coax crimp dies would be a close fit, but none were. So I chose to solder in two spots near the end to fix the metal case solidly.

Next the longer black plastic outer cover slides on. Originally it was secured in place by metal fingers on the inner metal case that gripped the plastic inside, but it wasn't designed to come off. Since it wasn't as tight now, I used a little CA glue to fix it in place.

The same thing applies to the smaller black part near the probe tip. Since the metal fingers were in good shape (you can see them near where the ground clip attaches) I simply rotated the case 90 deg. so the fingers gripped new plastic, and it held tight.

Last, put on the red switch cover, probe hook, and attach the ground lead - presto we have a rebuilt probe. No more noisy fiddly bits now, it works just like new.

[Zoli]

...
3/ Foam cleaner and a electric toothbrush (use an old head or somebody elses)

If you have to deal often, or heavy stuff once:
https://www.rubbermaid.com/home-organization/cleaning-supplies/hand-held-cleaning-tools/reveal-power-scrubber/SAP_2124405.html
I have one, and puts any/all toothbrush to shame; it worth every penny, and you can get it often on sale. Best of all, it has household management approval for double use.

[Peter_O]

I think someone on here 3D printed some button caps, alternately look for something else in poor condition to take them from, of course this can backfire & you have another item of TE you decide to keep. 

I have in the past - they don't look too bad. You can get real anal if you want to and paint the plastic with acrylic paint to perfectly match color.

Cannibalizing ist no option, no. Finally, this 3312 would have been a candidate starting from the seller's pictures. 

FDM 3D printers are available, yes.

Do you possibly have the STL at hand, please?!
This one?:
https://www.thingiverse.com/thing:3057821

We have some guys at our 3d print forum over here who are really into acrylic painting. maybe i should make an order. 

[McBryce]

Though sometimes, only a new toothbrush will do...

But then the wife gets all angry when she finds out what you've been using her new toothbrush for!

McBryce.

You haven't even learned to hide the evidence yet? What happens when you need to use her nail products?

It was only after she discovered that toothpaste doesn't actually taste like isopropanol that she started getting suspicious.

McBryce.

[m k]

I have a small 'had to be bought' on my table.

What is your best method to get 40 years of gunk out of HP knob grooves?

Either sopay water or isopropyl alcohol and an old toothbrush.

I had to google it.

point taken.

Not you, the other guy from your area.

The search was also a bit thin, only some old Lakota stuff was revealed.

[m k]

Also been busy working on the HP blanking BCD decade counter IC problem, took me quite a few evenings & weekends trying & failing to get a replacement design to work, then by chance I found a block diagram in the 8443A manual, showing five JK MS flip flops, one AND gate & some NAND gates.

I was wondering why my earlier search lacked XOR chips, so I tried again, and it seems that some numbers are reused.
Farnell offers 74LV1T87, a single gate version, single flops and inverters are also available.
With those singles your board area can be a bit smaller, 5 flops, 6 nands and an inverter.

[xrunner]

We have some guys at our 3d print forum over here who are really into acrylic painting. maybe i should make an order. 

The reason I say paint it is if you want to do a "museum quality" job. For example I had a Boonton power meter that was missing a button. I made the F360 file and printed the button in black, and put a white dot on it. In the picture it's the right button. But I wanted to see if I could make it appear as good as the original button. I put on about three coats of glossy acrylic paint and sanded in between. The paint is high quality modelling paint shown below.

It's the left button in the picture. You can't tell it apart from the original button when installed. So for some jobs, when just printing the button is not good enough, you can do more if you want to. 

[TERRA Operative]

My TM506/5006 mainframe rescue continues...

I went to install the new IEC fused mains inlet/voltage selector o my keeper TM5006, and while the pinout is the same, the size is not, the new one is shorter. So I'm waiting on an adapter cover plate to reduce the hole size down. Using Misumi Meivy machining service. Upload a STEP file, set the options and parameters, and 20 bucks and a wait for delivery later I'll have my part.
That should have me up and running with that one.

The other TM5006 has a fault in the switch mode board used for stepping down the voltages from the 350V rectified DC bus. Have the PCB n the bench and am currently poking at it to see what's up.

And the TM506 is all reassembled with new capacitors, but after testing with my TM500 mainframe tester, it looks like someone went and busted up a bunch of bays.
I bet it's the usual "Hmm, this bay doesn't work, lets's try the next... Still not working, let's try the next, hmmm, and the next......." Not stopping to realise that its a bad plugin that's POPPING ALL THE NPN PASS TRANSISTORS!!     
All up 3 bays of the 6 test bad... Good thing I have my tester or I would have risked killing an expensive plugin or two...

Anyway, I can get a replacement 2N3055 from Akihabara for a buck each for the high power bay, and I ordered a full compliment of TIP31C and TIP36C transistors for the other bays, so I'll be able to replace them soon and get this mainframe working perfectly again. It cleaned up real nice too.

[factory]

Also been busy working on the HP blanking BCD decade counter IC problem, took me quite a few evenings & weekends trying & failing to get a replacement design to work, then by chance I found a block diagram in the 8443A manual, showing five JK MS flip flops, one AND gate & some NAND gates.

I was wondering why my earlier search lacked XOR chips, so I tried again, and it seems that some numbers are reused.
Farnell offers 74LV1T87, a single gate version, single flops and inverters are also available.
With those singles your board area can be a bit smaller, 5 flops, 6 nands and an inverter.

I've been using parts that are available in HCT or LS versions, for the TTL compatibility to avoid problems with existing parts in the HP kit.

The HCT single gates available from Nexperia in 74 series GV package (SOT-23, 5/6 pin) are 00, 02, 04, 08, 14, 32, 86, 125 and two gates are 04, 14, 17, 34 (some are reused numbers that don't match up to old 74 series). These are what I have chosen, already have a few 02 ordered from Farnell, as Mouser don't have the GV version.

There is little choice in JK master/slave flip flops, I bought a few different types in DIL packages for my prototypes, that are still available in LS and/or HCT versions in SOIC package, I was unable to find modern HCT single JK MS flops, the 7490/92/93 counters contain four MS flip flops according to TI.

Three inverter gates were required, for the reset & count inputs and the divide by ten carry output, fun with old stuff that uses negative logic again. 

The layout problems are more due to numbers of tracks between devices, rather than the size of the ICs, it's looking more likely that I will need four layers to get all the supply rails connected.

Suggestions of some low cost tweezers for positioning/assembly of these surface mounts parts would be good, hopefully to avoid losing too many parts when they ping away never to be seen again, 0603 resistors are the smallest parts I have chosen.

David

[m k]

I thought 4 layers is a norm, also remember reading something about JLCPCB and 2 layer quality.

Can you use fine line style, 2 tracks between legs?
At least internally connected tracks should be ok.

Some consumer boards are also using jumpers quite heavily.
Under zero ohms you could use multiple tracks for sure.

Have you tried artistic freedoms with component positioning?
5 pin singles are also quite adaptive.

I generally don't like tweezers, one moment something usually zaps away.
How about those small manual suction pens, they at least have less spring force.

[Vince]

OK I have ordered the caps I need to complete the PSU recap of the 7603 scope. The cap situation is still not resolved, could hardly buy what I wanted or needed, had to make do with I could find. Now is not the best of times to be picky about electronic components 
35 Euros for just a handful of caps... add the 2 I just fitted yesterday, I am already in the 40+ Euros just to recap the PSU of this thing.. which is what I paid for it IIRC. Recapping old TE is expensive (and usually a pain in the butt to do as well). One more reason for which I will slow down on boat anchor harvesting, and will be very picky about which ones in my stock, will get to be restored, and which ones will be for parts...
I most definitely won't be restoring my 25/30 hollow state Tek scopes, that's for sure...

Anyway, today I switched to the Tek RM17. I bought it 9 months ago and since that time it's been waiting for caps too.
Had problems with the vertical pre-amp, all of which were due to a 3 section can.
I meant to replace it with a little adapter PCB, so I had pulled the can out of the scope back then, to take measurements... but never got round to doing it, begin overwhelmed by the details, that I wanted to get right the first time. The devil is always is always in the details. No, I won buy ready made ones from Ebay, I don't like them and way too expensive for what it is. I want to design my own the way I want them.

Anyway, so no PCB for now, but still wanted to button up the scope.  So instead I thought I would reuse the old can, this way the new caps would still mount on the existing cap base/socket, problem solved. I never messed with can caps, so I thought cool, good excuse to do my first one. I saw it open centimeter above its base, so as to be sure to get the seal intact.



Inside it's not as messy as I was told... it's all dry as a bone.




I carefully unrolled the foils to get to the metal plates/pins without damaging them.




The 3 new caps fit peeeerfectly inside the base, could not be happier....



.....I thought. Except I could not solder the leads of the new caps to the inner terminals of the old can.... solder just WILL NOT STICK !    No matter what.
Geez... but I thought people DID restuff these old cans eh ?! So how do they do ?!   
Maybe they remove the old terminals, leaving holes in the seal, and just thread the leads of the new caps through these holes... yeah maybe. Could have done that... but didn't think of it. Next time maybe.

So instead I soldered them directly to the wires that used to go to the can. The new caps are so small and light, no big deal. Put a bit of heatshrink tubing to cover the solder joint on the hot lead of the 3 caps.



Then to connect the common ground / negative leads, lucky me, right next to the old can was power resistor, mounted vertically on the chassis, solidly anchored. One of it's terminals was connected to ground via a tab, a tab which had an extra hole, waiting for me to shove all 3 leads of my caps through it ! So the 3 caps are firmly held in place, perfectly safe, they are going nowhere. Looks good enough to me.




Might fit a PCB in 20 years when I finally design one, but for now it's good enough.


Gave it a quick test ride to make sure it still worked.. it does, but I noticed a new problem. Sitting for 9 months is never good for any old TE eh...

Video of the problem below.

It looks like a problem in the horizontal section this time. The trace does not go all the way to the right edge of the graticule. IT depends on the sweep speed setting. Up to about about 10ms/DIV, the trace goes up to about 7.5 DIV. Then from that point, as you keep cranking the speed up, the trace extends a bit more to the right, each time you crank the knob one step. Once you reach 50µs/ DIV it almost extends all the way to the right edge.
When you set to not extend fully, you can see the right tip of the trace  trembling a little bit, it's not rock solid.. and once, all of a sudden it fixed itself and jumped all the way to the right... and a split second later came back to its former shorter self again 

So it looks like a bad contact somewhere, an intermittent issue..... or a bad cap again. That's one of the reasons I replaced this 3 sections can. Of of the 3 sections was causing horribly noisy signals when using the most sensitive settings of the vertical amp. Looked like a bad contact but was that cap.   

I will try to fix that problem right now if I can, while it's on then bench and while I am waiting for 7603 caps to arrive.
I does not look like an amplifier issue. Looks more like a sweep length issue. So will look at the amplitude of the sawtooth signal first...

[tautech]

My TM506/5006 mainframe rescue continues...

All up 3 bays of the 6 test bad... Good thing I have my tester or I would have risked killing an expensive plugin or two...

Anyway, I can get a replacement 2N3055 from Akihabara for a buck each for the high power bay, and I ordered a full compliment of TIP31C and TIP36C transistors for the other bays, so I'll be able to replace them soon and get this mainframe working perfectly again. It cleaned up real nice too.

That is one beast the don't make like they used to and in this application if they are loaded at all 2N3773 might be a better solution.
Just saying.