HP 1744A analog storage oscilloscope - repair thread

[c4757p]

I don't have much to post tonight, but I'm off from school tomorrow and will spend some time with this.

Just got an old HP 1744A for repair. Doesn't power on.

Hmm... looks like this one's had a rough life. Filth, rust, and what's this on the inside? Dead spiders, dead insects, and a (thank god!) vacant mouse nest. Bastard chewed a hole in the fan dust cover and crawled right in.

It's been exposed to quite a bit of moisture already, so I cleaned it out in the sink. This washed off some of the Aquadag coat on the exterior of the CRT, but it was already flaking and needs to be reapplied. Not a problem. As far as I can see, the tube hasn't been cracked, though I haven't had it all the way out yet. The mount is broken... will have to find a way to replace it. I doubt epoxy will be strong enough to repair the existing one.

I'll take more, better pictures as I take it apart piece by piece. These storage scopes are getting pretty rare - I am not going to risk powering it on until I've inspected every board.

[dr.diesel]

Ahhhh, old equipment porn, be back in 5 minutes....

[c4757p]

More like Floridian spider porn...

This thing was truly horrifying on the inside. Shame people treat their equipment like this...

I'm really hoping the CRT didn't suffer any damage from the broken mount. The ass I bought it from packed it about as well as you'd pack a football... I couldn't figure out how to remove the CRT though, and didn't want to mess with it since they're pretty much irreplaceable. I'll read the service manual before trying to get it out.

I have a slight suspicion that the only reason it wouldn't power up is a slightly misaligned power button.

[edpalmer42]

I have a 1744A sitting on my bench that was manufactured in 1983.  Since it just keeps on working, I haven't been able to talk myself into buying a digital scope.  Mind you, the fan is getting a little noisy.........

Ed

[Kilroy]

Do be carefull of that storage CRT...it is a very special tube. Too bad the scope looks so neglected because it is a very capable 100Mhz instrument. If it offers you any hope at all, I have seen some scopes that looked like they were thrown out of a moving vehicle and the tubes were fine, amazingly.

Incidently...the 1744A is basically a 1741A, sharing the same 100Mhz bandwidth but having a more sophisticated storage CRT. HP called it an "expansion storage" CRT. It used a small, high presicion storage mesh combined with an electronic lens to magnify and projected the image. And it was really quick...a write speed of 1800cm/uS.


I have a 1727A which has an even faster expansion storage CRT at a write speed of 2000cm/uS. It uses an improved CRT based on the 1744A tube but with added high speed vertical plates to match its high bandwidth 275Mhz amplifiers. This 1727A CRT is a beautiful piece of engineering...using transmission line type deflection plates which, to the vertical amplifier, appear as a resistive rather than a capacitive load.

[c4757p]

I still haven't taken the time to remove the CRT (time for bed! ) but I think it's OK. There was a lot of water in and around the whole instrument while I was cleaning out its former residents... and there is no evidence of moisture on the display surface at all. If it were cracked I'm sure it would have taken in water.

Transmission line deflection plates... I'm going to have to read about those.

Damn... thing still smells like a hamster box. The individual boards are going to need very thorough cleaning

[c4757p]

Warning: NSFW.

The CRT is good. Very good. Pornographically good.

I've linked the pictures clickably so they don't get loaded a million times by people clicking past this thread. 1.5 MB total...

The tube is very large. It goes all the way across this pad of paper, which is the same size as those large desktop calendars.
Pic

Handwriting inside the tube...
Pic

Here's the storage assembly. Intricate construction - look at those glass rods and pins...
Pic

And the storage mesh. Most storage CRTs had a mesh directly behind the display surface; this one has it halfway down the tube, with a lensing system between the mesh and display.
Pic

All of these pins are in addition to the whole bunch at the neck socket. These are dedicated to the storage assembly and horizontal deflection (and probably vertical as well, though I missed that bit in the manual). This looks like a nasty weak point on the tube...
Pic

The cathode electron guns:
Pic

Rusty neck socket. Does anybody know of a good way to clean that out? Or, should I just leave it?
Pic

The Aquadag coating on the tube is stripping. This is a conductive graphite material which is either grounded (on the outside) or connected to the anode HT (on the inside). The two overlap to form the plates of a glass-dielectric capacitor to decouple the HT. I'll strip it off the rest of the way and reapply.
Pic

The front of the CRT.
Pic

[c4757p]

Pre-diagnosis: Failed capacitor. One of the huge ones in the PSU is covered in a sticky, chemical-smelling substance.

Not mouse piss. You know your scope has had a hard life when you give it a literal bath to clean out the mouse piss. It was making my lab smell like a poorly kept pet store.

[PaulAm]

That's not rust on the socket but cement used to bond the plastic base to the tube.  As long as the base is solid on the tube, don't disturb it.

[c4757p]

That's not rust on the socket but cement used to bond the plastic base to the tube.  As long as the base is solid on the tube, don't disturb it.

Oh, is it? That makes sense, there's nothing there to rust Except the pins themselves, which are perfectly clean.

It's not quite solid. Wiggles a bit. Should I do anything about it?

[PaulAm]

If it's loose it should probably be bonded.  You can might be able to avoid that if you are VERY careful putting the socket back on.  If it's loose you run the risk of breaking one of the leads inside the socket, or worse yet at the glass seal on the tube.  During manufacture, I think the base was bonded first and then the wires soldered into the pins.

I am not familiar with vacuum tube restoration, so I wouldn't hazard a guess on what type of cement to use.  You need to get it right; that tube is irreplaceable.

I'm sure you'll get lots of advice on what cement to use, but unless it comes out of experience, it's worth what you're paying for it.  You should probably check in with the hp_agilent_equipment yahoo group or the Antique Radios forum.  Either one of those are likely to have people who not only actually know what they're doing but have done that kind of repair.

[HiTech]

I wouldn't bother inquiring in the ARF forum - most of those goobers there mess around with crap scopes. The mention of a Tek or HP ruffles feathers on some of them. I think you could safely use a small dab of construction adhesive like Liquid Nails to secure the socket to the crt. Or perhaps hot glue--- provided the glass isn't heating too much. I've never bothered to test the CRT temp. on my HP 1741A-- no reason to.

[c4757p]

I will test it with the socket loosely, gently connected, and if everything seems to be working (no broken connections) I will wash out the gap with alcohol and put some high-temp epoxy in there.

Checked around and the question has been asked already loads of times, with a wide range of answers. Kind of pointless for me to get yet another.

If a connection is broken, I will very carefully attempt to desolder the socket and see whether there's any wire length remaining behind it. Of course, I'm screwed if it's broken at the glass. I did find one of these tubes for sale for $40, which I will buy as a last resort (yeah, it's expensive, but this is a nice scope and I'd like not to see it die forever)

[HiTech]

I bought a working 1741A to keep on hand as a "parts" scope should mine go down. After having it a year I decided to sell it due to storage space needs. My main 1741A had been faithful and a person can have too much redundancy on hand whereby it never gets used and takes up needed space. Did the same with my Tek 475..... I have just the two scopes now.

[PaulAm]

Looks like a high temp RTV silicone would work.  There are 2 problems: bonding to glass and sustained high temps.  You just need something to hold it in place without putting any pressure on the glass.

Curiously, I ran across references to a tool that was used to cut off the ends of the pins so tube bases could be reset.  That was from a time when tubes were rare and expensive.

[IO390]

Oh my word... That tube is just incredible.

[c4757p]

High voltage PSU. This is a pretty normal flyback-type CRT HT board, though it has a couple more protection bits than most due to the precious CRT.

Schematic in the attachments.

Here's the whole thing:
Pic (218 kB)

Seems like this board got some of the worst of the moisture problems. Soldermask had peeled in a couple spots - I cut it the rest of the way off and remasked with polyurethane:
Pic (61 kB)

Many of the capacitors needed attention - but all except one were able to be reformed. I had a nice Rubycon on hand to replace the bad one - 105C and double the original voltage rating:
Pic (66 kB)
Curiously that capacitor was not in a high-stress position at all, and yet it was the faulty one It is C3.

I love these parts going down to the transformer assembly. Looks a bit bodgey at first, until you realize they have 3kV on them and it's the neatest way to handle that:
Pic (113 kB)
CR3 and C6. That's a ceramic tube capacitor, for those who haven't seen one.

[c4757p]

Quick update. Not a whole lot to see here. This is the gate assembly - responsible for controlling the display intensity. I'm not even going to post the schematic, it's basically just a couple amplifiers (even the "custom HP" chip isn't interesting, it's just a rebranded CA3045 matched transistor array). Of minor interest is the fact that this particular version has a CRT beam overcurrent sensor, helping to prevent burning of the storage mesh.

Front. You can see that this board suffered even worse moisture damage; quite a few traces were losing their solder mask and had to be repaired.
Pic (148 kB)

Back. Even more moisture damage...
Pic (113 kB)

Kind of interesting - I'd file this under "acceptable bodges". Look how they attached the rod for the beam find button to the switch:
Pic (103 kB)

The trigger assembly is coming next - that seems to have some good stuff on it. I'll try to put together some text to explain the functionality.