Cracked HP knob (3582A spectrum analyzer)

[Rax]

Continuing my instruments repair/restore saga.

This is an HP3582A with a cracked "span" knob. I seem to have all parts, but just wondering on what have others use to bod it back together. I'd probably err on the side of something like Gorilla Glue, though it tends to bubble a bit and not sure it would make a very "easy on the eyes" repair.

Pictures below.

[Rax]

The rest of the pics...

[Rax]

One more...

[Rax]

For anyone wondering, this is HP part # 03582-07904. I am also missing the small knobs #5040-8832. See images below.

I haven't been able to track down  source for either, though the former I hope I can just repair.

[Tomorokoshi]

Model glue? There are various kinds for use with plastic. Usually they go on thin and are relatively clear.

[Rax]

Interesting idea, Tomorokoshi... Are they typically strong? I used to do assemble models when a child, but haven't used model glue in decades. Would you think they'd withstand the exigencies of frequent use of those buttons?

[fuzzoli]

I'd recommend CA glue (i.e. CrazyGlue).  I've used it a few times to repair knobs.

On a similar note, if you want to send me the broken knob, I can try to design a 3D model of it that can be printed.  PM me if you're interested.

I've recently modeled and 3D printed several knobs for some Tek scopes I'm repairing.  Please see the attachments for some samples.  The top knobs in each photo are the originals, while the bottom ones are the printed versions.  I even mixed a custom color for one of the knobs.  It's pretty close -- not exact though.

-Frank

[MarkL]

I would try to glue the pieces back together with a plastic solvent welder such as Tenax 7R or Ambroid ProWeld, if you can find them.  If you can't, there's this product from Micromark which they claim is chemically the same:

  https://www.micromark.com/Same-Stuff-Professional-Plastic-Welder-Refill-2-fl-oz-_2

To use it, you would hold the pieces together and touch the applicator to the crack between the pieces.  Capillary action will draw the solvent into the crack and melt the plastic.  After holding it together for a minute or two, you can release it, but I wouldn't try to do anything with the knob until it has overnight to dry.

Use it sparingly and be careful not to get it where your fingers are holding the pieces.  The capillary action will draw the solvent between your finger and the plastic and ruin the plastic surface.  It's also probably not so great to have it absorbed into your skin.

Because it melts the plastic, the bond between the pieces is pretty strong.

[lugaw]

Model cements are solvent and melts the plastic to join them together. They are strong but be careful when using them.  When you over apply you might melt what's left of your knob. Check out the Tamiya cement lines for different kinds of plastic cement.  I'll personally use CA glue.

[Mario87]

Model it in CAD and 3D print a new part, easiest and probably most reliable long term fix other than getting a new part.

[jonpaul]

Bonjour, all these old knobs are of a material that tends to crack and craze over te decades.

Inside is a metal bushing.

No glue can hold with the torque required on certain controls of these old analyzers

A 3D needs to have a very special material, most common 3D are very soft amd low tensile stregnth plastic.

Best solution: Wait for some to appear on bay  OR use EEV WTB in buy sell.

You may fimd a donor unit or junker.

Bon chance


Jon

[Rax]

Thank you all for your input.

I provisionally grabbed an epoxy (JB Weld - Clear Weld) locally, but have not yet used it. I'd appreciate contrasting thoughts on Epoxy vs. Model glue (which I gather, from previous posts, is similar to other "plastic melting" solutions such as "MicroMark Same Stuff"). It seems CA glue is yet a third category, or a superglue (cyanoacrylate).  I am curious on comparative thoughts on what seem to be three different categories:

• Epoxy
• Model glue (plastic melting agent)
• CA (superglue)

In my mind, this far epoxy seems the better solution (due to its higher shear strength, and capacity to fill voids, the latter being to me pretty critical for this case where there's a metal post to attach the two plastic parts to).

Thanks!

[Tim T]

I have a 3582a and there are fairly strong detents on that span control. So just gluing the plastic knob together at point of failure  with adhesive of choice might not work. But worth a try and if no joy then you are probably looking at bonding the plastic to what appears to be a brass sleeve. Removing the sleeve from the shaft is recommended, and keep the adhesive away from the set screws.

tim t

[mag_therm]

Yes,  I had a similar cracking ( to Rax's photos)  failure of a similar unobtainable knob on the vert V/div of a Tek 466.
That was about 3 years ago. I think it was caused by stiff detents.

I clamped and glued it, I recall with JB Weld ( 2 part epoxy)
I did not try to close up the crack completely, thinking that might cause another crack.
So far it is holding.

Also the detents on one probe AC-Gnd-DC switch got so stiff I removed them, thinking that switch would break.

The Tek 466 was purchased damaged  about 15 years ago with both the B and A trigger concentric controls with fractured knobs and bent pot shafts.
( They are protruding at the right edge of the panel)
Not finding replacements, I looked at earlier models parts photos  (eg Sphere Research in Canada, they have Tek and HP parts)

I purchased some older model  10 k pots with similar inner concentric switches , along with older model knobs.
Turned bushes and nuts out of  brass  on the lathe and purchased a tap to thread the nuts.


I also turned replacements for the  rear feet on this 'scope ( combo case clamp, rear stand and cord wrap)
out of Al. -but without the cord grip grooves.

The repairs are functional but not original, scope is working OK.

[Rax]

I ended up using the JB Weld - ClearWeld stuff. I think attachment to the metal sleeve and filling of gaps is essential in this application.

Now, the big question is... Where does the indicator point when, for instance, the control is all the way CW? Or CCW, if easier to explain. This is because I never saw the knob attached and I have no idea. I will try to figure it out from the manual, but a clear direction on this would likely help me save a ton of time.

Thanks!

[Rax]

I have a 3582a and there are fairly strong detents on that span control. So just gluing the plastic knob together at point of failure  with adhesive of choice might not work. But worth a try and if no joy then you are probably looking at bonding the plastic to what appears to be a brass sleeve. Removing the sleeve from the shaft is recommended, and keep the adhesive away from the set screws.

tim t

Pinging Tim... Hopefully he can point me to the right direction on how to orient the knob (wait. did I just make a pun?...) on the shaft before securing. Thanks much in advance!

[Rax]

I think I figured it from the screen notations (once I started the instrument). If anyone has input to confirm my path, I do welcome it, but I think I'm in good shape.

[Tim T]

When the shaft is turned counterclockwise as far as possible the tab of the knob should be over 0-25 KHz Span.

tim t

[Rax]

When the shaft is turned counterclockwise as far as possible the tab of the knob should be over 0-25 KHz Span.

tim t

Thank you very much, Tim.
Radu.

[Rax]

This far, the epoxy seems to hold well. I've been playing with this and seems the button is solid.