HP 8644B Attenuator Question

[George_Race]

I have an HP 8644B Signal Generator that has developed a no output condition at the -70 db and below point of operation.  Has been acting up, sometimes working sometimes not, at that level for about the past year or so.

Have traced it down to the center section of the 33321-60029 attenuator being open when actuated by the control.  Took it apart and checked it out, can see the that 3 of the small "O" rings are missing on that section of the attenuator.  I really don't want to take it all the way down and attempt to repair it, causing still further damage then is there already.

Replacements for that particular piece are several hundred dollars.  But, I have found what appears to be a similar unit under the part number 33321SC, that appears to have all the same specifications as the original 33321-60029 one has.  It is CONSIDERABLY less to purchase, from several sources.

Does anyone know specifically what may be different between the two part numbers, if any?
Have the German Manufacturers specification sheet on the 33321SC, but can't find one on the 33321-60029.

Everything I can find so far makes them look identical in construction and operation.

Any help or suggestions would be greatly appreciated,
George



 

[MadTux]

https://www.google.ch/search?q=33321-60029&client=ubuntu&hs=n7l&channel=fs&source=lnms&tbm=isch&sa=X&ved=0CAgQ_AUoAmoVChMIzfe61_3yyAIVQdkUCh11wwOC&biw=1024&bih=607

33321-60029 seems to be the standard HP/Agilent RF attenuator type. Used all across their RF/microwave instruments.

You can screw apart both end caps, then you can slide away the metal cage. Then you'll see 3 or 4 electromagnets that operate tiny metal spring contacts in the upper section. That upper section is screwed shut with lots of screws, but you can take it apart without breaking anything. You have to remove the plate on top by unscrewing those 2 nuts on SMA connectors to before you see those screws.

Then you'll see those tiny gold plated metal spring contacts. On the part you screwed apart are 3 or 4 tiny ceramic or sapphire plates with dual Pi-attenuator elements (30dB/40dB elements are stacked Pi elements)

Even those ceramic/sapphire plates are sometimes cross compatible between instruments. I know that because I exchanged the 10dB and 20dB attenuator plate in one of my cheaply acquired 8566B that some  previous owner had toasted with overload.

Because I didn't have money to buy new old one from ebay for 275$ (almost as expensive as the complete 8566B), I started to look in my trashed instruments. HP 8657B (in which battery has exploded) has dual 3 element attenuators (but with non compatible attenuation factors). I found good elements in 8672A attenuator (also bought cheaply, but YIG was dead, so parts unit for me). It has 1x10dB, 1x20dB and 2x 30 or 40dB elements. So I took 10dB and 20dB plate and 8566B was working like new.

The ones for HP-8657B have parts number 33321SC and 33321SD. If you take picture of your attenuator top section (attenuator elements), you can see what attenuator elements it uses. That's important, because 30/40dB elements are longer than 20dB element which is also longer than 10dB element.

If you took pictures of yours, I'll take pictures of my 33321SC and 33321SD so you can see whether they are compatible. Then you can buy cheap 8657 attenuator.

[George_Race]

Thanks for all the detail on taking it apart.  I have been to the point where the next step was to remove all the many screws from the top plate.  That is where I stopped and put it all back together.

I don't think the elements are defective, it is the operating mechanism that has the problem.  When the armature is pulled down, there are two small pins, with micro sized "O" rings that have broken up and fallen off the pins.  That prevents the armature from closing the two ends of the switch leaf that connect the element into the circuit path.

So it is not a problem with the elements, but the actuator that is causing my problem.  Both the -60029 and the SC are 70 db Attenuators and are three stage.  Both are also 10 db steps.

So I am still thinking  that the SC could be an exact replacement for the -60029 attenuator.

[MadTux]

So, I quickly took my 33321SC and 33321SD apart and took pictures for reference. Also found data sheet for 33321xx in German.

My idea was to take your original upper attenuator section and pair it with a cheap relay lower section from 33321SD/SC. Only works however, if the attenuator pads have the same length. Else the spring contacts are to long/short

[George_Race]

Now I understand your idea!  Change out the relay section with one that has the same spacing of the contact pins as the one that is causing the problem.

[idpromnut]

@George_Race: have a look at the repair videos I did of my HP8642B:

Troubleshooting the problem:


Fix + testing:

[George_Race]

Your second video really helped!  I now know exactly how to take it apart, and actually measure and or see what the problem is with a section that is not working.  A big thanks for all the effort you put into the videos, really helped a lot.
I have one of the less costly ones on the way.  Will try to take some close up pictures of the repair process and if it all comes out the way I think it should, will post them to this thread.  Hopefully with what you have shown, and what I may come up with, there will be a choice of solutions that others with attenuator problems can follow to bring their instruments back to life as well.

[idpromnut]

Just remember that you can just trans-plant the attenuator pad assembly from your current unit (as it is the mechanical bits that are broken) into the "new" unit, since the electro-mechanical parts are the same (most likely).  I don't know if I mentioned it in that video, but the replacement I bought did not have the same attenuator pads in it.

[George_Race]

Yup, I am hopeful that it will indeed be just that easy.
Thanks again for all the encouragement.

[idpromnut]

@George_Race: I just ordered some of those rubber gaskets to fix the attenuator unit that I have; I will see if I have extras and can send you a few.

[Kean]

Just sticking my nose in here as I have an 8657B with a problem where the RF output drops in and out every 5 dB - so this info and the linked videos look really helpful.
My first step will be to try verify if the issue is with the attenuator drivers or one of the attenuator themselves.

[idpromnut]

Just sticking my nose in here as I have an 8657B with a problem where the RF output drops in and out every 5 dB - so this info and the linked videos look really helpful.
My first step will be to try verify if the issue is with the attenuator drivers or one of the attenuator themselves.

I'd zoom to checking the attenuators themselves first; the problem with mine was not the driver, but that the mechanical bits that would take the pads in and out of the signal path were not activating. Turns out that if those o-rings deteriorate, the pad actuators can become stuck and exhibit the behaviour you are seeing.

[Kean]

Agreed.
Actually once I finished watching your two videos I bought a reasonably priced replacement pair of attenuators that are supposed to be good.  While it doesn't look too hard to do the repair, this piece of equipment has been on the to fix list for far too long.  Once I've swapped them out I can look at repairing and repurposing the old ones.  If the new ones exhibit issues, then at least I have spare parts to work with.
Thanks for the informative videos - subscribed.

[George_Race]

@George_Race: I just ordered some of those rubber gaskets to fix the attenuator unit that I have; I will see if I have extras and can send you a few.

That would really be great, that way I could hopefully have a spare attenuator.  I found a nice collection of the mini O rings from a source in Hong Kong for less than $10 U.S.  Was thinking about ordering a set, but would end up with many, many, more than I could ever begin to use.

Let me know when you need my mailing address.
Cheers,
George

[idpromnut]

Agreed.
Actually once I finished watching your two videos I bought a reasonably priced replacement pair of attenuators that are supposed to be good.  While it doesn't look too hard to do the repair, this piece of equipment has been on the to fix list for far too long.  Once I've swapped them out I can look at repairing and repurposing the old ones.  If the new ones exhibit issues, then at least I have spare parts to work with.
Thanks for the informative videos - subscribed.

 

[vtp]

I found a nice collection of the mini O rings from a source in Hong Kong for less than $10 U.S.  Was thinking about ordering a set, but would end up with many, many, more than I could ever begin to use.

The small O-rings rot away due to ozone. HP/Agilent apparently made a screw-up around turn of the century and supplied a lot of attenuator plungers with NBR rubber.

If you want the O-rings to last they need to be fluorinated rubber like FKM, Viton etc.

Let us know if you find those, I also would be interested in few hundred pieces.

[George_Race]

Not sure if these are fluorinated rubber, but probably would last for quite some time.

Quite a few for the price, from Hong Kong.  Here is the link:

http://www.ebay.com/itm/Watch-Waterproof-Gasket-Seal-Eraser-Washer-Boy-DIY-Repair-Tools-New-Useful-Kits/171986543416?_trksid=p2047675.c100005.m1851&_trkparms=aid%3D222007%26algo%3DSIC.MBE%26ao%3D1%26asc%3D20131003132420%26meid%3D2c2c354310784f5f8e2dbe5ad91be6cc%26pid%3D100005%26rk%3D3%26rkt%3D6%26sd%3D361215176600

Looks like a good assortment and many of the sizes would fit the purpose.  Even includes a pair of setting tools!

Looks like a good source for just what we need to do the job.  I just ordered a set, will let you all know how they look in about 2 weeks or so.

[George_Race]

I removed the earlier post that was the Technical Data Sheet on the 33320 series of HP Attenuators.

Here is a link to the direct Keysight file download.  It is very high quality, unlike the poor quality copy that I posted earlier.

http://literature.cdn.keysight.com/litweb/pdf/5959-7857.pdf?id=638180