Repairing HP-8640B Step Attenuator
My HP-8640B rotary step attenuator lost the detents at every ten dB and was just rotating freely. Having already repaired a different set of gears a few years ago I recognized the likely cause of this new problem. This post will describe how I repaired the attenuator shaft and cams.
The rotary attenuator is easily removed by following the service manual. I used TM 9-4935-601-14-7&P, page 8-70. The attenuator has a mechanical side and an RF side where the five attenuators are. I did not need to do any attenuator repair to fix the cracked cam on the mechanical side so I did not open the RF side. See note below.
I strongly advise taking digital pictures or making drawings as you remove and disassemble the step attenuator. Make note of the RF output level setting when you remove the attenuator. I would advise leaving the attenuator at +20 dBm (fully clockwise from the front). The reason is that +20 dBm, and +10 dBm, are two positions where the attenuator cams are NOT engaging any of the switches on the RF board.
When I opened the mechanical side I could easily see the long shaft with the clear plastic cams. The detent cam was cracked in half and one piece had fallen away from the shaft. The other piece was about to come off. I could see that other cam segments were cracked part way through so I used a small amount of epoxy and a small cable tie to extend its lifetime. Make sure that the cable ties are not going to hit the RF switch springs as the shaft rotates.
First repair. Being terrified of disassembling the entire attenuator my first repair attempt was to epoxy the detent cam back together while the shaft was still in the attenuator housing. I used a dental hemostat to hold the detent wheel out of the way of the cam. I epoxied the detent cam pieces together around the shaft and used two cable ties to hold it in place overnight.
After I reassembled the attenuator and reinstalled it I looked at the output of the 8640 on a spectrum analyzer. The RF output was skipping some power levels by 20 dB, or 0 dB. In between the normal detent sounds I could hear the little RF switches clicking slightly before or after the detents. Clearly I had not adequately lined up the detent cam with the attenuator cams.
Second repair. I removed the attenuator again. This time I took the shaft completely out of the housing. You???ll have to remove the switch assembly on back of the attenuator so CAREFULLY remove the ???C??? clip and all the associated hardware. (If the C clip goes flying off into space I think they are available at a good hardware store.) Also, be careful with the clear plastic plate and the switch fingers.
Once all the restraining hardware is out of the way you can carefully pull the shaft toward the ???front??? of the attenuator and remove it from the attenuator housing. Depending on the position of the cams, they MAY snag on the RF switch springs as you slowly move the shaft out of the assembly. I used another dental tool to carefully push down on the spring(s) to get the cam past it. Repeat as necessary to get the whole shaft out of the housing. (My dentist was kind enough to give me a few instruments that he could no longer use. Otherwise you could make curved tools out of #12 or #14 gauge solid wire.)
Alignment of cam ???teeth???: This is the secret to good cam operation! This second time around I noticed that the cams are lettered from rear to front ??? A, B, C, D, E, F. There are five attenuators, operated by five cams A, C, D, E and F. Cam B is what I???ve been calling the detent cam. It does not operate any attenuator. Apparently when I tried to epoxy the detent cam the first time I didn???t have it seated just right. Rotate the shaft to a certain position and you will see the letters, A-F, lined up. They are very hard to see on the clear plastic. The letter on Cam B did not line up, possibly because there is less room to emboss the letter B. With the shaft out of the housing I was in a much better position to remove the detent cam, clean off the previous epoxy and line it up better on the flat of the shaft. Apparently that???s the secret of good alignment.
Success: After reassembly the sound of the detents was back to how it should have sounded, more ???solid??? with no extra little ???clicks??? in between the 10 dB steps. The spectrum analyzer showed ten dB steps over all the ranges that I could see.
For those who may have to repair attenuator sections, the RF side, here is a little help. The attenuator has sixteen positions from -130 to +20 dBm. There are five attenuators switched in or out by the cams on the shaft to cover a 150 dB range. Here is the sequence I have pieced together from a few schematics and my visual observations. See Photo 4. The external switch on the rear of the attenuator accommodates the extra gain between +10 and +20 dBm. If you???re going to repair the individual attenuator sections I don???t think you???ll have to do the mechanical work I described above. Just open the other side of the attenuator.
Please see photos in album "Repairing HP-8640B Step Attenuator"
Photo 1: Mechanical side of attenuator
Photo 2: Alignment of cams.
Photo 3: Showing RF switch activating springs.
Photo 4: Table of Step Attenuator sequence
Jerry DeHaven ??? WA0ACF
