Keyboard Shortcuts
Likes
Search
8566B attenuator fix
|
Hi Folks, Looks like the step attenuator (5086-7814) in my 8566B is busted. I took it out, and measured it on the bench. The 10dB stage I can switch on and off correctly, although in one of the states it keeps drawing current. The attenuation latches fine, but it doesn't open the circuit to the coil. I read here about the plunger o-ring issues, so I was planning to take it apart. In the 40dB and 50dB states that I can select the response is flat and matching is good from both sides. Thanks! |
|
Always worth taking it apart and attempting the repair on your own. Many people (myself included) have had good luck repairing these even outside of o-ring issues (for instance I¡¯ve found a little oxidation on the switch contacts that prevent the plunger from moving through the solenoid). When you say that on one pad the attenuator can switch but the coil still draws current, are you sure that¡¯s right? That seems strange to me. Feel free to post follow up questions when you pull it apart. But the operation is relatively straight-forward and they¡¯re easy to operate on the bench with a current-limited power supply. Anyway, it¡¯s worth figuring out what¡¯s wrong before you condemn it as irreparable or send it off to someone else to repair who would probably use the same troubleshooting and repair techniques that you would. Matt On Thu, Mar 9, 2023 at 7:52 AM <bagojfalvibagoj@...> wrote:
|
|
You need to get the cover off the attenuator and inspect the coils. I strongly suspect some damage. I would NOT operate the attenuator for any extended period as this is likely to damage it.
toggle quoted message
Show quoted text
After removing the cover, manually work the plungers to free them, then see if they will switch. If they work after "freeing" them, it is reasonable to repair the attenuator. If the coils are burnt, repair is much more complicated. Don Bitters is the go to resource on these attenuators (IMO). Cheers! Bruce Quoting bagojfalvibagoj@...: Hi Folks, |
|
I know, I can always try, it's more of a question of does it worth my time, or am I just going to get angry? :D To be honest I didn't double check the coil still drawing current, but it seemed plausible to me. I was working from the 3332x datasheet and I only found it here since that the 5086-7814 replaces the mechanical disconnect with electronics. My hypothesis was that if the plunger or something is blocked that could cause other kind of mechanical failure, for example blocking the coil disconnect only on one arm. |
|
You absolutely should try - it will be educational (at least) and potentially frustrating. Welcome to ttest equipment maintenance :-)
toggle quoted message
Show quoted text
Cheers! Bruce Quoting bagojfalvibagoj@...: I know, I can always try, it's more of a question of does it worth my time, or am I just going to get angry? :D |
|
Check "thesignalpath" on youtube.? Shariar did a complete teardown and repair of an HP step attenuator a few years ago.? It should give you all the guidance you need.? I did a quick search through his videos but didn't find it. I know it's there, but it might be embedded in a larger repair video of an HP sig gen or spectrum analyzer.? I'll keep looking but I know it's somewhere in his collection.
BTW, Shariar produces excellent "teardown and repair" videos plus some really good tutorials. BRgds....Frank |
|
This video does not address the hardest (IMO) part of the dissesembyl - detaching the coil assembly from the attenuator.
toggle quoted message
Show quoted text
In one of his other videos, he does the disassembly IMPROPERLY, by simply pulling the actuating pins put ot the switch contacts. It you watch his other video DO NOT try to remove the coils B4 disengaging the pins from the contacts. Cheers! Bruce Quoting "tom_iphi via groups.io" <iphi@...>: This should be the video: |
|
Anyway, the plunger thing is going to be a tomorrow problem. I noticed two things upon disassembly. This thing looks brand new inside. I have never ever in my life saw such healthy looking o-rings. Second, there's a lot more electronics than I anticipated. I took the magnet off one of the solenoids and tried to make it toggle. It seems to be very very weak and sometime it doesn't even move. In such a case the current draw stays high which I don't really understand. The shaft seems to be moving reasonably freely. Here's a video I made. I'm turning the psu on and off like 3 or 4 times. You can see it slightly move at 2 sec, until for the last cycle it actuates properly. Ideas? |
|
Need to see more pictures. Generally, you can manipulate the plunger w/o removing the magnet - but I have not worked on many of the attenuators with the PCB inside - most the simpler variety.
toggle quoted message
Show quoted text
You can recover the leaves by GENTLY pulling them up using tweezers (to lift, not pinch) or (as I do) a dental pick of the correct shape to slide under and lift. I do not see the switch (fingers that are interrupted by the plunger) in this picture - could you show more ? Cheers! Bruce Quoting bagojfalvibagoj@...: Anyway, the plunger thing is going to be a tomorrow problem. |
|
One thing I forgot to mention is that the magnets are necessary for proper operation of the plungers - I would not try to operate the device with them removed.
toggle quoted message
Show quoted text
Again, I will defer to Don Bitters on this as he has much more experience. Cheers! Quoting bagojfalvibagoj@...: Anyway, the plunger thing is going to be a tomorrow problem. |
|
I already managed to free the leaves. I think I was lucky. They seem ok to me. I would guess that the magnet is necessary to make the system bi-stable. The shaft sticks to it on one side and stays there until the coil shoots it over to the other side to stick there. The solenoid still should work as a solenoid I believe, but it doesn't look like doing a very good job at that. |
|
I think that is correct.
toggle quoted message
Show quoted text
I don't have the problem you do - I bought one of the HP relay drivers and built a small board with a DIP socket, that lets me plug in the attenuator and drive it using the switchwes. The drivers are cheap ($25) but the cables with matching plugs can be a bit harder to find. Cheers! Bruce Quoting bagojfalvibagoj@...: I'm looking at the board now and other than the ULN2003A drivers, there are 4 D flipflops which makes me wonder if this needs continuous supply in order to retain state correctly. |
|
I ordered a breadboard overnight. Never have one when you need it.
I have a frightening suspicion, which is that this attenuator works perfectly I just couldn't drive it properly. I hope at least I didn't make it worse. Now I kinda understand what the circuit is doing. Hopefully tomorrow I can check properly. The original symptom was about 50dB loss on the input in both frequency modes, so that makes it either the connector, cables, the attenuator or the switch. I took the attenuator out first as I read that is likely to have issues and since I couldn't turn on 40dB attn on it I concluded this is the culprit. But maybe not... Rabbit holes rabbit holes everywhere. |
|
Well good luck - keep us informed. For more information on proper drive, you could look at the schematic of the DP attenuator driver - or masure what the 8566 does - I believe it is a pulse from the CPU but not certain.
toggle quoted message
Show quoted text
Cheers! Bruce Cheers! Bruce Quoting bagojfalvibagoj@...: I ordered a breadboard overnight. Never have one when you need it. |
|
If you guys want to watch me unscrew screws for 30 or so minutes feel free:
I also added a scribble about what I believe the schematic of the control looks like. ? |
|
WELL DONE !
toggle quoted message
Show quoted text
It would have been nice to get better closeup of recovering the contacts after pulling them through, but overall excellent ! Did you note any deformation of the leaves due to pulling the pins through? Did you have much trouble re-engaging the pins? Thanks! Cheers! Bruce Quoting bagojfalvibagoj@...: If you guys want to watch me unscrew screws for 30 or so minutes feel free: |