|
Hi all,
?
So I've removed the attenuator and stripped it down to a point where I can see the important bits and I cannot, from careful inspection alone, see anything amiss with it (photo attached). The pads all look nice and there are 4 of them! I'm guessing one 10dB and the rest 20dB ones. The longer ones measure about 82 ohms each across their terminals with minor variations between them and the shorter measures 52 ohms approx. The springs underneath look like brand new and the plungers do flip from on to off with just a bit of pressure from a screwdriver but I haven't tried to actuate them under power (24V) yet. I thought I'd best pause and get some suggestions before I do something which could cause damage.
Some of you may recall from a previous thread that when the attenuator was in circuit and I attempted the calibration procedure, I saw no signal visible with the 0dB and 10dB settings. The 20, 30, 40, 50 and 60dB settings seemed okayish (some a bit noisy) but 70dB was *very* noisy.? Any suggestions as to how to proceed?
?
Thanks,
?
J.
|
Hello Jinxie,
It seems you have indeed one 10dB, one 20 dB and two 20dB combined to make 40.
From what you say I'd think the 10dB is the culprit. 60 ok, 70 not?
Remember for the test what someone above suggested: Why not try it with low voltage CC? Very easy to measure the attenuation and to see on a scope if it's noisy!
Not much, just my 2�cents!
Best,
Renaud
|
That's an old one! The newer versions have a PCB with one-shots on board. These attenuators are pulsed on/off so I'd be warry of using a bench supply to activate the solenoids; you might burn them up and then you're toast!
The better option is to plug it in to the unit and activate that way. You should be able to verify that the spring fingers are moving up and down.
With the pad assemble in place (just a few screws to hold it) you can do a few things. Place an ohmeter between the two ports and verify you have resistance (pretty low) for all combinations. Even better is to used an RF generator and a spectrum analyzer to
measure it. (a scope would also work)
You can try cleaning the springs and pads with IPA but be very careful!
Good luck.
Mark
toggle quoted message
Show quoted text
Hi all,
?
So I've removed the attenuator and stripped it down to a point where I can see the important bits and I cannot, from careful inspection alone, see anything amiss with it (photo attached). The pads all look nice and there are 4 of them! I'm guessing one
10dB and the rest 20dB ones. The longer ones measure about 82 ohms each across their terminals with minor variations between them and the shorter measures 52 ohms approx. The springs underneath look like brand new and the plungers do flip from on to off with
just a bit of pressure from a screwdriver but I haven't tried to actuate them under power (24V) yet. I thought I'd best pause and get some suggestions before I do something which could cause damage.
Some of you may recall from a previous thread that when the attenuator was in circuit and I attempted the calibration procedure, I saw no signal visible with the 0dB and 10dB settings. The 20, 30, 40, 50 and 60dB settings seemed okayish (some a bit noisy)
but 70dB was *very* noisy.? Any suggestions as to how to proceed?
?
Thanks,
?
J.
|
Hard to be sure from the picture but how do the O-rings look, the
lower one on the far left actuator rod (the 'ribbon cable' end)
looks a bit odd?
Sometimes the O-rings go gooey and sometimes they harden, either way
it can affect the 'throw' of the actuator solenoid - especially the
lower one - which might lead to the symptoms you describe.
On 14/09/2024 17:03, Jinxie via
groups.io wrote:
toggle quoted message
Show quoted text
Hi all,
?
So I've removed the attenuator and stripped it down to a
point where I can see the important bits and I cannot, from
careful inspection alone, see anything amiss with it (photo
attached). The pads all look nice and there are 4 of them! I'm
guessing one 10dB and the rest 20dB ones. The longer ones
measure about 82 ohms each across their terminals with minor
variations between them and the shorter measures 52 ohms approx.
The springs underneath look like brand new and the plungers do
flip from on to off with just a bit of pressure from a
screwdriver but I haven't tried to actuate them under power
(24V) yet. I thought I'd best pause and get some suggestions
before I do something which could cause damage.
Some of you may recall from a previous thread that when the
attenuator was in circuit and I attempted the calibration
procedure, I saw no signal visible with the 0dB and 10dB
settings. The 20, 30, 40, 50 and 60dB settings seemed okayish
(some a bit noisy) but 70dB was *very* noisy.? Any suggestions
as to how to proceed?
?
Thanks,
?
J.
Attachments:
|
There seem to my eye to be ridges on some of the gold plating on those pads.? This can cause issues.? Gentle use of a paper eraser can help here (but the gold is thin, don't over do it). I've also found deoxit gold can be helpful in cleaning the contacts.
|
Well done, gentlemen, loads of good suggestions here; plenty for me to get on with tomorrow when I next get a session in.
So here's where I am now. From the advice here it seems the best course is to replace the front panel of the attenuator so the contacts can bear against the pads. Then re-attach the ribbon cable to the motherboard. Both of these steps I have now taken. Tomorrow I'll squirt some RF and some DC (at different times!) through the input and monitor the output in the time and frequency domains. Seems that RF and DC show up different issues, so fine, I'll try both.
Now, the sharp-eyed (extremely sharp-eyed!) fellow who pointed out that O ring was perfectly correct. It is indeed heavily perished. One or two of the others are also showing signs of age, too, but that one is the worst by far.
Now I've had to remove A6A9 to get the attenuator out and detached some of its cables in the process. Will I be okay to turn the SA on with this section in this state? I would have thought so, but don't wish to fry something at this stage!
|
Yes, I did notice those marks on the gold plating. I'll take a better look tomorrow under the stereoscope to see if they're invasive. I have some 24kt gold wire which I could use to sputter those pads if they're beyond repair - assuming that would adhere, of course. 25 mil wire coil @ 1000 amps should do it surely?
|
Jinxie, don't touch those gold pads, just fix the o-rings. Gold is soft and it's supposed to smear a little by design.
?
Vladan
|
Vladan,
You're the boss! Okay, I'll not do that.
More tomorrow... I'm off to bed now.
|
You can operate the attenuator with a bench supply. Just make sure to operate the solenoids through the breaker contacts in place so the current will be interrupted. But also if you don't, if you break the current manually within a second or so, no harm will be done. Have done this many times.
?
Tom
|
Hi again....
?
1. I have the o-rings if you want some...
2. do NOT put DC if the attenuator is attached to the Spec An input. It would fry the mixer diode for the low band if the 0-2.5 GHz button is selected. Certainly it would take out the limiter also.
?
Dan in Chandler, AZ
|
Dan, you are quite right of course and it's good to point that out for anyone who may be reading this thread for advice in the future.
For the DC test I was only going to use 200mV and of course - IRO DC - there'd be no point examining that on a spectrum analyzer. I'll be using an oscilloscope for the DC testing.
Many thanks indeed for the offer of the o-rings. I have some nitrile rings which could work if they can be sliced through the 'equator' as it were. They're a bit too thick to use unmodified. That'll be fun with my eyesight and shaky hands!
I'm still reeling from Adrian being able to spot that faulty o-ring from the rotten photo I took! So much for my claimed "careful inspection"!
|
Incidentally, if anyone can explain why partially perished o-rings which are still in-situ can cause problems to the functioning of these attenuators I'd like to be disabused of my ignorance on the matter!
|
I believe it's because the 'push' and 'pull' forces from the
solenoid to the actuator rod are applied axially through the
O-rings, so if the ring is too squishy or thin (or hardly there at
all!) then you will not get the full range of movement on the rod
and thus not the correct movement and contact pressure on the switch
blades.
On 15/09/2024 09:40, Jinxie via
groups.io wrote:
toggle quoted message
Show quoted text
Incidentally, if anyone can explain why partially perished
o-rings which are still in-situ can cause problems to
the functioning of these attenuators I'd like to be disabused of
my ignorance on the matter!
|
Okay. Well I have an 'intermediate update.' Cycling through 0 to 70dB I can see all solenoids switching seemingly normally - apart from the second from left in the photo. Commencing with all attenuators in the 'up'/off position, It flicks 'on' okay but then remains stuck there and won't release again. And clearly its coil is remaining energized as it can't be manually flicked off, either. Not sure if this could account for the problem I've had or whether there are other things amiss (such as the perished o-ring) also.
So, anyway, once that first 20dB attenuator is switched in, it cannot be switched out again and there's a minimum of 20dB attenuation permanently in-circuit. Could there be an issue with the control logic?
|
Okay, are you sure the internal switching contacts are working as
intended?
It's been a few years since I was inside one but my memory is that
there are two per solenoid?
Effectively a N/C one when the solenoid is in the 'de-selected'
state and a N/O one that closes when it is selected.
They function in a break before make way where the power to the
'select' coil is routed through N/C contact and the power to the
de-select coil through the N/O contact.
So, if the actuator rod is gunked up with gooey o-ring and not?
moving freely over it's full travel I guess that might explain what
you are seeing - which coil is remaining energised?
On 15/09/2024 12:20, Jinxie via
groups.io wrote:
toggle quoted message
Show quoted text
Okay. Well I have an 'intermediate update.' Cycling through 0
to 70dB I can see all solenoids switching seemingly normally -
apart from the second from left in the photo. Commencing with
all attenuators in the 'up'/off position, It flicks 'on' okay
but then remains stuck there and won't release again. And
clearly its coil is remaining energized as it can't be manually
flicked off, either. Not sure if this could account for the
problem I've had or whether there are other things amiss (such
as the perished o-ring) also.
So, anyway, once that first 20dB attenuator is switched in,
it cannot be switched out again and there's a minimum of 20dB
attenuation permanently in-circuit. Could there be an issue with
the control logic?
|
Hi, just realised that the attached page 5 of the 33320 service
manual will probably explain it way better than me!
On 15/09/2024 12:20, Jinxie via
groups.io wrote:
toggle quoted message
Show quoted text
Okay. Well I have an 'intermediate update.' Cycling through 0
to 70dB I can see all solenoids switching seemingly normally -
apart from the second from left in the photo. Commencing with
all attenuators in the 'up'/off position, It flicks 'on' okay
but then remains stuck there and won't release again. And
clearly its coil is remaining energized as it can't be manually
flicked off, either. Not sure if this could account for the
problem I've had or whether there are other things amiss (such
as the perished o-ring) also.
So, anyway, once that first 20dB attenuator is switched in,
it cannot be switched out again and there's a minimum of 20dB
attenuation permanently in-circuit. Could there be an issue with
the control logic?
|
One of the most common causes I’ve seen of stuck attenuator sections is decomposing o-rings caused by attempts to clean the attenuator contacts with chemicals that attacked the o-rings. They become sticky like glue and will latch an attenuator section inline no matter what the solenoid does.? A broken or damaged push rod will have the same effect.?
toggle quoted message
Show quoted text
On Sep 15, 2024, at 3:40?AM, Jinxie via groups.io <paul666@...> wrote:
? Incidentally, if anyone can explain why partially perished o-rings which are still in-situ can cause problems to the functioning of these attenuators I'd like to be disabused of my ignorance on the matter!
|
Yeah, I get that, Steve, but this is not the issue with this pad's solenoid. It flicks back and forth freely when there's no power applied. But once it's energized to 'on' it's stuck there and can't be pried away again. That coil is remaining energized. Once that happens, there's no changing it. If I enter 0dB or 10dB, that 20dB one is always in circuit.? So for example, entering 10dB will result in me getting 30dB of attenuation and 0dB will get me 20dB. No wonder I couldn't calibrate it!
Adrian - it's the? second one from the left in the photo.
|
toggle quoted message
Show quoted text
On Sep 15, 2024, at 7:29?AM, Jinxie via groups.io <paul666@...> wrote:
? Yeah, I get that, Steve, but this is not the issue with this pad's solenoid. It flicks back and forth freely when there's no power applied. But once it's energized to 'on' it's stuck there and can't be pried away again. That coil is remaining energized.
Adrian - it's the? second one from the left in the photo.
|
Jinxie
