开云体育

  1. HP-Agilent-Keysight-Equipment
  2. Topics

8566B Attenuator Service

 

开云体育

I want to make sure we're on the same page here.
In the image below, the RED circles are the switch contacts CLOSED.
The YELLOW circles are the switch contacts OPEN.

If you get no continuity between the two spring contacts when closed (red) then yes, that's a problem.
If that's the case and cleaning is no help, you could try come light abrasive on them.

From: [email protected] <[email protected]> on behalf of Jinxie via groups.io <paul666@...>
Sent: Sunday, September 15, 2024 9:40 AM
To: [email protected] <[email protected]>
Subject: Re: [HP-Agilent-Keysight-equipment] 8566B Attenuator Service
?
Thank you, Greg. I would follow your suggestion, but I'm not about to give up all hope at the prospect of fixing a pair of leaf springs!!
I've cleaned those springs with three different solvents but they still won't make electrical contact. There is plenty of 'spring' in them to make a good contact with each other, but it's just not happening. I've never had this before. Ever. Ideally I'd like to remove this pair and examine them under the stereoscope, but that's not possible. The only other thing I can try (and I'm fast running out of ideas) is abrasive paper. I have the full range of grades here from coarse to ultra-fine. Does anyone know what these springs are made from? I don't wish to rub through some delicate plating and make the situation even worse!
 

If the coil is staying energised than the control contact is falty or not being actuated. When the armature move it shoul cause the circuit to the coil should be broken. The coils are not designed to be be permanently energised and can be burnt out if left energised.
?
Robert.
 

In fact, the tiny pin that pushes the two leaves apart is coated in glass and the pin is then an insulator when placed between the leaves...
?
Dan in Chandler, AZ
 

I wish I knew! Unfortunately I cannot see from any available angle if the leaves are parallel with one another. But even if they're not, there should still be some degree of contact sufficient to provide a short. Anyway, they look fine visually. Thanks for the tip about using paper. I hadn't thought of that. I'll give it a try with some IPA soaked in.
The weird thing is that this fault is so reliable and repeatable. It's not an intermediate contact as one might expect. It's just there 100% the time by default it seems.
 

@Robert:
Well, I would have thought so too. However, it seems quite happy to remain energized indefinitely (fortunately!) Perhaps this (somewhat less common) attenuator is designed to operate that way? After all, it would provide a far more definite on/off than relying on some sort of spring mechanism.
 

@Mark
Yes, we are on the same page! I'm going to try Adrian's suggestion of alcohol-soaked paper and if that doesn't work, I'll consider using very fine abrasive paper - say 2500 grade to begin with.
 

开云体育

Thumbs up! And Robert is correct about burning these out. They are designed to be pulsed to one direction or the other.
And if it's always energized, then something is clearly wrong!

From: [email protected] <[email protected]> on behalf of Jinxie via groups.io <paul666@...>
Sent: Sunday, September 15, 2024 10:45 AM
To: [email protected] <[email protected]>
Subject: Re: [HP-Agilent-Keysight-equipment] 8566B Attenuator Service
?
@Mark
Yes, we are on the same page! I'm going to try Adrian's suggestion of alcohol-soaked paper and if that doesn't work, I'll consider using very fine abrasive paper - say 2500 grade to begin with.
 

开云体育

I thought you had a problem with a coil staying energised as though its switch was not opening, as opposed to one not being able to switch back as the switch that needs to be closed to allow the current though it is not closing properly?? Maybe I’m getting confused!

?

I have been following this thread with much intertest and I am impressed by your determination and the amazing help here!

?

73, Mike,

Mike Stevens,

G8CUL/F4VRB.

?

From: [email protected] <[email protected]> On Behalf Of Jinxie via groups.io
Sent: 15 September 2024 18:46
To: [email protected]
Subject: Re: [HP-Agilent-Keysight-equipment] 8566B Attenuator Service

?

@Mark

Yes, we are on the same page! I'm going to try Adrian's suggestion of alcohol-soaked paper and if that doesn't work, I'll consider using very fine abrasive paper - say 2500 grade to begin with.
 

For anyone that's not read the thread, a little bit of summing up:
This attenuator has four pads: 10dB and 3 x 20dB. One of the 20db ones is sticking on. Once actuated it remains energized and can't be taken out of circuit again by programming in a lower attenuation value. It has been operated for extended periods and hasn't shown any signs of burning out (or even getting warm!)
I have detected a problem with the associated solenoid's lower pair of switch springs. Though the leaves of these springs are closed, there is no electrical connection between them for some reason.? Cleaning the contacts has thus far proved fruitless, but I haven't yet tried any abrasive method. If anyone knows whether this issue with the switch could cause the solenoid to be permanently energized, that would be useful.
 

Could have there been silicon based lubricant used at some point?
?
I have heard that it's use in switches is a no-no as the arcing can cause it to form insulating and extremely hard silicon carbide deposits.
?
I realise there shouldn't be arcing here, but maybe something of a somewhat similar nature has coated the contacts?
 

开云体育

Hi (again!) looking at the schematic Fig 5 on P5 of the 33320 manual, I see there is a TTL output driven through an inverter off output B of the driver.

Now come the 'ifs':

If that TTL signal is feedback to the driver control to tell it that the solenoid is in position B, and if both contacts are are effectively open? (one because it should be and one because it's faulty) then there is no connection through the coil(s) to the -24VDC line and the feedback to the driver (and I'm not 100% on this!)?may be wrong?.

If that is the case, a quick test is to operate the attenuator into the fault condition, i.e. select first a setting where the section should be selected then one where it should not, and them short across the faulty contacts and see if the solenoid operates?
...I think....

Adrian

On 15/09/2024 20:05, Jinxie via groups.io wrote:

For anyone that's not read the thread, a little bit of summing up:
This attenuator has four pads: 10dB and 3 x 20dB. One of the 20db ones is sticking on. Once actuated it remains energized and can't be taken out of circuit again by programming in a lower attenuation value. It has been operated for extended periods and hasn't shown any signs of burning out (or even getting warm!)
I have detected a problem with the associated solenoid's lower pair of switch springs. Though the leaves of these springs are closed, there is no electrical connection between them for some reason.? Cleaning the contacts has thus far proved fruitless, but I haven't yet tried any abrasive method. If anyone knows whether this issue with the switch could cause the solenoid to be permanently energized, that would be useful.
 

On Sun, Sep 15, 2024 at 09:05 PM, Jinxie wrote:
For anyone that's not read the thread, a little bit of summing up:
This attenuator has four pads: 10dB and 3 x 20dB. One of the 20db ones is sticking on. Once actuated it remains energized and can't be taken out of circuit again by programming in a lower attenuation value. It has been operated for extended periods and hasn't shown any signs of burning out (or even getting warm!)
I have detected a problem with the associated solenoid's lower pair of switch springs. Though the leaves of these springs are closed, there is no electrical connection between them for some reason.? Cleaning the contacts has thus far proved fruitless, but I haven't yet tried any abrasive method. If anyone knows whether this issue with the switch could cause the solenoid to be permanently energized, that would be useful.
?
?
Hi Jinxie,
Solenoid coils aren't actuated in these attenuators but for a very short time during switch-over. Normally, even a (wrongly) permanently electrically "actuated" solenoid won't draw power and therefore stay cool because it is normally disconnected.
It's not clear what you mean by "energized". Do you main power remains on (malfunction) or is it permanently held in one position (expected behavior)?
?
I've only recently started looking at this thread and I'm not sure if the operation of these electromagnetic attenuator switches is clear to all, especially when looking at the use of the words "actuated" and "energized".
Each attenuator element (10dB, 20dB) is controlled by its own solenoid unit and each solenoid unit is *bistable*, meaning it stays in its position *without power being applied*.
A reversing permanent magnet (system) is employed for this functionality. (actually, this isn't completely correct but it is simple and will do for this explanation. In truth, a small magnetic force holds the solenoid in the position that it has been thrown into by a temporary solenoid current).
Each solenoid unit contains two coils and each coil polarizes the permanent magnet system in its own (opposite) direction.
In series with each coil, a switch either allows current through or interrupts it. Both switches are moved (on/off) by the same solenoid action but one provides continuity in the position where the other interrupts it, in a break-before-make fashion.
Each coil is actuated by applying a voltage across its series-connected coil-switch combo. The arrangement is such that each coil is interrupted by its switch, once its actuation is finished. In fact, each coil/switch set may remain "energized" forever, since after the current has moved the actuator, the switch is opened and no current flows.
Effectively, each attenuator "remembers" its position after power-down.
Perhaps worth noting, apart from the prominent O-ring failure mechanism:
- I have encountered attenuator modules with loose or even broken (ferrite) magnets, causing weird behavior.
- Applying an actuating voltage across both coil/switch elements of one solenoid results in continuous switching back and forth of the solenoid.
?
Raymond
 

?Applying an actuating voltage across both coil/switch elements of one solenoid results in continuous switching back and forth of the solenoid.
?
I have never tried that! How fast does it move?
?
Does anyone have the schematic? (8566B) What driver does it use?
?
 

> I have never tried that! How fast does it move?
?
I might have added “at the same time”, of course.?
I guess 3 - 5 times per second, from memory.
?
Raymond
 

Adrian,
Thanks for your comments. I did wonder if there might be some feedback involved in this. Not sure if I'll have time tonight to try your idea out, but should find some tomorrow to fit it in.
 

开云体育

I doubt there is any feedback. No circuitry in the attenuator for that.

From: [email protected] <[email protected]> on behalf of Jinxie via groups.io <paul666@...>
Sent: Sunday, September 15, 2024 2:02 PM
To: [email protected] <[email protected]>
Subject: Re: [HP-Agilent-Keysight-equipment] 8566B Attenuator Service
?
Adrian,
Thanks for your comments. I did wonder if there might be some feedback involved in this. Not sure if I'll have time tonight to try your idea out, but should find some tomorrow to fit it in.
 

Raymond,
That's a very useful explanation which I'm sure I'll fully understand tomorrow when I can get a fresh look at things. I will also double check my findings and ascertain whether current is still live in that solenoid (but I'm sure it will be).
If anyone else has any suggestion to add, now would be a good time to leave a comment and I can follow up with any findings tomorrow.
 

When I was a child we used to make 'electric shock machines' using this make & break principle. We'd wire up old Post Office relays to behave in this way and you could get several thousand volts from the back EMF. I'm pretty sure the cycle rate was a lot faster than 3 per second, though.
 

开云体育

One last thing: I have seen enough of the schematic to see that the common pin is +22VDC. (it's usually 24 or 5 but OK)
Opposite polarity of that shown in the datasheet. (figures 4 & 5) It think that's pin 6. (Table 4)

Mark

From: [email protected] <[email protected]> on behalf of Mark Bielman <mbielman@...>
Sent: Sunday, September 15, 2024 2:24 PM
To: [email protected] <[email protected]>
Subject: Re: [HP-Agilent-Keysight-equipment] 8566B Attenuator Service
?
I doubt there is any feedback. No circuitry in the attenuator for that.
From: [email protected] <[email protected]> on behalf of Jinxie via groups.io <paul666@...>
Sent: Sunday, September 15, 2024 2:02 PM
To: [email protected] <[email protected]>
Subject: Re: [HP-Agilent-Keysight-equipment] 8566B Attenuator Service
?
Adrian,
Thanks for your comments. I did wonder if there might be some feedback involved in this. Not sure if I'll have time tonight to try your idea out, but should find some tomorrow to fit it in.
 

开云体育

I would use a DVM (DC volts) and check if those coils have power. (referenced to common/brown wire)
They should not!

From: [email protected] <[email protected]> on behalf of Jinxie via groups.io <paul666@...>
Sent: Sunday, September 15, 2024 2:35 PM
To: [email protected] <[email protected]>
Subject: Re: [HP-Agilent-Keysight-equipment] 8566B Attenuator Service
?
Raymond,
That's a very useful explanation which I'm sure I'll fully understand tomorrow when I can get a fresh look at things. I will also double check my findings and ascertain whether current is still live in that solenoid (but I'm sure it will be).
If anyone else has any suggestion to add, now would be a good time to leave a comment and I can follow up with any findings tomorrow.

41 - 60 of 107