Author Topic: ESI KVD wafer repair  (Read 2870 times)

0 Members and 1 Guest are viewing this topic.

Offline BradCTopic starter

  • Super Contributor
  • ***
  • Posts: 2104
  • Country: au
ESI KVD wafer repair
« on: September 08, 2018, 06:21:31 am »
G'day all,

I've picked up an old ESI RV722. Electronically it exceeds my ability to test it (as in my limited testing indicates it does what it was designed to do and is more accurate than my measuring equipment). Mechanically it appears to have had a knock.

All switch wafers, shafts and associated mechanical elements are fine. One indicator disc was misaligned and rubbing on an adjacent disk, but a gentle poke with a couple of wooden sticks has seen that right.

The non-switch wafers that sit at the back of the unit holding the resistor ends haven't been so lucky. The wafer on decade 1 has been broken into 3 bits, and the wafer on decade 2 is cracked.

I don't want to go barging in getting finger grease over everything, so right now it sits uncovered until I can figure out the best way to tackle it.
What I had intended to do was use super-strength (24hr) Araldite. The 2nd decade wafer just needs the cracks filled to stop it getting worse. The first decade I had intended to coat the broken edges with Araldite and draw the wafer back together with waxed lacing twine which I can remove later. I'm at a bit of a loss as to what I can use to give it some better mechanical reinforcement to prevent it popping apart in the future without providing a potential leakage path.

Any ideas, or a better approach? I can't get the back panel off without some desoldering/resoldering of the connectors used for calibration, and if I can avoid using the soldering iron in there I think I'd rather.







 

Offline meandeev

  • Regular Contributor
  • *
  • Posts: 66
  • Country: de
Re: ESI KVD wafer repair
« Reply #1 on: September 09, 2018, 08:40:40 pm »
Hi,
this must have been a hard hit to the device :-(

I also bought a RV722 "as is" several days before, but had a litte bit more luck with the mechanical parts....
(see https://www.eevblog.com/forum/metrology/kvd-why-1-1-value-on-first-decade/)
If I were you, I would remove the "frame" with the 4 screws at the front (unfortunately you have to
desolder all wires from the calibration sockets, as you wrote...).
But then you have plenty space for your repair without touching the precious parts.
On the other hand you have to be more careful without this "safety frame".

Araldite (or 2k epoxy) only in the line of breakage would also my preferred glue, but why a waxed twine?
I would use a normal cotton twine to prevent contamination.

For the second decade: I cannot imagine that the ceramic is not broken into at least 2 parts,
a break only on one side is very uncommon. It is not visible at the pictures. And if you have two pieces,
you can bend it away to fill the lines of breakage.

I would not be afraid of popping apart in the future, because of the Araldite. Normally it will not break
at the same position, especially in that case of porose ceramic.
If it will be sit on your bench, it will outlive you, but if you plan to play football with it, I dont´know. :-)
As I said, it must have been a hard hit. 

I have also questions: can you reach the 1.0 and 1.1 values at the first decade of your device?
Is there a c-clip at the shaft under the wheel like at the other decades (shown in my first picture)?
Or can you push the axis and reach the 1.0 and 1.1 values?
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2380
  • Country: de
  • Sounds like noise
Re: ESI KVD wafer repair
« Reply #2 on: September 09, 2018, 09:06:29 pm »
I don't know Araldite, but my suggestion would be "UHU Endfest" or if you want something close to the properties of ceramic "Roth Thermokitt".

-branadic-
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline meandeev

  • Regular Contributor
  • *
  • Posts: 66
  • Country: de
Re: ESI KVD wafer repair
« Reply #3 on: September 09, 2018, 09:21:57 pm »
Hi branadic,
"Araldite" and "UHU endfest 300"  (I´m more familiar with this, too)  are brands, substance is the same, thats why I answered "2k epoxy" :-)
 

Offline BradCTopic starter

  • Super Contributor
  • ***
  • Posts: 2104
  • Country: au
Re: ESI KVD wafer repair
« Reply #4 on: September 09, 2018, 11:26:10 pm »
Thanks guys. Yes the more I look at it, the more it looks like desoldering those sockets is the best way to tackle it. I'll have a look at the best way to tackle that. I suggested the Araldite because that's my go to and I keep it handy. It sounds like that might be the way to go.

No, I can't reach the 1.0 or 1.1 settings on mine either, and looking at the schematic you posted, they are not set up the way Fluke does it. In fact, I can't really see how they'd work properly being wired that way.

Thanks for the feedback guys.

(edit) Looking at the specs, the UHU Endfest 300 has twice the ultimate strength of the Araldite (150kg/cm3 vs 300kg/cm3). I wonder if it really matters given I plan to treat this device like a piece of precision test equipment and not a football?

I can get it in Aus, although it's 4 times the price of the Araldite.
« Last Edit: September 10, 2018, 02:16:05 am by BradC »
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2380
  • Country: de
  • Sounds like noise
Re: ESI KVD wafer repair
« Reply #5 on: September 10, 2018, 06:01:28 am »
You can bake UHU Endfest to increase final strength. I'm not sure if this is true for Araldite, too?

-branadic-
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline BradCTopic starter

  • Super Contributor
  • ***
  • Posts: 2104
  • Country: au
Re: ESI KVD wafer repair
« Reply #6 on: September 10, 2018, 09:02:18 am »
You can bake UHU Endfest to increase final strength. I'm not sure if this is true for Araldite, too?

-branadic-

Yes it is. Not quite sure how much heat I'm willing to put into those parts though. I usually let it set up where it no longer sags and then poke a 50W halogen over it to finish the cure. On smaller and less fragile/precise devices I pop them in the oven for a post-cure, but I think I'd have trouble getting this one past the wife. Plus I'm pretty sure our oven isn't big enough.

I'm more interested in a long term outcome for this one, so I think I'll glue it up and leave it for the full 72 hour cure under light to moderate heat before touching it.

I guess I'm more concerned about desoldering parts than the actual mechanical repair. Principle of "it's actually working, lets try and keep it that way".
 

Offline BradCTopic starter

  • Super Contributor
  • ***
  • Posts: 2104
  • Country: au
Re: ESI KVD wafer repair
« Reply #7 on: September 11, 2018, 03:12:28 am »
Oh dear.




Better go and buy some latex gloves before I get my fingers in there.
 

Offline BradCTopic starter

  • Super Contributor
  • ***
  • Posts: 2104
  • Country: au
Re: ESI KVD wafer repair
« Reply #8 on: September 11, 2018, 08:13:41 am »
Cable ties are just tight enough to hold the wafers in alignment / together and not tight enough to bend anything. I may have a problem with leakage to the chassis if it turns out this epoxy is vaguely conductive as both decades have a break where the epoxy is touching both the mounting bolt and the adjacent lug. At least with the decades isolated from the case I can measure any potential leakage once it's all set up.




I have a 50W halogen above the area warming it up, and I'll leave it there for the 8 hour cure then put the cover back on and let it sit for the remainder of the 3 day cure period.

Fingers crossed it holds together and doesn't leak (sounds like a boat).

I was originally reticent to get the soldering iron in there for fear of contamination and flux issues, but there is flux on every joint anyway, so I really needn't have worried. I certainly wouldn't have been able to get those wafers back together with the chassis in place.
 

Offline BradCTopic starter

  • Super Contributor
  • ***
  • Posts: 2104
  • Country: au
Re: ESI KVD wafer repair
« Reply #9 on: September 15, 2018, 06:20:13 am »
Right. Let it set, then used the rework gun to heat each joint to 120C for an hour each for a final post-cure heat up.

I wanted to try and measure leakage to chassis but I don't have a megger.
What I do have is a Tek PG506 which can source 100V dc. So I put a 1M resistor across the input of a HP3478a as a shunt, hooked that in series with the PG506 and put one cable on the chassis and the other on one of the KVD binding posts. I measured 8mV across the shunt which varied over a day or so with temperature and humidity. I calculate that at about 8nA leakage. Not good.

I grabbed the gloves, a pack of new cotton buds and some isopropyl alcohol and went over each wafer and associated connections. That solved it. The measured voltage across the shunt is now just noise. So I'm happy now any leakage is below my ability to measure it. The induced noise on the meter is actually the opposite polarity to that which would be caused by leakage in the KVD and varies furiously as I move around the bench, so I'm pretty sure it's just noise.

Result! No pics because there's nothing notable to show and it's all back together.

I did spend an extra hour with a bright light and tweezers removing the remaining cotton fibres from the cleaned area. Next time I'll get chamois sticks or something. The cotton buds catch on everything, particularly the fibreglass washers and lugs.
 

Offline branadic

  • Super Contributor
  • ***
  • Posts: 2380
  • Country: de
  • Sounds like noise
Re: ESI KVD wafer repair
« Reply #10 on: September 15, 2018, 08:32:51 am »
Fingers crossed the glued joint will hold together when switching around. By now, well done!  :-+

-branadic-
Computers exist to solve problems that we wouldn't have without them. AI exists to answer questions, we wouldn't ask without it.
 

Offline BradCTopic starter

  • Super Contributor
  • ***
  • Posts: 2104
  • Country: au
Re: ESI KVD wafer repair
« Reply #11 on: September 15, 2018, 08:48:43 am »
Fingers crossed the glued joint will hold together when switching around. By now, well done!  :-+

-branadic-

There is no dynamic load on it. This is the wafer that has no shaft or moving contacts. It's entire purpose is to support the resistors and associated wiring, so in theory it *should* be ok. If not, I'll deal with it when it happens.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf