Author Topic: Investigation on ... (Read 12707 times)

branadic · « **on:** December 02, 2016, 08:05:33 pm »

This thread is about the investigation on ...

Well, no words about it yet. It's up to your guess. Here are the first pictures of some parts a friend of mine currently fabricates for this project. The next pictures will follow soon and after a while you will get an idea of what this will end up in.

tautech · « **Reply #1 on:** December 02, 2016, 08:09:46 pm »

Waveguides ?

2N3055 · « **Reply #2 on:** December 02, 2016, 08:48:59 pm »

Either resonant cavity or oven...

Andreas · « **Reply #3 on:** December 02, 2016, 09:03:43 pm »

Hello,

looks more like one of those zero ohms resistors or a 4 wire shunt.

with best regards

Andreas

Gyro · « **Reply #4 on:** December 02, 2016, 09:04:45 pm »

Hmmm,

It's in the metrology section, high thermal mass enclosure, structurally stable..... Resistance standards? No fixing holes on the top surface though and they look finished. Something potted? Strange if so. $:-\$

Theboel · « **Reply #5 on:** December 02, 2016, 11:51:23 pm »

I bet its a resistance standard

VintageNut · « **Reply #6 on:** December 03, 2016, 02:59:08 pm »

Quote from: Theboel on December 02, 2016, 11:51:23 pm

I bet its a resistance standard

+1 Metrology individual resistors look similar to this.

TiN · « **Reply #7 on:** December 03, 2016, 05:03:27 pm »

Resistor tempco test boxes?

I'm gonna build some too, as my current TEC jig used for LTZ reference module test is just too painfully slow for TCR test.

Gyro · « **Reply #8 on:** December 03, 2016, 06:19:18 pm »

Come on branadic, give us a break - is there some kind of lid???

ManateeMafia · « **Reply #9 on:** December 03, 2016, 06:48:57 pm »

My guess is that they will laser weld the lids and install kovar leads or some hermetic connector out of the sides.

branadic · « **Reply #10 on:** December 03, 2016, 07:43:04 pm »

There is a lid on top and some feedthrough capacitors at the sides.

mimmus78 · « **Reply #11 on:** December 03, 2016, 07:48:04 pm »

Too much pieces for a simple investigation ...

Inviato dal mio Nexus 6P utilizzando Tapatalk

tatus1969 · « **Reply #12 on:** December 03, 2016, 08:19:30 pm »

~~RF directional coupler for a VNA...~~

kxenos · « **Reply #13 on:** December 03, 2016, 08:55:34 pm »

I would guess voltage reference using low noise chinese zeners or oil filled resistors

acbern · « **Reply #14 on:** December 03, 2016, 08:58:21 pm »

hermetic resistance?

manganin · « **Reply #15 on:** December 03, 2016, 09:10:27 pm »

Quote from: acbern on December 03, 2016, 08:58:21 pm

hermetic resistance?

If you look closely there seems to be a naked bulk metal foil resistor inside the box.

2N3055 · « **Reply #16 on:** December 03, 2016, 09:18:41 pm »

4w for resistor, and 2 for thermistor or RTD....
hm...

manganin · « **Reply #17 on:** December 03, 2016, 09:34:38 pm »

Worth noting that the 4-wire feedthrough for the main resistor is actually not necessary. Simply calculate how much the very short resistor copper wires affect the total temperature coefficient.

The SR104 uses 2-wire feedthrough but there is also 15 cm 2-wire cabling from the feedthrough to the binding posts (2w/4w interchangeability was important at the time it was designed).

KerryW · « **Reply #18 on:** December 03, 2016, 09:40:31 pm »

Copper chips?

TiN · « **Reply #19 on:** December 04, 2016, 04:54:16 am »

I'd buy few samples of such neat boxes for my unhappy foils

branadic · « **Reply #20 on:** December 04, 2016, 09:43:01 am »

Quote

If you look closely there seems to be a naked bulk metal foil resistor inside the box.

Good observation

Yes, it is a hermetic case for Vishay VAR resistors in 4W connection and an additional PT1000, using PCap01 a like chips to measure and digitize temperature. The case will be filled with Galden, to avoid humidity effects. The lid is soldered to the bottom part.
It's just to find out how Vishay VAR can perform. Because at the institute I work for our main topic is packaging of integrated circuits. What I've learned so far is, that every highend sensor can perform bad if package and assembly is bad. So take the naked part, use an adequate packaging technology and the part will perform much better.

Andreas · « **Reply #21 on:** December 04, 2016, 10:47:13 am »

Hello branadic,

one question arises to me.
Does Galden expand with temperature?
How is the relation of the expansion coefficient against the housing.
Or is there some pressure relief planned?

With best regards

Andreas

branadic · « **Reply #22 on:** December 04, 2016, 10:54:47 am »

I plan to use something like LS230, so no big deal within normal operation.

Theboel · « **Reply #23 on:** December 04, 2016, 11:22:59 am »

Hi Branadic.
Could you consider to use more sensor like BME280
So you can measure humidity and pressure

acbern · « **Reply #24 on:** December 04, 2016, 11:38:13 am »

why galden? just because its handy? I would think others such as fluorinert or silicone transformer oil (ok, soldering issues, but thats after the fact anyway) are more itended for that purpose, and partly also show higher insulation resistance. I do not know galden very well, other than its used for VPS, maybe you can give us some details why you selected it.

branadic · « **Reply #25 on:** December 04, 2016, 11:45:18 am »

There is no need to measure humidity if the case is fully filled with LS230, Galden does not absorp water. And changes in pressure will be very low, presumably to low to measure it with BME280. Furthermore BME280 will not like the contact with Galden, especially its humidity sensor. To integrate a really suitable pressure sensor is a lot of effort.
The only thing I could consider is some kind of pressure membrane made of CuBe that can balance small pressure differences between inside and outside, but that would be a bigger effort too.

The knowledge why I use Galden is some ip that I can't share with you, so you have to take it as it is.

SeanB · « **Reply #26 on:** December 04, 2016, 12:14:36 pm »

I would guess because it is very inert, contains no moisture and is not going to react with the components of the casing and the resistor chemically, and will only have a constant effect on AC characteristics as it is a dielectric liquid.

SF6 would probably also work there, but is a lot more difficult to handle. Would require a fill port to use it, and a vacuum proof seal as well.

I guess you do the fill simply by using some pretinned case lips, a tinned cap edge and solder without flux, by simply immersing in the hot liquid, after filling the case with cold liquid and placing the lid on. Solder melts below the boiling point, and then the lid is sealed perfectly with no gas bubbles inside, and with a hermetic seal as well.

zlymex · « **Reply #27 on:** December 05, 2016, 12:16:46 am »

Quote from: manganin on December 03, 2016, 09:34:38 pm

Worth noting that the 4-wire feedthrough for the main resistor is actually not necessary. Simply calculate how much the very short resistor copper wires affect the total temperature coefficient.

The SR104 uses 2-wire feedthrough but there is also 15 cm 2-wire cabling from the feedthrough to the binding posts (2w/4w interchangeability was important at the time it was designed).

Agreed.
That 30cm(15cm * 2) is only 10mR in resistance(1mm² is assumed), which is only 1ppm of the total resistance, and the variation of which can be safely ignored.

zlymex · « **Reply #28 on:** December 05, 2016, 12:50:16 am »

Quote from: branadic on December 03, 2016, 07:43:04 pm

There is a lid on top and some feedthrough capacitors at the sides.

Very interesting. Are those feedthrough capacitors be soldered to the case?

babysitter · « **Reply #29 on:** December 05, 2016, 06:14:58 am »

Hi Branadic,

I too was thinking about a prefilled solderable (Otto Schubert) enclosure and Novec 7100 for a board, I think I also mumbled about it on volt-nuts or here. For pressure/expanson relief, I was thinking about a nickel bellow (e.g. from servometer mfg.) but someone pointed me to the "pressure capsule" (Druckdose...) used in liquid-filled thermostats, any old fridge might give you one.

Gyro · « **Reply #30 on:** December 05, 2016, 10:15:57 am »

Quote from: SeanB on December 04, 2016, 12:14:36 pm

I would guess because it is very inert, contains no moisture and is not going to react with the components of the casing and the resistor chemically, and will only have a constant effect on AC characteristics as it is a dielectric liquid.

SF6 would probably also work there, but is a lot more difficult to handle. Would require a fill port to use it, and a vacuum proof seal as well.

I guess you do the fill simply by using some pretinned case lips, a tinned cap edge and solder without flux, by simply immersing in the hot liquid, after filling the case with cold liquid and placing the lid on. Solder melts below the boiling point, and then the lid is sealed perfectly with no gas bubbles inside, and with a hermetic seal as well.

Isn't a sealed rigid looking container, completely liquid filled, going to subject the contents to very large pressure changes with temperature? That sounds like something that could impact drift.

Totally inert (Noble) gas filling sounds a safer bet, even if it does involve fill ports. These wouldn't be difficult to arrange, taking a tip from the refrigeration industry. A couple of thin copper tubes soldered into the case, fill through one, exhaust through the other. Crimp and solder seal the exhaust one and then crimp the inlet one, remove the gas supply and solder seal that too. Argon would be the cheapest and most readily available option.

SeanB · « **Reply #31 on:** December 05, 2016, 07:09:01 pm »

Galden is very non reactive, as it is a fluorinated hydrocarbon. It takes a lot of energy ( like a high current arc) to break it down.

SF6 however is around the most stable compound you can get, it will not decompose under any condition short of having lightning hit it. Dangerous though as it is invisible, has no odour and displaces oxygen in the ambient with a leak.

Kleinstein · « **Reply #32 on:** December 05, 2016, 07:45:24 pm »

If really 4 wire feed troughs are needed, depends on the resistor value. For a 10 mOhms resistor you will definitely need it, at 1 M you don't. Also the wires at the feed through can be thin and maybe something else than copper.

For a constant pressure, it might be good to have a little gas left. If this is just the vapor of the liquid used one would just follow the vapor pressure curve. So stable pressure at stable temperature.

branadic · « **Reply #33 on:** December 07, 2016, 07:25:22 pm »

Currently the copper parts are on their way to me... Lids are finished and the surface of the copper parts blasted with glas beads. Thanks to Hugo Lindner, he did a very great job, didn't he?

Andreas · « **Reply #34 on:** December 07, 2016, 07:46:19 pm »

Hello,

where do I get this golden christmas tree?

with best regards

Andreas

acbern · « **Reply #35 on:** December 08, 2016, 12:00:14 pm »

Nice units. Did not know Hugo Lindner, bookmarked. What did you have to pay appr. per piece? I feel I am often ripped off by my mechnical guy.

Secondly, VAR resistors are not that easily available from stock, standard Z-foil are. Why not use these?

TiN · « **Reply #36 on:** December 08, 2016, 12:26:21 pm »

Sorry, can't help it. That face with triple holes looks too much like :

branadic · « **Reply #37 on:** December 09, 2016, 07:37:30 pm »

Gift giving

The parts arrived today, what beauties. One of them had a date with some feedthrough caps and a resistor. No, nothing dirty just a fitting

Next step, Ni/Au finish of the surface!

branadic · « **Reply #38 on:** January 11, 2017, 07:39:55 pm »

Surface is now processed with nice looking Ni/Au.

cncjerry · « **Reply #39 on:** January 14, 2017, 05:06:53 am »

you gave it away too soon... I was going to say "a snow cone maker."

Nice work by the way.

Dubbie · « **Reply #40 on:** January 14, 2017, 07:17:25 am »

Did you get those cast or just machined from solid bar? That's a pricy piece of metal!

Sent from my phone using Tapatalk

branadic · « **Reply #41 on:** January 14, 2017, 10:50:10 am »

Quote

Did you get those cast or just machined from solid bar? That's a pricy piece of metal!

They are maschined from solid copper bars I had laying around for years. But they are smaller then they look.

branadic · « **Reply #42 on:** August 20, 2018, 07:14:11 pm »

The first 100R - PT1000 combination is almost done. Need to solder the lid and fill the case with fluid.
Also finished two assemblies with 1k - 33k NTC combination for a forum member. As I only need another 2 cases for 1k and 10k there are 5 units left. Someone interested in them?

-branadic-

Magnificent Bastard · « **Reply #43 on:** August 20, 2018, 08:08:25 pm »

Quote from: branadic on August 20, 2018, 07:14:11 pm

The first 100R - PT1000 combination is almost done. Need to solder the lid and fill the case with fluid.
Also finished two assemblies with 1k - 33k NTC combination for a forum member. As I only need another 2 cases for 1k and 10k there are 5 units left. Someone interested in them?

-branadic-

Beautiful!

IRFP · « **Reply #44 on:** August 20, 2018, 08:44:01 pm »

Good evening,
Superb achievement!

These are capacitors crossed? There are simple, non-capacitive and hermetic feedthroughs. Here is the manufacturer link: https://7doloy.micronor.fr/produits/#traversees-isolantes-unitaires
Pour ce qui est de l’huile, moi j’ai utilisé pour stabiliser en température mon LM399 de l’ISOVOLINE de chez TOTAL, c’est de l’huile diélectrique pour transformateur HT/BT.
Attention à l’étanchéité de la chambre, cette huile est extrêmement fluide.

branadic · « **Reply #45 on:** August 22, 2018, 08:08:34 pm »

Thanks for the flowers.
Yes these are low value feedthrough capacitors. Hermetic feedthroughs are nice, but my problem is, I don't speak french and their website seems to be very limited with information in an international language.

-branadic-

branadic · « **Reply #46 on:** August 23, 2018, 06:12:14 pm »

Seems like the five units found two new owners

-branadic-

TiN · « **Reply #47 on:** August 24, 2018, 04:04:16 am »

I'd be interested in one to run to do 8508/01s comparison, if you have more for sale. My main interest is in 10/100/1K values, as I have 1R/10K and highers to better standards already.

branadic · « **Reply #48 on:** August 24, 2018, 08:02:47 am »

You are to late boy, they are all sold.

-branadic-


EEVblog Main Site	EEVblog on Youtube	EEVblog on Twitter	EEVblog on Facebook	EEVblog on Odysee

Author Topic: Investigation on ... (Read 12707 times)

Share me