Beautiful precision from the golden age - Fluke 720 oil can teardown

[RoadDog]

[MrYakimovYA]

What is this?

[xrunner]

What is this?

Looks like precision resistance values created with wire-wound parts, hence the naming Rxxx ...

[MrYakimovYA]

What is this?

Looks like precision resistance values created with wire-wound parts, hence the naming Rxxx ...

It's very interesting are those resistors bifilar wounded to minimize parasitic inductance?

[Mickle T.]

It looks like part of an old Fluke differential voltmeter.

[guenthert]

What is this?

Looks like precision resistance values created with wire-wound parts, hence the naming Rxxx ...

It's very interesting are those resistors bifilar wounded to minimize parasitic inductance?

I don't know, but I noticed that those 'coils' come in pairs, so they could be wound in alternating orientation for the same effect.

[xrunner]

What is this?

Looks like precision resistance values created with wire-wound parts, hence the naming Rxxx ...

It's very interesting are those resistors bifilar wounded to minimize parasitic inductance?

I don't know, but I noticed that those 'coils' come in pairs, so they could be wound in alternating orientation for the same effect.

I found a few that were clear enough that I think I can see how they were wound. I believe every other one alternates the winding direction. I can see (starting at the left) the winding's exit wire, say this one was clockwise. Then I believe I see the next one's exit wire coming around in the counter-clockwise direction, then the next again clockwise. Can't see the last one for sure.

But I may be wrong ...

[Kleinstein]

Alternating the winding direction makes absolute sense. This is not only to reduce the inductance, but also to reduce the hum pick up / inductive coupling between resistors.

[Conrad Hoffman]

If you take apart (smash) one of the big plastic Julie resistors, they're wound exactly like that, alternate directions of each section of the bobbin.

[TimFox]

I wonder what the four resistors (labeled Rxxx A,B,C,D) at each location do in the circuit?

[jeffjmr]

Are you gonna tell us?   

[RoadDog]

They are resistors! 

I’m waiting to see if anyone comes up with a good guess.

[MrYakimovYA]

By the way, do I understand correctly than wire-wound resistors have minimal additional (not Johnson) noise? I'm not a good expert in noise issues. So I'm learning those questions now.

[JimmyJo]

Finally, someone opened a 720a oil tank, and took pictures.

Now please tell us the whole story, what did this poor box do?
What i'm really interented in is what kind of solvent you used to clean off the oil, looks clean!

[RoadDog]

One of the resistors is open. I used alcohol. I don’t think any solvents would be a good idea.

It was quite the project to get it apart.

[RoadDog]

So a few observations and thoughts.

Don’t open one of these oil cans unless you want to deal with a ton of solder. I mean a ton of it. I desoldered it as best I could and finally was able to get one corner to start moving up. It was a slow and tedious process as all four edges are heavily soldered.

Each 9.896k resistor is actually four 2.474k resistors in series. Most of the strings have drifted very little although a couple of them have drifted up over 9.910k. I haven’t done the math but I believe they would still be able to be trimmed although maybe with some modification to the trim circuit like Dr Frank performed on his 720.

I’m not sure about bifilar winding as mentioned. Does that matter with DC? I’m no expert or engineer. Just a self taught guy and pretty new to this level of precision.

Often I see these type of Fluke resistors on eBay they’re often in an envelope that says they’re a matched set. I’ve never seen these value resistors but others for similar era equipment. So I’m guessing they split each A decade resistor into four resistors for heat distribution and I’d imagine this whole set is matched somehow. Each string is labeled A, B, C, D and I don’t think that would matter if they just took a pile of resistors and installed them randomly.

The wire is soldered on the resistors so it’s repairable and I think that’s the only way I can return this to a working unit. Even if I had another bad can and tore it apart I think swapping a resistor from that can into this one would mess up the matching and it wouldn’t be stable.

I can’t see anything wrong with the open resistor even under magnification. This may just end up as a piece of wall art unless I can figure it out and repair it. Of course even if I find the issue and hit it with a soldering pen it could shift from what I’ve read here. 

[Kleinstein]

The usual resistor material is a NiCr alloy and is very hard (nearly impossible) to solder. In most cases some kind of spot welding is used for those wire wound resistors.

Low ohms resistors sometimes use Manganin (CuMn) or a similar copper based alloy that is relatively easy to solder, but more sensitive to higher temperature and less stable over time. With low ohms the connection gets more critical and thus the preference for a material that is better to connect.

The 720 has methods to do a fine trim an the resistors. So if the resistor is broken at one end, which is likely point of failure, it should be OK to repair that resistor even if the wire gets shorter by some 5 mm. Soldering / welding would effect mainly the first few mm and thus only a small part of the resistor and should thus be acceptable. The original resistors also have the ends treated in a simular way.

The parasitic inductance would not matter very much, but low inductance also does not hurt. The construction with seprate winding chambers also helps to keep the peak voltage seen by the wire insulaton small. Than changing the direction is only a small addition.

[RoadDog]

The usual resistor material is a NiCr alloy and is very hard (nearly impossible) to solder. In most cases some kind of spot welding is used for those wire wound resistors.

These are soldered. Same with the card resistors that are part of the trim circuit. Not sure what alloy they are but very obvious it’s solder and not welded so I think it’s possible to repair it. Super fine wire though and it’s tough to see. I really need a microscope at some point. Something better than a magnifying glass.

I’ll take a closer look at the windings later today and see if they reverse direction and report back.

[PartialDischarge]

If you take apart (smash) one of the big plastic Julie resistors, they're wound exactly like that, alternate directions of each section of the bobbin.

Ohmite HSP resistors are the same

https://www.ohmite.com/assets/docs/res_hsp.pdf

[alm]

One of the resistors is open. I used alcohol. I don’t think any solvents would be a good idea.

It was quite the project to get it apart.

Alcohols can be quite potent solvents . What kind of alcohol did you use? Ethanol? Isopropanol? Methanol?

[RoadDog]

I suppose you’re right. I meant any other types of solvents that could react with the wire somehow. I figured alcohol was the safest bet. I used 99% isopropyl alcohol from the drug store.

[RoadDog]

Also I let it dry for quite a while on top of my refrigerator which gets warm and a small fan blowing on the open port just to be safe before hitting it with the soldering iron.

[RoadDog]

What is this?

Looks like precision resistance values created with wire-wound parts, hence the naming Rxxx ...

It's very interesting are those resistors bifilar wounded to minimize parasitic inductance?

I don't know, but I noticed that those 'coils' come in pairs, so they could be wound in alternating orientation for the same effect.

I found a few that were clear enough that I think I can see how they were wound. I believe every other one alternates the winding direction. I can see (starting at the left) the winding's exit wire, say this one was clockwise. Then I believe I see the next one's exit wire coming around in the counter-clockwise direction, then the next again clockwise. Can't see the last one for sure.

But I may be wrong ...

You are not wrong. Each section of the bobbin is wound the opposite way.

Also Kleinstein was right I think they are welded not soldered. I thought for sure they were so soldered but looking again at some I believe they were welded. They have a coating over them that looks like epoxy to hold the wire to the end of the bobbin. I think I need glasses soon. Eyes are getting old 

[MrYakimovYA]

So, the main advantages of those resisor set, if I understand it correctly:
1. Low tempco.
2. Low additional noise (not Johnson).
3. Very low parasitic inductance.
4. May be low ratio if the set can be used as resistor devider.
5. Good long time stability.

I can't see any trim section in the string/or each resistor to get very very very low initial tolerance. I could be wrong here.

I just tried to bring it all together) To sum it up...

[RoadDog]

So, the main advantages of those resisor set, if I understand it correctly:
1. Low tempco.
2. Low additional noise (not Johnson).
3. Very low parasitic inductance.
4. May be low ratio if the set can be used as resistor devider.
5. Good long time stability.

I can't see any trim section in the string/or each resistor to get very very very low initial tolerance. I could be wrong here.

There is a trim circuit in the 720. It’s not inside the oil can. It’s a couple resistors and a 20 turn potentiometer in a series/parallel arrangement that allows very fine tuning when self calibrating it. You can take a look at the manual which includes a schematic towards the end.

https://download.flukecal.com/pub/literature/720A____imeng0200_0.pdf