case positive electrolytic axials

[coppercone2]

if you want some axial capacitor that has a positive case, for use on negative rails, so the capacitor body is tied directly to ground, what do you search for?

[Conrad Hoffman]

Not even aware that such a thing exists. I'd search for shrink tubing instead!

[Ian.M]

If the can was bonded to the positive terminal it would be more susceptible to electrolytic corrosion, so they don't do that.  You could pay the premium in size and cost to use non-polarised capacitors . . .

[Gyro]

The oxide dielectric film is formed on the Anode foil, as in anodize, with the electrolyte being in contact with the negative foil and Aluminium can. Unless the manufacturer is prepared to anodize the inside of the can, then it will always be at the negative pin potential. They're not going to bother to do that while insulated sleeving is available and fitted, especially if dielectric film it is likely to degrade if the can is left floating (by far the normal situation as people use the legs / pins).

I don't think it would even be wise to ground the can of a bipolar electrolytic. The result could be rather unpredictable.

Invest in some heatshrink.

[coppercone2]

No, I mean so that the outside of the cap is shielded.


When you hook up axial cap to the positive rail, the capacitor body is connected to ground.


However, when you hook up a axial cap to the negative rail, the capacitor body is tied to Vee, instead of GND.

So the only thing that can change it would be tape.

Its just one of those annoying things I noticed.


Some old designs used foil caps, that have a shield can. You can obviously hook it up "backwards" but its incorrect to the design, and you can pick up interfernece with a sensitive probe (mr. carlson shows this). So he made a tester to identify where the body of the cap is connected to with a special testing machine, so you can use covered (dipped) caps properly, i.e. always tie outer layer to ground. But with a polarized cap, you can't do it!

[TimFox]

Many axial film or paper capacitors marked the “outside foil” end with a bar around the body.
It’s not an internal shield can.
Not so common anymore.
If you use an LCR meter such as the DE5000, with a third “guard” connection, it is easy to find it on an unmarked device by temporarily wrapping it with copper tape and trying all three permutations of the three LCR terminals and the three capacitor connections.
For a foil capacitor with one end grounded, obviously the outside foil should go to ground.
For a coupling capacitor, neither side grounded, some people identify the lower impedance node and connect the outside foil there.
I learned that it was a safety practice (especially in tube circuits) to connect the outside foil to the node at lower DC voltage.

[TimFox]

From my post of 2020 about determining outside foil with a DE-5000.


A quantitative measurement.  I take advantage of the three-terminal measurement (high, low, and guard) capability of the DE-5000 LCR meter.  The capacitor being measured is a 715P (originally Sprague, then SBE and Vishay, since sold to CDE), which is film-foil construction, single unit.  The value is 100 nF, 600 V.  The manufacturer did not mark the "outside foil", nor specify it on the data sheet.  (I have an extensive collection of polypropylene capacitors, both film-foil and metallized film, but could not find one with the old bar marking the outside foil.)  I therefore arbitrarily marked the ends "1" and "2" before the measurement.  The capacitor is about 0.6" diameter, 1.2" long, with 1.0" lead spacing.  I wrapped a piece of 0.5" wide copper tape around the midsection, with a tab sticking out for the test connection.

Here are the tabulated values, all measured at 10 kHz.

DE-5000                         
Red             2              1               Cu          Cu         1           2
Black           1              2                2            1         Cu        Cu
Guard         Cu            Cu               1            2          2          1
 
C            99.38 nF     99.38 nF     4.1 pF     25 pF    17 pF      2 pF
Q            > 1000       > 1000        22           75         14       very low

Obviously, "1" is the outside foil, showing roughly 20 pF to the copper tape.  I made no effort to control the stray capacitance at the pF level, so the small capacitances vary.  The low Q of the small capacitances results partly from the mediocre dielectric constant of the epoxy shell.

[coppercone2]

I remember now, I was excited that someone found a use for the guard terminal


It actually seems like a pretty serious design decision to put that capacitor in correctly for sensitive circuits, given all the other lengths people go to.

I remember shielding capacitors in precision charge balancer circuits is a important design consideration. I was going to fool around with it, but it actually turned out to be alot of work to figure out how to nicely shield a capacitor to the same quality as a manufactured one. I was going to buy some but I guess other things got in the way. I think he actually had a grounded shield on a capacitor that was 'floating' on both terminals there, maybe in a pre-ltc 6943 circuit.

I imagine being able to make precisely hydroformed soft copper enclosures that fit capacitors would be nice in that case. It seems like maybe I can give it a shot to try to plate them !

I think I saw a type of foil capacitor that had a sheet metal body, with solder tabs for ground in some old picture somewhere.

[TimFox]

Military paper capacitors, WWII vintage, to avoid ingress of humidity, were often assembled in "bathtubs" or cylinders with hermetic terminals, sometimes with one side grounded to the metal case (tin-plated brass or steel?).
Old Sprague "Vitamin Q" paper capacitors were often made as axial lead cylinders with one or two hermetic terminals (one if the capacitor was grounded to the case).
Here's an old Sprague from a surplus dealer:  https://electronicparts-outlet.com/en/home/4962-sprague-vitamin-q-audio-grade-paper-in-oil-capacitor-mil-specs-033uf-300vdc.html
Another old Sprague (I think the wire away from the camera goes to the case):  https://www.tedss.com/2020124450
Here's a modern re-make of the two-terminal with floating case variety:  https://jupitercondenser.com/products/vitamin-q?srsltid=AfmBOopIDn5ydSIHYsGEUMVqbBPCyKT2JXDoaopnLbBlUCSz9b-hdsl8
Here's a surplus "bathtub" multiple capacitor with a floating "common" terminal for the two capacitors:  https://www.tedss.com/2020006143
If you need to shield a modern axial capacitor, you could put it in a reasonably small die-cast metal box with feed-through terminals at opposite ends.
If the case fits too tightly, there will be different stray capacitances at each end.

[coppercone2]

hmm damn because I wanted to keep it as a 'component' not make a whole assembly, if you plate them, you actually get a neat electronics-manufacturer like component in the end. If you have to make hermetic boxes its just over the top


I noticed the #1 thing with the analog microelectronics, you want the parts reasonable. Even when you need to put shields , its always a massive hassle and layout is difficult and it quickly starts to feel like you have massive amounts of compromises and decisions to make.

Modifying parts with a coat or skin has minimal impact, and its often enough trouble to use the part as it is on a reasonable size PCB.


I am really curious now, I have some big axial foil capacitors. I want to give it a coat of conductive paint or electroless copper and then try to plate a shield on it!


I think the key will be to put a little bead of putty around the corners, to make them into a rounded 'ball' before plating, so you don't have obnoxious cusps that present difficulties to painting

[Ian.M]

Covering the shrinkwrap on the sides of the capacitor can with copper foil tape and jumping it to the positive lead where it exits the bung, and thus to ground, would be good enough for most low voltage applications where you don't want the can acting as a large antenna on your negative sensitive node. The extra shielding from covering the ends mostly isn't worth the difficulty of doing so,

[coppercone2]

I really think it might not be too much effort with electroplating and painting, I need more experiments but I am getting better at it


the last thing I shielded was my current sense inductor, which was telfon wire, wrapped in copper and painted with solder. Maybe after a decade I can make a more professional coating

[Ian.M]

You are mad if you want to do wet chemistry on it! Even cutting a kapton tape washer to insulate the negative can end, masking the can to lead joint with a blob of bluetack and spraying the f****r with nickel screening spray is too much work vs wrapping with self-adhesive copper tape + a wire jumper and two solder joints.

[coppercone2]

It does not seem bad to me but I actually implement the nitty gritty of difficult designs and this one does not seem that bad to me!

being able to manufacture precise washers is quite important and I learned there is a entire method to making the means of production for dielectric precision thin washer production and its valuable


The plastic washer is a under valued engineering object, it comes up again and again professional in different forms

its one of those things where the material cost is un-countably low but the value of the finished object is unreasonably high, and I say this professionally not hobbyist

I have seen 4 of these annoying paper thin bastards delay something like $300,000 of HEAVY machinery from working, and cause all sorts of chaos in a small lab

[coppercone2]

also some people hate taping things so much that working with a tank of corrosive material feels less stressful 

sometimes tape is more then just a occasional solution and you get SICK of it, I had a non tape masking method in mind

[TimFox]

There is conductive shrink tubing:  https://ph.parker.com/us/en/series/electrical-heat-shrink-tubing
I have used it for other applications, with a "drain wire" under it to connect the tubing.
Rather than electroplate, you could start with a layer of normal polyolefin shrink tubing (and maybe a plastic washer at each end) and cover it with a layer of conductive.