WARNING: incorrectly labelled, but EXPENSIVE, V-Cap and Solen PTFE capacitors

[RJKflyer]

I have just replaced the output tube grid signal caps, and the ground return caps in each channel in my MV-55, with Solen FPE (Tin/PTFE) and V-Cap ODAM respectively.

I do a lot of valve equipment restoration and test every film cap before installation as I am used to minimising noise in high-impedance circuits.

The warning is that the manufacturers' declarations regarding which connection was the outer foil proved to be wrong on one of two V-Caps and one of four Solen.

At this price point (GBP500 for the six), completely unacceptable, and more to the point a huge fail for those without the equipment to verify such claims.

[admin]

Do those caps have polarity?

[RJKflyer]

It is not about the 'polarity' of the capacitor - it is about its behaviour as an aerial.

Nearly all film (or foil) caps are essentially a rolled up number of alternate sheets of dielectric and metal foil, or a metallised dielectric. The foil or metallised layers are the 'plates' of the cap. The leads are attached to alternate films or layers.

This means when finished one of the 'plates' of the capacitor formed by one conductive foil or layer is on the outside.

Thus your capacitor, so simply shown on a schematic by two parallel plates and a gap, is really like an open ended drinks can with a conductive layer exposed to the outside (to which one wire is connected) and the other layer 'shielded' within the can, and connected to the other wire.

If you install such a cap into e.g. the very high impedance, and thus sensitive, grid feed to a tube, you want the 'can' connected to the SOURCE not the GRID. The former being hopefully the lower impedance end of things. You do not want that nice big conductive can sweeping up all the RF and available noise and plumbing it into the grid of your amplifier tube.

The outer foil should go to the ground or lowest impedance.

Way back, ALL film caps had the 'Outer Foil' marked by either words or a black line. Nowadays, in the semiconductor world, and often with PCBs with ground planes, it is usually less important. See random picture found courtesy of guitar.com attached.

You cannot rely on the printing - most even high quality film caps, tumble down the production line and thus the print 'direction' is usually random.

However, these supposedly audiophile caps, sold at vast expense, always state which way the outer foil is orientated. such as "direction or writing is towards outer foil connection" or similar.

But clearly this is not always the case as they're not bothering to do the requisite QA.

See the Solen Tin/PTFE examples attached - the black dot I have added when testing shows the foil end and they are the opposite to each other. On the other channel (not shown) they were both the same direction and according to Solen's claims.

[admin]

I'm familiar with how a capacitor functions. I don't recall this information on the solen caps I have bought before (similar to the black solen cap you have in the picture). Makes me wonder if I installed them correctly (last time was a cap replacement in the crossover for my magnepan n 20.1 speakers)?

Question, how are you able to test which connector is for the external outer plate?

[RJKflyer]

I use a bespoke foil side tester - proprietary design to Mr. Carlson's Lab on Patreon.

This unit is a very sensitive amplifier with a front end that electronically switches the two capacitor connections between input and gnd in the sensing amplifier.

LED bar graph indicates the received 'noise' which is 'injected' simply by holding the cap between finger and thumb - you act as a nice fat aerial to pick up local mains-borne 50/60Hz.

Least LEDs are illuminated when the outer foil is grounded as it is shielding the input.

Here's a polystyrene cap under test - nice example as they are usually marked with a red end for foil side. I am not touching leads although that isn't clear in pic. Left hand end is coloured red - sorry hard to get it all captured whilst unit operating.

Yes, all film caps have an outer foil - those Solen MKP FC are replacements and were also tested.

[RJKflyer]

An alternative is with your 'scope.

See with red end to the 'scope sense connection (red croc clip) we have outer foil 'exposed' and it picks up the 50Hz in my case.

With red end to black croc clip ('scope gnd) we are now 'shielding' the red input and it picks up almost nothing.

[RJKflyer]

I thought this practical test might be informative, and hopefully persuasive for anyone who considers orientation unimportant.

I have knocked together a simple 12AU7 two-stage cathode-follower amplifier with about 20dB of gain, running at 200V plate.

First video shows the input capacitor (which couples to the grid via a 470k VR) installed correctly (you can see the black dot I apply to identify outer foil end). Note how the black needle moves as I touch the cap.

Second video shows the cap installed incorrectly with the foil 'aerial' towards the grid. Spot the difference when same experiment is performed.

Yes, this would all normally sit in a screened case, but in our type of equipment we'll likely have all manner of AC floating about INSIDE the case.

[admin]

Interesting videos. Question, is the noise only introduced if there is something coming into physical contact with the backwards cap?

[RJKflyer]

No contact required: (a) as I get close to the cap you see noise increase, and (b) it's insulated!

It's a metal can which acts as an aerial. Although harder to see you may be able to spot that the residual noise level is higher with the 'backwards' cap.

[admin]

Do solen caps even have markings for polarity? I'm trying to remember if I saw anything on the caps themselves to indicate direction and I just don't remember seeing anything on them. Maybe I'm just forgetting?

[RJKflyer]

The lesser Solen PPE FC polypropylene range do not purport to be marked for polarity, and my experience is that they are truly random.

The vastly expensive FEP Tin/Teflon ones were stated on my supplier's website as:

"S3Q series have the outer foil on the right of the marking. The Outer foil end should be connected to the lowest impedance path to the ground."

However, going to Solen.ca the datasheet makes no such claims: https://solen.ca//storage/media/uHemulp ... NlsKbL.pdf

All a bit of a lottery...

[admin]

In all honesty, I think your supplier is making claims that are not supported. The data sheet from Solen is going to be the final word. I've only bought Solen from Solen directly so can't speak about 3rd party suppliers.

[RJKflyer]

Yes, you may well be right.

It does emphasise that as modders we need to get this right when we install them, else large sums of money may well end up making the equipment far worse.

[admin]

It's odd that Solen would not view this is a problem to the extent of properly (if at all) marking their caps.

[RJKflyer]

I buy only best quality caps for all HV and instrumentation work and they are randomly labelled - they tumble down the production line and the print ends up where it ends up.

These days, with semiconductor circuits a bit less susceptible to pickup due to lower impedance maybe, a general lack of AC ripple around the board due to lower voltages, lower currents, ground planes on PCBs and so on, it is clearly not considered important enough to merit production marking.

But for us, as I have explained, and now proven, it matters.

We live in the World where marginal gains matter, and we pursue them relentlessly and often at considerable cost.

A lovely weave on your solid gold mains cable is a bit pointless if your grid coupling cap is back to front...!

[Wildcat]

Interesting info! Good to know if I ever take on a refurbishing project, or have to repair one of my pieces.