Tuning capacitors

[Crackle]

Would minor surface corrosion on the aluminium vanes of a standard air spaced AM type tuning capacitor effect the capacitance by much.
It seems I always have to adjust my radios by opening up the oscillator trimmer to the limit of its travel in order to get the HF end of MW covered.

I have another 1934 KB radio where the LW coverage is way out. R4 is appearing at 1700 on the dial, reception is very loud on LW with just its own loop aerial, and MW is almost as good.

I am not sure how to get LW back onto 1500 on the dial, the LW tracker adjustment does not seem to have very much effect.
Circuit is here.
       

thanks
Mike

[Murphyv310]

Hi Mike.
I think it will do.
I had one that was badly corroded and it crackled and was very rough. After pulling it to bits I cleaned it with IPA then managed to carefully fold some wet and dry paper in between the vanes and fixed parts, dipped the wet and dry in IPA and then blew it out with the air gun. On assembly it was perfect and I did notice the tracking was much better on the parts that actually worked before.

[Craig]

I've just done some sums, and come up with a typical vane/stator spacing of about 0.3mm, which is about right for a 10 vane 500pf tuning capacitor. Now suppose that there is 0.005mm of corrosion (ie 5 microns), so the total on both sides of the gap will be 0.01mm. The corrosion will be aluminium oxide typically, which has a relative permittivity of 10. This increases the effective gap by 0.01 x 10 = 0.1mm. So the effective gap is now 0.3mm+ 0.1mm = 0.4mm. So the capacitance will go up from 500pF to 750pF, and the tuning frequency by the root of that - a shift to longer wavelengths of ~20%.

Now that is an over-estimate in capacitance shift. The oxide layer could be thinner, which would change the values, and of course the aluminium is a tiny bit thinner because some of it has been converted to oxide.

But the principle remains - could a visible oxide layer change the value of a tuning capacitor? Yes.

[Murphyv310]

As they say the proof of the pudding is in the eating and certainly in my case Craig removing the corrosion sorted out the bad tracking and the other issues.
I've another old tuning cap in the drawer and I'll check its capacitance fully closed, its another corroded one and I'll dismantle it and check the after capacitance and see how much of a change we get.

[Terry]

I 've been thinking about the effects of the corrosion although without Craig's numerical analysis - which puzzles me.

Craig says the corrosive material effectively increases the gap by 33%, then claims that this increases the capacity by 50%!

Surely it needs an effective reduction of the gap to increase capacity?

My thinking was that the increase comes about because the corrosive layer alters the effective dielectric in the gap to give the increase but, either way, as the plates are fully unmeshed at the high frequency end of the band, where does the increase  in capacity come from there as the corrosion no longer plays any part in the process?

(Signed)

Puzzled of North Hykeham.

[Murphyv310]

Hi.
I agree with your reasoning Terry. The corrosion is certainly not a smooth surface its full of pits and bumps so the surface area increases and when the plates are "meshed" then the increase in effective surface area ups the capacitance. A bit like brake discs if grooved, they have a larger contact area with the pads and it increases the effective braking.

[boater sam]

There was a discussion recently about the affect of the moving vanes not being in the centre of the gaps in the fixed ones. The concensus was that it altered the capacitance.
So any change in the gap or the dialectic property will do the same.

[pwdrive]

I used to think that moving off center would alter the capacitance but I'm not so sure now.

Lawrence.

[Murphyv310]

Hi.
The dual gang one here that is in a bit of a state reads fully meshed.
591pf & 618pf, its a 500x500pf type.

[Cleggy]

I've just done some sums, and come up with a typical vane/stator spacing of about 0.3mm, which is about right for a 10 vane 500pf tuning capacitor. Now suppose that there is 0.005mm of corrosion (ie 5 microns), so the total on both sides of the gap will be 0.01mm. The corrosion will be aluminium oxide typically, which has a relative permittivity of 10. This increases the effective gap by 0.01 x 10 = 0.1mm. So the effective gap is now 0.3mm+ 0.1mm = 0.4mm. So the capacitance will go up from 500pF to 750pF, and the tuning frequency by the root of that - a shift to longer wavelengths of ~20%.

That's a very interesting calculation but I thinks it's hard to quantify what the OP describes as "minor surface corrosion". Bright, shiny Al (which has had no chemical treatment or corrosion) usually has a surface oxide layer which is <<1 micron. But anodised Al can have several microns of oxide on it and still look shiny. I'd guess that this is one of those issues where the visual appearance might be quite misleading and you'd have to measure the C values to know what they really are.

I have a 1941 HRO, which I bought in the 1960's, and the 4-gang tuning cap on that looks as though it has had some form of a thin coating of some waxy substance applied to it. It tunes as expected, and although it does not look particularly nice, I've never attempted remove the coating. I cannot recall having seen such a coating on any other radio (though I know they used to "tropicalise" military equipment), nor have I seen the same thing on various pictures of HRO's; can anyone comment on it?

C