The AM Forum

Howdy, guys and gals.

Who knows the amount of voltage an Air Variable capacitor is rated at, given the distance between the plates?

Is there a formula?

Reason is, I've got a dual-section split-stator air variable capacitor that's got 5/16" spacing between the plates.  3 plates on the rotor and 2 on the stator.

E. F. Johnson model number 5BD8  (ser 365)

Google doesn't spit anything out.

73 = Best Regards,
-Geoff/W5OMR

There is a lookup table in the ARRL handbooks.  I built an Excel spreadsheet based on that data but can't get to it right now.

If no comes forward with the info, I'll post the spreadsheet when I get a chance

GL es 73, Al

I recall from memory that 3/16" gives a nominal 7000 volt spacing.  1/2" gives 9 kv.  The the chart should be in any handbook (assuming they haven't discontinued it as they have done with a lot of other useful information in recent editions).

It is not exactly a linear function.  Doubling the spacing gives slightly less than double the voltage breakdown rating.

You'll have to do the math, but here ya go...

Spacing   Voltage (PEAK)

.015          1000
.020          1200
.030          1500
.050          2000
.070          3000
.080          3500
.125          4500
.150          6000
.175          7000
.250          9000
.350          11000
.500          13000

You're at .3125, so I'd say at least 9Kv


--Shane

and now, boys and girls, the idiot savant ham radio operator will attempt to CORRECTLY read a tape measure.
(...geeze...)

As measured on my Digital Volt meter, the cap is ~40uf (or 0.040nf, if you prefer) per section, with voltage spacing at/near 9/16" of an inch (better'n 10kV).

anyone interested in it, before it goes on e-pay?

AS I'm in the same boat as you, I found an EASY way.

Run down to Harbor Freight.  Pick up a cheap digi micrometer.  One that does BOTH english and metric.

Problem solved.  You'll still have to solve for the decimal eq. of the fraction, but otherwise, you're good to go. :)

I wouldn't trust the HF mic's to be accurate enough to do ANY real precision work (machining parts for an engine, etc), but for cap ratings, they should be more than adequate.

--Shane

I think you likely mean 40pF

or   0.04nF

 ;D

Interesting that doubling the spacing from .015 to .030 only gives a 50% increase in voltage rating (1 kv to 1.5 kv), but doubling it from .125 to .250 doubles the rating from 4500 to 9000.  But doubling it from .250 to .500 gives only a 44.4% increase from 9kv to 13kv. If that chart is true, not only is it a non-linear function, it does not vary consistently according to the percentage increase in spacing.

Here is the formula for breakdown voltage with a gas dielectric, as a function of the pressure of the surrounding gas. The non-linearity has to come from the log functions, plus the "Secondary Electron Emission Coefficient", however that would be calculated.

http://en.wikipedia.org/wiki/Breakdown_voltage#Breakdown_in_vacuum

40 pf is pretty useless but 400 pf would make a fine tuner cap.

I think my DVM only goes up to maybe 40, or 400 - not much more...

so... 0.040nf according to the meter...
and there's two sections.  Might be "just enough" to get you over that hump to get a tank circuit like say... a B&W CX-77 butterfly plate tuning cap on 75m.
;-)
Check the 'for-sale' area, lest I get my wrists smacked by the moderators...

Here's what I have. The data was lifted out of the ARRL handbook.  There was a little atypical bump in the data so I smoothed it out and did a paranormal trend on it.

I'm thinking someone some time did a "high school science project" and actually used a lab to gather the data.  There are formulas out there but this is more or less what I use

Shane - that's the same data I got out of the handbook.  I've included the curve from that.  Note the atypical sag around .125"  That's why I did the normalization.  I've included that chart too FYI

Al

Old engineers rule of thumb - Dry air is 30KV/inch breakdown voltage - Thats not what you design to - that's the breakdown voltage.

I'll dig my Orr's Radio Handbook out later, I seem to remember the spacings in it where a bit different than the ones in the ARRL manual.  I always took the most conservative ratings.

--Shane

Sounds good to me.  Lemme know what your results are.  I'm all for the conservative approach.  I'm dealing with the same issues while building a transmitter - also (maybe) a balanced tuner. If I can get some data, I can draw that curve in Excel for the in between points.

Al

Johnson KW match box uses 7000 volt caps. My Fugly tuner uses 1/4 inch plate spacing

Here's the reference chart from Orr:

.030          1000
.050          2000
.070          3000
.100          4000
.125          4500
.150          5200
.170          6000
.200          7500
.250          9000
.350          11,000
.500          15,000
.700          20,000

Incidentally, he has a SEPERATE chart for use when NO dc blocking cap is present.  The ARRL says you need DOUBLE the Vrating of the DC component, Orr is a bit more pessimistic.  And, he points out they are ratings for CW/SSB..  No data for AM...

Spacing    DC Volts
.030          400
.050          600
.050          750
.070          1000
.070          1250
.078          1500
.100          2000
.175          2500
.200          3000
.250          3500

Orr does state to use double the spacing necessary for plate modulation, but then he states that you need 1.5 times the voltage when plate voltage is present on the cap...  Of course, his chart shows different.  I'd go with at least double.



--Shane
KD6VXI

My first high power AM rig used a single 304-TL, running 2500 volts on the plate @ 400 mills.  The plate cap had 0.175" spacing, and would occasionally flash over on voice peaks even when everything appeared to be properly tuned up and fully loaded.

My present HF-300 rig runs only 2 kv on the plate at maximum, and also uses a 0.175" spaced cap in the plate circuit.  In never flashes over unless I forget to properly tune something and fully load the final, even with positive modulation peaks well in excess of 100%.

Both rigs use(d) a series fed tank circuit so that the DC did/does not appear across the plate cap, but eliminating the need for a plate blocking cap/ rf choke.

OTOH, the first time I tried to load the 80m dipole as a 1/8 WL dipole on 160, using a parallel tuned circuit with the OWL tapped down on the coil, it not only was very inefficient, but the tuning capacitor would flash over on voice peaks with more than about 100 watts carrier, even when adjusted for a perfect 1:1 SWR on the coax link from the transmitter to the tuner.  The tuning cap was split stator, a BC-610 plate cap 150/150 pf (also with 0.175 spacing). The problem was that I was trying to feed the OWL at its most reactive spot, exactly midway between a voltage loop and a current loop.  I added an additional 60' of OWL to the feed line so that it came out to a voltage loop, and it tuned up perfectly using parallel feed; I have never seen the tuner arc over on voice peaks at any power level.

Maybe Al can graph this also.

Johnson/Cardwell contribution to arcing:

Peak withstanding voltage for the 153 and 154 series.

Spacing      Volts

.030            1000
.045            2000
.075            3000
.080            3500
.125            4500
.175            7000
.250            9000

It might also be a good subject for the tech section

Carl

The actual data looks a little "squiggley"  I did a trendline to make it into a "predictable" curve

Al

Hi Guys:
Shane - KD6VXI.
Which edition of Bill Orr, W6SAI, handbook did you use?
73,
Chuck - AA5WG

Now, how does Frequency effect those numbers?

Seems like altitude AMSL and humidity also have an effect.

73,  Jack, W9GT

You noticed  :o  Im sure Johnson rounded off the values a bit to the nearest 500V ;D

Next overlay all the graphs and see which are conservative vs optimistic. Take the mean average and it probably realistic.

There is a mention of RF vs 60Hz voltage and also humidity, maybe buried somewhere on W8JI's site or a discussion on one of the amp forums.

Carl

23rd edition, 1987 edition, 1993 printing date, if that makes a difference.


--Shane
KD6VXI

It's also in the 1959 version.

Shane:
Thank you for the info. regarding W6SAI handbook.
73,
Chuck

That can be done. Maybe  I'll take a look at things after I finish my #$%^& weekly finances. ARGHHH

Yes, these graphs are to be done with all the above variables, humidity etc.  There's a lot of Kentucky windage involved.

Carl

I took my overlay data from three sources: W6SAI, Johnson, Cardwell, and ARRL.  I had to do some interpolating on some points. Actually they tracked pretty good.  I have included the spreadsheet this time so my data can be reviewed.  I'll include a snapshot for those who don't have Excel.  Actually, I think the numbers track pretty good. As before, the solid line is a "trend line" The actual data is the dotted blue line.

Added the overlay.

Al

Lowe's sells a digital vernier caliper made by General that does fractiondown to 1/64, decimal inches out to 4 places and metric out to 2 places for $25 or $30 and it's not cheap looking either.   Works quite nice and feels good in the hand. And it seems fairly accurate too.

All you guys have been awesome with your answers and thank you, Al, for the Data.

9/16" of an inch is pretty high-voltage spacing.  But, yes.. 0.040nf does equal 40pf.

Getting back to the actual capacitor, there's 3 plates on the rotor, per section, and 2 on the stator.
So, it's a split-stator 40pf capacitor, with somewhere in the neighborhood of ~15kV spacing.

Might be good for a 3CX3000 6M amp. What LWH dimensions overall and shaft size does it have?

The Frame is 10" long, 6 3/4" wide and around 4 1/2" tall.

Maybe good for a 20-10M amp with the sections in parallel, will also need a low output C tube for 10M.

Or use it for a 20-10M ATU for that 3CX3000 or YC-156 amp.

Carl