Mike Silva wrote:
Duncan Munro wrote in message
. ..

The metal film 33R measures 6.5uH and the oxide 22R measures 4.5uH on the
aade.com meter. Both values (if the readings are correct) would represent
a high ratio of X to R at HF frequencies...

I wonder, though, if the AADE meter is not getting confused by the
resistance of the resistor. The fact that the measured inductance is
just about proportional to the resistance might be evidence for that.

Duncan has kindly sent a couple of samples, with duplicates that have
had the paint scraped off. I just arrived home from a few days away, so
haven't had time to measure them yet.

Each one is only an open spiral of about two turns along the whole
length of the 3W resistor body, so you can see immediately that there's
no way the inductance can actually be more than a few hundred nH.

This actual, physical inductance is in *series* with the resistance.
What seems to be happening is that the AADE meter displays the
resistance and reactance in their equivalent parallel form, which is a
function of the measurement frequency (which varies, but is understood
to be in the order of a few MHz).

Guessing a frequency and then doing the parallel - series
transformation on 22 ohms in parallel with 4.5uH produces results in the
right ballpark: R is still around 22 ohms but the *series* inductance is
100-200nH.

I will try to measure the resistors tomorrow.


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

On Tue, 30 Nov 2004 21:04:28 +0000, Ian White, G3SEK wrote:

Guessing a frequency and then doing the parallel - series
transformation on 22 ohms in parallel with 4.5uH produces results in the
right ballpark: R is still around 22 ohms but the *series* inductance is
100-200nH.

There is an additional complication in that there is another inductor in
the box itself of 680uH, LX (or should I say RX) is in series with that.
It's late now, but I will try and work out what's going on tomorrow night.

I will try to measure the resistors tomorrow.

If you get the opportunity, it would be much appreciated.

--
Duncan Munro
http://www.duncanamps.com/

Duncan Munro wrote:
On Tue, 30 Nov 2004 21:04:28 +0000, Ian White, G3SEK wrote:

Guessing a frequency and then doing the parallel - series
transformation on 22 ohms in parallel with 4.5uH produces results in the
right ballpark: R is still around 22 ohms but the *series* inductance is
100-200nH.

There is an additional complication in that there is another inductor in
the box itself of 680uH, LX (or should I say RX) is in series with that.
It's late now, but I will try and work out what's going on tomorrow night.

I will try to measure the resistors tomorrow.

If you get the opportunity, it would be much appreciated.


Over a range from 50kHz to 50MHz, Duncan's two resistors measure about
22 ohms + 31nH, and 33 ohms + 23nH.

To avoid having to construct a special test jig, I measured each
resistor with about 30mm of bent wire leads, which would account for
about 20nH of those measured inductance values.

The very low inductance of the resistor body is completely consistent
with the physical construction. On closer inspection, the metal film is
an almost continuous tube, with a very narrow spiral gap of about 1.5
turns. The gap adjusts the resistance by slightly increasing the overall
electrical path length, but it adds very little inductance.

If you used very short leads instead of the longer lengths I had to use,
these resistors would have a low SWR up to at least 144MHz.

Thanks again to Duncan for supplying the resistors.


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

Ian, if it's not too much trouble, I'd be very interested in how the
measured inductance compares to that of the resistor body/leads only.
I'd think that could be done by coating a similar-size resistor with
conductive paint or foil and measuring with the same method. The
difference between this measurement and the one you made would then show
how much inductance is due to the spiral element, and would represent
the minimum possible inductance for that resistor body type and lead length.

Roy Lewallen, W7EL

Ian White, G3SEK wrote:


Over a range from 50kHz to 50MHz, Duncan's two resistors measure about
22 ohms + 31nH, and 33 ohms + 23nH.

To avoid having to construct a special test jig, I measured each
resistor with about 30mm of bent wire leads, which would account for
about 20nH of those measured inductance values.

The very low inductance of the resistor body is completely consistent
with the physical construction. On closer inspection, the metal film is
an almost continuous tube, with a very narrow spiral gap of about 1.5
turns. The gap adjusts the resistance by slightly increasing the overall
electrical path length, but it adds very little inductance.

If you used very short leads instead of the longer lengths I had to use,
these resistors would have a low SWR up to at least 144MHz.

Thanks again to Duncan for supplying the resistors.

Roy Lewallen wrote:
Ian, if it's not too much trouble, I'd be very interested in how the
measured inductance compares to that of the resistor body/leads only.
I'd think that could be done by coating a similar-size resistor with
conductive paint or foil and measuring with the same method. The
difference between this measurement and the one you made would then
show how much inductance is due to the spiral element, and would
represent the minimum possible inductance for that resistor body type
and lead length.

I could certainly do that, because Duncan has supplied pairs of
resistors: one in original condition, and the other with the coating
cleaned off, just ready for painting.

However, the difference in inductance is going to be very small, and I'd
need to build a test jig that can keep other stray inductances under
control.

Over the weekend, maybe...


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

The idea was to measure the coated resistor under exactly the same
conditions as the regular one, so couldn't you just use exactly the same
setup as before?

Roy Lewallen, W7EL

Ian White, G3SEK wrote:
Roy Lewallen wrote:

Ian, if it's not too much trouble, I'd be very interested in how the
measured inductance compares to that of the resistor body/leads only.
I'd think that could be done by coating a similar-size resistor with
conductive paint or foil and measuring with the same method. The
difference between this measurement and the one you made would then
show how much inductance is due to the spiral element, and would
represent the minimum possible inductance for that resistor body type
and lead length.

I could certainly do that, because Duncan has supplied pairs of
resistors: one in original condition, and the other with the coating
cleaned off, just ready for painting.

However, the difference in inductance is going to be very small, and I'd
need to build a test jig that can keep other stray inductances under
control.

Over the weekend, maybe...

Roy Lewallen wrote:
The idea was to measure the coated resistor under exactly the same
conditions as the regular one, so couldn't you just use exactly the
same setup as before?

Afraid not... In the present setup, the main contribution to the total
inductance comes from the long, floppy resistor wires, and I couldn't
guarantee not to disturb their configuration while applying conductive
paint to the resistor.

To get a reliable answer, I'd need to reduce the lead length and make
the whole thing mechanically more stable.


Ian White, G3SEK wrote:
Roy Lewallen wrote:

Ian, if it's not too much trouble, I'd be very interested in how the
measured inductance compares to that of the resistor body/leads only.
I'd think that could be done by coating a similar-size resistor with
conductive paint or foil and measuring with the same method. The
difference between this measurement and the one you made would then
show how much inductance is due to the spiral element, and would
represent the minimum possible inductance for that resistor body type


I could certainly do that, because Duncan has supplied pairs of
resistors: one in original condition, and the other with the coating
cleaned off, just ready for painting.
However, the difference in inductance is going to be very small, and
I'd need to build a test jig that can keep other stray inductances
under control.
Over the weekend, maybe...


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

On Thu, 2 Dec 2004 23:20:40 +0000, "Ian White, G3SEK"
wrote:


What does a typical "test jig" consist of Ian? That's one thing that's
always puzzled me about these kind of measurements; particularly at
UHF++.
--

"What is now proved was once only imagin'd." - William Blake, 1793.

Paul Burridge wrote:
On Thu, 2 Dec 2004 23:20:40 +0000, "Ian White, G3SEK"
wrote:


What does a typical "test jig" consist of Ian? That's one thing that's
always puzzled me about these kind of measurements; particularly at
UHF++.

One that minimizes unwanted or uncontrolled lead lengths. In general,
one that is based on solid lumps of metal and large, broad,
low-inductance conducting surfaces.

I had simply bent the resistor wires so that one end pushed into the
centre of the VNA's N socket, and the other wire was literally tied onto
the body of the socket. However, the measurement showed that most of the
small inductance could be accounted for by those two wires - which you'd
never leave as long as that in a practical layout.

To home in on the inductance of the resistor body itself, I'd have to
build a jig that allows the wire lengths to be reduced almost to zero.
Harold W4ZCB sent a picture of something he uses, which is just a brass
plate soldered to the back of an SMA connector. The Device Under Test is
then soldered directly between the centre pin and somewhere on the
plate.

But I'm afraid my only visit to the workshop last weekend was to dump
yet another cardboard box on the floor.


--
73 from Ian G3SEK 'In Practice' columnist for RadCom (RSGB)
http://www.ifwtech.co.uk/g3sek

On Thu, 2 Dec 2004 12:01:03 +0000, Ian White, G3SEK wrote:

Thanks again to Duncan for supplying the resistors.

No need, it's thanks to you for taking the time to measure them and post
the results!

At least I now know my linear is not going to fail because the carbon comps
have been replaced with the metal oxide and metal film jobs. Thanks again
Ian.

--
Duncan Munro
http://www.duncanamps.com/