开云体育

Originally posted on 2007/05/05 at 8:09:04 PM Central Daylight Time
One correction made.


Hello Nigel, Paul et al

AFAIK no schematics for the Wellbrook have ever been published. However the
harware content which was discussed here, and other information, leads me to
the conclusion that the amplifier is a push pull grounded base design, as Paul
suggests.

It is not that difficult to work out a possible configuration, as follows:

1. Basic physics of a small loop

The basic physics of a small receiving loop is well understood (see eg
Jasik's Antenna Engineering Handbook, 1st ed, 1961, p 6-2)

The unloaded voltage induced in a loop antena =
2.Pi.A.N.E.Sin(theta) / Lambda
where
Pi = 3.14
A = the area of the loop
N = the number of turns
E = the electric field strength
theta = the angle measured from the axis of the loop
lambda = the wavelength

Let us simplify this, assuming that N = 1, and theta = 90 degrees. Let us
also replace Lambda by c/f where c = the speed of light, and f = the frequency.
Then
V = 2.Pi.A.E.f / c

It can readily seen that for a constant field strength E, that the voltage
ouput is proportional to frequency. There is a simple method to overcome this
difficulty.

The untuned receiving loop can be considered as a voltage source in series
with the radiation resistance (very small, ie a small fraction of an ohm), the
loss resistance (probably somewhat larger), and the inductance of the loop,
which is the largest component. Take for example a (Wellbrook) loop with a
diameter of 1 meter, and a tube diameter of 25mm. Then the the inductance (L) is
ca. 2.33uH, with an impedance of ca. 14.6 ohms at 1MHz.

If one terminates the loop in an impedance in a resistance (R) much lower
than the inductance, then the current output is determined by the inductance, and
equals
V / (2.Pi.f.L) = (A.E) / (c.L)
which is frequency independent

Then the power output from the loop is i.i.R
= (A.A.E.E.R) / (c.c.L.L)
which is again frequency independent.

As the frequency goes down, the ouput goes down for constant E, with the -3dB
frequency equalling
R / 2.Pi.L *** corrected ***

2. Loop termination resistance for a ALA1530

To determine an approximate value for R, it is necessary to look at the
performance figures of a Wellbrook loop. Fortunately, there are performance
figures for an ALA1530 on the Wellbrook web site. Looking at the numbers the -3dB
frequency is ca. 550kHz, which would give an R of ca. 8.1 ohms.

It is that simple - terminate a one meter loop with a resistance of ca. 8
ohms, and you will get approximately the frequency response performance of an
ALA1530. Clearly the response falls away at very low and very high frequencies,
but that is probably due to various transformer effects.

3. Amplifier gain for a ALA1530

We must now consider the amplifier gain. Assuming a unity antenna factor, ie
one volt out for a field strength of 1 Volt per meter, then the gain required
is
power out / power in
Assuming the ouptut load is 50 ohms, then the required gain
= (c.c.L.L) / (A.A.R.50)

Sticking in the numbers which we already have, I calculated that for a unity
antenna factor, that the gain required would be ca. 33dB. The actual antenna
factor is ca. -7dB, and so the required gain is ca. 26dB.

To recap, we know know that to duplicate a ALA1530, we would need an
amplifier with an input impedance of ca. 8 ohms, and a gain of ca. 26dB.

4. Amplifier configuration

Assuming that the ALA1530 uses two transistors in its push-pull amplifier,
then the only way to do is through a common base amplifier with a transformer
coupled input and output. A Norton amplifier would not have sufficient gain.
It should be remembered that the input resistance of a Norton amplifier depends
upon the output load, which we do not want in this case.

The small signal input resistance of a common base amplifier is ca. 26mV
divided by the emitter current. To get some good linearity, a high quiescent
current is essential. Removing the heat from the transistors has to be
considered.

Let us assume that pretty much all of the stated 120mA supply current goes to
the two transistors, ie ca. 50-60mA each. Then the input resistance of each
transistor at 60mA bias is ca. 0.43 ohms. The two are effectively in series
and so the total input resistance is ca. 0.86 ohms.

The required turns ratio of the input transformer is SQRT( 8 / 0.86) = 3.04.
This is conveniently very close to 3.0, which is what you would wind. This
means that the transformer can be wound with a trifilar winding.

The amplifer gain effectively is approximately the output load resistance
divided by the input resistance. We need a gain of ca. 26dB, ie ca. 400.

Assume that the gain is 27dB, ie 500. The the required output resistance is
500 x 0.86 = 430 ohms. Assuming that the transformed output load is 50 ohms,
this would give a turns ratio on the output transformer of 2.93, ie very close
to 3.0, which could also easily be wound with a trifilar winding.

My conlusion is that a push-pull amplifier with the transistors each biassed
at ca. 50-60mA, and a 3:1 step down transformer on the input and output would
work give approximately the measured performance of the ALA1530. This should
be a fairly straightforward amplifier, if some attention is paid to the
detail. 2N5109 transistors should work quite well. There are, of course, other
choices.

5. Observations

Many people have wondered how the broadband matching was done in the ALA1530.
It is simply done by having an amplifier with a constant input impedance
lower than the inductive reactance of the loop at frequencies for which a
constant gain is desired.

Many people have thought that the amplifier must in some way be exotic. I
suggest that it is actually fairly simple, and that this might be the reason for
potting it.

Later versions of the ALA1530 have an amplifier with better intermodulation
performance. They also have a higher quiescent current, and a higher gain.
This suggests to me that there are more transistors, probably in a two stage
push-pull amplifier.

There was a broadband loop available in the UK before the Wellbrook ALA1530
was introduced. It was designed by Edward Forster, and sold by his company -
Phase Track Ltd, which was based in Reading. His patent - GB2235337 - gives
some useful background on broadband loop antennas and their amplifiers.

Finally, broadband loop antenas are a compromise solution. At lower
frequencies, the radiation resistance is very low, and hence they are very
inefiicient. This means that you will not be able to hear signals near the noise floor.
Of course, they do appear to offer better noise rejection than an active rod
antenna, unless you are very careful (eg see some of Dallas Lankford's recent
writings). They are also convenient, and can offer some directionality.

HTH and 73

John KC0G


In a message dated 5/5/07 12:28:19 PM Central Daylight Time,
nonlinear@... writes:

Dear Nigel

forgot that getting old, I remember that now, I remember seeing the pictures
but was fuzzy on whether for sure bipolars were used

I believe it was not known from the autopsy whether it was a Norton design
or not
otherwise it is likely a push pull grounded base design for low noise and
wide bandwidth

what do you think?

was there a schematic ever published for the Wellbrook

best
Paul

----- Original Message ----
From: "gandalfg8@..." <gandalfg8@...>
To: loopantennas@...
Sent: Saturday, May 5, 2007 12:36:40 PM
Subject: Re: [loopantennas] Wellbrook ALA1530

In a message dated 05/05/2007 17:28:55 GMT Daylight Time,

nonlinear@rogers. com writes:

highest gain and lowest noise more likely with jfets
ie more likely in the design than bipolar although not necessarily true if
one considers say the David Norton noiseless design with bipolar
however the Norton design is limited to smaller gains

------------ --------- --------- --------- --------- ---------

"Is more likely" is a bit different to the earlier "must have".
I was just intrigued as to your reasoning, considering that the hardware
content of a dismembered 1530 was discussed here in some detail only a few
weeks
ago and the active devices were obviously bipolar.

regards
Nigel
GM8PZR

***
See what's free at .

In a message dated 06/05/2007 04:41:30 GMT Daylight Time, crabtreejr@...
writes:
AFAIK no schematics for the Wellbrook have ever been published. However the
harware content which was discussed here, and other information, leads me to
the conclusion that the amplifier is a push pull grounded base design, as
Paul
suggests.

It is not that difficult to work out a possible configuration, as follows:

1. Basic physics of a small loop

The basic physics of a small receiving loop is well understood (see eg
Jasik's Antenna Engineering Handbook, 1st ed, 1961, p 6-2)

------------------------------------------------

Hi John

Many thanks for the well reasoned analysis.

I haven't worked throught it in detail yet, will certainly do so later, but
a couple of points occur to me....

I'm aware of Jasik but don't have a copy, I tend to use Kraus (1950) or
Lapport (1952) as my main theoretical antenna references.
Kraus carries out quite a detailed analysis of small and large loop
performances, incidentally also ch6:-), including calculations of radiation
resistance.
The relationship involves quite a complex power series, although
approximation to a single term for small loops gives reasonable accuracy.
His plotted results indicate a 1 metre loop having a radiation resistance
that is indeed likely to be fractions of an ohm at 1MHz but rising to 3 ohms at
30MHz and 100 ohms at 100 MHz.
I'm not sure that this would have any great significance, it certainly
shouldn't at LF where the Wellbrook loops are most favoured, but will again
consider that more later.

Re your comment............
---------------------------------------------------
Sticking in the numbers which we already have, I calculated that for a unity
antenna factor, that the gain required would be ca. 33dB. The actual antenna
factor is ca. -7dB, and so the required gain is ca. 26dB.
-------------------------------------------------------------

I might be missing something here but with an antenna factor of -7dB
wouldn't you expect that to require an increase in amp gain rather than a decrease?

I'm pretty sure your conclusions re the type of amp are correct.
At first I had some doubts as the posted photos seem to show more than two
transformers but related postings do indicate two.
Originally at least it would seem that the transistors used were TO92
ZTX337s.
Although rated for quite a high max collector current and power dissipation
I can't see these being continuous figures for a TO92 package and am intrigued
regarding heat dissipation in that potting compound.
Having shared the general enthusiasm for 2N5109s, and other TO5 or TO39
style devices using good heatsinking, and having spent a fair amount making sure
I have ongoing stocks, this has given me some cause for reflection.

I doubt the 2N5109 enthusiasm will wane but some comparative measurements
might be interesting.

regards

Nigel
GM8PZR

In a message dated 06/05/2007 17:22:12 GMT Daylight Time,
nonlinear@... writes:

Is it absolutely confirmed there were only 2 transistors in the autopsy?

best wishes
Paul

At first I had some doubts as the posted photos seem to show more than two

transformers but related postings do indicate two. Originally at least it
would seem that the transistors used were TO92 ZTX337s.



-----------------------------------

Hi Paul

Are we confusing transformers and transistors?:-)

Pat confirmed two transistors and two transformers.
Initially though the rather fuzzy photos in the group photos section seemed
to show a couple of can enclosed transformers still on the board with what
might have been another one detached.

Pat, if you see this, I wouldn't mind higher res copies of those photos if
you have them, and wondering if you managed to trace the circuit from the
remains?

regards

Nigel
GM8PZR

Dear Nigel

Is it absolutely confirmed there were only 2 transistors in the autopsy?

best wishes
Paul

At first I had some doubts as the posted photos seem to show more than two

transformers but related postings do indicate two. Originally at least it would seem that the transistors used were TO92 ZTX337s.
<<<

N
apologize for the interleaving answers and further questions
P

In a message dated 5/6/07 5:55:03 AM Central Daylight Time, gandalfg8@...
writes:

In a message dated 06/05/2007 04:41:30 GMT Daylight Time,
crabtreejr@...
writes:
AFAIK no schematics for the Wellbrook have ever been published. However the
hardware content which was discussed here, and other information, leads me
to
the conclusion that the amplifier is a push pull grounded base design, as
Paul
suggests.

It is not that difficult to work out a possible configuration, as follows:

1. Basic physics of a small loop

The basic physics of a small receiving loop is well understood (see eg
Jasik's Antenna Engineering Handbook, 1st ed, 1961, p 6-2)

------------------------------------------------

Hi John

Many thanks for the well reasoned analysis.

I haven't worked throught it in detail yet, will certainly do so later, but
a couple of points occur to me....

I'm aware of Jasik but don't have a copy, I tend to use Kraus (1950) or
Lapport (1952) as my main theoretical antenna references.
Kraus carries out quite a detailed analysis of small and large loop
performances, incidentally also ch6:-), including calculations of radiation
resistance.
The relationship involves quite a complex power series, although
approximation to a single term for small loops gives reasonable accuracy.
His plotted results indicate a 1 metre loop having a radiation resistance
that is indeed likely to be fractions of an ohm at 1MHz but rising to 3 ohms
at
30MHz and 100 ohms at 100 MHz.
I'm not sure that this would have any great significance, it certainly
shouldn't at LF where the Wellbrook loops are most favoured, but will again
consider that more later.

Re your comment............
---------------------------------------------------
Sticking in the numbers which we already have, I calculated that for a unity

antenna factor, that the gain required would be ca. 33dB. The actual antenna

factor is ca. -7dB, and so the required gain is ca. 26dB.
----------------------------------------------------------

I might be missing something here but with an antenna factor of -7dB
wouldn't you expect that to require an increase in amp gain rather than a
decrease?

I'm pretty sure your conclusions re the type of amp are correct.
At first I had some doubts as the posted photos seem to show more than two
transformers but related postings do indicate two.
Originally at least it would seem that the transistors used were TO92
ZTX337s.
Although rated for quite a high max collector current and power dissipation
I can't see these being continuous figures for a TO92 package and am
intrigued
regarding heat dissipation in that potting compound.
Having shared the general enthusiasm for 2N5109s, and other TO5 or TO39
style devices using good heatsinking, and having spent a fair amount making
sure
I have ongoing stocks, this has given me some cause for reflection.

I doubt the 2N5109 enthusiasm will wane but some comparative measurements
might be interesting.

regards

Nigel
GM8PZR

Thanks for the comments. At least people are reading this with a critical
eye...

Re. implications of the very low radiation resisitance, even at say 1 MHz -
that should be the subject of another posting. I need to give the matter some
careful thought. I need to go through the maths, and get it right. An off
the cuff comment would not be appropriate.

Re the amplifier gain. If the antenna factor were unity, ie the antenna
output were 1 volt into 50 ohms for a field strength of 1 volt per meter, then the
required amplifer gain would be 33dB. The actual antenna factor is about
-7dB, ie ca. 0.45, ie the antenna output is ca. 0.45 volts for a field strength
of 1 volt per meter. The required antenna gain is then +33 - 7 = 26dB.

I too share your concerns about getting heat out of a transistor when it is
in potting compound. The other question is just how much current does a 2N5109
need before you get little or no improvement in IMD performance.
Conventional wisdom says as much as possible. However in his article 'Wideband Amplifier
Summary', in the November 1979 issue of Ham Radio Magazine, Dr. Ulrich Rohde
suggested that for a 2N5109 that the IMD performance was best at a collector
current of ca. 10mA. He then went on: "A typical CATV transistor, such as the
2N5109, has a flat curve of intermodulation distortion between 20 and 80mA."

Dallas Lankford has found pretty much the same using the MRF581A. You can run
the MRF581A at 25mA collector current without a heatsink. Given that
MicroSemi/APT (ie non-Motorola) devices are cheaper than the 2N5109 on this side of
the Pond, it would be my first choice of device.

HTH and 73

John KC0G




***
See what's free at .

In a message dated 5/6/2007 6:32:18 PM Eastern Standard Time,
crabtreejr@... writes:
Re the amplifier gain. If the antenna factor were unity, ie the antenna
output were 1 volt into 50 ohms for a field strength of 1 volt per meter,
then the
required amplifer gain would be 33dB. The actual antenna factor is about
-7dB, ie ca. 0.45, ie the antenna output is ca. 0.45 volts for a field
strength
of 1 volt per meter. The required antenna gain is then +33 - 7 = 26dB.


I'm not getting the math here. If the gain of the loop were -7dB from unity
and for unity gain you need 33dB gain, then wouldn't you need more gain for the
same output i.e. 33 + 7 = 40dB? Less gain 26dB would give you even less
output?
73 Todd WD4NGG



*** See what's free at .

In a message dated 5/6/07 6:12:23 PM Central Daylight Time,
toddroberts2001@... writes:

In a message dated 5/6/2007 6:32:18 PM Eastern Standard Time,
crabtreejr@... writes:
Re the amplifier gain. If the antenna factor were unity, ie the antenna
output were 1 volt into 50 ohms for a field strength of 1 volt per meter,
then the
required amplifer gain would be 33dB. The actual antenna factor is about
-7dB, ie ca. 0.45, ie the antenna output is ca. 0.45 volts for a field
strength
of 1 volt per meter. The required antenna gain is then +33 - 7 = 26dB.


I'm not getting the math here. If the gain of the loop were -7dB from unity
and for unity gain you need 33dB gain, then wouldn't you need more gain for
the
same output i.e. 33 + 7 = 40dB? Less gain 26dB would give you even less
output?
73 Todd WD4NGG

Think of the loop element and amplifier as a complete system. With the
components as described, the amplifier needs a power gain of 33 dB to get an
antenna factor (system gain) of 0dB.

The actual measured antenna factor (system gain) is -7dB. Hence you need
less amplifier gain, ie 33 - 7 = 26dB.

HTH and 73

John KC0G

To have a


***
See what's free at
.

gains are always positive

so -33 + (+7) = -26 db to be overcome

therefore positive 26 dB gain is needed
is the way I am looking at this

Paul

In a message dated 06/05/2007 23:31:54 GMT Daylight Time, crabtreejr@...
writes:



Re. implications of the very low radiation resisitance, even at say 1 MHz -
that should be the subject of another posting. I need to give the matter
some
careful thought. I need to go through the maths, and get it right. An off
the cuff comment would not be appropriate.
I need to give this further consideration too.
Given the relatively low input impedance of the amp I would have expected
that increase of radiation resistance with increasing frequency might have been
more of a limiting factor but suspect that's far too simplistic.



Re the amplifier gain. If the antenna factor were unity, ie the antenna
output were 1 volt into 50 ohms for a field strength of 1 volt per meter,
then the
required amplifer gain would be 33dB. The actual antenna factor is about
-7dB, ie ca. 0.45, ie the antenna output is ca. 0.45 volts for a field
strength
of 1 volt per meter. The required antenna gain is then +33 - 7 = 26dB.
That's what I assumed you meant before but, as previously written, it didn't
seem to make sense, with confusion perhaps mainly being due to the use of
"required".
I'd have probably still have written it here "the actual amplifier gain is
then.....26dB", rather than "required" but perhaps that's just me being
pedantic as usual:-)




I too share your concerns about getting heat out of a transistor when it is
in potting compound. The other question is just how much current does a
2N5109
need before you get little or no improvement in IMD performance.
Conventional wisdom says as much as possible. However in his article
'Wideband Amplifier
Summary', in the November 1979 issue of Ham Radio Magazine, Dr. Ulrich Rohde
suggested that for a 2N5109 that the IMD performance was best at a collector
current of ca. 10mA. He then went on: "A typical CATV transistor, such as
the
2N5109, has a flat curve of intermodulation distortion between 20 and 80mA."

I've always tended to follow the conventional wisdom, as have many major
manufacturers, but last year I set out to make some comparative measurements on
medium power devices, buying small batches of 2N3553s, 2N3866s, and 2N5109s
from the same UK supplier.
I was foolish enough to assume I was buying NOS devices but suspected when
they arrived that all had been recently branded from the same generic stock.
It took a while but I eventually got confirmation from the supplier that I was
right.
This practice has been going on for years, certainly it was common in the
60s, with both valves/tubes and transistors, among many of the "own brand", but
non manufacturing, companies in the UK so I should have known better and
asked first.
At least I didn't waste my time looking for non existent differences:-)

I will pursue this eventually, once my test gear comes out of storage, as I
do have enough genuine devices now to make it worth the effort but
diminishing stocks may make that more of an interesting exercise than of much practical
value.

Given the number of loops used mainly at LF I also suspect there's many
devices with lower ft that would be adequate in such circumstances but have been
ignored in the past.


Dallas Lankford has found pretty much the same using the MRF581A. You can
run
the MRF581A at 25mA collector current without a heatsink. Given that
MicroSemi/APT (ie non-Motorola) devices are cheaper than the 2N5109 on this
side of
the Pond, it would be my first choice of device.

I haven't seen those results but am certainly aware of genuine NOS Motorola
2N5109s hitting "premium" prices.

I've managed to stock up with what should be enough for my future needs, and
at what I consider a good price, certainly significantly better than the low
quantity prices being asked here, but still with an "ouch" factor buying in
quantity.
It'll be rather sad then if I eventually demonstrate that a 10 cent TO92
device could be just as good:-)

regards

Nigel
GM8PZR

Whooops........
My previous posting of this looked ok when sent but a total mess when it
arrived back from Yahoo.
I'm not saying changing it will make my ramblings any more sensible.....
but perhaps a bit more legible:-)
-----------------------------------------------------

In a message dated 06/05/2007 23:31:54 GMT Daylight Time, crabtreejr@...
writes:


Re. implications of the very low radiation resisitance, even at say 1 MHz -
that should be the subject of another posting. I need to give the matter
some
careful thought. I need to go through the maths, and get it right. An off
the cuff comment would not be appropriate.

-----------------------------------------------------------------
I need to give this further consideration too.
Given the relatively low input impedance of the amp I would have expected
that increase of radiation resistance with increasing frequency might have been
more of a limiting factor but suspect that's far too simplistic.
-----------------------------------------------------

Re the amplifier gain. If the antenna factor were unity, ie the antenna
output were 1 volt into 50 ohms for a field strength of 1 volt per meter,
then the
required amplifer gain would be 33dB. The actual antenna factor is about
-7dB, ie ca. 0.45, ie the antenna output is ca. 0.45 volts for a field
strength
of 1 volt per meter. The required antenna gain is then +33 - 7 = 26dB.

------------------------------------------------------------------------------
------------
That's what I assumed you meant before but, as previously written, it didn't
seem to make sense, with confusion perhaps mainly being due to the use of
"required".
I'd have probably still have written it here "the actual amplifier gain is
then.....26dB", rather than "required" but perhaps that's just me being
pedantic as usual:-)
------------------------------------------------------------------------------
-----------

I too share your concerns about getting heat out of a transistor when it is
in potting compound. The other question is just how much current does a
2N5109
need before you get little or no improvement in IMD performance.
Conventional wisdom says as much as possible. However in his article
'Wideband Amplifier
Summary', in the November 1979 issue of Ham Radio Magazine, Dr. Ulrich Rohde
suggested that for a 2N5109 that the IMD performance was best at a collector
current of ca. 10mA. He then went on: "A typical CATV transistor, such as
the
2N5109, has a flat curve of intermodulation distortion between 20 and 80mA."

-----------------------------------------------------------------------------
I've always tended to follow the conventional wisdom, as have many major
manufacturers, but last year I set out to make some comparative measurements on
medium power devices, buying small batches of 2N3553s, 2N3866s, and 2N5109s
from the same UK supplier.
I was foolish enough to assume I was buying NOS devices but suspected when
they arrived that all had been recently branded from the same generic stock.
It took a while but I eventually got confirmation from the supplier that I was
right.
This practice has been going on for years, certainly it was common in the
60s, with both valves/tubes and transistors, among many of the "own brand", but
non manufacturing, companies in the UK so I should have known better and
asked first.
At least I didn't waste my time looking for non existent differences:-)

I will pursue this eventually, once my test gear comes out of storage, as I
do have enough genuine devices now to make it worth the effort but
diminishing stocks may make that more of an interesting exercise than of much practical
value.

Given the number of loops used mainly at LF I also suspect there's many
devices with lower ft that would be adequate in such circumstances but have been
ignored in the past.
------------------------------------------------------------------------------
-------------------------

Dallas Lankford has found pretty much the same using the MRF581A. You can
run
the MRF581A at 25mA collector current without a heatsink. Given that
MicroSemi/APT (ie non-Motorola) devices are cheaper than the 2N5109 on this
side of
the Pond, it would be my first choice of device.

------------------------------------------------------------------------------
------------------------
I haven't seen those results but am certainly aware of genuine NOS Motorola
2N5109s hitting "premium" prices.

I've managed to stock up with what should be enough for my future needs, and
at what I consider a good price, certainly significantly better than the low
quantity prices being asked here, but still with an "ouch" factor buying in
quantity.
It'll be rather sad then if I eventually demonstrate that a 10 cent TO92
device could be just as good:-)

regards

Nigel
GM8PZR

In a message dated 08/05/2007 00:42:07 GMT Daylight Time,
mcqueen_34@... writes:
Potting compound actually helps the heat issue by
creating more mass that the transistor must heat and
providing more surface area to radiate the heat it
does produce.

------------------------------------------------------------

Sorry to ruin your fantasy, but, in the majority of cases, if not all, that
doesn't make any sense whatsoever.
Potting compounds generally have relatively poor thermal conductivity and
the heat is much more likely to build up in the vicinity of potted devices
rather than be carried away to the surface.
Having more mass to heat, without that mass being able to conduct the heat
elsewhere, just leads to localised temperature rise and potential destruction.
In other words, your potting compound is likely to be a good insulator that
keeps the heat in rather than a good conductor that lets it out.
Heat what can't reach the surface can't radiate so the enhanced surface area
in that case means sweet FA.

It isn't so much mass that's important as being able to conduct the heat
away quickly and then disperse it over an increased surface area which can
radiate it more effectively.
That's why heatsinks are made of metal and exposed to free or forced air
supplies and why efforts are made to ensure the thermal resistance between
device and heatsink is minimised.

Fluid cooling is, of course, another option but I don't think potting
compound cooling is ever gonna catch on:-)

regards

Nigel
GM8PZR

"I too share your concerns about getting heat out of a
transistor when it is in potting compound."

Potting compound actually helps the heat issue by
creating more mass that the transistor must heat and
providing more surface area to radiate the heat it
does produce.

--- gandalfg8@... wrote:

In a message dated 06/05/2007 23:31:54 GMT Daylight
Time, crabtreejr@...
writes:



Re. implications of the very low radiation
resisitance, even at say 1 MHz -
that should be the subject of another posting. I
need to give the matter
some
careful thought. I need to go through the maths, and
get it right. An off
the cuff comment would not be appropriate.
I need to give this further consideration too.
Given the relatively low input impedance of the amp
I would have expected
that increase of radiation resistance with
increasing frequency might have been
more of a limiting factor but suspect that's far
too simplistic.



Re the amplifier gain. If the antenna factor were
unity, ie the antenna
output were 1 volt into 50 ohms for a field strength
of 1 volt per meter,
then the
required amplifer gain would be 33dB. The actual
antenna factor is about
-7dB, ie ca. 0.45, ie the antenna output is ca. 0.45
volts for a field
strength
of 1 volt per meter. The required antenna gain is
then +33 - 7 = 26dB.
That's what I assumed you meant before but, as
previously written, it didn't
seem to make sense, with confusion perhaps mainly
being due to the use of
"required".
I'd have probably still have written it here "the
actual amplifier gain is
then.....26dB", rather than "required" but perhaps
that's just me being
pedantic as usual:-)




I too share your concerns about getting heat out of
a transistor when it is
in potting compound. The other question is just how
much current does a
2N5109
need before you get little or no improvement in IMD
performance.
Conventional wisdom says as much as possible.
However in his article
'Wideband Amplifier
Summary', in the November 1979 issue of Ham Radio
Magazine, Dr. Ulrich Rohde
suggested that for a 2N5109 that the IMD
performance was best at a collector
current of ca. 10mA. He then went on: "A typical
CATV transistor, such as
the
2N5109, has a flat curve of intermodulation
distortion between 20 and 80mA."

I've always tended to follow the conventional
wisdom, as have many major
manufacturers, but last year I set out to make some
comparative measurements on
medium power devices, buying small batches of
2N3553s, 2N3866s, and 2N5109s
from the same UK supplier.
I was foolish enough to assume I was buying NOS
devices but suspected when
they arrived that all had been recently branded from
the same generic stock.
It took a while but I eventually got confirmation
from the supplier that I was
right.
This practice has been going on for years, certainly
it was common in the
60s, with both valves/tubes and transistors, among
many of the "own brand", but
non manufacturing, companies in the UK so I should
have known better and
asked first.
At least I didn't waste my time looking for non
existent differences:-)

I will pursue this eventually, once my test gear
comes out of storage, as I
do have enough genuine devices now to make it worth
the effort but
diminishing stocks may make that more of an
interesting exercise than of much practical
value.

Given the number of loops used mainly at LF I also
suspect there's many
devices with lower ft that would be adequate in
such circumstances but have been
ignored in the past.


Dallas Lankford has found pretty much the same using
the MRF581A. You can
run
the MRF581A at 25mA collector current without a
heatsink. Given that
MicroSemi/APT (ie non-Motorola) devices are cheaper
than the 2N5109 on this
side of
the Pond, it would be my first choice of device.

I haven't seen those results but am certainly aware
of genuine NOS Motorola
2N5109s hitting "premium" prices.

I've managed to stock up with what should be enough
for my future needs, and
at what I consider a good price, certainly
significantly better than the low
quantity prices being asked here, but still with an
"ouch" factor buying in
quantity.
It'll be rather sad then if I eventually
demonstrate that a 10 cent TO92
device could be just as good:-)

regards

Nigel
GM8PZR











[Non-text portions of this message have been
removed]

__________________________________________________
Do You Yahoo!?
Tired of spam? Yahoo! Mail has the best spam protection around

Don't want to get into a p...ing contest but thermally
conductive potting compounds generally have Thermal
Conductivity of 10 Btu-in/oF- ft. Normal potting
materials are 2 Btu-in/oF- ft when filled with calcium
carbonate and 1 Btu-in/oF- ft when unfilled. Comparing
that to air at 0.01 Btu-in/oF- ft. Still air is a
perfect insulator (It's the way fiberglass insulation
works)

Since it is unlikely that the loop amplifier uses
forced air cooling and is typically enclosed in an air
tight weatherproof enclosure; potting is a great
solution to disipate the heat while ensuring a
weatherproof seal.



--- gandalfg8@... wrote:

In a message dated 08/05/2007 00:42:07 GMT Daylight
Time,
mcqueen_34@... writes:
Potting compound actually helps the heat issue by
creating more mass that the transistor must heat and
providing more surface area to radiate the heat it
does produce.

------------------------------------------------------------

Sorry to ruin your fantasy, but, in the majority of
cases, if not all, that
doesn't make any sense whatsoever.
Potting compounds generally have relatively poor
thermal conductivity and
the heat is much more likely to build up in the
vicinity of potted devices
rather than be carried away to the surface.
Having more mass to heat, without that mass being
able to conduct the heat
elsewhere, just leads to localised temperature rise
and potential destruction.
In other words, your potting compound is likely to
be a good insulator that
keeps the heat in rather than a good conductor that
lets it out.
Heat what can't reach the surface can't radiate so
the enhanced surface area
in that case means sweet FA.

It isn't so much mass that's important as being able
to conduct the heat
away quickly and then disperse it over an increased
surface area which can
radiate it more effectively.
That's why heatsinks are made of metal and exposed
to free or forced air
supplies and why efforts are made to ensure the
thermal resistance between
device and heatsink is minimised.

Fluid cooling is, of course, another option but I
don't think potting
compound cooling is ever gonna catch on:-)

regards

Nigel
GM8PZR

____________________________________________________________________________________
Never miss an email again!
Yahoo! Toolbar alerts you the instant new Mail arrives.

In a message dated 08/05/2007 03:12:14 GMT Daylight Time,
mcqueen_34@... writes:
Don't want to get into a p...ing contest but thermally
conductive potting compounds generally have Thermal
Conductivity of 10 Btu-in/oF- ft. Normal potting
materials are 2 Btu-in/oF- ft when filled with calcium
carbonate and 1 Btu-in/oF- ft when unfilled. Comparing
that to air at 0.01 Btu-in/oF- ft. Still air is a
perfect insulator (It's the way fiberglass insulation
works)

Since it is unlikely that the loop amplifier uses
forced air cooling and is typically enclosed in an air
tight weatherproof enclosure; potting is a great
solution to disipate the heat while ensuring a
weatherproof seal.

----------------------------------------
In the cold light of day my earlier response looks harsh and somewhat rude,
sorry about that.

I should have explained my terms of reference better.
I've used some compounds specifically for their heat transfer properties but
the more efficient of these were generally of low to medium viscosity.
We'd been specifically discussing hard setting compounds, used here for
their ability to conceal and difficulty to remove.
They fill the case, in those I've looked at anyway, enough to conceal the
components with some overlap but not with any attempt at completely filling it.
I still doubt strongly that this compound, I'll keep it a bit more
specific:-), would aid heat transfer, and the outer casing is plastic anyway and
doesn't exhibit any obvious rise in temperature during use.

If the transistors, or any other point source, were generating significant
heat in this unit I would be concerned about reliability.

Air cooling, of course, even in "still" air, does occur, through convection,
which means that the air doesn't remain "still" if not constrained, so the
fibreglass analogy is cheating a bit:-)

regards

Nigel
GM8PZR

In a message dated 08/05/2007 22:41:16 GMT Daylight Time,
kbrowning@... writes:
I f the potting is being used primarily for protection, please think again.
There are other products that will offer possibly better protection, yet
allow repairs or changes to
be made.

---------------------------------------------------------
Hi Keith

In the instance we were discussing the manufacturer, Wellbrook, is using a
hard potting compound to prevent investigation of their products.
I agree a conformal coating would be more user friendly but that's quite the
opposite of what they intend:-)
--------------------------------------------------------

Here in New Zealand, my enquiries for conformal coatings either bring on a
blank expression or
various other unsatisfactory products.
I offer this advice in case this option is not known, AND more to the point
does any one know what
the product is and where it may be purchased.

----------------------------------------------------
As you comment the term "conformal coating" is generic and describes a type
of product rather than one specific product.
One manufacturer's definition is given here......
_.
asp_
()

I'm not sure why you should be having so much trouble locating product, Dow
Corning is just one manufacturer of such products and a google search brings
up a lot of information.
In the past I've bought conformal coating products from Farnell and RS, both
distributors in the UK, and I'm sure there would have been plenty of other
outlets too.

Both also used to carry a spray on transparent plastic coating for PCBs in
an aerosol can, and I've used this succesfully too on PCBs having some
atmosperic exposure.
PCBs could be reworked just by using a hot soldering iron to melt through
the plastic coating....although the smell was a bit off putting at times:-)
I don't know if either are still available but both catalogues can be
browsed online and google again will find them.

regards

Nigel
GM8PZR

I f the potting is being used primarily for protection, please think again.
There are other products that will offer possibly better protection, yet allow repairs or changes to
be made.

One product comes under the generic name of conformal coating.
Through my work, I see and repair a lot of dog training and tracking equipment.
At least one brand use a clear material that looks to be poured over the boards to be protected and
has set to a rubbery texture very similar to coating it with clear, non acidic silicone sealant. The
result is almost identical.
It can be easily picked off to access components as it tends to fracture if stretched enough. I have
found CRC Electra-clean will make it crumble and able to be brushed off with a stiff tooth brush.
(I'm not advocating using CRC-Electra clean haphazardly)
This coating offers excelent protection as displayed with some of the damaged items I service.

I have tried to find the source of the product, but the items, while theoreticaly are made in the
USA, actually originate in Korea.
The people I deal with in the USA have no idea what the product is, other than "conformal coating".

Here in New Zealand, my enquiries for conformal coatings either bring on a blank expression or
various other unsatisfactory products.
I offer this advice in case this option is not known, AND more to the point does any one know what
the product is and where it may be purchased.

Regards
Keith browning
----------------------------------------


Since it is unlikely that the loop amplifier uses
forced air cooling and is typically enclosed in an air
tight weatherproof enclosure; potting is a great
solution to disipate the heat while ensuring a
weatherproof seal.

----------------------------------------
In the cold light of day my earlier response looks harsh and somewhat rude,
sorry about that.

I should have explained my terms of reference better.
I've used some compounds specifically for their heat transfer properties but
the more efficient of these were generally of low to medium viscosity.
We'd been specifically discussing hard setting compounds, used here for
their ability to conceal and difficulty to remove.
They fill the case, in those I've looked at anyway, enough to conceal the
components with some overlap but not with any attempt at completely filling it.
I still doubt strongly that this compound, I'll keep it a bit more
specific:-), would aid heat transfer, and the outer casing is plastic anyway and
doesn't exhibit any obvious rise in temperature during use.

If the transistors, or any other point source, were generating significant
heat in this unit I would be concerned about reliability.

Air cooling, of course, even in "still" air, does occur, through convection,
which means that the air doesn't remain "still" if not constrained, so the
fibreglass analogy is cheating a bit:-)

regards

Nigel
GM8PZR


















If you've got links, post them in the Links section!


Post files here. If the file comes from a website, please put it in the Links rather than uploading
the file.


You can now view images at higher resolution in Photos. Upload JPG and GIF into Photos. Please
convert BMP or TIF to JPG or GIF before uploading.


And please trim all this when replying!
Yahoo! Groups Links





__________ NOD32 2249 (20070508) Information __________

This message was checked by NOD32 antivirus system.

PLEASE!
What is this "Wellbrook ALA1530-corrected" continuance?
You, I suppose almost all of you, have the experience and wisdom to produce a lot better
Active Loop Antenna, but obviosly have not realised your full potencial to do it, to do it together?!
Why to loose your precios time to concentrate on irrelevant details?
Why not see your possibility to create something new?
Man was created to be a little creator, was he not?
According to the image of the Creator!
Why just try to study from month to month something out of our reach, something hidden, produced to make a few dollars or sterling pounds?
Why not try to do something worth while together for common benefit, just for fun?
To give and to be given.
Wasn't this group grounded for that purpose, yes or no?
Why are we here?
To have satisfaction doing things to be enjoyed together, or just to earn money?
Let us study our attitudes, brothers....................................................

gandalfg8@... wrote:

Dear Aimo

because this particular device is such a good performer some of us
want to understand why

clearly we all want to invent new things, for sure we want to do this

but for many of us, we are curious why the Wellbrook performs so
well for a relatively small capture area

it is this we would prefer to understand for fully from the antenna
theory to the electronics

to go ahead we must understand what we have in the present
right now I suspect only a few of us do

my 2 cents worth
Paul

Hello Rob

The left hand scale is in dB, and is the correction which would need to be
applied to get a unity antenna factor. So the actual performance is the inverse
of the scale. One would expect the actual performance to decrease with
frequency.

So the actual antenna factor is ca. -7dB or 0.45 (ie 10^(-7/20)) over most of
the hf frequencies.

The cross check is the value of 0.5 which Andy Ikin gave in his article in
Medium Wave News in March 1998. See:
A. Ikin, Broadband Loop Aerials (part 1), Medium Wave News, Jan 1998, pp
13-16
A. Ikin, Broadband Loop Aerials (part 2), Medium Wave News, Mar 1998, pp 9-12

Medium Wave News is published by the Medium Wave Circle. See:
www.mwcircle.org
I have not checked, but suspect that these articles are not available online.

You will note that a couple of the other Wellbrook loops have a slightly
higher gain, with a quoted antenna facro of 1.0 and/or 0dB.

HTH and 73

John KC0G



In a message dated 5/15/07 1:54:28 PM Central Daylight Time,
rmoore5@... writes:

John,

At the bottom of the Wellbrook ALA-1530 information page, there is a
calibration chart which show the Aantenna factor vs frequency. It
indicates that the antenna factor is roughly 8 over most of the range.

How did you come up with a .45 factor?

Rob

--- In loopantennas@..., crabtreejr@... wrote:

The actual antenna factor is about -7dB, ie ca. 0.45, ie the antenna

output is ca. 0.45 volts for a field strength of 1 volt per meter. The
required antenna gain is then +33 - 7 = 26dB.

***
See what's free at .