Variable inductor

[John Popelish]

It's for reactively coupling a 200 Ohm output impedance amplifier, into a 50
Ohm load.
Although many interesting comments have been made (and thanks for all of you
who've contributed), the acceptable solution is to change the inductor for
different frequency ranges.

Have you considered using a 4:1 transformer (or a transmission line
transformer) to do most of the matching, so that tuning would not be
necessary?

 

[Ken Smith]

Ken,



I realize I actually have a fundamental misunderstanding of how this is
supposed to work.

Note that I said "two inductors". The voltages in the lower circuit
cancel (to a first order) so that the control doesn't have to have a high
impedance.

BTW: If you do this with a couple of 12V transformers, you can control
the brightness of a 25W light bulb with just a flashligh battery.

 

[Phil Allison]

"John Popelish"

Have you considered using a 4:1 transformer (or a transmission line
transformer) to do most of the matching, so that tuning would not be
necessary?

** Of course he has not !

The know nothing fuckwit TROLL is having a wonderful time *pissing on* his
technical superiors and sucking up to all the usual gullible NG dupes.


Mooooooooooo.............




....... Phil

 

[Joel Kolstad]

Note that I said "two inductors".

Ah... those are actually two inductors up there so the fields don't couple; I
did miss that -- thanks for the correction.

---Joel

 

[Grumps]

Have you considered using a 4:1 transformer (or a transmission line
transformer) to do most of the matching, so that tuning would not be
necessary?

Yeah, I had a look at some transformer offerings from Mini-Circuits. Maybe
I'm wrong, and this is why I'm posting here, but to convert 200 to 50 Ohms
impedance would require a turns ratio of 2 ( Zpri = n^2 Zsec ). This would
produce a 3dB loss, and then you have about 1dB insertion loss too.
Simulations of the tuned LC approach have a loss of <2dB. I'd like to save
every dB that I can.

 

[Joel Kolstad]

Yeah, I had a look at some transformer offerings from Mini-Circuits. Maybe
I'm wrong, and this is why I'm posting here, but to convert 200 to 50 Ohms
impedance would require a turns ratio of 2 ( Zpri = n^2 Zsec ). This would
produce a 3dB loss

Where do you think that 3dB loss is coming from? Ideal transformers have no
losses; real have have the dB or so insertion loss you describe.

A transformer is just an inductor or two or so where the coupling is very
close to unity, after all...

 

[[email protected]]

John Popelish wrote:

Yeah, I had a look at some transformer offerings from Mini-Circuits. Maybe
I'm wrong, and this is why I'm posting here, but to convert 200 to 50 Ohms
impedance would require a turns ratio of 2 ( Zpri = n^2 Zsec ). This would
produce a 3dB loss, and then you have about 1dB insertion loss too.
Simulations of the tuned LC approach have a loss of <2dB. I'd like to save
every dB that I can.

I think you aremistaken. ant transformation from 200 ohms to 50 ohms
must drop the voltage by at least 6 db (half) , but it also must raise
the current by almost 6 db (double), so the only loss, from a power
standpoint, is the insertion loss.

This transformer looks like a fair match to your application (about 1/2
db insertion loss over 50 to 100 MHz), if the power capability is
acceptable:
http://www.coilcraft.com/misc/wbtf4.html
http://www.coilcraft.com/wbt.cfm

 

[Chris Jones]

Yeah, I had a look at some transformer offerings from Mini-Circuits. Maybe
I'm wrong, and this is why I'm posting here, but to convert 200 to 50 Ohms
impedance would require a turns ratio of 2 ( Zpri = n^2 Zsec ). This would
produce a 3dB loss, and then you have about 1dB insertion loss too.
Simulations of the tuned LC approach have a loss of <2dB. I'd like to save
every dB that I can.

With a transformer, you can basically get nearly all of the power
transferred from the source to the load, with a good transformer certainly
within less than a dB of no loss at all. If you were able to get 2dB more
with your reactive matching circuit than would be available from the
transformer, then you would have to have invented some kind of free energy
perpetual motion thingy.

If you want to develop 1 Volt RMS across a 50 Ohm load, you need an amount
of power equal to:
V^2/R = 20mW

Now if you make this 50 Ohm load look like a 200 Ohm load from the point of
view of the signal source (because you attached some fancy matching
network, or a transformer) then you still need 20mW from the signal source,
you can't get power out of nowhere.

For your source to provide 20mW into 200 Ohm, you will need 2V:
((2V)^2)/200Ohms = 20mW.

There is no way to avoid this voltage ratio without some kind of amplifier
having its own power source.

Chris

 

[Grumps]

With a transformer, you can basically get nearly all of the power
transferred from the source to the load, with a good transformer
certainly within less than a dB of no loss at all. If you were able
to get 2dB more with your reactive matching circuit than would be
available from the transformer, then you would have to have invented
some kind of free energy perpetual motion thingy.

If you want to develop 1 Volt RMS across a 50 Ohm load, you need an
amount of power equal to:
V^2/R = 20mW

Now if you make this 50 Ohm load look like a 200 Ohm load from the
point of view of the signal source (because you attached some fancy
matching network, or a transformer) then you still need 20mW from the
signal source, you can't get power out of nowhere.

For your source to provide 20mW into 200 Ohm, you will need 2V:
((2V)^2)/200Ohms = 20mW.

There is no way to avoid this voltage ratio without some kind of
amplifier having its own power source.

You are all, of course, correct. I must've had a brain fart.
So I wonder why the datasheet for the amp in question (AD8367) suggests an
LC rather than a transformer?

 

[Chris Jones]

You are all, of course, correct. I must've had a brain fart.
So I wonder why the datasheet for the amp in question (AD8367) suggests an
LC rather than a transformer?

If you only want it to work at one frequency, LC or transformer would both
be OK and the LC might be cheaper. For a wide range of frequencies the
transformer is much easier to use.

Chris

 

[[email protected]]

You are all, of course, correct. I must've had a brain fart.
So I wonder why the datasheet for the amp in question (AD8367) suggests an
LC rather than a transformer?

It may have something to do with the input having a lot of reactance,
andnot being a purely resistive 200 ohm inpedance. See figure 15 on
page 9 of:
http://www.analog.com/UploadedFiles/Data_Sheets/477808511AD8367_a.pdf

A resonant impedance conversion takes thisreactance into account, while
a transformer would pass this reactance upstream (though transformed by
the transformer's impedance ratio).

 

[[email protected]]

The inductor can be tapped and the taps switched by pin diodes..
Made a n oscillator (for a synthesiser) years ago with varicaps to
interpolate the tuning between tapped steps. The taps were switched
when the varicap control voltage was above or below a window value.
An alternative to a multiloop pll synthesiser for wide band., cant
recall ssb noise results or total tuning range.

 

[Michael A. Terrell]

I'm trying to remember how much variability coils with *copper* slugs
had (or were they brass?)

I've haven't seen them since the early days of TV sets (mid '50's).

...Jim Thompson

Have you ever looked at the variable inductor tuner used by Delco and
some other car radio manufacturers in the '60s & '70s? It tuned from <
540 KHz to >1600 KHz. The LO was offset by + 262 KHz, and tracked with
the tuned RF sections. That is a 3 to 1 tuning range.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[Winfield Hill]

Michael A. Terrell wrote...

Have you ever looked at the variable inductor tuner used by Delco and
some other car radio manufacturers in the '60s & '70s? It tuned from
<540 KHz to >1600 KHz. The LO was offset by + 262 KHz, and tracked
with the tuned RF sections. That is a 3 to 1 tuning range.

That's right, I had forgotten those radios with inductor tuning.
Didn't they use a long coil with a ferrite rod that was pulled
through on a string, or something like that?

 

[Michael A. Terrell]

Michael A. Terrell wrote...

That's right, I had forgotten those radios with inductor tuning.
Didn't they use a long coil with a ferrite rod that was pulled
through on a string, or something like that?

Yes, but Delco used a rack and pinion mechanism to allow the use of
pushbuttons. I checked several of my Sams AR series manuals, but
couldn't find a decent photo of the tuner that showed the coils. I know
someone who repairs car radios and I can see if I can get some pictures
for you if you need them.

It was a long thin coil with a long ferrite slug with a threaded
brass insert that was used to align the tuner at the bottom end of the
dial, and trimmer caps for the top end. A string was used on some
radios to pull the pointer across the dial but most were mechanical
pointers from the rack & pinion drive.

Fair Radio used to have new surplus inductive tuners from AM pay
radios made for hotel rooms for about $20. I don't know what they have
these days.


--
Service to my country? Been there, Done that, and I've got my DD214 to
prove it.
Member of DAV #85.

Michael A. Terrell
Central Florida

 

[joseph2k]

Yes, but Delco used a rack and pinion mechanism to allow the use of
pushbuttons. I checked several of my Sams AR series manuals, but
couldn't find a decent photo of the tuner that showed the coils. I know
someone who repairs car radios and I can see if I can get some pictures
for you if you need them.

It was a long thin coil with a long ferrite slug with a threaded
brass insert that was used to align the tuner at the bottom end of the
dial, and trimmer caps for the top end. A string was used on some
radios to pull the pointer across the dial but most were mechanical
pointers from the rack & pinion drive.

Fair Radio used to have new surplus inductive tuners from AM pay
radios made for hotel rooms for about $20. I don't know what they have
these days.

Poxy hell. I seem to remember an old Philco B&W tv with a rotary switch
thingy to select channels with ferrite slugs for VHF low band and brass
slugs for VHF high band with two or three slugs per channel (that had to be
tuned before sales).
My, how much the relative cost between labor and materials has changed.

 

[Phat Bytestard]

Poxy hell. I seem to remember an old Philco B&W tv with a rotary switch
thingy to select channels with ferrite slugs for VHF low band and brass
slugs for VHF high band with two or three slugs per channel (that had to be
tuned before sales).
My, how much the relative cost between labor and materials has changed.

Electronics. It gets better and better, and cheaper to make and
cheaper to buy with no declination in quality. (ideally)

It is one of the last businesses that follow the American dream.

Some sleazeball landlord, or stickitinyourass lawyer certainly
doesn't.