Hartley oscillator stability

[dave vanhorn]

I need a little advice.

I'm building a design that was done using an NPN bipolar in a hartley
configuration, with a ferrite tuned coil, running between 700kHz and
1.1MHz
There are four identical oscillators, built and used as two pairs.

Schematic is at http://www.paia.com/theresch.htm

It's having several problems, but I'd like to fix them without a major
redesign.

First, is stability. The osc drifts a fair bit.
This osc is being mixed with another identical osc, and I was hoping
that component thermal variance would be minimized, but I can actually
detect blowing on the circuit from 3' away..

The osc is also hard to tune, just a tiny twitch on the coil and
you're past the proper point. I'd like to tame that down to even as
little as 1/8 or 1/4 turn.

The other, is some odd behaviours that I'm seeing.

1: jumps in frequency, maybe 10-400Hz. These occur at random times,
and are not "wobbles" so much as discrete steps.

The input voltage is shunt zener regulated, and the zener is quite
cool.
I may need to increase current into the zener to make it more stable,
but I don't think this is the problem.

2: When the two oscillators are near the same frequency, I can see
large changes in the waveforms of both oscillators, and in the mix
product.
I can email a scope capture to those interested.

The mixer is a simple pair of diodes, and I am feeding in the output
of the oscillator coil tap, through 2.2k. I've tried as high as 10k,
which reduces the level, but does not particularly help. Note the
mixers are not implemented as shown in this schematic, they use a pair
of diodes only. I've even gone as far as to use matched pair diodes
here.

I also tried 2N4401 transistors, but no noticable difference.

The inductor is a mouser part, 42IF110, 760 kHz

I added an additional 0.01uF across the base bypass cap, which made
the control wiring relatively immune to "hand effect", but had no
effect on these other problems.

The board layout looks good, and the board is mounted over a ground
plane, with rigid standoffs, and wiring is taken away from the
oscillators, not laying over them..

There are no antennas attached, so we are just talking circuit issues,
nothing beyond the PCB or power supply should be a problem.

 

[Joe McElvenney]

Hi,

About your random frequency jumps, are any of the oscillator
capacitors silver-mica as they are known to cause this problem?
This was the subject of a discussion in one of the news-groups
just recently.

As for the changes in waveform, I would guess you have
insufficient isolation between the oscillators and the mixer
circuits - i.e. they are trying to sync-up each other as they
approach the same frequency.


Cheers - Joe

 

[colin]

Hi,

About your random frequency jumps, are any of the oscillator
capacitors silver-mica as they are known to cause this problem?
This was the subject of a discussion in one of the news-groups
just recently.

As for the changes in waveform, I would guess you have
insufficient isolation between the oscillators and the mixer
circuits - i.e. they are trying to sync-up each other as they
approach the same frequency.


Cheers - Joe

hmm im having similar problems but with twin crystal osilators as i posted
..., some of the things i found might be usefull ... , are the inductors
screened or beter still toroidal (although u cant tune those without
unwinding a turn )? are the caps polystyrene? rather than ceramic ? is there
any flux anywhere in sight ? is the whole thing in a screened enclosure? are
the coils firmly secured or does knocking it cuase it to jump?

i would of said small heatsinks on the transistors might help with temp
stability although with those 56k emiter resisters i doubt they disipate
anything, but if your using a zener regulated supply i sguest u ditch that
straight away and use a lm78xx, or at least pre regulate the supply to the
zener, do any of the other components gethot and might heat up the
transistors unevenly ? or is there anything that might be cuasing cross
interference ? (signals with high speed edges for example.)

are u using metal film resistors or beter?

interesting about silver mica tho i never knew that, (looks with lower
regard on box of silver mica caps removed from eqpmnt that seldom get used
anyway)

if u want to reduce the tuning sensitivity u can arange it so the ferrite
plug is closer to the center of its travel where it usualy has minimal
efect, by either reducing the capacitance or inducatnce (try spreading the
coils out a bit by or pulling some of the wire off, of course this might
mean u cant tune it to the desired freq anymore, but u can just usualy
unwind a bit without actualy disconecting or breaking it, try a little bit
at a time, but if it has many turns might need to pull quite a bit)

or maybe if you lowered the frequency of both oscilators in the pair then
the diference wld be less afected by change.

Colin =^.^=

 

[Fred Bloggs]

I need a little advice.

I'm building a design that was done using an NPN bipolar in a hartley
configuration, with a ferrite tuned coil, running between 700kHz and
1.1MHz
There are four identical oscillators, built and used as two pairs.

Schematic is at http://www.paia.com/theresch.htm

It's having several problems, but I'd like to fix them without a major
redesign.

You will never get this circuit to work with that transformer-IIRC the
tempco of those cheap things is running at several 1000 ppm and the Q is
not particularly high so that relatively small external component phase
shifts cause relatively large shifts in tuned tank frequency. As for
tuning, you have (d/dL)F= -Fo/L so that you have ~760K/360u or 2KHz/uH
so that even 1%/turn gets about 760Hz per 1/10th turn- I'm sure it is
larger- Then those jumps are caused by complex thermal hysteresis
interaction with core loss vs frequency vs amplitude vs large tempco. As
an IF transformer the component is perfectly satisfactory- just don't
get excited over center frequency shifts of a few KHz or so- but as a
tuning element for an oscillator requiring stability to 10Hz say- which
is 10/760E3 x 1E6= 13ppm- you can forget it.

 

[John Jardine]

I need a little advice.

I'm building a design that was done using an NPN bipolar in a hartley
configuration, with a ferrite tuned coil, running between 700kHz and
1.1MHz
There are four identical oscillators, built and used as two pairs.

Schematic is at http://www.paia.com/theresch.htm

It's having several problems, but I'd like to fix them without a major
redesign.

First, is stability. The osc drifts a fair bit.
This osc is being mixed with another identical osc, and I was hoping
that component thermal variance would be minimized, but I can actually
detect blowing on the circuit from 3' away..

The osc is also hard to tune, just a tiny twitch on the coil and
you're past the proper point. I'd like to tame that down to even as
little as 1/8 or 1/4 turn.

The other, is some odd behaviours that I'm seeing.

1: jumps in frequency, maybe 10-400Hz. These occur at random times,
and are not "wobbles" so much as discrete steps.

The input voltage is shunt zener regulated, and the zener is quite
cool.
I may need to increase current into the zener to make it more stable,
but I don't think this is the problem.

2: When the two oscillators are near the same frequency, I can see
large changes in the waveforms of both oscillators, and in the mix
product.
I can email a scope capture to those interested.

The mixer is a simple pair of diodes, and I am feeding in the output
of the oscillator coil tap, through 2.2k. I've tried as high as 10k,
which reduces the level, but does not particularly help. Note the
mixers are not implemented as shown in this schematic, they use a pair
of diodes only. I've even gone as far as to use matched pair diodes
here.

I also tried 2N4401 transistors, but no noticable difference.

The inductor is a mouser part, 42IF110, 760 kHz

I added an additional 0.01uF across the base bypass cap, which made
the control wiring relatively immune to "hand effect", but had no
effect on these other problems.

The board layout looks good, and the board is mounted over a ground
plane, with rigid standoffs, and wiring is taken away from the
oscillators, not laying over them..

There are no antennas attached, so we are just talking circuit issues,
nothing beyond the PCB or power supply should be a problem.

When running close in frequency the oscillators will naturally try and
'lock-in step' or synchronise with one another. This is evidenced by a
linear increase in 2nd harmonic distortion as the oscillators get closer
together in frequency and by sudden jumps in frequency as the oscillators
jump in and out of synchronisation. The jumps and lock ups are initiated
from any one of an inordinate number of what would be normally insignificant
effects. This is a standard problem but has non-trivial solutions and
there's a whole list of 'fixes' to go at.
Up at the top of the list, the oscillators need to be electrically and
physically identical. Each oscillator need a buffer (or two!) adding. Each
oscillator pair need its own heat isolation enclosure but seperately
screened within it.
The Theremins are working at audio frequency differences which maybe implies
a rough second to second stability of 1 part in 10,000 from the 750k
oscillators. This is OK for one osc' but actually a tall order for a pair of
them running into each other.
Be prepared for much tinkering

 

[Tim Wescott]

I need a little advice.

I'm building a design that was done using an NPN bipolar in a hartley
configuration, with a ferrite tuned coil, running between 700kHz and
1.1MHz
There are four identical oscillators, built and used as two pairs.

Schematic is at http://www.paia.com/theresch.htm

It's having several problems, but I'd like to fix them without a major
redesign.

First, is stability. The osc drifts a fair bit.
This osc is being mixed with another identical osc, and I was hoping
that component thermal variance would be minimized, but I can actually
detect blowing on the circuit from 3' away..

The osc is also hard to tune, just a tiny twitch on the coil and
you're past the proper point. I'd like to tame that down to even as
little as 1/8 or 1/4 turn.

The other, is some odd behaviours that I'm seeing.

1: jumps in frequency, maybe 10-400Hz. These occur at random times,
and are not "wobbles" so much as discrete steps.

The input voltage is shunt zener regulated, and the zener is quite
cool.
I may need to increase current into the zener to make it more stable,
but I don't think this is the problem.

2: When the two oscillators are near the same frequency, I can see
large changes in the waveforms of both oscillators, and in the mix
product.
I can email a scope capture to those interested.

The mixer is a simple pair of diodes, and I am feeding in the output
of the oscillator coil tap, through 2.2k. I've tried as high as 10k,
which reduces the level, but does not particularly help. Note the
mixers are not implemented as shown in this schematic, they use a pair
of diodes only. I've even gone as far as to use matched pair diodes
here.

I also tried 2N4401 transistors, but no noticable difference.

The inductor is a mouser part, 42IF110, 760 kHz

I added an additional 0.01uF across the base bypass cap, which made
the control wiring relatively immune to "hand effect", but had no
effect on these other problems.

The board layout looks good, and the board is mounted over a ground
plane, with rigid standoffs, and wiring is taken away from the
oscillators, not laying over them..

There are no antennas attached, so we are just talking circuit issues,
nothing beyond the PCB or power supply should be a problem.

Along with all the other recommendations:

Think about using one common heterodyne oscillator instead of two. I
would make it crystal stabilized, probably with some common
microprocessor crystal like 4MHz. You can use a 74HC4060 for an
oscillator/divider to get you down to 1MHz, or you can use a colorburst
crystal for something more like 780kHz.

Use the extra circuit space to add _good_ buffer amplifiers to your
pick-up oscillators. It may not be a bad idea to divide those down as
well; using a TTL divider should do a good job of isolating them, if you
can get a good enough signal (heh heh) for the divider to work.

Finally, I didn't look at the details of your text but from Bloggs's
comments you're using IF cans for the oscillator coils. Amateur radio
practice is to use #6 mix iron-powder toroidal cores from Amidon (who
resells the Micro-Metals cores). Toroidal cores have very good
characteristics in oscillators, and the #6 mix is very temperature
stable. You'll have a big coil at those frequencies, though.

 

[Tam/WB2TT]

I need a little advice.

I'm building a design that was done using an NPN bipolar in a hartley
configuration, with a ferrite tuned coil, running between 700kHz and
1.1MHz

An additional comment would be that if you are using capacitors in your
oscillator tuned circuit that are ceramic, meant for bypassing, get rid of
them quick. Use ceramic NPO, dipped mica, or even mylar. I might also worry
about the slugs in your inductors moving.

Tam

 

[dave vanhorn]

Hi,

About your random frequency jumps, are any of the oscillator
capacitors silver-mica as they are known to cause this problem?
This was the subject of a discussion in one of the news-groups
just recently.

No, I specifically avoided SM for that reason.

As for the changes in waveform, I would guess you have
insufficient isolation between the oscillators and the mixer
circuits - i.e. they are trying to sync-up each other as they
approach the same frequency.

Different problem, and I have that as well, but this one is different.
Picture my desired output at about 50mV, with 200mV of a square wave,
at about 1/10 to 1/20th frequency added in.

No idea where this is coming from..

 

[Tim Shoppa]

I need a little advice.

I'm building a design that was done using an NPN bipolar in a hartley
configuration, with a ferrite tuned coil, running between 700kHz and
1.1MHz
There are four identical oscillators, built and used as two pairs.

Schematic is at http://www.paia.com/theresch.htm

You could've *told* us it's a theremin .

It's having several problems, but I'd like to fix them without a major
redesign.

First, is stability. The osc drifts a fair bit.
This osc is being mixed with another identical osc, and I was hoping
that component thermal variance would be minimized, but I can actually
detect blowing on the circuit from 3' away..

The oscillators have been designed to be unstable so that just
waving your hands will change the frequency. Hand capacitance is
only a few pF at most so that the components are chosen so that a few
pF of change will shift the frequency by kHz. Strays are of the order
of a few pF, too... I suspect you're seeing strays shifting about.

1: jumps in frequency, maybe 10-400Hz. These occur at random times,
and are not "wobbles" so much as discrete steps.

Capacitances (both stray and intended) can exhibit this sort of effect.

The inductor is a mouser part, 42IF110, 760 kHz

Is this the type used in the original design? I would think that the can
around the inductor would minimize the effect of hand capacitance,
which is what you want to maximize. Is the can grounded, left floating?

IF transformers aren't necessarily the most stable... perhaps the
most unstable part is that tuning core, which seems to be a fundamental
part of this design.

My gut feeling is that what you're seeing is entirely typical for this
type of theremin circuit.

Tim.

 

[John Woodgate]

I read in sci.electronics.design that dave vanhorn <[email protected]>

Different problem, and I have that as well, but this one is different.
Picture my desired output at about 50mV, with 200mV of a square wave,
at about 1/10 to 1/20th frequency added in.

No idea where this is coming from..

Isn't it the beat frequency between your two oscillators? If it's
amplitude is that big, I'd say that your mixer is seriously misbehaving.

 

[J M Noeding]

You could've *told* us it's a theremin .


The oscillators have been designed to be unstable so that just
waving your hands will change the frequency. Hand capacitance is
only a few pF at most so that the components are chosen so that a few
pF of change will shift the frequency by kHz. Strays are of the order
of a few pF, too... I suspect you're seeing strays shifting about.


Capacitances (both stray and intended) can exhibit this sort of effect.

what is the tapping of emitter vs collector? shouldprobably be around
15-20%

 

[dave vanhorn]

You will never get this circuit to work with that transformer-IIRC the

tempco of those cheap things is running at several 1000 ppm and the Q is
not particularly high so that relatively small external component phase
shifts cause relatively large shifts in tuned tank frequency. As for
tuning, you have (d/dL)F= -Fo/L so that you have ~760K/360u or 2KHz/uH
so that even 1%/turn gets about 760Hz per 1/10th turn- I'm sure it is
larger- Then those jumps are caused by complex thermal hysteresis
interaction with core loss vs frequency vs amplitude vs large tempco. As
an IF transformer the component is perfectly satisfactory- just don't
get excited over center frequency shifts of a few KHz or so- but as a
tuning element for an oscillator requiring stability to 10Hz say- which
is 10/760E3 x 1E6= 13ppm- you can forget it.

So in the short version, I'm hosed?

So, let's say I wanted to redo the design.
I want something that will wobble nicely with hand capacitance a
couple feet away from a 30" or so antenna rod, but with an actual
tuning range, no serious thermal issues, and not particularly
sensitive to VCC problems.

I was wondering about the transformers. They are sheilded, and the
sheild is grounded, but they are just too @#$% twitchy to use IMHO.

Can anyone reccomend a replacement that would do better here, but
retain the same basic circuit? I'm not opposed to winding my own, but
coil forms and such are more scarce than hen's teeth here in farm
country. (not counting toilet paper rolls and PVC pipe.)

I would like to push the operating frequency down, around 200kHz seems
ideal.
No idea why they put this in the middle of the AM BCB.

 

[Fred Bloggs]

So in the short version, I'm hosed?

So, let's say I wanted to redo the design.
I want something that will wobble nicely with hand capacitance a
couple feet away from a 30" or so antenna rod, but with an actual
tuning range, no serious thermal issues, and not particularly
sensitive to VCC problems.

I was wondering about the transformers. They are sheilded, and the
sheild is grounded, but they are just too @#$% twitchy to use IMHO.

Can anyone reccomend a replacement that would do better here, but
retain the same basic circuit? I'm not opposed to winding my own, but
coil forms and such are more scarce than hen's teeth here in farm
country. (not counting toilet paper rolls and PVC pipe.)

I would like to push the operating frequency down, around 200kHz seems
ideal.
No idea why they put this in the middle of the AM BCB.

Check out Micrometals for a low-mu temperature stable powered iron
toroid- I believe there is material in the 50ppm range. As an example of
how bad those ferrite interstage transformers are- the similar TOKO
types are 7000ppm/oC.

 

[J M Noeding]

So in the short version, I'm hosed?

So, let's say I wanted to redo the design.
I want something that will wobble nicely with hand capacitance a
couple feet away from a 30" or so antenna rod, but with an actual
tuning range, no serious thermal issues, and not particularly
sensitive to VCC problems.

I was wondering about the transformers. They are sheilded, and the
sheild is grounded, but they are just too @#$% twitchy to use IMHO.

Can anyone reccomend a replacement that would do better here, but
retain the same basic circuit? I'm not opposed to winding my own, but
coil forms and such are more scarce than hen's teeth here in farm
country. (not counting toilet paper rolls and PVC pipe.)

I would like to push the operating frequency down, around 200kHz seems
ideal.
No idea why they put this in the middle of the AM BCB.

LOL.
750kHz is an ideal frequency provided you know what you are doing, not
least knowing what sort of components you use, but if you want a
stable reference oscillator you could use ceramic resonator -
available on almost any frequency, see
http://www.noding.com/la8ak/12345/n19.htm

 

[Tim Wescott]

Check out Micrometals for a low-mu temperature stable powered iron
toroid- I believe there is material in the 50ppm range. As an example of
how bad those ferrite interstage transformers are- the similar TOKO
types are 7000ppm/oC.

Amidon carries the Micrometals stuff in the US, and they'll mailorder
anywhere there's a mailbox.

 

[James Arthur]

I read in sci.electronics.design that dave vanhorn <[email protected]>


Isn't it the beat frequency between your two oscillators? If it's
amplitude is that big, I'd say that your mixer is seriously misbehaving.

I'd expect a beat note to be sinusoidal and of comparable
amplitude, especially if the oscillators are latching onto
each other (as has already been suggested and a virtual certainty).

To me these amplitude-modulating 'jumps' sound like squegging,
an undesired oscillation on top of the intended one. The feedback
network C11 * R9 = 470pF * 56K yields a time constant of 38KHz --
suspiciously about 1/20th of your oscillator's fundamental, just
like your bogey, and a prime suspect. Q2's base bypass network
has a time constant that's only about 220nS -- too short. It
oughtta be lots longer.

Try C5,C9 = 10nF, and consider temporarily disabling the other
oscillators (Q3, Q4) to minimize interactions. That done, the
remaining oscillators should then start locking solidly onto
one other near zero-beat, plus other annoying behaviors!

James Arthur

 

[dave vanhorn]

The oscillators have been designed to be unstable so that just

waving your hands will change the frequency.

Tuneable by hand capacitance, yes. Unstable, no.
Also, I have no antennas connected, so we are just looking at the
oscillators themselves.

Hand capacitance is
only a few pF at most so that the components are chosen so that a few
pF of change will shift the frequency by kHz. Strays are of the order
of a few pF, too... I suspect you're seeing strays shifting about.

Shouldn't be. It's pretty stable mechanically, I took all the wiring
off the board straight away from the oscillators, and the whole thing
is mounted over a copper ground plane.

Is this the type used in the original design? I would think that the can
around the inductor would minimize the effect of hand capacitance,
which is what you want to maximize. Is the can grounded, left floating?

This is what was specified by PAIA.
The cans should be, and are grounded. You play it by influencing an
antenna, not by interacting with the oscillator coils.

IF transformers aren't necessarily the most stable... perhaps the
most unstable part is that tuning core, which seems to be a fundamental
part of this design.

I'm coming to appreciate that.

My gut feeling is that what you're seeing is entirely typical for this
type of theremin circuit.

That's what I'm afraid of.
I'm not in a position to buy a ready-built instrument, but I was
hoping that this one, from a fairly reputable musical kit maker, would
serve.
It has taught me a lot about the subject so far.

 

[Roger Lascelles]

Firstly, your oscillators must produce clean sinewaves. Any clipping or
distortion of the sinewave means the LC circuit is not in control of the
oscillation, and that makes your circuit *much* less stable against other
influences such as transistor variations and coupling from the other
oscillator. The secret of reducing clipping is to reduce bias current until
the amplitude drops well below clipping level - actually, you are looking
for a peak to peak swing of well under twice the supply rail less bias
drops, so that collector voltage never drops below emitter.

An ordinanary bipolar can produce a very stable oscillator, practically
regardless of exact type at your frequency.

Once you have a clean waveform, you have to reduce coupling between the
oscillators. You can use separate LM78Lxx regulators or R and C decoupling
on the power rails of the two oscillators. A CMOS inverter makes a good
simple buffer.

Your project sounds just like a simple metal detector I built my kids, where
the search coil is part of an oscillator. A second oscillator can be tuned
to produce and audio beat. I used simple LC oscillators with bipolar
transistors running without clipping. The signals were buffered and squared
up through CMOS inverters then mixed in a CMOS gate. The two oscillators
will tune to within less than 10 HZ of each other before locking. It is
built on perf board with no special shielding.

There is no need to buy fancy capacitors in the 1nF and up range for this
job - inexpensive Greencap capacitors are stable enough for the job.

Roger

 

[Tim Shoppa]

You will never get this circuit to work with that transformer-IIRC the
tempco of those cheap things is running at several 1000 ppm and the Q is
not particularly high so that relatively small external component phase
shifts cause relatively large shifts in tuned tank frequency. As for
tuning, you have (d/dL)F= -Fo/L so that you have ~760K/360u or 2KHz/uH
so that even 1%/turn gets about 760Hz per 1/10th turn- I'm sure it is
larger- Then those jumps are caused by complex thermal hysteresis
interaction with core loss vs frequency vs amplitude vs large tempco. As
an IF transformer the component is perfectly satisfactory- just don't
get excited over center frequency shifts of a few KHz or so- but as a
tuning element for an oscillator requiring stability to 10Hz say- which
is 10/760E3 x 1E6= 13ppm- you can forget it.

The tunable-transformer-setting-the-oscillator-frequency scheme
seems to be a fundamental part of this design. It's a permeability-
tuned-oscillator but the IF transformer core, as you point out, is
pretty bad.

There are indeed far more stable tunable transformers (I seem to recall
some with a brass slug that are substantially more stable, along the lines
of permeability tuned oscillators, with quite respectable Q's) that
might be used. But these will
not be as cheap or readily available or as small as the little IF cans.
They usually aren't used in a Hartley configuration but I don't see
why they can't be, especially if you're building the PTO from scratch.

PTO's actually get some respect... do a Google search and you can
see some nifty little homebuilt ham radio designs. The one gotcha
for a theremin is that PTO's usually get the LC product with less L
and more C... meaning that you'll be less sensitive to hand capacitance.
Normally that's a good thing, except when you're building a theremin
More turns, of course...

Tim.

 

[colin]

Check out Micrometals for a low-mu temperature stable powered iron
toroid- I believe there is material in the 50ppm range. As an example of
how bad those ferrite interstage transformers are- the similar TOKO
types are 7000ppm/oC.

i think you just need to methodicaly track down the source of the problem.
ie just disconect everything exept one oscilator on its own so nothing els
is powered and get that to work, then do the next one on its own then get
them to mix etc.. or if thats too dificult just disconect the mixers and see
if that makes any diference etc ... i dont think you should just asume its
not going to work, some good advice here and there,

incidently that micrometals website is fascinating, they even have a program
to calculate your inductor performance for you listing the performance for
an inductor made from each of their cores. even have a rf program too, tells
you resonant frequency etc, wich is usefull, i just spent a day trying to
replace a 2.2mh choke that was wound on a bobin type ferrite using a toroid
becuase it was cuasing too much interference used in a resonant 12v-250v
converter.

took me ages to wind it , then i find its capacitance is far too high >.<
the bobin type is only tiny yet it seems i need a huge toroid to get the
same performance ... main problem is trying to avoid saturation i even tried
making a gaped toroid ( by carefully breaking one then super gluing it back
together with a piece of paper in the gap.)

im always making my own coil formers, just get a rod, i often use the blunt
end of a drill cos got a wide selection to chose from, or screwdriver etc
... wrap some parcel tape round it to get it up to the right size (also so
its less likely to stick to it) then wrap a piiece of ordinary paper around
it once maybe twice, then secure that with tape at each end, then wind one
layer of wire and then while holding the wire tight use your third hand to
put some super glue on it and blow on it to make it dry quicker, then if
your carefull you can wind the second layer over that then super glue that
layer ..., i can get about 3 neat layers that way depending on wire size.
more than that and you have to make end caps wich is fiddly, ive found u can
just wind lots of tape to do that tho. if u super glue each layer itl still
hold its shape when u take all the tape off.

by now the paper shld of absorbed quite a bit of super glue and be fairly
rigid. if u realy clver u can make the tube just the right size so a slug
just fits snugly down it. if its too slack put a slither of parcel tape
inside. ive made quite a few that way, ocaisonaly i get them stuck to my
fingers tho ... if it gets stuck to the rod then just heat it up with a
soldering iron till the parcel tape melts ...

they can be quite delicate tho but i usualy end up gluing them to the pcb. i
got an old book wich gives gives me tables etc for air cored inductance
based on coil dimensions and turns.


Colin =^.^=