Any chance for a diy autoformer volume control?

[merlin el mago]

Guys inductance plays an important role to adapt the AVC to any system. Playing airgaps is the key.....

 

[abraxalito]

Inductance is a parasitic property like DCR - an optimized trafo has the lowest DCR and the highest inductance. Introducing an airgap will only lower the inductance and thereby degrade the LF response.

 

[gpapag]

I wound up to 8k turns with a 0.07mm wire on the 49%Ni core.
I did some THD vs freq at 2.1Vrms signal.
550 Ohm source, 20kOhm load
George

 

[gpapag]

Here are some distortion measurements at five taps (1st, 2nd, 6th, 12th, 18th) of the 1k4 turns AVC from post #14.
This time at 2.4Vrms (Motu M2 sound card)
George

 

[gpapag]

I wound a 7k turns autoformer with a 0.07mm wire.

 

[gpapag]

I did on it the same distortion measurements as with the 1k4 turns.
It seems that it behaves better.
Compare the attached with those of post #24

 

[gpapag]

And here is the THD step frequency on all 24 taps

 

[Zeta4]

I'm pleased to see that your making very good use of the lams. With 7k turns I would think that your primary inductance will be very high
and probably more than you need for good LF response. So I think there is plenty of scope for the "pinstriping" we talked about which would
reduce 3rd harmonic distortion for a given i/p signal level. Keep up the good work.

 

[gpapag]

Hi Zeta
I enjoy experimenting with your lams.
I tried the high primary inductance for the low THD and the clean multitone FFT they manage.
The stacking I have used is the 1 to 1 alternate long e's and short e's, thus the smallest air gap.
Even with this minimum gap, I haven't managed to approach core saturation. I used a headphone amplifier to raise the drive of the AVC to 7.34Vrms and only when I lowered the freq to 10 Hz I managed to increase the THD from 0.005% to 0.01%.
I will bring-in a hefty power amp to raise the drive further.
It seems there is plenty of oppportunity to test increasing the air gap by changing the stacking order ( your "pinstriping"). This is what I will try next.
Regards
George

 

[Zeta4]

George

I'm impressed that you managed to wind 7k turns on that bobbin. Do you have a tensioning device on your winder? With my eyesight I would
have a hard time even seeing the wire !

The more turns the lower the flux density in the core. I did a rough calculation and at 10hz with that size core and 7k turns a signal of
20V rms produces a flux density of about 0.85T which is well below the saturation limit.

Getting the distortion below 0.01% at 10hz with a significant signal level with restacking could mean that there is little advantage in going to 80% nickel. Interesting.

Have you done a frequency response plot above 20khz yet ?

 

[gpapag]

My "winder" is primitive. I rely on fingers feel for the tension "adjustment".
The bobbin window can take some 9k turns with the 0.07mm wire if not much thread interleaving is allowed.
I was borderly sober when I wound this, so I can attest that 7k is a feasible target.

FR to 96kHz
Loopback and @1 tap curves are indistinguishable, they overlap.

George

edit. FR dB and phase degrees at the legend, refer to cursor position at 20kHz

 

[gpapag]

Now, the heavy stuff
I used Greg Ball's GB150 power amplifier (it's PSU good to some 70W) to drive the AVC at higher levels.
The amplifier can play 10Hz at 22.9Vrms relatively clean (0.018% THD)
When I connected it to the AVC,at the tap #1 (i.e. the whole 7k coil) distortion increase is negligible. It starts increasing from tap #2 and on.
I didn't reach hard clipping but significally distorted waveform.

George

 

[Hearinspace]

A set of frequency sweeps to determine the attenuation of the AVC taps showed good freq response and confirmed the attenuation steps are very close to the target.
Freq. sweep 0-24kHz 5.88Vpp.
Rsource 550 Ohm.
Rload 20kOhm

George

Hello ! Would you say what you used to generate that graph? Very easy to read. Thanks

 

[gpapag]

Hi.
All testing is done using the excellent REW
https://www.roomeqwizard.com/

John Mulcahy the developer and maintainer of the software never ceases to impress me with the immediate and effective response to my mails (questions, suggestions, bug detection).

George

 

[gpapag]

Happy to announce I managed to evaporate a 10mm length of 0.07mm wire during some high level tests
That wire was the one going to the innermost winding.
I will rewind that ex-innermost winding around the outside of the whole coil and see if anything changes dramatically.
If it does, I will scrap the whole thing and rewind it again.

 

[gpapag]

I rewound the damaged innermost winding , now around the outside of the whole coil.
Fortunately there are no changes to the basic parameters: attenuation ratio, inductance, frequency response, distortion.
Only a slight pick-up of mains subharmonics on tap 23 and 24. Sounds logic (the low level coil is not "shielded" anymore by the large copper mass of the whole coil).
I will continue the testing with this configuration.

 

[Celsius]

Very interesting post, I was thinking too in TVC, but is very complicate wired a small bobbin with 24 secondary wires with a diameter smaller than 0,1mm.
I have a binary remote control with 8 bit give 64 steps, but is possible to do with 8 secondary autoformer ?

 

[gpapag]

If you have patience and persistence you can do it.
I would advice you though to start with the 24 steps, see how it goes.
You may find that 24 steps are enough.
George

 

[Celsius]

But is not only the difficult, use my remote control is very important too,

 

[Celsius]

But I am not sure if mix secondaries in autoformer is a good idea for binary combinations.