Bal to SE and AVC

discussion of magnetic volume control design for both line and speaker levels.

Post Reply
avcLover
Posts: 6
Joined: Sun Feb 21, 2021 9:02 am

Bal to SE and AVC

Post by avcLover »

Hi all,

Been using a 28 step AVC with great success for a 10Y preamp/driver setup. However the input transformer placement is bugging me and thinking of redoing it.

Current setup:
20Vpp balanced source into a 15K bridging transformer for SE conversion with a 15K load, AVC to the tube grid with a 100K grid leak.

It occurred to me lowering the voltage before the input transformer would simplify the circuit and I can do away with the grid leak since it will have a low impedance path to ground through the transformer loading. No chance of grid current.


Advantage:
- no grid leak
- AVC doesn't have 15K source
- ???

Disadvantage:
- extra AVC pair cost
- ???

Is this a sensible approach?
Attachments
AVC first
AVC first
2.PNG (32.88 KiB) Viewed 4396 times
Input transformer first
Input transformer first
1.PNG (33.43 KiB) Viewed 4396 times
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

Hey,

first minor correction... you mentioned that the autoformer following the 15K:15K bridging transformer would see a 15K source and this is not correct. It would be driven by the source impedance plus the winding DCR's in parallel with the 15K load resistor.

As to the chicken or the egg situation you have at hand I do like the idea of the balanced autoformer since then the Bal>>SE conversion transformer can be eliminated. In theory a single autoformer could be used from pin 1 or 2 to ground but then the loads can get unbalanced. Placing the balanced pair of autoformers will keep the load seen by the source relatively constant if you terminate the unused half with a resistor = to the ST input impedance. This arrangement will also allow for a penalty free phase inversion and to drive balanced loads if needed.

One other option you can try if your source is to the balanced standard is a single autoformer across pins 1&2 and the secondary feeding the 1:1 15K transformer. The good news is you can try this hookup to see if it works with the parts you have on hand. This has the added benefit that assuming reasonable attenuation, a 600:600 transformer can be used and that typically will have much better bandwidth than the 15K:15K. do not terminate the 600:600 with a 600Ω resistor. The important thing to realize with this hookup is there can be no connection to ground of the autoformer, it must be floating and that the source must have symmetrical ground references for pins 1 & 2. In this case the following transformer is needed to allow for the ground reference for the following stage. Attempting to look at the input of the 600:600 transformer can only be done with a differential probe or a pair of scope probes. any variance from a symmetrical balanced source or load will throw the measurements off and make you scratch your head. Traditional means of measurement can be used on the secondary of the 600:600.

dave
Attachments
Screen Shot 2021-02-21 at 9.26.30 AM.png
Screen Shot 2021-02-21 at 9.26.30 AM.png (35.59 KiB) Viewed 4385 times
Last edited by dave slagle on Sun Feb 21, 2021 7:56 pm, edited 1 time in total.
Get Your Fix
www.hifiheroin.com
avcLover
Posts: 6
Joined: Sun Feb 21, 2021 9:02 am

Post by avcLover »

Small mistake in the schematic, pins 1 and 3 are reversed, I need to tattoo this on my arm, 1- Ground, 2 - Positive, 3 - Negative


Didn't take long to make the changes as suggested with the floating AVC.
Just for the hell of it I left the existing grid leak and gave it a try, a bass tilted signature, similar to what was bugging me but worse.

Without the grid leak, it all came together and what I was hoping the change would bring.

I didn't initially notice the weighty sound with the previous setup but comparing Bal with input trafo and AVC to just SE out from the same DAC with digital volume adjustment made me realise something's off. Now SE with digital volume lacked the super fine detail an AVC brings to the table but frequency response isn't something that should be affected and got me thinking.
I'm sure not having the grid leak is a big benefit as well.

Thank you Dave, you've spared me from buying another AVC so I can get a silver pair in the future. You rock!
dave slagle
Posts: 2085
Joined: Sat Jan 22, 2005 3:54 am
Location: NYC
Contact:

Post by dave slagle »

I redid the sketch... please check to make sure it is correct.

If your source is true to the balanced standard and can drive 600Ω a 600:600 will likely be a big improvement over the 15K:15K. This is assuming you are in the range of 12dB or more of attenuation which will take the source Z driving the transformer vanishingly low.

I am not very well versed at the ins and outs of the variations on the balanced outputs found in consumer audio and the success rate of this type of hookup is around 50% so you won that coil toss. Pro audio will almost universally work in this.
Attachments
Screen Shot 2021-02-21 at 3.03.09 PM.png
Screen Shot 2021-02-21 at 3.03.09 PM.png (82.31 KiB) Viewed 4384 times
Get Your Fix
www.hifiheroin.com
avcLover
Posts: 6
Joined: Sun Feb 21, 2021 9:02 am

Post by avcLover »

The sketch is looks good and what I've been using, I guess if the DAC wasn't to BAL specs it wouldn't see the expected impedance and start humming.

Also took your advice from a 2005 post on a bifilar choke setup and it works surprisingly well. Hooked up a Magnaquest cobalt 10K otp with headphones and it sounds heavenly, super acurate and clean. Fillament bias with Coleman's so no caps in the signal path. I'm ignoring the PSU cap for convenience. Ditched the grid stopper for style points, can't hear a difference without it.

Even with my best parafeed caps, mundorf sgio, claritycap cmr, duelund etc. pure iron is still superior once the different components are matched to synergise.
Attachments
Bifilar no cap headphone amp
Bifilar no cap headphone amp
bifilar.PNG (15.73 KiB) Viewed 4372 times
Last edited by avcLover on Tue Feb 23, 2021 10:46 pm, edited 1 time in total.
avcLover
Posts: 6
Joined: Sun Feb 21, 2021 9:02 am

Post by avcLover »

And because you've pointed me in the direction of a new 600:600 input transformer a crazy idea came about. A 600 1:1+1 input transformer capable of doubling the input voltage to 40Vpp max, apply a negative voltage for a fixed bias 3W PX4 amp without caps, standard 40H 5K gapped OTP.
My DAC has a class A output buffer with 0.5ohm output impedance and should be able to cope with the load, PX4 is not as demanding as a 300B for miller capacitance.

200Vpp output swing would do nicely for my needs and I can always revert to a 2 stage setup for more gain and use the input transformer as 1:1
Attachments
PurePath spud PX4
PurePath spud PX4
PurePath amp.PNG (19.62 KiB) Viewed 4371 times
Post Reply