V72 Humbucker Choke

[CJ]

decided to try a humbucker choke for the V72,

this was done by reverse winding the center section of the bobbin with 10,500 turns of #48

and forward winding 5,250 turns for the outer two sections and wiring the finish of the first section to the finish of the center section, then the start of the center section to the start of chamber #3 and using the start of section 1 and the finish of section 3 for our leads,

21,000 turns of wire can pick up hum if exposed to an EMF source of say, 60 or 120 cycles which you might have near the coil if you have your power supply nearby,

happy to report (easier winding for me) that there was no real gain in hum reduction with the trick coil,

a hair trimmer was used as the EMF source, and those things sure do put a lot of stuff,

splicing #48 is never fun, so this means that instead of 6 splices, we only need 2 splices at the start and finish to build a choke,

just thought i would let you know so you do not have to do this experiment yourself,

also, you lose a slight amount of high end due to increased capacitance from having a higher voltage gradient between turns near the center of the coil,

here is the V72 test jig that i used with a DIY output and it is propped up on a Reddi DI output which was done on some Telefunken pwr supply snap in lams, can't wait to try that out,

 

[gyraf]

Thanks!

 

[CJ]

interesting note:

if you measure AC current through the choke, it will be different depending on which side of the choke you put your current meter on,
this is if using the choke in the actual tube circuit,

the tube plate side will read higher current than the power supply side,

so you will get two different readings for inductance if you divide the voltage across the choke by the current and use the reactance formula to get the Henries,

so current through a series circuit is not always the same, for AC current that is,

using the supply side current will yield a larger value for Henries as the AC current will be smaller and therefore when you divide voltage by current, reactive ohms will be higher, which when divided by 2pif will yield  more Henries,

now if you use  signal generator only to measure AC current, then it should not matter which side of the choke you use for current readings,

 

[emrr]

Hmmm, struggling to immediately grasp the meaning of this difference.  RMS meter reading differently due to smoothing effects? 

 

[abbey road d enfer]

interesting note:

if you measure AC current through the choke, it will be different depending on which side of the choke you put your current meter on,
this is if using the choke in the actual tube circuit,

the tube plate side will read higher current than the power supply side,

Eat your heart out, Gustav Kirchoff!

 

[abbey road d enfer]

decided to try a humbucker choke for the V72,

this was done by reverse winding the center section of the bobbin with 10,500 turns of #48

and forward winding 5,250 turns for the outer two sections and wiring the finish of the first section to the finish of the center section, then the start of the center section to the start of chamber #3 and using the start of section 1 and the finish of section 3 for our leads,

happy to report (easier winding for me) that there was no real gain in hum reduction with the trick coil,

That's because while you have effectively put one section in opposite polarity with the other, it has also reduced the overall inductance.
Humbucking works well only when the coupling factor between coils is small.

 

[My3gger]

decided to try a humbucker choke for the V72,
21,000 turns of wire can pick up hum if exposed to an EMF source of say, 60 or 120 cycles which you might have near the coil if you have your power supply nearby,
just thought i would let you know so you do not have to do this experiment yourself,

Thanks for letting us know, i was wondering for some time if mumetal is really necessary for chokes. Guess Germans used it because V7x were small, self contained amps with psu and power trafo pretty close to the rest of it.
Do you think we still need mumetal around choke in typical 2U chasis with enough space to move things around, maybe even add a steel plate to separate amp from psu/power trafo? Are chokes more like OT's that mostly won't get EMF inside and if they do we can easily do what i wrote?
In my test with open frame V72 choke i didn't have problems with 50/100Hz or something else. That was with toroid, well thought layout and specially ground.

 

[CJ]

"...that's because while you have effectively put one section in opposite polarity with the other, it has also reduced the overall inductance..."

no, if you read the post you will see that the reverse wind leads were flipped so that inductance would be the same,

i believe that the reason why the humbucking effect was nil in this situation is because guitar pickups are air core inductors, save for the pole pieces which do not really add up to much as far as a core is concerned, so they re more likely to pick up hum, where as the EI core is semi self shielding, the EMF field gets shorted out by the core before it hits the windings,

you can use mu metal for the chokes, but as soon as you hit them with DC, the inductance drops down to the see value as steel due to the saturation effect,e

 

[abbey road d enfer]

"...that's because while you have effectively put one section in opposite polarity with the other, it has also reduced the overall inductance..."

no, if you read the post you will see that the reverse wind leads were flipped so that inductance would be the same,

In that case, there is no humbuckinh effect. Humbucking can only happen if the voltage generated by the incoming flux is in opposition to the other coils. Since they are on the same core, voltages combine vectorially. If in the same polarity, inductances are added and no humbucking OR opposite polarity=> umbucking but inductance decreases.
So my new comment is: "that's because the overall inductance is unchanged, but there is no humbucking"

 

[CJ]

if you leave out the core, the incoming EMF will hit the windings equally, since one is reverse wound, this will cancel hum, this is how pickups work,

stayed up too late so no Nobel prize for the AC current readings being different, so Kirckoff's Law is still good,  ???

but this meter does read different depending on if it is on the cap side or the plate side at upper freqs so i will have to sort that out, probably a capacitance thing, it's a Fluke bench mode too, so WTF , over? should have bought HP,

 

[abbey road d enfer]

if you leave out the core, the incoming EMF will hit the windings equally, since one is reverse wound, this will cancel hum, this is how pickups work,

Indeed. But as I said earlier, coupling factor is very loose on humbuckers. In P100 style, coupling factor is higher, that's why they have to increase the number of turns like crazy in order to regain output (18k instead of 6-7). The top and bottom coils really fight each other in a P100.

but this meter does read different depending on if it is on the cap side or the plate side at upper freqs so i will have to sort that out, probably a capacitance thing,

Very likely. The stray capacitance of the start of coil is much higher than the end of coil.

it's a Fluke bench mode too, so WTF , over? should have bought HP,

I don't believe the meter is the issue, it's just the way things are.

 

[emrr]

Off topic:  any deconstructions of UTC O-13, A-30, etc?  I don't see those out there, would be among the 200H (max) range single tube options that are common for folks to build some things with. 

 

[CJ]

never done the A-30, is it made for DC?

noticed that the V72 OPT was picking up more hum than the choke,

might experiment with rev wind on the sec to see if there is any difference,

yes i see the point about coupling, hard to get EMF into the windings without it hitting the core which attracts flux,

i have a ferrite pot core version of the V-72 choke that i will rewind to see what it sounds like and how much hum it picks up, s/b better as thhe windings are incased in the core,

good thing my electronics teacher is on permanent vacation, he would have whipped me good for that series circuit gaff, 

new years resolution is to give up the crack, :

 

[CJ]

here are some inductance curves for some chokes,

Winston O boogie sent me an original so that would be a pedigree to compare others with as 2 are rewinds on the orig Tele lams and 1 is a DIY done on EI625 steel,

Ni core has fewer turns, DCR is about 10K, steel original and Winston choke are about 21,000 - 22,000 with DCR at about 15 K, mine is 21,000 with DCR at 25 k as #49 and #48 were used,

max Q occurs after 1 K hz as you can see,

ditched the ammeter and used a 1k resistor as a current shunt to make measurements off of using voltage instead of milli and micro-amps, this was to eliminate the weird variations that the ammeter was displaying by being either above or below the choke,

i think the nickel choke is 50/50 Ni,

this test used an actual V72 test so we could apply a DC offset in the correct matter,

next will be a ferrite version and a 625EI non reverse wound then we are done experimenting with V72 chokes,

 

[CJ]

here is the spreadsheet for the original V72 choke so you can get a closer look of what goes on  with Inductive Reactance, AC current  and Henries as the freq varies,

signal generator injects a signal into the V72 preamp jig so that we get about 20 volts AC across the choke which is held constant,

max Q is where I-ac thru the choke is lowest and Reactance is highest, after 1200 Hz, the core starts to get lossy thus the decrease in inductance, but as freq goes up, so does Reactance so Henries do not really matter as much,

at 20 K Hz, you have an air core inductor as the core is completely dead at this point,

 

[abbey road d enfer]

here are some inductance curves for some chokes,

I guess the inductance is calculated from the impedance curve.
In fact the inductance doesn't change that much. What we see is the counteraction of stray capacitance.
Below resonance, the circuit is inductive, above resonance it is capacitive.
The consequence is that it reduces the HF response after a boost at the resonant frequency.
In the V72, this has to be considered in view of two inductors in parallels, the choke itself and the primary of the output xfmr.
After your measurement, the 20kHz impedance of the choke is about 150-500k.
The 20kHz impedance of the xfmr primary is probably about the same.
The reflected load accounts for 38k  (Zload x [N1/N2]²=300r [11.2]²)
It is clearly the dominant load, so the frequency response of the output stage should be reasonably flat up to 20k. Any deviation would be taken care of by the NFB from the 2nd tube's anode to the first's cathode.
The plate impedance is very high (about 2Meg) so is not a factor.

EDIT: I actually typed that before seeing your spreadsheet, but it doesn't change a thing...

 

[CJ]

parafeed info>

http://www.siteswithstyle.com/VoltSecond/Parafeed_fun/Parafeed_fun.html

test results:  effect of  OPT/Cap on choke current>

 

[emrr]

Awesome link there CJ, thanks. 

 

[bruno2000]

Quote from the "conclusion" of the link:
"Take all measurements with a grain of salt. We are talking tubes here. If measurements meant something actually sounded better, we'd be listening to Solid State."

Best,
Bruno2000

 

[abbey road d enfer]

I wrote previously: "The plate impedance is very high (about 2Meg) so is not a factor."
I must add that regarding the low frequency hump mentioned in the link above, the actual output impedance at the plate of the second stage is much lower due to the overall NFB.
It may seem a little counter-intuitive but the lower the output impedance the higher the hump.