K&K Audio / Lundahl Transformers

Welcome! Need support, you got it. Or share you ideas and experiences.

Return to K&K Audio / Lundahl Transformers

Message Sort: Post Order or Asylum Reverse Threaded

LL1620CFB in SE Operation

71.164.127.111
Posted on May 28, 2012 at 22:40:15
danlaudionut
Audiophile

Posts: 4777
Location: Schenectady
Joined: June 6, 2002
 Thinking of making a SE triode KT120
with cathode feedback and
the LL1620CFB came to mind.
If I ordered it gapped for 60mA
and paralleled the primaries
for 120mA current capability,
paralleled the CFB windings,
used the C variation on secondaries,
it should give me a 2.25K:8 OPT
with 25% cathode feedback.
Should be good for a KT120.
Are my calculations correct
or did I miss something?

DanL

PS I realize this would be a special order.




 

Hide full thread outline!
    ...
RE: LL1620CFB in SE Operation, posted on May 30, 2012 at 06:42:33
KevinC
Manufacturer

Posts: 2540
Joined: April 19, 2001
 Well, if I do the math, I get a different result, perhaps meaning that I misunderstand what you propose to do. I assumed that you would series connect the larger winding and the smaller winding on each coil and then parallel the two resulting windings to make the primary. (You would parallel the remaining small winding on each coil for the feedback winding.) If so (and C' is chosen for the secondary), the turns ratio is 19.2+9.6:3, so the impedance ratio is about 92. with an 8 ohm load, the primary impedance then calculates to be about 737 ohms, ignoring the effect of winding resistances.

My brain functions best between noon and 1PM, so I could be overlooking something...

Kevin Carter
K&K Audio
www.kandkaudio.com

 

RE: LL1620CFB in SE Operation, posted on May 30, 2012 at 07:56:23
danlaudionut
Audiophile

Posts: 4777
Location: Schenectady
Joined: June 6, 2002
 Kevin

I tried posting another option that never
got posted onto the site. I will try again ...
How about I get it gapped for 120mA for about 30H.
Use it normal 3.3K/D wiring for a 2.5K ratio -
with the 25% loss on primary winding.
60mA rating gives me 60H but I lose 25%
from the CFB winding so that lowers it to 45H
and doubling the current halves the L
so that leaves me with about 22.5H.
The KT120 works out to a 680 ohm rp.
So F3 is below 5Hz and a 50 watt rating.
Works out better - lower F3, higher L
and higher power rating.
Also I can use it with 33% UL tap.
I don't know if I want triode or UL.
Maybe a switch so I can try both.

I always liked the idea of cathode feedback.
There was a thread on the DIY board about
the SAC Thailand Silk Minute SE UL EL34
which got me thinking of doing it but
I was never fond of the EL34 in SE.
So I thought of trying it with the
KT120 instead, for more power.
I saw LL1620CFB listed and started
playing around with that idea.
Should be good for about 20 watts.
Link below is a similar Lundahl design
but uses 50% UL tap for feedback.
I'll use a 6GK5 for the driver though.
Maybe fixed bias with DIY HiFi Bias Supply.

DanL



 

RE: LL1620CFB in SE Operation, posted on May 30, 2012 at 13:38:20
KevinC
Manufacturer

Posts: 2540
Joined: April 19, 2001
 Your newest idea sounds better, but the calculated primary impedance is still not correct. You assume that if you remove 1/4 of the turns on the primary, you lose 1/4 of the primary impedance. However, because impedance goes as the square of the turns ratio you lose impedance faster than you do turns. Without taking into account the winding resistances (which are significant in these calculations) due to laziness, your idea results in a calculated 1660 ohms, lower than you planned.

Kevin Carter
K&K Audio
www.kandkaudio.com

 

RE: LL1620CFB in SE Operation, posted on May 30, 2012 at 16:08:40
danlaudionut
Audiophile

Posts: 4777
Location: Schenectady
Joined: June 6, 2002
 Kevin

So I am better of with the 6K/C variation
which will get me just over 3000 ohms.
Still really good at 25W and 6.5Hz F3.

Just thought of something ...
This CFB tansformer won't work for tubes that have a gain of 4 or more.
If you take 25% of the output of a tube with a gain of 6 (KT120)
you will have 1.5 times more signal on the cathode's feedback
than there is on the grid so that can't work.
Or am I missing something again.

DanL



 

RE: LL1620CFB in SE Operation, posted on June 1, 2012 at 05:24:33
KevinC
Manufacturer

Posts: 2540
Joined: April 19, 2001
 You are forgetting that the cathode feedback winding is acting in opposition to the plate winding where there is about 10 times the signal that there is on the grid of a KT-120. The grid has nothing to do with this action directly.

Kevin Carter
K&K Audio
www.kandkaudio.com

 

RE: LL1620CFB in SE Operation, posted on June 1, 2012 at 05:43:18
danlaudionut
Audiophile

Posts: 4777
Location: Schenectady
Joined: June 6, 2002
 Kevin

OK Assume a gain of 10.
25% of 10 is 2.5.
Now you have 2.5 times more signal
on the cathode as you have on the grid.
The circuit will have negative gain.
The feedback will dominate the operation.

DanL



 

RE: LL1620CFB in SE Operation, posted on June 1, 2012 at 12:17:38
KevinC
Manufacturer

Posts: 2540
Joined: April 19, 2001
 I stand by the outline concept I wrote in the last post, but will restate it more completely. If you use 25% of the primary winding for cathode feedback then that signal opposes the other 75% that is in the plate circuit, yes? It opposes the plate derived signal, because it's out of phase therewith. If the signal on the grid is 1V, then the signal on the plate is 10V, and the signal on the cathode is about 1V. The plate signal is present in the 75% of the winding that is in the plate circuit. If one third of that signal is opposed by the signal present at the cathode, then 25% cathode feedback will leave some net gain in the output stage (not counting the output transformer step down).

If you want to see a mathematical treatment of primary side cathode feedback in output transformers, check out the appendix at the back of the classic paper in the link.

Kevin Carter
K&K Audio
www.kandkaudio.com

 

One last question ..., posted on June 1, 2012 at 15:43:48
danlaudionut
Audiophile

Posts: 4777
Location: Schenectady
Joined: June 6, 2002
 Kevin

What is the "m" variable in the equations?
Or is it a constant?

DanL



 

RE: One last question ..., posted on June 1, 2012 at 17:40:36
KevinC
Manufacturer

Posts: 2540
Joined: April 19, 2001
 m is defined at the beginning of the Appendix as the screen voltage divided by the grid voltage.

Kevin Carter
K&K Audio
www.kandkaudio.com

 

RE: One last question ..., posted on June 2, 2012 at 09:25:52
4689


 
Kevin, do you have some 1620 CFB/PP in stock, and what is the price?
Also, do you have any experience (castumers feedback, at least) about it?

 

RE: One last question ..., posted on June 2, 2012 at 11:50:19
KevinC
Manufacturer

Posts: 2540
Joined: April 19, 2001
 Sorry, but I don't stock any of the LL1620CFB variants. I can usually get them within 5 weeks, though. The price is $280 each.

I have had feedback (pun intended) from one user who liked the results in a PP amp.

Kevin Carter
K&K Audio
www.kandkaudio.com

 

Page processed in 0.010 seconds.

[ Top Top of Page ] [ Contact Us ] [ Support/Wish List ] [ Copyright © 2013, Audio Asylum, all rights reserved ] [ Take5 ]
[ General ] [ Speakers ] [ Tubes ] [ Vinyl ] [ Digital ] [ Hi-Rez ] [ Video Asylum ] [ Cables ] [ Tweaks/DIY ] [ Music ] [ Films ]