KISS 134
Noise Reduction with
Lundahl’s Double-Coil Cut-C Core Chokes
a pictorial essay
by Andre Jute
KISS 134
Noise Reduction with
Lundahl’s Double-Coil Cut-C Core Chokes
a pictorial essay
by Andre Jute
My 300B amps routinely use Lundahl LL1638 cut-C core chokes in various configurations. However, it is a mistake to consider a choke a simple lump of iron. There is no such thing as a simple transformer: a choke is a one-winding transformer. The Lundahls aren’t even that simple: they have two windings which can be connected at least four different correct ways resulting in different inductances, current handling capabilities, different degrees of mutual coupling, and effect on noise at various frequencies of interest in noise reduction.
Note in particular that the effects are not mutually exclusive. Per Lundahl, the winder of these chokes (and the other transformers in the KISS amps), told Jute on Amps: "The LL1638 consists of two coils and a common core. At low frequencies (100-120 Hz) the coils are fairly well coupled and thus act as one choke. At high frequencies, as a result of the poor coupling between the coils, the LL1638 acts as two separate HF chokes, thus blocking/reducing HF common mode noise." (1)
***********
Differential mode noise rejection
Before we can connect the windings we need to ask what it is we want to achieve. Differential noise flows along the signal or power line and returns on the ground line. The dominant element in differential noise is double the mains frequency of 50Hz or 60Hz, so 100-120Hz. At a lesser level there may also be higher frequencies present; they are known as differential EMI (electromagnetic interference) and arise from differential noise on the AC mains being coupled through the power transformer, or from noise generated by the rectifier diodes.
***********
Common mode noise rejection
Common mode noise flows on both the signal line and the ground line in the same direction; it finds a separate return path along the ground reference.
It is theoretically possible to configure a Lundahl LL1638 (and the others Lundahl catalogues of the same type) as a real common mode choke but I have never done it and don't expect to. (Note that the "Serial connection for improved common mode rejection" connection on the Lundahl spec sheet is in fact a differential mode hookup! (1))
***********
The distinguished tube DIYer John Byrns has provided a simple non-technical description in words of how all this works. I strongly advise you to read it after you think you have grasped the points made in the illustrations in this chapter.
You can see this knowledge about double coil chokes applied to gain an extra margin of silence in the circuits of the KISS 300B T39 "UltraFi", the KISS 300B T44bis "Populaire" and also in the T68bis "Minus Zero" 417A potato amp.
***********
Acknowledgements
***********
Footnotes
(1) The Lundahl spec sheet describes this hookup for differential mode rejection as one which offers "improved common mode rejection".
The winder of the LL1638, Per Lundahl. explains the mechanism to the KISS Amp Project: "The LL1638 consists of two coils and a common core. At low frequencies (100-120 Hz) the coils are fairly well coupled and thus act as one choke. At high frequencies, as a result of the poor coupling between the coils, the LL1638 acts as two separate HF chokes, thus blocking/reducing HF common mode noise."
And elsewhere Per Lundahl has said: “Our chokes are made up from two separate coils. At 100 - 120 Hz the coupling between the two coils is reasonably firm, thus for this signal the choke functions as expected as a differential mode choke. At higher frequencies the inductive coupling between the two coils is small. Thus for HF signals, using the proposed connection, each coil can be viewed as a separate choke.”
(2) Some people worry about the possibility described in this extract from the Electrical Engineering Series by John D. Ryder PhD, who was Professor of Electrical Engineering and Dean of Engineering at Michigan State University. I've never seen hundreds of volts of noise, as one would in a tube amp if Professor Ryder is right. Doug Bannard doesn't expect that I ever will: "Any mains current in the choke and resultant contribution to the ripple by placing the choke in the negative lead will be extremely small — so small in fact as to be virtually negligible for any reasonable value of primary to secondary transformer capacitance and reasonable inductor value."
THE VOLTAGES IN THIS AMP WILL KILL YOU.
GET EXPERIENCED SUPERVISION IF IT IS YOUR FIRST TUBE AMP
All text and illustration is
Copyright © Andre Jute 2006
and may not be reproduced except in the thread KISS xxx on rec.audio.tubes.
