Does position of the CM choke make a difference?

Discounting stray capacitance of the components in question,

Yes.

In one position, the choke's leakage inductance acts in series between the reg-touching "1" and the JFET-bypassing "1". This creates a three pole lowpass, but whether it's truly good or bad depends on the damping, which isn't obvious here (it may oscillate like a seesaw -- good reason to use a tant at the reg).

If we include stray capacitance, then one is better isolated from the JFETs, while the other is better isolated from the input cable. Which... really isn't saying much at all (if it's a reciprocal filter, which it should be).

Tim

-- Seven Transistor Labs, LLC Electrical Engineering Consultation and Contract Design Website:

I think you went over my head. Which configuration is more likely to keep noise and RF off the 12 volt line running to the Drain of the J310 and Source of J271. This amp could be at the end of a 100 ft run of wire.

Thanks, Mikek

Yep, both have some useful properties.

In the absence of a common mode choke on the regulator input, there can be antenna-like currents in those power wires. The regulator isn't intended to deal with RF inputs at the power terminals.

In the absence of a common mode choke on the regulator output, RF coupled to that output wiring (in the circuits being powered, with the power traces operating as a pickup loop) would show up on the regulator's output terminal (and the regulator doesn't have RF gain, so you can't expect the regulation to work right).

It doesn't matter in theory but it can in practice. That is because, while a linear regulator isn't creating noise, it's physical size can present enough of an antenna to pick up internal noise and carry it out. Or carry external noise in.

A common mode choke should ideally be right at the metal enclosure. Not away from it, especially not inside.

Oh, if this is a tiny box at the end of a shielded cable, then put the ferrite bead, around the cable, at the source! It won't do much at the end.

How good is the cable shield? How well filtered is it at the head end? How low noise do you need?

Try to define your problem as an attenuator and filter problem, and solve for the losses (and impedances as needed to stick things together properly). Your problem seems to contain many more variables than you have shown here, so you are the best judge of it right now. :)

Tim

I'm on the wrong computer to send an example the multiple possible antennas the may be used on. Most common would be a Delta Loop with a terminating resistor. The antenna(s) may be up to 200ft from the shack, ideally the battery would be at the antenna with the 2 FET amp. Then only the feed line to the shack is the concern. The author gave up on coax, although has used twinax with success. His go to feed line is a twisted pair of speaker wires. Signal Ingress will degrade the nulls with phasing multiple antennas. His 2 FET amp has about 150 output impedance, the speaker cable about 100 ohm and the receiver 50 ohm. He sometime does impedance matching other times not. His big concern is signal to noise ratio. I hope to run power from my shack and avoid the battery problem. The discussion has lead me to think I will need additional Chokes, external to the amp box to prevent signal ingress on the power line.

Mikek

I found a link with the first iteration of the use.

Also used with Flag antennas and short BOGs. (Beverage on Ground) At some point four antennas are phased for a very directional signal, with a high front to back ratio.

Why not put a bias tee inside the shack (and another inside the amp if necessary), and obviate the need for a separate power cable?

Tim

-- Seven Transistor Labs, LLC Electrical Eng>>> I think you went over my head.

One step at a time. I want to get a working system, then see if I get signal degradation with that type of a change. Signal integrity is very important when phasing multiple antennas. The author had problems with builders not following his instructions, and then complaining that his antenna system doesn't work. It doesn't take much to degrade a 40 db null. I'd like to try Cat 5 with 4 antennas, but I suspect they would talk to each other. It does seem that a bias T would have less ingress problems rather than more. The author is adamant that a feed line of coax has too much ingress and twisted speaker wire is desirable as it is less likely to act as an antenna.

Mikek