The AM Forum

I have run my R390A for years by grounding one side of the balanced input and feeding directly with 50 Ohm coax. I have also used the C connector directly. They both work very well.

Today I wound up a Guanella Balun, a 4:1 style on a Type 43 and one on a Type 77 torriod.

I fooled with the trimmers - the result is that both were at least 20 dB down compared to the unbalanced. I think I had better look inside the R390A input box. Ideas?

Mike WU2D

OK I got it to work. I had a shorted turn. Not ready to make any judgement, I also wound up a proper dual core version. The dual core version of the Guanella takes two identical bifilar wound coils and essentially connects them the same way.

It worked across the entire band. I noticed no advantage to the dual core.

I then wound up a 1:1 Voltage balun. This also worked but there was a perceptable decrease in signal level.

Finally I made a 4:1 conventional trifilar transformer. This worked very well on 75M but not as well on 10M.

Results:

1. I saw no real improvement nor degradation in signal strength over the unbalanced feed either as I started by simply grounding one side of the balanced input and feeding or by coming into the unbalaced C connector.

2. The trimmer tuned much sharper on signals with the balun. This indicates a possible higher Q situation which could be an advantage.

3. I saw no advantage in common mode noise reduction. Of course my coax feed is bead choked as it come into the basement. (that balanced feed is purported to have a noise advantage if the trimmers are adjusted propersly). I could not prove this.

Advice:

If you really want to take advantage of the balanced input of the R390A, use a sheilded Balanced feedline!

Mike WU2D


 

IIRC there's enough range to align the thing so either will work fb.

It would be interesting to see what an antenna analyzer to the input would show and then using the various impedance transformers.

The results might vary:

#1: Are the RF coils (transformers) peaked using  50 ohm unbalanced coax and sig-gen with a 50 ohm unbalanced output?

#2: Or are the RF coils peaked with a balanced 125 ohm input? I believe the stated input impedance of the balanced antenna connection is 125 ohms.

Since my R-390/A's are after a tuner trying to make a transmitter happy, if that is possible; the tuner is presenting a 50 ohm unbalanced load to the TX. I ground one side of the balanced antenna connection for the R-390/A, peak the RF coils, etc. & treat the beast as though the antenna input is 50 ohms unbalanced using either a URM-25D or a HP 8640B. Both sig-gens give the same results, a little better than a 20 DB sig + noise to noise ratio. YMMV


Craig,

At what input level?

Steve,

The notes are out in the shack. Several years ago, on the R-390 e-mail reflector (http://mailman.qth.net/pipermail/r-390/), I asked how did the military do it? Roger Ruszkowski, one of the more knowledgeable that post there did several write-ups. He worked on these in the military and taught the classes.

So the short version with out the notes, the best as I remember. Disconnect the speaker and hang a 600 ohm resistor on those terminals. Connect a volt meter across the resistor; get the calculator ready if the meter does not have a db scale. Crank up the sig-gen till the  volt meter reads a certain (x) value; write down that value. Turn on the modulation, think it was 400Hz at 30%; write down the value........................do the math. DB= 20 times the log of V2 divided by V1. Do this on all the bands, if you get that 20DB number with just about the same output from the sig-gen, the receiver is gud to go. If not, you have more work to do.

My thoughts are to treat the RF section according the impedance of the antenna. So if I'm going to feed the RF section apples, then peak all 6 octaves, (18 RF coils) for apples. I use the same antenna for TX & RX. Tunna keeps the TX happy with 50 ohms unbalanced.

Using the balanced antenna input puts the signal through "all" the IF stages. The unbalanced doesn't, if my memory serves me this evening. More plus reasons to go though the balanced than the unbalanced antenna connection. But I doubt many TX antennas present a 125 ohm balanced load to our transmitters, nor would they like it. So I peak the RF section; balanced antenna connection one side grounded, 50 ohm coax, from the 50 ohm connection on the sig-gens.

Anyway, I could be all wet behind the ears and really don't care much. The plastic radios (YeaComWoods) seldom read less than a S9 RFI level, the R-390/A's any where between 20-40 DB on the carrier meter. I can't hear piss weak stations anyhow.

Craig,

TNX OM.

Mike WU2D,

Another thought comes to mind. Those little coax cables connected to the antenna relay inside of the R-390/A; are they on the correct connection? Or have they been swapped, that is the balanced coax connected to the unbalanced antenna input and the unbalanced coax connected to the balanced antenna input.

They have been swapped by some previous owners due to availabilty of input connectors & wanting the signal to pass thru all IF stages.


Craig,

I came to the same conclusion a few years back. I tired both inputs using WWV and my own Sig gen.  With and without a 4 to 1 balun.  I found that my home made balanced to unblanced connector or the adapter on ebay work exactly the same. I did not test at 10 meters.  Only from 20 down.

C

I've been told to use the balanced input because of the extra coil winding in the circuit but there's no human-noticeable difference here with the present lash-up. The set has more than enough gain. I'm guessing the antenna and local noise level are more important.

Exactly! And true for any receiver.

MDS (Minimum Discernible Signal) in dBm has become the defacto standard for CW and SSB evaluation which is a CW signal that is 3dB above the receivers noise floor with the corresponding IF filters in use.

For boatanchors and AM Ive been using a 30% modulated signal 10dB above the noise to give an indication of the receivers capabilities.

I use 10M for all HF testing as that is where external HF noise is minimal and a receiver that can hear well below that noise has a good chance of digging as deep as possible during exceptional band conditions and needing less maintenance from slowly aging tubes.

For my 390A (all Collins 1955 version) I use one of those commercial 50 Ohm to Twinax baluns that show up at times and it sweeps well and also compared to a home brew version. As many have noted the 390A does fall off a bit on 10M which is no surprise considering the tubes used and circuitry and I havent bothered to hot rod it.

Carl