The AM Forum

It was brought to my attention that this transmitter was emitting spurs, and it is true.  It puts out spurs over abt a 100Khz bandwidth.  I have not dug into it yet outside of tuning, which is functioning normal.  Anyone run into this?  Thought I would bounce this around before digging into the xmitter.  Possibly a parasitic. 

Peter

Hmmm. I wonder if you have a switch-mode power supply somewhere in the shack that's getting into the audio. Just a thought.


Don

I have seen this respond to switch and control cleaning .... especially ground wipers on variable caps ... clearing corrosion on ground conections ... all this helps

Hi Peter -

If the spray cleaning doesn't help - then could you give a little more info?

How are you seeing the parasitics… on a spectrum analyzer, listening to a receiver?  You said +- 50 KHz...  how many peaks do you see?

What causes them to appear.... modulation, tuning, on all the time?

What do they sound like?  Can you make them appear/ disappear or change freq by tapping on various modules in the rig?  Wiggling wire harness? Does wiggling the bandswitch or other knobs change them?
Look for bad grounds, especially the variable cap rotor shaft fingers at the ends. Clean them all.

Does a heat lamp or cooling freeze on various areas cause any change?  Focused heat and cold tests can create a temporary repair and expose bad caps and other components.


Tom, K1JJ

Hi Tom and Bob,

I'll try some cleaning and screw tightening.  Spurs are there with no modulation.  Not all the spurs get modulated when audio is applied.  See screenshot.  Tuning has no affect on spurs.
TNX
p

hmmmm   ...  a prickly forrest

Hi Peter,

From what I see, the spurs are down only about -30 DB from the main signal. So considering they are being reported by friends on the air and so strong, looks like they are definitely real and not a fake out.

Usually these kinds of problems are easy to find and are related to one failure. If it were a homebrew rig it would be a different story, but a commercial rig means it worked at one time and you will find it. Was it a kit (screws in tube sockets) or a factory build? Been modified?

If you do the spray cleaning, tightening routine and swap tubes with no success, get in there with a scope and do the "divide and conquer" approach.  Move thru the RF chain starting with the VFO and see where the spurs begin. Disconnect stages and check them individually if you see spurs everywhere.

Does it happen on other bands besides 80M?

This should be an easy fix.  If you have no luck, there are other things we can try later.

T

Some Questions:

Does this occur with the VFO only, with a crystal only, or both?

Have you considered using the Manual's instructions for Neutralizing the Valiant's RF Final?

Have you checked the bias and power supply voltage values at each stage?

Have you scoped the bias and LV power supply to see what kind of ripple and or noise  you might have?

What do the Meter readings show for each stage?


Phil - AC0OB


 

Johnsons have trouble with meter shunts and indicated currents are not accurate .... give them a gander

Hi Peter
The spurs are quite evenly spaced, except perhaps mixed products. It seem to me that an oscillator is quenching, over-oscillating. Get a fine coax with a small coupling loop connected at the display (SDR receiver?) and couple all the oscillator signals one by one (VFO and crystal oscillators) into it to see if there is one with the spurs.

Yes, I see it too.  I used a pair of dividers and see a smaller signal every 6 KHz and a larger one every 12 KHz all the way up and down the band.

It is too orderly and too populated to be a drifty parasitic generated by a poor neutralization feedback job.  And an audio parasitic would generate smaller and smaller peaks up the band.

Nico, could you elaborate more on what you mean by an oscillator stage "over-oscillating" to generate a periodic pattern like this?"  What component failure could cause this?

T

True, but we can't rule it out entirely.

The only time I have seen the VFO oscillator go gaga is when I had a cracked OA2 at the base and it wasn't regulating the 6AU6's screen.

If somehow the Keyer tube V11 is oscillating, this could introduce spurious pulses at its grid and trash at the plate.

Problem is, there are tuned circuits at the plates of V1, V3 and V4 that should minimize any spurs.


Phil

TNX all,

I have not dug into it as yet, no testing, scope work, meter readings etc.  The radio performs normally in all pother respects.   As this issue was not there before I think it can be dealt with.  The model was factory built.  I'm suspecting lower stages and will confirm with SDR receiver in spot mode.

p

I was just reading a 2002 edition of ER. There was an article about 6146 types, and the mention of "spurious emissions" when 6146B was used when the design called for 6146.

They mention that the neutralizing setting (if used) had to be adjusted after going to the 6146B's. Also mentioned that some Collins gear needed modification in order to get a N/C null. The author also mentioned that the N/C setting with the 6146B was not always stable...it could shift.

Jim
Wd5JKO

I did a search for pictures of amplifier parasitics and this picture showed up. It's similar to Peter's shot.

These parasitics were cured with the proper suppressors in the final amplifier's  anode lead. There was also mention of improper neutralization causing these kinds of problems.

So it will pay to keep an open mind as the troubleshooting begins.

T

Hi Tom
Over oscillation or quenching occurs when the positive feedback  of an oscillator is excessive and / or the time constant of the grid leak is too large. You use that with self quencing superregenation receivers.
The reason may be a grid leak resistor that is too large of a non decoupled screen grid adding feedback. Oscillators that have a feedback from the cathode to the grid may over oscillate when the capacitive voltage divider is not correct, when one of the capacitors failed. 

Its before the finals.  Spikes are there in spot.  Looks like the VFO.  I'll put the radio on the bench, but probably not for a few days. 
TNX
p

sounds like a gassy 6au6

Good to hear Peter!

The AMFone crime squad has narrowed it down to the VFO OA2 regulator tube, so let's bet on that for now.

Try a crystal instead of the VFO and see what happens. From the schematic below I see the xtal does not use the 6AU6 VFO tube circuit and uses the following 6CL6 instead, so a xtal is a legit test to isolate the VFO.

Do you have a tube pin extender for the 6au6? It will save a lot of time to measure the screen voltage at pin 6 without taking the VFO apart. If not, then a thin wire stuck into the pin/socket will work.

I just love these troubleshooting mysteries. Everyone pitching in to fix a rig remotely.  ;-)

Nico:  Thanks for the info about over oscillation.  An oscillator needs some positive feedback to oscillate. So you're saying if the positive feedback gets too high, the oscillator loses control of the single frequency generated and starts going spurious. Maybe the OA2's dropping resistor heated up, dropped value and burned up the OA2. This caused the regulated screen voltage to hit the rail and put the VFO into spurs due to the increased screen voltage and more gain.  Your original suggestion was right on if this is the case here.


T

for a stable oscillator closed loop gain needs to equal and stay at 1.  instabilities such as voltage regulation and gassy tube can cause the closed loop gain to vary above and below 1.  If closed loop gain rises significantly distortion or clipping of the sinewave can result.

Hi Tom
I see that the G2 has a OA2 regulation but with a capacitor to ground. That makes you a relaxation oscillator if the positive resistance (G2 impedance) is bigger than the negative resistance of the OA2. A solution is to add a series resistor of 1 - 5 kOhms in between the OA2 and the G2 / capacitor. or change the OA2 to a semiconductor, a zener. A series resistor increases the positive resistance and minimises the change to make a relaxation oscillator. The reason that the oscillation starts may be a weak tube (an increased impedance of G2. So a new tube may be a solution but this is a design flaw.