The AM Forum

I am working on a Valiant and am looking for some help.
Problem is, When I speak into the microphone, the output frequency jumps down. If I whisper it doesn't jump much, a few kc's. But the more audio, the bigger the jump. Up to a few hundred Kc's on a good Yeaaaaa. This condition is on all bands
Here is a description of the transmitter.
This is a factory wired radio.
Someone has removed all of the chokes at the accessory jack J8.
I changed R3 (18K) to a 25K 10 W
I have solid stated the L.V. and Bias supplies as shown on AB2RA's web page, and performed Tim, WA1HLR's level 1 mods. I have replaced all of the old style paper caps. Beefed up the filter caps, 40 MFD in the L.V., 22 MFD in the bias, and 100 MFD in the HV supply.
I am seeing -272 Vdc on C93B side of R50 (1K resistor) but only -223 Vdc on the C93A side. Low per the schematic, which says it should be -265Vdc
Lv is 329Vdc, Hv is 655 Vdc.
HV rectifiers are 3B28's
ka1sbn

How are you measuring the frequency?
If you are using a counter on the output of the xmtr in any manner the modulation will add to and subtract from the carrier frequency.
You need to measure the frequency from a stage before modulation  like a driver stage before the finals

Fred
Ka2dzt

Hi Fred
That is exactly what I am doing, I have a freq. counter coupled to the R.F. output at J6. I will look at it before the finals and see if it is stable.
John Thank you

I have a couple of Valiants and some experience with them.
I always remove at least one of the calibration adjusters from the top of the VFO box so I can stick a RF sniffer probe down in there for calibration purposes. Truthfully....I remove all of them because those clips that fit into the slot on the trimmer can come off and cause a lot of trouble....then I install those push-in buttons to close the holes.

Try using a xtal to help identify if the VFO is causing the problem.
Likely the VR tube/18k Chernobyl resistor is causing a problem.

Did it work OK before you did all the mods?

Hi Dave, I have changed out R3 chernobyl resistor with a 24K 10W and re-located it under the chassis. What do you use for an R.F. sniffer probe?

Hi Pete, it had all the original electrolytics and paper caps installed before I started working on it. I should have just changed them and tried it, but I fooishly did all of the mods first.

I always test the equipment before I start replacing anything. Unless, I know for sure a part is bad when received, i.e. electrolytic has the top of it blown off, resistor broken in half, or some other indication, I found it a lot less aggravating to do a quick test before doing any mass replacements of parts, doing mods, etc. 
Now, you have no idea if something got screwed up when replacing parts, doing mods, an accidental drip of solder, accidental movement of a part lead now touching another part lead, etc.

So where do you start - was the problem there when you got the rig - did it start after the capacitor replacements - did it start after the mods - I would suggest checking every area you worked on

I've never found a shotgun replacement of all capacitors in a piece of equipment was necessary. 
I think this idea was started by people selling capacitors on the web  :D

I have a Pagel Counter preamp....it is in line with the counter.
It is a box with BNC connectors and 3 AA cells....the input has a 6" piece of RG-58 with 2" of the open end stripped back leaving the center conductor and it's insulation free. I put heat shrink over it as well.
These Pagel amps are a bit hard to find but they come up occasionally on EBay for about $30.
I think it has a 20db gain so you can sniff weak signals that show up on the counter....pretty slick.
You can determine if a stage is working in seconds.....count xtals and quickly calibrate VFO/PTO without any hassle.
It's one of the most important tools I have.
I think some of the old handbooks have a ckt that is easy to build....but might not have enough gain.

I think there is an access hole on the bottom of the chassis right underneath the VFO so you might be able to insert an insulated wire in there and see if there is enough signal to trigger the counter.
If not....there is a slot on the left rear top plate of the VFO where you can drop an insulated wire in right over the osc tube. There may be enough signal to start the counter this way.

Also....I would go back to 18k 5w cement tub wire wound and put the resistor back inside the VFO box.
The heat from the resistor helps to stabilize the VFO....both of mine are pretty stable..

If you decide to do this...you can also make a hole over the VFO tube for the sniffer, just remove the side plate and remove the calibration stick near the osc tube.
It has a C clip on it that you can remove...then take out the bakelite tube and the wire clip that fits inside the trimmer. Later if needed you can adj the trimmer with a non-conductive tool....while you are sniffing!! See attachment.

John,

You can add a counter feed from the driver stage.  Connect a resistor 10k-100k to the hot end of the driver coil.  In series with the resistor add a small cap 25pfd-35pfd.  Add a bnc connector somewhere on the back of the xmtr and connect the small cap to the connector.  Doesn't matter if you connect the cap or the resistor to the coil.  You need to reduce the affects of the coax cable running to the counter,  the capacitance of the cable will de-tune the driver slightly,  the purpose of the resistor is to limit these affects.

You can also do this using a few-turn coil to the cold end of the driver coil.  Ground one end of this added coil and run the other end the BBC connector.

You may have to experiment a little but all this is pretty straight forward.

One last words of wisdom,  if you see smoke you made a mistake. LoL

Fred
Ka2dzt

Thanks Fred, I will do that

An easy way to monitor your signals is with a websdr in your region.  This one or others like it:  http://sebagolakesdr.us:8901/

Peter

Hi Fred,
Since I don't plan to use this rig on SSB, I disconnected the coax end from SW4, soldered a short piece of buss wire onto the center conductor, insulated it with heat shrink tubing, and tucked it into the shielded section near V4. I put a BNC adapter on J5, and have this connected to my Freq. counter. The Vfo is stable.
Chasing a couple other problems now.
I can't quite get 8 mills of drive is one problem.
And when I tune up on 40M, the plate meter is pegged, and I cant dip it.

Hi Peter, I haven't played around with the web SDR's yet. Is this one yours? Once I get this transmitter working, I will try it.
John

Going to a higher value resistor to bias the VR tube may not be great.  The VR tube forms a shunt regulator.  If you decrease the current through it and then pull load current, you may not have enough current through the VR tube to allow it to regulate.  Generally folks increase the wattage of the resistor, but not increase the ohmic value.  I would use a scope and see how much the voltage moves as you modulate the transmitter.  Not saying for certain that is the problem, but increasing the value of the series resistor takes you in the wrong direction for voltage stability.  

Joe-GMS

Ok Joe, Thank you, I will take a look at it with my scope. BTW, I heard you on the air at my QTH about a week or so ago. You had a very nice Big Signal.
John

Thanks for the signal report John!  Lets just say, I try!!
 

Yes.  There are several in the region.  You can search the websdr network at www.websdr.org

Happy to see posters using the quote function  ::) correctly  ::)
I think I got them all fixed

I suspect your "problem" is your counter's lack of sensitivity.   Looking at the pic, you can see the minimum envelope is very nearly ZERO.   In your case it may not be zero, but the valley is close enough to zero that your counter can't count the carrier, essentially missing some of the carrier signal, so it will read lower than if it were counting the carrier alone. 

Pick up the signal at the buffer's output or earlier, and I think your "problem" will go away.

(https://photos.smugmug.com/Miscelaneous-/BIGGIES/i-Rbz66MP/0/4a9a1a6c/L/AAAA-L.jpg)

listen to the carrier as you talk on a receiver with the bfo on or in ssb mode. (no ant. connected) beat note should be steady. Alot of older rigs might change a tiny bit. maybe 10 or  20 cycles. No problem,
                                                                                                                        good luck, jay

"...Chasing a couple other problems now.
I can't quite get 8 mills of drive is one problem.
And when I tune up on 40M, the plate meter is pegged, and I cant dip it."

Check the padding capacitors that pad SW3 and SW6 and make sure they are 1kV or greater mica's. I have had numerous original padding capacitor failures that cause tuning and loading problems on Valiants.

Also make sure you have good 6CL6's and 5763's in circuit. Of course, having a sufficient voltage and low ripple on the LV supply is paramount.

Phil - AC0OB

When you are attempting to dip and load it on 40, are you connected to an actual antenna or a dummy load?  Asking in case perhaps you have an issue with that antenna vice the transmitter, or the feed line or connectors to your 40 meter antenna.  I’ve had issues where the center conductor pins of PL-259s are a bit undersized and do not make good contact with the mating SO-239 center contact sleeve.  Or the problem has been with the SO-239 center sleeve being worn oversize.  The problem seems mostly associated with the cheap solderless type PL-259s with the white removable plastic center insulator with the tapered crimpable center conductor tip.  A small sliver of some copper foil into the SO-239 sleeve eliminates the problem, just be sure the foil doesn’t come out when you unplug the PL-259.


Did you ever audibly verify with a receiver or an SDR if you actually have frequency instability while
modulating.  As others have  commented, a digital frequency counter display will go nuts under normal modulation.