The AM Forum

I had a nice running Valiant.  I took it out of line and used my Globe Chump for 6 months.  Hooked the valiant up and had some trouble.

In short.. The back part of the band switch was arcing.  I was greatfull when Robert Sent me one! The old one had a carbon track on it.

With the bandswitch installed. No more arcing.

Transmitter tunes up and loads to 300 MA MAX into a Dummy load (wont load higher).  Problem is I get only 90 watts out. Grid checks out perfect. Freq stable.  PEP is about 300 watts out.

Checked all the tubes with the TV7.  Found all where good.

Checked each 6146 in the final one at a time and then checked each socket.  I got about 50 watts per socket.
Found one 20 watt tube.. Replaced it.. All 5 tubes now make 50 watts each.

Put two in the modulator section.

Put three in the output.

Tuned it up.. 90 watts out.. Hmmm.  All tubes are good. All socket positions are good.

Run one tube.. 50 watts.
Run two tubes. 90 watts.
Run three tubes 90 watts.


Checked HV.  760 volts. 866s are out and diodes are in.  700 solid under load. Adjusted Bias and clamper tube.

Any ideas guys?

Tubes are good.
Voltages are good.
Audio is clean on scope.
Sounds good on the air.
Output is low.

Clark
Ke7trp

Are the tune and load knob setting the same as when it worked right?

Clark,

300 ma X 700V =210 w input.   90w output /210w input = 42% efficiency.  Should be closer to 75% or so.

Question is, where is the power going?  Are the finals showing serious plate color?   That's 40w of heat dissipating per tube when a 6146 is rated at only 25w. (35w with a 'B')   Are other components heating up?

Or maybe the finals aren't inputting 210w as you think - a clue.

Also, check that third  6146 socket and associated wiring/components  carefully since the addition of that final makes no difference. Does that tube show more color than the others?

Let us know.

T

Trying to run 160 AM right now is nearly impossible with the CW contest going on..   I am down on 1885 if anyone is up late and has the guns to get over the CW.

I left it keyed at 300MA and 90 watts output for 15 solid minutes into the dummy load. The power did not change one watt and the plates where not glowing red.

I should have been more clear... You can add the third tube in any of the sockets.  I can run two tubes in the back two.. 90 watts.   Add a third to the open socket. Nothing gained but a little peak power.  I can then take one of the back tubes out.. 90 watts..

It is not tube or socket dependant.

I dont see anything getting hot and I would think after 15 minutes something would be damn hot. The meter is a bird 43 that I use everyday. I even tried three slugs.. 100, 500 and 1000. All show a bit less then 100 watts carrier with about 250 to 300 PEP.

I also want to let you guys know that I checked Bias on modulator and final and set as per manual. I also changed and set the clamper tube and adjustment.

Thanks alot for helping me!

Clark

You and your boat anchors.

Good luck.  :)

--Shane

Check your wattmeter

I moved this from the handbook section (which is for completed projects, papers, etc) so it would get more exposure.

I believe the key is in this statement


SHould be able to load it up to much more than that.  Check the B+ (both HV and medium V).   Check the screen voltage, and finally check the idling grid bias voltage.

Also the shunts on the meter are a known trouble spot and I woulnd't trust it.   Try giving it some more grid current, could be the grid current shunt is bad and causing the meter to read high.  When you adjusted the bias, if you did it using the internal meter, it could be wrong if the Valiant's meter is reading incorrectly.

I am leaning toward one of the Tune  or Load caps being bad.

Why would leave it keyed 15 minutes. You may have damaged the tubes by beating the crap out of them.

Clark

You provided a very comprehensive list of symptoms and data that rule out a lot of possible causes. The fact the the rig modulates properly, and that the B+ is ok with 3 tubes in place, says a lot. For example, when the modulator modulates the plate voltage, the rf output amplitude follows (modulates properly). Etc.

It's a long shot... but check the filament wiring. If there is a high resistance somewhere... perhaps the filament voltage on the 6146 rf tubes is dropping too much when 3 tubes are plugged in.

Stu

Good suggestion!

How many times has something run well, then we make a change and it craps out. We sometimes forget to back-track and undo or closely review our change first.

T

Good point.

Even if its not a connection error... if the rf load impedance is too high, you will run out of plate voltage on rf peaks... and that will limit the current. However, the rig will still modulate.

Make sure that you have not used too much loading capacitance. As the loading capacitance is increased, the r.f. load impedance on the tube will increase.

Try reducing the loading capacitance (which most people would refer to "heavier loading"... which I always think is confusing).

Stu

Hi Shane.. Fix the 250s?

Switch wiring is correct. This problem existed before the switch replacement. The old switch has a carbon trace on it to a rivet that was causing it to arc. I have gone back and double checked it as per your suggested.

Plate volts are 700 under load with solid state rectifiers.
Screen volts is 230 under load with solid state rectifiers.

If this watt meter reads low, I have the worlds strongest Ranger....

I will check the fil volts on the rig today. Thanks for the tip.

Took all tubes out and swapped in tubes from a working spare valiant. Same power out. 90 watts.

You should be able to key this all night if needed at 300 ma.  10 min is typical on AM if I get long winded. Nothing is getting hot. I was using a heat gun to test.

From all the help here I am greatfull!

I think my next step is to hand check all the loading caps. Robert sent me some (what a nice guy).

The reason is that only loading 5, 6 and 7 work. The others pin the plate current. I thought this was because of the band I am on which is 80 or 160 but others have told me that the loading should not have high current on each band.

Thanks again to all of you that are helping me!   My Globe Champ after 54 years of use is losing its final tubes. I am trying to source another set and I want to get the valiant up and running as a backup.

Thanks

Clark
ke7trp

I have had a similar problem with my old beat valiant....pull the clamp tube out of the socket (6AQ5 or something like that) and load her up again measuring the outpoot.

Pulled the clamper out. Same 90 watts.  Thanks for the tip.

I am going to start testing the caps soon. My meter is dead and wont read them. Another ham is comming over with his meter in an hour. Hopefully its just a bad cap or two.

What value variable will work?  Any links or schematics? 

Thanks alot guys!  Sitting on 3870 with the Globe champ if anyone wants to chat.

Clark
ke7trp

Clark, With one tube at 50 watts out, what is input? If it is correct then it has to be the additional tubes with the increase of current in the pi network breaking down a loading cap. In this case testing the caps with a meter may not show anything. I would replace them all or as suggested put in a variable around 1500 mmf.

Well, in the minimum loading position there is a total of 3,000 pf of fixed capacitance in the Valiant as designed (all coarse loading cap sections in parallel).  If you want to retain all band coverage at minimum you are going to need around 2,400 pf of loading in addition to the variable loading cap already present and this will still require you to start out at fairly heavy loading (so tune quickly since the off resonance plate current is going to be pretty high).  It isn't that difficult to find standard fixed mica caps in the 1200, 900, 600, and 300 pf sizes and since you have already replaced the switch I would vote for keeping it stock.

 At least your Valiant must not have been in CB service because on 11 meters the higher capacity loading caps would not have seen any action!

Rodger WQ9E

a T tuner is a high pass while an L is low pass. You can make a T a low pass by using 2 inductors and 1 cap.


da wrong subject

just FWIW my valiant makes around 100-105w out when only loaded for 275-290 ma of plate current. I never load it up to the max 360ma. It helps keep the undersized mod tranny from saturating. If your output is low, check to make sure you have wnough grid drive to the final, and that your tank coil taps are on the right turns. If the L/C ratio is wrong you wont get mush outpoot. Also check your neutralization cap. If your max output point is way off of the plate curent dip, suspect something in the neutralization circuit. Also like Dave said, give it a quick try with 6AQ5 clamp tube completely pulled out. Check your final bias as well as valiants used a combination of both fixed and grid leak bias on the finals. If the grid leak resistor is way out of tolerance the clamp tube will do strange things.

                                                                          The Slab Bacon

Before you go completely nuts place separate milliameter in series with the PA plate current. I went through the same deal on a Viking 500 and finally found that the meter shunt inside the plate current meter was broken off. What I thought was full current was actually much less. It may be either a bad meter shunt or the "gimmick" silver mica loading cap is toast. If you can't load any heavier and get more output it is the cap.
BTW what does the screen voltage measure?

Don't throw away the arced bandswitch wafer. I have fixed dozens of them by glass beading the carbon path off. If the contact is fried I replace them with spares from donor switches.

  I know this is a long shot but isn't the final on a Valiant neutralized? If so the grid 1 current should be rather constant as the plate is tuned through resonance when the neutralization is correct. Maybe the neutralization circuit has shifted such that adding more tubes makes things worse because the drive is self limiting with increasing plate to grid capacitance.

   Also, you say the rig modulates upward just fine, but have you looked at this with a scope configured in X:Y mode so you can observe the trapezoid pattern? This will require adding a HV resistive divider to sample the modulated B+ for the X axis. The trapezoid pattern will show the Class C 6146 final RF linearity when modulating and poor linearity if present will be very apparent. Poor linearity can have all kinds of root causes. One of them is poor neutralization.

good luck,
Jim
WD5JKO

While rebuilding my Valiant, I found the shunts for final plate current and modulator plate current to be way off calibration.  Rather than deal with trying to adjust such a small value of resistance wire, I replaced the shunts with 0.47 ohm 5 watt fixed resistors, then I added a resistor of about 500 ohms in series with the meter movement for those two measurements.  For the modulator plate current reading, the series resistor is between the shunt resistor and the meter switch + terminal.  For the final plate current reading, the series resistor is in the negative return, between the meter switch - terminal and ground (this switch terminal was previously connected to ground.)  I calibrated the series resistors out-of-circuit, using a small battery, a pot, and a calibrated milliameter, current verified by a Beckman DVM.  Now I can trust the readings. 

After reading this thread, I am considering replacing the tuning and loading caps in my Valiant, and reducing dependence upon all the fixed caps.  Lots of good info in this thread, surely helpful points to consider while bringing my Valiant online!

Worked on the valiant tonight.

ALL the caps are good and test about what they should.  C38 and C39 (the two groups of four) measure a bit higher then stated.  one is 150 but measures 188.  The other is 330 total but measures 400.  I guess thats ok.

I then measured the loading cap and switch.  600,900,1200 is what its supposed to be. It tested at 600, 950 and 1400.  Close enough?

All postions of the course loading switch are working. Everytime I switch it up, I get more cap. The total is 3000 or so.

My valiant had a 500PF door knob.  I installed a new 1000PF cap and found I could get a bit more loading. I then tried 2000PF but saw no real difference in power. Just where it loaded. I left a single 1000PF in.

At 330 MA I get 100 watts with 400 pep. 
At 400 MA I get 110 watts with 420 pep.
At 450 MA I get 130 watts with 420 pep.

I also tried another watt meter just to make sure my Bird was correct. Both watt meters matched. 

I did not measure or test the shunt. I think its working ok. The bias voltages are exactly inline with the indicated current on the meter for the plate and modulators.

I asked all the guys on the air about power output. They tell me the same thing.. 150 to 160 watts with about 500 forward.

I guess mine is weak. Not much else I can test unless anyone has anymore ideas. Thanks to all of you for the help and parts!  I really learned alot and I had fun working on it.

I sure wish I knew where the power is going. The volts and the current are there. The output is not. Nothing is getting hot or glowing even after a 10 minute key up at 100 watts out.

Clark

I'd guess metering, then.  If nothing is getting hot, then it isn't dissipating what you think it is.

I'll email you on the other. 

--Shane

As I recall the original shunts for the final and modulator meter circuits are supposed to be about 1/4" after tinning but since I don't have the manual handy lets say they are supposed to be 1/2" or .500 inch of actual resistance wire net of the amount tinned.  If the original constructor actually ended up with .600 after tinning (only a 1/10 of an inch error) then your meter is going to be reading significantly higher than actual.  Somehow the idea of precision shunt and having them cut and tinned by a kit builder doesn't seem to make sense.

In general, the tendency is for the shunts to be too long which is on the safe side.  It meant the owner never loaded the final to rated power and realistically the difference between 100 watts out and 150 is not significant at the receiving end.  However, if your final shunt is incorrect but your modulator shunt is correct then you will over modulate running the modulator at rated (which in this case would be actual) current.

I bought a bag full of little low value precision resistors from Mouser and it only takes a couple of minutes to replace the existing shunt with an accurate one.  Although you have a fair amount of leeway with the 3 finals an inaccurate shunt in the modulator will lead to either excessive distortion from too little resting current being set or excessive dissipation from too high an actual resting current.

I replaced mine in my original Valiant after using a precision Wheatstone bridge and measured the shunt as being 30% high.  If those shunts are accurate it is more due to luck than anything else.

You definitely have a Valiant that at least started out as an early model.  The 500 pf coupling cap wasn't big enough for proper tuning on 80 or 160.  If you plan to operate a lot on 160, depending upon the antenna you use you may want to go to the 2,000 pf value used in later models.  Your current coupling cap is fine for any antenna the Valiant will match on 80 meters and higher.

Rodger WQ9E

I am confident in my Meter and shunt working properly.

If I set the mod bias to the voltage specified, the meter reads the mod current specified in the manual.
If I set the final bias to the voltage specified, the mter reads the final resting current on SSB that is stated.

If I load past 350MA I do not get any more power. Just a bit more carrier. Loading higher is not really making any difference. Alot of people are suggesting I am loading it light because the meter is inacurate. I loaded it at 450 MA and had no real increase in power anyways. 

THats a good idea on the mouser caps for the shunt. I will save that information.

These caps where tested with a modern Tectronics meter. In and out of circiut with no voltage.  I dont have a leakage tester. I wish I did and I am looking for one now.

Clark

Ok.. I guess, I will make put some of the caps in that Robert Sent me. I should be able to get close.. these caps are 3500 working 7000 surge and 1000P. I also have some 39Ps.   

Thanks

Clark

Its not the caps.  I just put two 1000s to get 500 in the spot that needs 600pf. Close enough. No change. I then used a 1000 where it calls for the 900 section. No change in output. I then used a 1000 and a 100 to get 1100 for the 1200pf location. No change.. 90 to 100 watts.

I then inserted the meter to check the shunt. My meter is 20 MA low. Not bad.  I loaded it to 330ma of actual current at 700 volts. Output is 110 watts with 415 pep at 100%.  If I remove one tube and run two, I get the same exact output.  I removed each tube one at a time so I ran only two in every location. Same output. Add the third in and you dont get any increase in output. Filiment volts is 6.1 volts.

I guess I will run it like this. It makes 110 and its clean on the air. I wanted to get it up and running as a back up to my Globe. With your guys help, I have my back up transmitter. Its not full power but its working.  Nobody is going to tell I am missing 40 to 50 watts. If it ever fails completely we will know the problem.

I learned alot from you guys and if there is anything that I can do just ask.

Thanks to all of you for the help!!!!

Clark

Checks at 180 PF  Looks perfect.  Checked in and out of circiut.

The C39 network is supposed to be 330 but checks around 400PF.  Also looks perfect.

Clark

Clark,

Given the original size of your coupling cap your Valiant started life as an early unit and it may not have had all of the updates.  The way the metering was set up originally in the AM mode the final plate current (actually cathode current) reads 30 ma high because the current for the VR regulator tubes also passes through the meter shunt.  Later on the wiring was changed slightly so that the VR tube current no longer passed through the meter shunt.  Johnson's interim fix was to change the instructions to load to 360 mils instead of 330 on AM.

Since you get the same output with 2 tubes as you do with 3 it is also possible your grid excitation is a bit low; this might be one last thing to check with your meter.  You don't want to overdo it at the 6146 control grids aren't terribly rugged but if your actual excitation is significantly less than indicated your efficiency is really going to suffer.  To do a quick experiment, temporarily try a very short test with 12 mils of indicated grid drive and if your efficiency goes up significantly then check your grid metering circuit.

I believe the service bulletin is on BAMA but if not I will be happy to email it to you as an attachment.

Rodger WQ9E

How about the plate blocking cap. It could be open. Parallel it with 1000 pf and give it a try.

Plate door knob replaced. Was a 500 and now a 1000PF.

Ran grid up to 12 ma. Lost 5 watts doing that. Set it back at 7.5 ma.  Adding Grid is not increasing power so I guess thats not it.


The Neut cap is fine. I just checked it.

Grid:

Full load 350 MA on 80 meters is 118 volts.  Found that someone soldered in a resistor inline on the grid. I jumped it with a test lead. now 140 volts. No change in power output can be seen. Hooked test clip to terminal (bypassing both resistors) and put 150 volts of screen on. No real change in ouput. Right at 100 watts.

This valiant is a later one from what I can tell.  It did not have the small 500 PF cap for the door knob. It had a larger 500PF at 20KV.  I put in a 1000PF cap that is half the size but came out of another RF amp. Its working better with the 1000PF for sure. I can load to 100 watts on 160 now. Before 160 was way down.

Also want to mention. I loaded it on every band a few minutes ago. Output is about the same on every band 100 watts or so. No single band does significantly more or less then another.

Let me know what you guys think about the Screen volts.  I have no idea why anyone would drop the volts down with a resistor like that. Its def not supposed to be there. At any rate.. The added Screen did not increase the power


Clark

Check the cap. Its perfect. 1003PF Swapped it for another one. No change.

I ripped out two resistors that some previous owner had added to the Screen. Before was 118 volts. That resistor was a bit charred.  I ran solid copper wire to screen junction over to the terminal.  Now I have a solid 155volts on the screens.

I now get 130 watts key and 490 watts PEP.  If I load it to an indicated 390 which is really 360.  I can pin the 500 watt slug. I think this means its fixed unless someone has anything else to add here. It wont load to 150 ot 160 watts but its close and its making 500 watts peak out. I am at least in the ball park here..

I am going to recheck the bias, neutralization, and set the clamper one last time to 10 ma and put it back in service. Should be fun to talk on it tonight for the AMI net on 75.  Also going to talk down on 1885 wiht it to my friends.

Now I can focus my time on the Globe 300.

Are we all in agreement that this radio is about where it should be? 

Thanks again for all the help!!!!!!  I learned a hell of alot with all of your help!

Clark

Hey, that's good news, Clark!

So it was the screen voltage loading down. That was a bear of a problem to find.  Bob/W1RKW had a similar problem with his 813 rig last week and squared it away too.

I must admit, I was baffled with the symptoms after a while.
Hindsight is 20-20, of course, but what can we learn from this for next time?

Perhaps one of most powerful troubleshooting techniques is to exhaustively measure voltages with the VOM and scope. Watch what they do under dynamic conditions on the scope and compare this with the specs, or a working unit... or our best estimate of what each measurment shud look like.

Quick story:  When I was a young tech, just out of school, I had trouble with a repair job. I went to the head engr and asked for advice. His first response was, "Have you measured all the voltages and compared them to the manual?"  I said no, I haven't gotten to that yet. He just shook his head... :-)

That lesson stuck with me and I keep learning it over and over in my own work.


Well, anyway, yours was a tough problem created by some hambone doing a bad modification to the rig - the worst kind of problems...and took a good effort to discover. 

Congrats to you and all the guys who helped here!

Tom, K1JJ

Clark,

Glad you found the problem! Congratulations!  There is nothing like a rig that has been modified (especially poorly) to generate hard to find problems. 

Enjoy your "new" Valiant, it is a fun rig to operate.  As a novice I used one at the 75 watt level and used the hum of the plate transformer as my keying monitor.  Those were the fun days.

73 Rodger WQ9E

I am talking on 1885 now with the valiant. Loaded at an indicated 360MA with 150 watts out! PEP is about 500.  Its talking around 450. 

Thanks to all of your for the help. I am greatfull for the learning experience.  I am going to use the valiant on 160 for a while and then off to 75 for the AMI net.  Hope to hear some of you guys on the air!


Clark
Ke7TRP

Can someone CONFIRM proper Screen Volts on the 6146s.  My manual clearly shows 152 volts on Pin three of the 6146. Tonight on a net, the guys are insisting this should be 200 or more volts. Can someone confirm this for us?

Clark

Clark,

My 1955 Handbook shows for telephony, one tube, 6146:

600 volts plate voltage  @ 85ma
150V screen voltage  @ 12 ma
-85v Bias  @ 3ma
52w output

OR

CW:
750v plate
160v screen

OR

CW:
400v plate
200v screen


It shows a max screen voltage of 250 V and 3w dissipation, so you have some wiggle room there.


73,
Tom, K1JJ

I have 153 volts of screen.  I get 155 watts on 160 and 135 on 80 meters.  I talked on it for an hour straight on 160 tonight with great reports. 3 hours on 75 meters at 135 watts. with good reports. Some of the guys said the screen is to low and that I have issues still.  The manual shows 152 volts on pin 3 screen.  I think this transmitter is working 100% with all the help I got. I really enjoyed the bands tonight with it. Some claimed the audio is thin.. But its a stock valiant.

Thanks again for all the help guys!!

Clark

I got better positive peak modulation when I dropped the screen voltage from 200 to 160 to 170. At 150 the output power drops off.

error

 "" It you run the plate current and screen voltage higher your running the risk of blowing the modulation transformer, etc. "


Cowards die many times before their deaths. The valiant never taste of death but once.

W.S.

congrats clark! observation is your most powerful tool to fix radios. 8 times out of 10 its previous "repairs" or mods from a former owner.
 

stock valiants need a fair amount of work before they swing the munky.

Yep. The difference in signal strength between 110-115 watts and 150 watts is not noticeable. The better quality audio at the lower power level will be noticed.

This radio is now 100% functional.  It makes full power on every band and is stable. I had some issues with the output wave form but when I ran a ground braid to the tuner this was resolved.

Thanks to all for the the help. I have been loading it at 330 actual and it seems happy. 

Clark

I have an ARC39 mod transformer that was the same line up as the valiant with 28 volt heaters in an aircraft rig. I think it would be a good spare for the V2 or Valiant. I think it came from Fair Radio. It is rated 50 watts RMS/ 100 watts peak. Nice sealed unit. It ran lower voltages but in a mil application. No idea how it sounds but looks nice.

It seems like the output section is at fault. The arcing contacts on the old bandswitch was a component (loading cap?? / components) breaking down and turning into a high ohmage resistor.

Fred

thank you Brutha!! ! ! ! ! ! ! ! !  At least I didnt have to say it this time! The weak link in a valiant is the undersized mod tranny. Reducing the current through the secondary will help the audio dramatically. Loading it lighter is the best "bang for the buck" audio mod you can do.
                                                            The Slab Bacon

Understood.  I will load it down to 110 watts or so and let it breath a bit.

Clark