downward modulation in plate modulated rig

[NR5P]

Well I finally had a qso on the apache.  Almost completely restored.  Some said I had punchy audio and some that I was distorted slightly.  Everything looks great as far as plate current and modulation current.  I've noticed that the power meter hasn't been showing much modulation at all.  I hooked it to the scope through capacitance coupling and I see I have negative modulation.  My peaks aren't raising any higher than my carrier is!  Last time I checked I had 100 percent upper modulation.  Man am I unlucky with this thing.  I'm tired of lifting it.  Maybe a shorted plate transformer could cause this?  I noticed a service bulletin about a switch for the ob2 tubes that could fail and cause them to fire on peaks and cause this.  Is there anything else I should check while I have it off the shelf AGAIN!  

[Jim KF2SY]

Maybe a shorted plate transformer could cause this?

If the plate transformer was shorted you would know it right away.

Are you sure it was working correctly before? Try swapping the primary leads from the mod transformer going to the EL34's underneath.

Or just pull and swap the EL34's with each other.    Ur description looks like it could be a weak tube on the pos. side.

[NR5P]

It's got a new set of modulation tubes(doesn't necessarily mean anything).  Plate current under modulation is dropping I just noticed down to about 180ma from 225ma idle under modulation.  Plate voltage stays steady the whole time. 

[NR5P]

sorry for wasting everyones time.  It was a bad solder to ground for the ground side of heater on one of the modulator tubes.  I need to quit getting frustrated so quickly and look at the obvious.  I noticed the heaters weren't operating on a tube when I took another look and there was 6.3v like there is supposed to be when I unplugged them.  Luckily before ordering a tube I checked and there was a ground that I soldered to the base that didn't take.  when there was current it was dropping voltage there.

[W2PFY]

Glad you found it Low grid drive will also cause that condition. Just in case you have more of the same Do you run 6-8 mills of drive? 

[w1vtp]

Glad you found it Low grid drive will also cause that condition. Just in case you have more of the same Do you run 6-8 mills of drive?  

That's something people always forget too. Inject an audio tone and adjust your grid drive for max xmtr power output. That's where your grid drive should be set with a Class C amp. Forget what the manual says, the right amount may be more or less than what it says.

YES!  Be sure the grid drive to the modulated stage is adequate

[NR5P]

I'll give that a try.  I'm actually having the same problem.  I wonder if I weekened a tube by turning the bias up to 50ma when there was only one tube operating.  I need to check the rest of my work to though.

[NR5P]

oh yes and I am running 6ma of grid drive but haven't tried peaking it with audio in just with the spotting pushbutton.  thanks for the tip

[NR5P]

something else that has occured to me is the lug I soldered the cathode resistor is bolted to the chassis.  It may not be making a great connection and I may not have the idle current that I think I do.  I cant solder to the tube base on these new sockets I have it doesn't take well.  the meter gets its current off the .02 ohm grid resistor so it wouldn't take much to mess it up.

[W1RKW]

I can relate to the grid drive issue.  On the 813 rig although things looked good with low grid drive they were not.  Driving the grids was essential to getting good modulation on GORT.  They had to be driven hard and stable.

[ke7trp]

Interesting discussion about Drid Drive.  On my King, The manual states 15 mils.  But I have noticed that there is way more peak power with increased Grid Drive. I wonder if this is safe to run it up to 20.  I will have to put an audio gen into the rig and check various levels.

c

[NR5P]

things are looking terrible now.  I soldered everything to a better ground.  the new tube sockets just aren't very good to solder to.  I went and started turning up the mic gain after tuning up and setting mod bias.  Looked about same maybe a little bit more modulation then fffffff....modulation current pegged meter I immediately turned off tx and removed mod tubes and modulation side looks ok but now very little almost no plate current.  Plate voltage is good.  Its just so frustrating I've learned alot working on this but I'd like to use it.  I've gotten it almost finished and I keep having problems here and there that keep me off AM!  This is driving me crazy

[Jim, W5JO]

Interesting discussion about Drid Drive.  On my King, The manual states 15 mils.  But I have noticed that there is way more peak power with increased Grid Drive. I wonder if this is safe to run it up to 20.  I will have to put an audio gen into the rig and check various levels.
c

Give a read of the Champion manual about how to set the grid drive.

[W2PFY]

This is driving me crazy

Respectfully, take a few days off from the project, the answer will come to you. It always worked for me. Sometimes I took a couple weeks to consider what I was doing wrong. When I went back at it, presto, things fell into place

[Pete, WA2CWA]

3.5 ma is the recommended maximum grid current for a 6146. Apache manual says grid current should be set to 6 ma maximum (because there are 2 of them). You also need to consider the accuracy of your meter shunt (I think it was 5.65 ohms) after 50 years unless it has been replaced recently. I still own my Apache which I built back in the "good old days" and have never seen a reason to increase grid beyond 6 ma. I actually set my grid current to a hair under 6 ma. Also, grid current should not fluctuate when modulating the 6146's. Once you straighten out the bugs, follow the tune up procedure for the rig in the manual. The rig will work fine and the components will last a long time. Also, as part of the the hardware assembly, most screws were attached with toothed lockwashers and nuts. Solder lugs attached to the chassis generally were toothed to dig into the chassis when the nuts were tightened down.

[NR5P]

when it was working for a little while I was setting grid drive to 6ma and getting full output.  It uses a .2ohm for modulators and a .1ohm for 6146s in cathode for the meter and i did replace both of those.  also I do have a new modulator tube shorted now.  put in another pair "unmatched" and it looks ok.  but now I don't think 6146s are getting voltage for some reason, weird:(

[The Slab Bacon]

just FWIW, I have seen weird problems like that with Apaches (I have redone 6 of them). They are sometimes caused by the screwy way they installed the ground lugs on the porcelain final sockets (and other places) with the grounding screw going through the porcelain and the ground lug as well. You get a crappy, intermittant connection to ground as the screws loosen with time from the heating / cooling cycles, or the original builder was afraid to tighten them hard enough for fear of breaking the porcelain.

CHECK EVERYONE OF THEM YOU CAN FIND FOR A LOOSE CONNECTION TO THE CHASSIS

I have been bitten by this on several of the Apaches I have done, so now it is one of the first things I check. Also keep in mind that a kit-built item is only as good as the one who builds it. I have found quite often kit stuff was so poorly built that I ended up just completely dissasembling it and re doing it from scratch. This way you know exactly what you got!!

                                                               The Slab Bacon

[ke7trp]

Ok.. Read it.. Says to set to 10 to 12 ma. No special procedure.. What did you mean?

C

Interesting discussion about Drid Drive.  On my King, The manual states 15 mils.  But I have noticed that there is way more peak power with increased Grid Drive. I wonder if this is safe to run it up to 20.  I will have to put an audio gen into the rig and check various levels.
c

Give a read of the Champion manual about how to set the grid drive.

[NR5P]

I'm going to put it on the kitchen table tomorrow.  I'll keep you all updated on my findings I may not get to look at it until friday night.  I don't think the 6146s are getting plate voltage but my power supply has output so it won't take long to track it down.  Maybe I'll get to learn how to rewind a mod transformer before I get it on the air...

[NR5P]

Good news!!  I hope.   I found out what is wrong with my rf amp section.  The apache uses the plate voltage and drops it down for screen voltage with tube v4.  There is a resistor pair(2 15kohm 10watt resistors in series) where it comes off the plate voltage.  These resistors are open.  I guess because of age I don't know.  I'm trying to think of the senario that would cause  modulator tube to short and a resistor in the rf screen voltage for 6146s go out at exactly the same time.  I'm sure they have to be connected though. 

Also there is another mistake I made when fixing it up.  I used a wirewound .1ohm in the rf amp for the cathode resistor that I replaced.  I think it is ok there because it is bypassed by capacitors.  I also replaced the .2ohm resistor in the el34 cathode section.  with a wire wound resistor.  I don't think it would have enough inductance at audio frequencies to cause the problems I'm having, but OBVIOUSLY im no broadcast engineer:)

[Pete, WA2CWA]

Also there is another mistake I made when fixing it up.  I used a wirewound .1ohm in the rf amp for the cathode resistor that I replaced.  I think it is ok there because it is bypassed by capacitors.  I also replaced the .2ohm resistor in the el34 cathode section.  with a wire wound resistor.  I don't think it would have enough inductance at audio frequencies to cause the problems I'm having, but OBVIOUSLY im no broadcast engineer:)

Several of my meter resisters are made with simple solid hookup wire . Used a drill bit as a winding form. When I achieved the correct resistance, removed the form, and soldered wires in place.

[WD5JKO]

Nathan,

   Losing both those 15K 10 watt resistors has to be a clue. Even if the series combo received full B+ (shorted 6146 G2 ckt), the wattage dumped by those resistors would be about 15-20 watts, and within ratings of two resistors in series. The EL34 upward mod tube shorting out when you heard ffffff is also a clue.

   It is also true that without that EL-34 working (no filament voltage), you did not have a problem, and you made some contacts on AM (downward mod only).

   I do not believe that it is just a coincidence that 3 parts died in one ffffffff event. Sure a WW resistor can go open at any time, but two in series go open at the same time when a DC overload cannot over rate the wattage ratings? No way!!

   I'm just waking up here, so bear with me. The scenario you describe sounds like an audio parasitic to me where the upward mod EL-34 took off. The resistors that opened acted as a lossy inductor, and possibly could get much hotter than from just a simple DC overload condition.

   So, can you post a picture of the socket wiring of them EL-34's along with a schematic of the modulator and RF PA sections?

   I suspect the EL-34's run class AB1, so a 10K 1/2 CC resistor (in series with grid drive) at each socket G1 pin should stabilize those tubes. Maybe look at G2 bypassing, and even something like a 22 ohm 2W CC resistor in series with each plate wire.

   Also as stated before, with ~ 800V B+, you do not want to exceed 25-30ma (max) modulator idle current (per tube) since .03 * 800 = 24W when 25W is the max Pd rating of a EL34. Remember that the E34L is rated at 30W.

Jim
WD5JKO

[NR5P]

I may be able to get a picture later but here is the schematic of the audio section as it is now.  http://amwindow.org/tech/htm/w3sccmods.htm  only instead of the 470k ohm I have a 10Mohm in the audio input, To make the d104 sound a little broadcastier?  I may replace this I'm going to see how I like it.  Just so it has a higher input impedance.  I don't see any hotspots on the resistors but I do see a few cracks.  Sorry i made it sound wrong about the resistors though, only one opened up and if I move it it makes contact again but as soon as voltage is applied it goes open of course.  I just decided to go ahead and replace all 3(those two and one in another part of the circuit)  as they all look cracked and old. 

[NR5P]

found the problem.  the 5k wirewound resistor in the screen circuit for the 6146 had a problem.  A little tiny hair of a wire was sticking out of were the lead comes out.  Must have been a part defect but it came into contact with the chassis and was shorting out, this was the cause of the modulation problem it seems and definately the rf amp problem that came later.  I'm getting what I should on the power meter tomorrow I'll check it on the scope when I get it in the operating position again. 

[KC2YOI]

Very good, I hope it pans out.
I'm coming up on an Apache repair and mod project and I learned a whole lot here.
                                                          DJ