The AM Forum

Hi All,

Sorry but I've got a stupid question ('cuz I'm stupid!)  I haven't played with AM very much and am just now getting my feet wet.  I picked up a HB 2X813 X 2X813 rig recently.  The 813 modulator tubes are triode connected.  The mod iron is a UTC VM4.  The mod deck has negative cycle loading.  The audio and the RF come from a Johnson Ranger.   I bought both the TX and the Ranger together, so they are "matched".  I witnessed the combo giving a full 100 percent modulation on a scope while putting out 350 watts carrier at the guy's QTH.  Now that it's at my place, I can only seem to get about 75 percent modulation or so (per the scope).  I just don't know where to start.  Any ideas of where to start looking or what could be different?  73, Mike WB0SND

How does the pattern look?  Did the leads get flipped from ranger to mod deck?  Maybe its going 100 neg now and 75 pos. Calibrate the scope.  The thing should go 4 lines high for 100 % if carrier is 2 lines.  The carrier should not close all the way and should still see a small carrier signal there at 75%. If it closes all the way to a thin line, Then its 100% neg.

Post some pictures. We all want to see the rig!

C

Final grid current low?

You could be getting RF into the audio. Make sure you have a good ground on the rig.

I'm using a Heathkit SB-610 monitorscope.  It's just in the output circuit of the amp.  I'm not using the trapezoid pattern, so it must be the "tuned sampler".  I'm not seeing the envelope hit the "baseline".  It's only about 75 perecnt or so. 

It think there is plenty of grid drive and I'm showing the typical 20 mA or so of grid current.
I'm also seeing about 30 mA of screen current and 300 mA or so of plate current.  It all looks normal to me, but just not 100 percent modulation.  Listening to the audio from the transmitter gives the feeling that it's "light". 

I will check the grounding as that may be the source of the problem.  My 120 circuit in the shack is an old 2 wire circuit, so I'm running a short piece of #10 out to a ground rod outside the shack as well as another one to a cold water pipe in the basement.  I will be rewiring the shack soon.

thanks for all of the suggestions so far.  Keep 'em comin' !

It wont hit baseline if it has neg peak limiting..  Calibrate the scope, Then hit the transmitter with audio. See what the pos peaks are. 

C

Ha.  I'm starting to think that maybe nothing is wrong.  It does have neg peak limiting.  I'll do as you describe in your first response and report back.  It might be a couple of days though.  I will also post a picture of it.  It's pretty cool, but nothing extraordinary.  I was doing some testing with a local friend of mine, W0HRO, Ken, but it might not have been a good test as he's about 25 air miles from me which is kind of iffy on 75 meters.  I may need to get some reports from 200+ miles away. 

    Something else might be at work here, as well...  
    You state your 110 is an "OLD 2 wire" circuit.

    Check your line voltage under load, could be
    the voltage is sagging.  Might need to upgrade there...

Ok cool..  Your on the right track now..  Thats also a good suggestion to check line voltage. The King only does 250 watts and about 80% mod on a wall socket.  On a seperate 30 amp line it does 300 to 350 and 110 pos.

C

Absolutely NO question my line voltage is TERRIBLE.  Sometimes it's as low as 105 volts under NO load.  The power supply has 3 taps on the primary.  105, 115 and 125 VAC.  It was connected to 125 VAC, so I changed it to 105 VAC.  Big mistake.  Shortly after that, I had a parasitic oscillation in the final and it blew the screen regulator transistor.  Maybe it was a coincidince.  At any rate, I have it back where it was.  I will definately be upgrading the wiring to the shack in a few weeks.....

Update.  I had the transmitter RF routed through switches and the scope was also in line but routed through switches and t/r switches.  Tonight I hooked the scope directly to the  transmitter and then to the dummy load.  I now observe I can easily get to 100 percent modulation, reaching baseline with very little driving audio.  So, it now appears to work as it should .....well maybe.  I do observe that the power does "dive" rather dramatically on modulation peaks.  If the transmitter is loaded to 350 watts carrier, I see the carrier dive to about 275 to 300 watts on peaks.  I also see the plate current take a corresponding dive. 

Is this due to the poor line voltage or could I have something else wrong?

(http://)

Nice looking rig!!

Bet it does better with a stiffer AC line and good grounding.

That is a good looking transmitter.  More pictures!

Rich

Ive run across many plugs and connectors holding on by a strand or two, does wonders for the regulation.

Put a meter on the wall socket to confirm.. But I am going to suggest that its simply the lack of power.  This transmitter cant run properly on a lamp socket.  I was in the same boat with the GK500C and the T368.  I had a 30 amp line run with number 10 wire to a 30 amp 120 volt plug. The King and T368 came to life. 

Your choice is to run a line for this rig OR you are probably better off loading it lighter.  Dont load it to light or you could damage the iron.  But experiment and try 230 to 250 watts carrier.  Adjust Grid Drive while putting audio into the rig. The second more Grid does not increase peak output power, Its set. You could be underdriving the Grid so use that method.

What a neat old rig you got man!

C

Oh, I know the power is inadequate.  During the heat in the past few days, line voltage was running about 105 volts.  I have an expanded scale line voltage monitor on the socket.  What I am doing temporarily is I'm running it off of one phase of my 220 circuit I have in the shack.  It runs much better off of that circuit.  I seem to be getting 100 percent modulation now, but still having some sag causing the power output to sag a bit on voice peaks.  This is at 350 watts carrier.  The transmitter will put out in excess of 400 watts, but I dont' want to zorch the thing.  Nonethess, I made a couple of QSO's with it and got good audio and signal reports, so I think all is well now.  I'm going to put in a real stout dedicated 110 volt line sometime in the next few weeks.  Thanks for everyone's help.  Boat anchors rule!!

#10 wire is your friend

If you can get a step up/down transformer it might help with your present wiring. Using your 220 line, step it down to 110 right next to the rig, then use a good heavy line cord on the rig. It will do wonders with your voltage regulation.

Look for 110/220/440 transformers, they are rated in KVA. I use a 5 KVA transformer with my T368 which is 110 only.

Thanks for all of the good suggestions.  I am going to put in a new circuit soon.  I will use #8 or #10 wire.  I'd like to use #8 because it's a pretty long run.  It will probably cost more that the transmitter for the wire!  I will report my results when I get it done.  It will probably be a few weeks though.......

Right now I am running it off of one leg of the 220V circuit in the shack.  Nothing else is on while this transmitter is on.  Seems to work much better.  There is a 105, 115 and 125 volt tap on the primary of the power transformer.  It was on the 125 volt tap and I switched it to the 115 volt tap since my line voltage is much closer to that value.

I now can get a full 350 watts carrier with 100 percent modulation, but there is that sag on peaks.  Power dips about 50 to 75 watts on mod peaks.  That should be much less when I stiffen the line.  Since the modulator is being run with the same power supply as the RF deck, I would expect some sag.  I have some highly regulated Kepco power supplies here but may not be quite enough for the job.  They are 0 to 1500 volts at 200 mA.   I thought about trying to run the modulator off of one of these supplies and seperate the modulator from the RF deck, but I don't think they are quite heavy enough. 

I will post some more pictures soon.  It really needs to be "repackaged" to be pretty.  At the very least, a paint job.  The transmitter was originally started as a project by one guy.  He did kind of a hack job and didn't finish.   It was then purchased by another fellow who finished the job and got it working very well.  It's kind of like my 1930 Buick.  It looks great from about 20 feet at 20 mph.  But when you get close, it's a little homely.  But, over time, I will try to pretty it up.

At present it operates on 160, 80 and 40.  I wouldn't mind putting it on other bands, but the fellow who completed the job claimed it would not work well on the higher bands.  I'm not sure why that is as the 813's are rated at full power up to 30 MHz.  Any ideas there guys?

73, Mike WB0SND

The 813 has a quite large plate capacitance that can make it difficult to tune under some operating conditions at high frequencies.

Look at the old SSB handbook on 813 amplifiers. on 10M a small amount of L was added between the coupling cap and plate tune to deal with high plate C.
High plate C drives the Q high on 10m and makes it hard to tune with a plate cap with high minimum C. You end up making the inductor small to get it to resonate. circulating current is high so there is high loss. Very layout critical but they will play on the higher bands

They will work on 10 like the guys say. I have seen LOADS of CB amps on 11 with 813s on Fleapay.  There used to be a guy on there that sold 1, 2 and 3 tube "no bandswitch" amps.

C

Thanks for the info, guys.  I may try to get it to play on the higher bands this winter.  No hurry yet as we STILL dont' have enough solar activity to do anything higher than 20 M.  There are 3 or 4 positions left open on the bandswitch, so adding extra bands should be able to be done.  I'll look up the earlier threads and the handbook articles.  I'll probably have it on the air tonight for the Collins CCA AM night on 75.  I normally use the 32V2, but it will be fun to use the big rig.  I'll use the 75A-4 or 51J-4 so I'll still be kosher for the Collins net.

If you make the 10M coil as calculated and then place the Tune cap dead center it will require 4X the C to resonate.  That method goes back ages for commercial SW amps. You can also locate the cap tap elsewhere as desired.

My 4 x 813 amp runs fine on 10M.

Modern hams are "rediscovering" that principle with amps using the YC-156 and other hi output C tubes.

Something to think about if you have this problem:

Depending on how far apart the switch tabs/contacts are on the bandswitch, you may have arcing between them under heavy plate modulation. If so, then use every other tab to increase the spacing. Since you have extra tabs, it might be the only way.

On my 4-1000A plate modulated rig, I used every other tab and STILL needed to paint on some corona dope to tame it. It seems shorting the coil from the loading capacitor end is the only way to get to 40M and above without arcing. If you tap the coil without shorting it, the Tesla coil effect is too much and arcing is sure.

I've run tests and find the slightly lowered Q and small power loss from shorting the coil is not excessive. The other alternative is to use plug-in coils. (All whirl class amplifiers use plug-in coils according to one amp guru I know...) ;D

T

The guy that built the amp said he tried to get it to work on the higher bands but it just didn't have much output.  I don't think the problem is arcing in the bandswitch.  The tank circuit bandswitch is a pretty hefty one.  The tank coil might be too big to work on 20 and higher.  Possibly need to switch in a different inductor rather than trying different taps on the existing one??? 

Yes, bandswitch tap arcing will be obvious if it happens. I mentioned it only so you will be ready with a solution if it happens once your rig is operating.

Yes, the tank coil may be too large and needing a smaller unit for the higher bands. If there is low power output, most likely the L/C tank ratio is off and the coil L/D ratio is too high.  A good coil is about as long as its diameter or maybe X2 longer than diameter.  You will need to calculate the correct tank L/C values for each band, then adjust the coil in or out a few turns (or stretch/compress it) until you find the best efficiency for a given band.

There are good calculators on the web for tanks.


Another possible problem is not having enough drive on a higher band. Be sure you can adjust the input tuning to the amp  near to 1:1 swr and you are able to get the correct grid and screen currents so that the tube can produce power. After that, the tank is the challenge. The tank values can be very critical on the higher bands above 20M.

T

Do you think I'd be better off switching over to a variable roller inductor for the tank coil?  I hate to disturb the already excellent performance on 160/80/40.  My friend said he had a difficult time achieving the correct Q for the tank circuit on 160.

I really don't know the reason he couldn't get it to work on the higher bands.  I'm figuring that you guys are right about the L/C ratios in the tank circuit, but I think my friend is sharp enough to know that.  He may have not tried too hard to make it work since his primary interest is on the lower bands.  There really isn't much activity yet on the higher bands so, there was no pressing reason to make it work at the time. 

I'm probably not going to try to tackle this until this winter as I have a lot on my plate at the moment. 

How about a picture of the tank circuit.
Sounds like adding a small inductor for the upper bands would be the way to go but it may force you to modify the main inductor or taps. Carl's idea of the tapped tank coil will work on all bands. My old 4-1000A was configured that way and had good efficiency on all bands. We actually added 160 after I sold it to a friend. The connection to the plate tune is critical for stability It needs to be a short wide lead

Could someone elaborate on this technique?  "Place the tune cap dead center" << What does this mean when using a pi network? Maybe someone cud draw it out.

I also have a linear that requires minimum C on 20M - no prayer of getting it on 15 or 10M without a trick like this.

T

Ya know, thatz a trick I never even thought of   :o  ???

"Tapping down" the plate tuning cap. Instead of hooking it directly to the plate lead of the tank, move it a few turns down toward the "cold" end of the coil. HUH!!

The only thing is that in a bandswitched amp or final, it could pose some switching
and/or lead dress issues. I also wonder how it will effect the Q of the plate tank ???

This amp claims full output on 10 meters with 813s.  1k out on 10.

C

Look at dem final tank values........................

Iz dey bird watts or dosey watts  ;D  ;D  ;D

PEP output will be about 2X dissipation if you want it to last. Efficiency around 65%. Frank I tapped the coil on my 4-1000A back in about 1976 still works great today. BTW I once ran 5500 volts on 10 meters. and it was nice and stable.

Big difference in stability using a 4X1 on 10M in grounded grid linear vs: grid driven, no swamping, plate modulated... ;)

T

Just trying to get this straight..Seems like Tom Vu had the same question.  You are saying just take the plate tuning cap lead off of the plate end of the Pi net tank coil and move it to the center of the 10 Meter portion of the tank coil and leave the plate coupling cap connected to the plate end of the coil?  This technique might be good for other amps with multiple tubes and high total output capacity???

73,  Jack, W9GT

   

From what I understand, I think you got it correct, Jack.


I'm wondering if a second independent bandswitch is needed to pull this off or can it be cleverly configured to be tapped all the time for say, 10M, and still work on the other bands with the main bandswitch??  The 10M coil is on the "hot" side near C1 anyway.... 

A schematic of a commercial bandswitched rig using this technique would help.


T

Tom ....look at early 60's ARRL handbook at the single band 813 rack mount amp for the 10 mtr scheme .... guess this is a dual pi/l output .... gnarly