twin 811As

[VE3LYX]

OK, Thanks. I took the feed thru out and looked at it with a magnifying glass. Eyesight is shot. It said 1500 on it which was probably 1500pfs so that was an issue. I eliminated it and could see the current fall as I tuned. Not as sharply as I would like but a SIGNIFICNT change. I hadn't thought much about it but you are right. Since I m in the B minus line measuring I am measuring everything not just plate current which my account for the softness in the definition of my reading. THEN I blew the cap I had subbed in for the old doorknob. When pushed it shorts and I saw over an Amp on the meter. I had a bunch of these 1.5 kv caps and had trouble with a similar on before in my ARC 5. Arc 5s use a similar tank to what I have and have a problem perhaps in old age with their shunt cap . I had to replace it twice in mine before I got one that would take the abuse.  I will take care of that this weekend and resintall the old doorknob ceramic. It could be a Q issue too . I hadn't thought much of that. Just used the coil cap I have because it was in range with the GDO. I would guess if that is so I have too much c and not enuf L. So close and yet so far it seems.
Don
Heater voltage/ Just excellent. I build a separate heater circuit(independant heater transformer) in my stuff so I can leave it live but switch the primary of the B+ transformer on or off  leaving the tubes glowing (brightly I might add) I can draw on the meter as high as 364 if I push it hard. The plates get orange in their centres . On Am with my DX 60 driving the just start to colour under heavy modulation. Everything seems right except  there should be more output. I shall investigate the Q thing. Must confess I hadn't thought much about it till now. 

[Opcom]

Somehow I mis-read about the feedthrough and RF choke or I would not have suggested the choke as a possible problem. I forgot about those capacitive feedthroughs having not seen one in a while.

[VE3LYX]

No pi . It is an old style link coupling with a shunt fed tank and a 2 turn link (5/16 copper tuning)
I made a new plate choke. Immediately saw a difference. Removed the old one and hard wired the  new in. Tuning was still goofy. Got a flashlight and went hunting. Down underneath the tank I found the problem and fixed it. Tunes sharply now. In shack FS meter pins when I push it. Also pins if I leave it sit with B+ on. Amp will kind of run on its own if left sitting. Got to tame it a bit. Will make some new parasitic chokes. Maybe a couple of neutralizing stubs , just in case to see if I can tame it. I could hear that clean air sound on a monitor when I powered  or drove the amp with my Dx60B which I throttle back. Found a point in drive level where it wont immediately pop into power so  run just above it. Should be able to figure out a way to bias the tubes so they rest at a similar spot and only pop into full service with a good push.
Still have a few questions to satisfy myself before I hook it to my skyhook but progress has been made today.
Don
 "forgot about those capacitive feedthroughs having not seen one in a while. "
Me too and now that  cant see too good didn't notice it had a value on it

[VE3LYX]

OK Got it working and behaving properly. Need a bit more output, probably play with the link adjustment but it is working, tuning, putting Rf into the dummy and sounds decent on air. Did an on air test about 10 minutes ago on 3725. Nobody around.
Don

[VE3LYX]

I have it working but am not thrilled with its output. I built it GG and wonder how it would perform as a class B which would not be a big change in wiring. Tubes seem to fail as they warm up. Output deteriorates after about 7 minutes of use. Quicker if I push it harder. I have the Chinese 811As (brand new) I am torn between thinking these tubes are suspect and the possibility it is something I did wrong. Open to suggestions.
Don
Tested it this morning after modifying the output link. Power out was good for about 6 minutes or so and then began to fall. Tubes run with just a orange spot in the centre of the plates. When power falls it is gradual. At first you think you are imagining it but with in a minute of so all shack rf indicaters tell you it is so.
Before I buy new tubes again I need to know if it just these cheapies cant take the load or am I doing something wrong or should I reduce the voltage (1549V last time I measured it.)Help!

[Opcom]

It doesn't seem normal to ruin even a Chinese 811A in 7 minutes without something more spectacular happening. Could you be ruining the filament emission somehow? 811's can be grid driven very easily.

Does the amp include a meter of some kind to measure grid current?

Is it possible as a long shot that the tubes are not being ruined but that something else in the input side is getting hot and changing value?

This output-dropping thing was happening to me and it was a 56 Ohm 1W resistor in series with a class C grid. It slowly burned up and was >200 Ohms before it opened. Should have not been put in there.

[VE3LYX]

As the amp lost power the grid current rose to a very high point. Before that it was very stable and predictable.
I have ammeter shunt in the circuit I was using to back the tubes up when idling. I notice from the front of the aluminum front panel it gets warm there indicating that it must be heating up and have been a bit suspicious of it so I bypassed it but the problem still existed it seemed. I have a plug in meter in the grid circuit but it is not calibrated  but just a relative device. I also use 6 volt lamps a lot as since I cant see well anyway close up I can easily tell from the bulb where we are as they respond very well. It is across the shunt.  I use an RF indicator lamp in the final tank to which I can use to tune. When the power goes so does the ability to tune the amp. When the amp is behaving everything responds as it should be. Just when you think it is ok it starts to act stupid and soon falls off. I will go over everything including solder joints to see if I can find something that could be heating up during operation introducing resistance. The goofy thing to me is (although while I have built many tube circuits I have very little experience with Hi - power GG amps ) the tubes still respond to modulation and such even when things are goofy. They turn the same amount of dull orange when I speak and act the same colour wise if I feed a cw signal in.. This leads me to believe the signal is getting in but not out. I have cut the drive way back to almost nill grid on my DX60B exciter but the problem still comes. Anyway I will check for anything similar to what you described and will monitor plate voltage as well to see if it goes nuts or drops before I go hunting for a set of 572Bs  to replace the suspicious 811As
Don
And thanks for jumping in. I am lost here at the moment.

[VE3LYX]

I am going to check the PS supply. Could be the voltage for the plate is decaying under hard use. It is a doubler circuit using an old TV transformer for hi voltage. It doesn't even get warm. Filament transformer though is running at about its peak current as it does get warm ,almost hot even. Filament brightness does not change though from when it works good to when it goes nuts.  I run my heaters on a separate transformer so I can shutoff Hv but leave the tubes "at the ready". 
don

[VE3LYX]

I tested voltage for 5 minutes with shunt in and 5 minutes with shunt bypassed. Small difference but definitely not the source of the trouble. Eventually I came to the conclusion that I must have had a parasitic osc develop after the amp got real warm.(runaway grid current and loss of tuning) I had built it with a link coupler and wasn't real happy with it so I revamped a bunch of the wiring in the tank and popped in a load cap (1000pf variable) while ungrounding the coil and putting that end to the coax centre and the load cap rotor. Right away the thing started to perform and even after ten minutes of operation would not bring back the phantom problem . You were right. It wasn't the Chinese 811As. I was able to back of the drive on the Dx60B to a whisper and still light the dummy load to full brightness. So the amp works fine. BTW I measured heater voltage too just to see. A bit high but ok at 6.86 volts. I doubt that will hurt anything. So the amp is now working properly. It parks well(idles calmly), no bad habits. That is the good news. Now for the bad. The 2 811s produce little if any more power then I can get out of my rubbed on Dx60B barefoot.  I have this little green deal working pretty good with a 6293 final and a sub in the audio chain and I have tweaked and retweaked everything till I got it working well. The addition of an amplified mike brought it round the corner. I also spent a lot of time learning to tune it which is the key to running a controlled carrier rig. The difference with them to close vrs "right on" is huge. So while my amp now works it is of questionable value for this deal since the difference is minor. I could add another 811A or a pair I guess. I could boost the HV to 1700 (1149v under load in full drive) But I think I will just leave it for now. It certainly amplifies. Someone suggested at the beginning it might be so. I think I will invoke" the cabbage rule" and quit while I am ahead. I like it the way it is for now.
Don
Early this morning I dug out my calculator and did some math. At 1149 V I came up with 103 watts out if everything was near perfect. Seems close to my real life results. I used 30% efficiency because I thought that was realistic in a real world. At 1500v I had 157 watts out which would make me happy. I will look around for a stout transformer like a 375-0-375 and ignore the centre tap.(750v end to end feeding my doubler circuit) That would feed my doubler well and might pop me into where I wanna be. I have one such transformer now feeding my ARC5 but not doubled. Since I steal my ARC5 vfo voltage from the centre tap I want to keep it there although I could do the voltage divider thing I guess. I think it would handle the 811As. And I have been worried I am pushing the 1625s a bit hard in the ARC5 at 770 Volts. May do a JUMPER WIRE trial first to see if it can stand the 811As without getting too warm

[KL7OF]

DON!..Right on....keep going..I applaud your efforts...cool stuff...  Steve

[VE3LYX]

I have tweaked this circuit and rubbed on it and am now making some better power . Will try it on air over the next few days. Got B+ up to 1492v.  
Don

Successful on air test sat oct 26 with VE6PG.

[VE3LYX]

while the amp worked I was not particularly thrilled with it. I went back at it this morning going back to a choke I had made earlier and tried but with a different coupling cap and some different parasitic suppressors. It was a bit unruly but I noticed smoke coming off the parasitics after a test. MMMMM interesting, must be some power around here somewhere. So on a hunch I took my single bridge tube clip and its single parasitic which I knew for sure worked properly with these tubes and married it to the other choke and coupling cap. AH HA! Right away it had its good manners back but some power was evident. I had it on my old school light bulb dummy load. I set the DX60B exciter to tune and set the grid drive at just above 1. Much to my surprise (I had left the linear on) The dummy load was glowing dimly. I quickly reached over and trimmed the tune on the linear and got a reasonable orange glow. I switched to AM. Even at resting carrier the light bulb had almost full brightness with no modulation. I switched to CW mode and there was a bright flash when I hit the PTT button and the light bulb was toast. Off to town for  new supply. Obviously the thing is working now, much much better then when I did the on air test above. Power appears to be where I had hoped it would be at the outset. How will it be on air? Don't know yet. I  was in the shack but no AM activity on 3725kcs while I was there. Probably will have to wait for the weekend  but I'm ok. It is obviously doing the job now. Glad I kept rubbing on it. Learned a lot in the process. And probably will learn more before I am done. 
Don

[VE3LYX]

Success. Had a very successful QSO with some of my 80M AM group this week. Using the amp with minimal drive from my DX60B I was heard for almost the first time by one member who is in my dead spot propagationwise. He said I was a bit light but he could for  a change copy me. That is all I wanted and why I built this since it isn't much fun when one of the group cant hear you during your turn. With all the bugs sorted out the amp just loafs alone and isn't straining at all. Not even colour on the plates. Judging best I an from tests in my 40s style dummy load power output was at that setting about 150% of barefoot. I may have to drive it a little harder so he can have a stronger signal but not much I think. Now, to add a 40m shorting switch to the PA tank. Hate to mess with success but I am more then pleased with the outcome. Does everything I wanted with ease. Idles with no bad habits without killing B voltage and doesn't bother the receiver during these periods at all.
Don

[Opcom]

Looks like you got it right where you want it! That's great!

[VE3LYX]

Yes. I am very content I guess would be the right word. I used it this morning again successfully. They noticed right away the extra. Said even my resting carrier is quite strong. I knew that because it will light a 100 watt light bulb almost too bright on AM with no audio. If I speak loudly the bulb is gone in  bright flash. I don't know what resting carrier is in a DX60B. I doubt 15 watts at best because it is a controlled carrier rig. Anyway it seems to be just loafing despite all the kafluffle over driving a GG amp with a AM transmitter. I am beginning to wonder of some are repeating what they heard and have never actually done it. My take is if all the tube parameter are below max spec how could on hurt them. I have 1407 VB plus. Current is way below max. Grid current is about right to the low side. Tubes do not even glow now even at 60% of full DX60B power driving them. HV transformer runs cool. Voice quality was no damaged by using the amp. I had thought about making it a 2 bander with a shorting switch on the tank coil (uses untuned input) but I rethought it. I have no need of it on 40M just needed it on 3725 on 80M for a particular purpose. It does that so well I m not going to monkey with it.
Don

[K1JJ]

You've come a long way in the last few years with these homebrew projects, Don.  This 811A project is the biggest and most sophisticated yet. 

Maybe a pair of 813s modulated by a pair might be a good one to try next.

Keep  up the good work, OM.

T

[VE3LYX]

Thanks T ,
 I had neglected my hobby for a bit when I was busy in the shop but re-entered with some vintage radio (a TNT , a Bare Essentials rig (50l6)and many regens including two transceivers. One AM and CW and one CW only. That wet my appetite for AM which had all but expired when I got my license. With encouragement from Tim VE6PG I bought the DX60B and matching rx so I could work more AM. Now it is my main mode. My favourite rig is my ARC5 T18 which I moved to 80M and built a cathode modulator (ok three cathode modulators) for. Finally getting them working well and using them on air from time to time.   Both radios are a bit shot on power in ruf conditions so I set out building this Linear. I kinda just drew it out from what I knew before worked. It fought every inch of the way but now it is up and running and seems quite happy at its task. It is so simple I am almost embarrassed. I drew out the final schematic yesterday just in case. I have successfully modulated my TNT too with my first cathode modulator but  1.5 to 2 watt AM rig is about useless here. And then late in the night I realized the exact same circuit I used for this linear would work for a pair of 45s. In fact they even fit the same sockets which came from a 1924 Harris neutrodyne as did the Tune cap. So my next AM project will be a dual 45 GG linear for the vintage rigs for AM use. I may even get 10 or 11 watts out . Who knows? Got the flamed maple wood already picked out from my fiddle making stash. Whoever said retirement would be boring was a fool!
Don  
Speaking of plate modulation. I have never do anything with it. Too many transformers I thought. However the final cathode modulator version I came up with is not much different except because it uses a transformer to modulate the B- input to the cathode it doesn't have the insulation issues being at or near ground potential. My favourite form is screen modulation with carrier control. I know a lot don't like carrier control but it gives the amp and the power supply a break.  I have built two such rigs. One for 80m and 40M which has about 2 watts out and one for 20M which is a bit more substantial with a well driven 6146B in the final. Looking for 20M AM up here where I live though is like hunting for sasquatsch. Never heard any, not even once.(YET)
  

[VE3LYX]

With your words fresh in my mind I headed downstairs this morning to studio B my AM station. I had been flogging my regen come tx creation and had it cooking pretty good on CW. I looked at it all ready to go and at my cathode modulators. What the hey I thougha nd pulled out the morse key and plugged in the best modulator switching on its filament. I found its freq on my portable grundig crank and took the radio to the far end of the house where my wife was.  I will be calling CQ I said. Let me know of you hear anything. A few minutes Later I heard her holler from the top of the stairs "hear you loud and clear (S9 I think that would be) I tried many different set ups all morning but the first was the best. (And I even tried plate modulation. for me a first. Takes a lot more power to modulate the plate then the cathode. I wonder why. Tis the same B circuit. I didnt have enuf stomp in my modulator to do a good job on plate modulation. It does work wonderful on the cathode though. I also made  modulator testor. A 8 ohm speaker with a 1/4 jack . I just plug any of my self powered cathode modulators in and speak into the mic. Very handy for setting them up.
Don

[ve6pg]

looks gud don..i think al is on 3725 around 5-6pm...i might try this afternoon, likely closer to 5:30-5:45pm...
Tim..
..sk..

[VE3AJM]

Glad that you are back online Tim. The computer must be working OK. You git that DX-100 going yet?

I've replaced the 6146s in the Valiant modulator with triode connected 6CD6 horizontal output tubes, driving them with the external audio amp thru the Hammond 1705 driver transformer. They're idling at 60ma with zener cathode bias arrangement.

Next will replace the mod transformer with the Hammond 2100 100w iron. It will just squeeze in there.

Al

[VE3LYX]

OK. Glad you're back online too. If I am around I will be there. Been working on my shack light duty power supply for the last two days. Even had smoke a couple of times but nothing serious. Trying to get it quiet enuf for the little transceiver which has a regen. Regens hate AC power (hummmmmmmmmmmmmmmmm at the regen fringe) but I succeeded yesterday in getting it useable and this morning went in for the kill. Now works pretty darn good. (5 section cap/ resistor filter)
Don

[ve6pg]

puter werking but real slow connection...hope to look at the dx100 this weekend..very busy week, and really bagged when i get home..no time fer wireless...
..sk..

[ve6pg]

don...that amp has alot of snot...this aftanoon u were very strong here..i've never had a problem hearing you, but i see/hear the difference..any idea on the rf output?
..sk..

[VE3LYX]

I think about 150 out. I can blow the bulb in my dummy at will and it is a 100watter.  Resting carrier lights it now bright enuf I wonder if it will survive the short test. It will glow dimly occasionally even when the DX60B is on tune.
I threw the manual for the Dx60B in the drawer and tune now only by the bulb. I discovered that tune on these is the whole ball game. Probably each radio has its own sweet spot. Anyway thanks . I am more then happy with the amp.
It parks well. (doesn't misbehave if I leave B+ on during receive. )I learned a bit about that from Larry NE1S a year or so ago . Something he told me turned my light on in my head. This amp is sort of a combo of the things I learned building 5 or 10 watters . I stole from several circuits. Took me awhile to work thru the problems with Plate chokes but I finally made one that works. The thing works so well I am afraid to monkey with it. I keep the heater alive with the main power switch using a 12 volt HD transformer (heaters are in series and fed (kathode 01 cap)in the middle) A second switch switches the 110 primary of the old TV transformer on or off as I wish. It now gives 1410 volts thru a double circuit under load. 811As now run cool (no glow)I don't have a meter on it. I have a 3 ohm shunt in the B-between the B - and ground with a 52 light bulb in parallel. I can use it for tune and it also monitors modulation nicely. I try and keep it just winking to about full brightness. If I get stupid on modulation or drive the bulb will blow but the amp survives because the shunt is fairly HD. For me lamps are easier then meters.
BTW. Picked up another ARC 5 same as I have exactly .My brother bid on it for me as he lives south in NY state. I moved the last one to 80M and have used it a couple of time with the group.I will move this one down to 160 which will require on an adjustment in the vfo coil and correction cap and an adjustment of the PA coil slug. How far down I can get it I don't know yet as it is still "on the bus"  so to speak. It tunes down to 2 mcs as is so I suspect I can get to 1.8 one way or another with a bit of creativity. Since the antenna tuner on them is designed to feed a too short trailing wire I wont even need an new antenna.
Don