twin 811As

[VE3LYX]

Last night about midnight I finished a homebrew 811A linear. It has a pair only. Wanted just a decent boost for my DX60B. Just enuf to get me out of the mud when I need it.
Now I am just working up the nerve to fire it up. I had tested it on a too light transformer .Now I have a old TV transformer in place which is about double the iron the other one was. After I check everything over again and am sure all is safe I will give it a go.
Don

[W3RSW]

Well you built it so give it a tryout.
Go for it.
 Looks like link coupling.  Your heavier transformer might surprise you in linear SSB service but may still sag in more constant AM service.  What partially loaded voltage were you finally able to obtain for the 811's ? 

Just remember that "safe" has several meanings including "safe" from bodily harm.

You have a lot of high voltage, both DC anf RF, exposed there.  Be careful adjusting the transmitter loading.  I assume your link coupling will be how you determine load along with possibly an outboard antenna coupler.  If so, Be sure the HV is off before pinching or moving the link coil.

[KL7OF]

I like it.....Plate caps with wrapped copper, the link coupled copper final, and wood chassis.....
Fire it up and let us know the output....Thanks for show and tell..Steve

[VE3LYX]

I have guards so I cant even accidently touch the HV. I forget what it got to before. I was thinking I had it made then the smoke came. Ran a couple of minutes though. I built some small linears for my PW stuff. Got so I liked the Link style coupling. No Plate voltage on the antenna and I seem to be able to make them well. Probably a holdover from my regen set days. I have a single parasitic just below the copper links. I have done that before but not with this much HV. I have tuned it best I could with a Grid dipper. Yes I have a serious Homebrew tuner between any rig in the shack and the longwire ant. I am thinking though of a 350 watt bulb for testing.. I like wood because it gives a measure more of safety . Ok on the advice on the Tv transformer. I guess there is only one way to find out.
Don

[KA0HCP]

Looks great!

[WD5JKO]

Don,

   What band, or bands is this amplifier intended for?

 The 811 is a little limited on plate dissipation (65 watts or so per tube), so normal class B AM amplifier mode (constant carrier) can be run at approximately 1/2 the plate dissipation of the tubes, so 1/2 of 130 is 65 watts. That means each tube dumps 65 watts of heat, and the output is 65 watts. So efficiency at carrier only level is about 33%, and often lower like 25-30%.

 Since the DX-60 is carrier control, then everything changes since the standing carrier drops without audio drive. This lowers the heat in the amplifier. You will need to experiment with the RF drive level, audio level, and linear coupling to find the right settings. This might take some experimentation, and you will need a scope.

  One thing to consider since you might have more drive then you need. The traditional method is to add an attenuator between the DX-60 and the 811 amplifier. Another idea I like to use is to move the amplifier bias point towards cutoff or a little further into class C. For AM this works well since the carrier is amplified class C to just the point of drawing grid current, and then the audio goes from there over the linear amplifiers linear class B range. This is known as class BC. With constant carrier AM, this boosts the efficiency up above the typical 30-35% upwards to 40+%. The signal is still clean, no splatter, etc. The further the bias goes into class C, then more drive is required to get the carrier to the edge of grid current.

  To add bias to the tubes, just string some silicon diodes from the filament CT to ground. About 5V (7 diodes) for 1KV plate voltage...In my quad 572B amplifier I use a 12V 50W zener when the plate voltage is 1800V.


Jim
WD5JKO

[VE3LYX]

Thanks for the tips. Yes I understand coasting in C , talking in B. It is for 40m but will tune 80m AM as well.
I have a scope I built in the 1980s. Still works. I had it on this spring on my slopbucket rig.
Since I don't listen on the same antenna as I transmit I didn't include a relay but switch the plate transformer primary manually. I have my other homebrew small linears working that way and am comfortable with it. Here are two I use that way on my PW rig. One has a 6L6 driving a pair of 6l6 tall boys. It works nicely. The other started out as a 6146B linear which morphed to a 6293. Then I put a octal base on a 829B style tube but pulse version which is a twin beam power tube and plugged it into the 6146B spot and had it working in no time. But these little guys are to get miliwatts to watts which they do nicely. Anyway when I am comfortable with the set up and how it is hooked up I will fire it up. right now I need to find it a safe spot next to the DX60B . Didnt expect it to be finished so quickly.  
Here are two of my PW ones.
Don

[Opcom]

I like your 811 amp a lot. Good going on the use of many kinds of material. About the first pic with the knobs, they appear to glow green. At first I thought you might have used fluorescent paint to redo the dial markings and added a UV illuminator. Very 1970's. What made the markings stand out so? Or was it the camera and lighting? The whole thing looks great anyway!

[VE3LYX]

Oh That is the rig underneath. It is a transceiver I built. Has 4 tubes . A 6sn7 in regen mode with a 6v6 audio PA and a 6bl7 osc/ modulator powered by an electrical/mechanical carbon mic circuit which feeds the control grid (osc section) and feeds the screen grid (modulator section.) of another 6v6 as a final. It also can do CW. It covers 80 and 40 am windows. Not a lot of power . About 1/4watt. I use it with the other linears and have had a couple of decent contacts (100 plus miles)but the low power even on the small linears and no one close running AM hamper my enjoyment with it. It tracks well with its single control tuning for both rx and tx. I also added SIT Sender incremental tuning. Yes I used a slime green luminous pen from Michels to redo all the markings and label it. It is labled in German because I also speak that at a basic functional level. Helps my language skills a bit to practice or think in the language. The 811A linear is the one on top. I fired it yesterday. Maybe my transformer wasn't the problem before. I have a shorted capacitor deck in the PS. I trouble shot it step by step and that is where the overload comes from even with tubes out. I have a lot of trouble with these new electrolytics. Their voltage ranges are suspect to me. I have 4 450 volt ones in this one for 1200 volts DC. They are the new little ones. I ordered Sprague but they were out of stock so sent me the cheapies (but didn't refund the price difference) I think I will double the number of caps to lower the voltage across each to a much lower level. If someone has a good idea for decent caps in the PS I am all ears. It is a diode bridge circuit with outside secondaries wired (centertap not used. )
Don

[Opcom]

The markings remind me of an SX-28 a very long time ago, someone had wiped red blacklight paint on the front panel and it filled in the grooves in that finish, and wiped the rest off. It was spooky looking.

[VE3LYX]

Well I always believed REAL RADIOS GLOW IN THE DARK.
BTW does anyone have specs for a LK576-1 transformer?
Don

[VE3LYX]

Talked to an engineer friend well known for his expertise in  all kinks of power delivery. He gave me several good solutions. IE softer start with some resistance in the supply, slow blow fuse to allow for the intitial surge.
Don

[VE3LYX]

It appears his advice was well founded. Power supply runs with no obvious issues now. Fuses don't blow. Will tie up the loose ends next week and get this thing in action.
Meanwhile I did get a little sidetracked. I got a socket for my 829B tube collection and dug around in the junk box. Now have a push pull HF amp all wired except for winding the two  coils for 80 and 40. New power supply for it as well.
Now for a really dumb question. Can you successfully screen modulate a push pull amp? I left a terminal between ine screen supply and the B+ just in case I could add a dropping resistor and insert some audio. If so I will give it an osc as well. I could run it parallel but have never tried push pull in a final before.
Don

[Opcom]

What make is the LK576-1 transformer? got a pic? let's discuss, as there are ways to figure some of that out.

[VE3LYX]

I haven't any idea. Got it in a junk sweep tube amplifier. Suspect it to be from an old TV. Yes I should have fed it 12volts pri and tested all the leads and did the math. However it worked in the old amp so I just removed it and transferred it. My good voltmeter died and I bought a cheapie but discovered it only goes to 750 AC or 1000 DC. Anyway it looks like I am ok there anyway. I added a 100ohm cushion to the Power supply before the diode string and added 2 more 40uf X450 volt caps along with the  equalizing resistors to the PS. I put in a bit more fuse as was suggested by my friend and it runs calmly with no strange noises or heat.
Hopefully that will be all I need. Got to make up some patch cables to hook it in and move my scope downstairs. I suppose I could just start off light and test it making sure it was not over the edge but I have the scope so I will use it. While I love pushing a race engine as hard as it can safely go I am the complete opposite with my radios and never want to see them sweat so to speak. I am not real comfy with high voltage and big power. Only want enuf to do the job. The DX60B is just a wee bit light. Hopefully this will help that. I have no trouble working east and south east but working northwest which I often do there are a couple who have trouble hearing me well.
I used to have trouble hearing them too but since I started using a beverage to hear that all disappeared. Unbelievable difference.
Don

[VE3LYX]

Tested no load plate voltage at 1549volts. discovered I had neglected to install the cathode to filiment resistors and bypass caps. Fired it up and it ran for a few minutes till the 2 Amp fuse died. Don't really know what fuse to use. 5 or 7 or 10amp? Was able to measure current flow at the meter jacks. One step closer. Double checked tank circuit with the grid dipper. BANG ON!
So I have voltage, I have current flow , I am tuned up on freq. Nothing got warm. maybe there is hope after all!
Don

[VE3LYX]

Got lots of RF now. Still working on the dummy load till I get it figured out. In fact the dummy load Packed in on the last test but I have it fixed. I smoked a padding cap in the tank circuit and since it has no direct dc feed (shunt fed) it can only be RF. I was wacking it with the ARC 5 and smoke started rolling up and quit when I stopped pushing it. I changed it out for a pair of 1KVs caps and it ran fine. A 6 amp fuse handles the transformer draw fine. I am still not comfy with it and until I am it will stay on the dummy. Transformer even after an hour of abuse was not hot so that is good.  Power resistors though in the front of the PS string were looking "WARM". Still progress is progress. I am not real comfy with this kind of stuff so forgive me if I don't get it right away. I keep a FS meter in the shack all the time. It was pinned despite being adjusted to its lowest sensitivity. Anyway I am trying bit by bit to figure this all out before making a mess on air.  
Don
I wonder if the ARC5 is too much drive for it. Is that possible? If so I could use the DX60B as it has a throttle (adjustable drive) which the ARC 5 does not.

[VE3LYX]

After checking the circuit thru and rigging up a different tank for a test I realized the Amp was generating its own RF and lots of it. I came to the conclusion there were only two possibilities. A parasitic caused perhaps by the use of parasitic plate choke from another amp but was built for 50Mcs instead of 3.7mcs. I also used only one with the plates twinned so that was also area I wondered about. I had done that before though with good success. The other concern was maybe my plate choke was making the amp into an osc since I built it myself. This morning during the morning AM group I worked between my turns at returning the tank to my original design and then removed the parasitic/ plate cap assembly. It was one large turn around a 10 ohm 10watt  . I removed that and made a 4 turn one winding it closer to the resistor body as well. I reinstalled not expecting any success but guess what. Shack RF disappeared and the meter (in neg B lead ) began reading normal. I let it run for 20 minutes and everything stayed the same. I could hear in a nearby receiver the faint sound of a null when turning the tuning control. It showed up where I knew 3.7 was on the scale so I hooked up a low power rig and tested it on AM. I could dip the meter on tune although not huge but distinctly noticeable. I spoke into the mic (B+ off) and could hear it on the monitor receiver. I turned on the B + and tried again. Voice was noticeably louder but signal was clearer too and lacked noise. I took a portable rx upstairs and had my wife listen without any instructions Her report was the same .Much clearer and quieter on test #2 (Power on. ) Maybe tomorrow will hook it to the main rig and see how it works on air. Hopefully that is it for sorting it out but if not it is at least closer. BTW Power supply runs cool.
Why is the signal less noisy thru the Amp?
Don

[VE3LYX]

Lost a cap in the capacitor string resulting in no voltage doubling so I had under load only 642 volts which will barely wake up  set of 811s. Looking for 1100 to 1200 volts. Have ordered enuf parts to build a different style rectifier filter doubler assembly. Should be here today I hope.
Don

[VE3LYX]

So how hard can 811As run? Dull orange dots, Yellow or no colour at all. Got it working.
Don

[KL7OF]

I always liked them to run with an orange glow on the plates .....in amateur service you can beat them pretty hard...Soundlike your power supply might be the limit....what metering do you have?   Grid drive ?? Ip? etc?


very cool stuff Don...

[VE3LYX]

Thanks. That is what I wanted to know. It is a Grounded Grid and I used a old meter shunt to back it up  when idling. I have two jacks from an old osc scope coming out either side of that shunt in the B-. One black and one red and insert a voltmeter to the proper polarity hole. Readings at this point are relative and not accurate. It does respond well though. Idling shows 40 on the meter(2000MVscale) and a light test signal pops it to 57 instantly. I have pushed it as hard as 134 but was scared to run it there. I am still on the dummy load  but strength is obvious on a nearby monitor rx with its RF gain restricted. I am not totally happy with the output tuning as it is less definite then I think it should be. IE not like tuning a VFO for instance.  I will recheck it with my GDO to be SURE it is tuning properly. Has been a struggle but its is coming together bit by bit. Its power supply is very quiet which pleases me. I am tired of dirty power supplies. Probably due to its hash filter in this one but it is actually cleaner thru the Linear then right off the TX (TX is noisier ) My wife who helped me do some of the monitoring also mentioned that under the higher power (louder transmission) there is less noise in the background.    Thanks for your tips and encouragement. It looks like it is close to on the air time for this project! It is an 80M only Linear.
Don

[VE3LYX]

Something still is not right. Either a high voltage feed thru is also a capacitor or the plate choke is no good. Everything responds as I think it should. Increase drive causes increased plate current. When Modulated current varies with modulation or as I think it should. This is a response to the controlled carrier of the Dx60B I am driving it with which cuts carrier when there is no modulation. Everything responds as it should. I  rechecked output tank circuit with the GDO. It is bang on. Somehow not much RF is getting from the plate circuit to the shunt fed tank. I subbed the shunt cap. No difference. When I built it I did use a hi voltage feed thru I had from an old Transmitter I had scrapped. I wonder if the feed thru is also a capacitor. If so that could be the trouble. If not then My plate choke is suspect but because I have made so many I doubt it. It has none of the signs I have learned to expect if it isnt good. One might ask why not a Pi output. I could change that but this style of output was used for many years with good success and I have several other Homebrew circuits that use it and they all work so there is no reason why it shouldn't be performing. Used to developing things in my racing endevours I do not get frustrated but stay cool knowing full well it is something stupid I have done unknowingly that is preventing it from working properly. I will find it step by step eventually. Right now it is almost working if you know what I mean.
Don  

[KL7OF]

Filament voltage OK under load?

[Opcom]

assuming a pi network,

What value of loading cap?

Try meshing it all the way in, then tuning. Just a little drive is needed.
Is the tuning still not sharp? It ought to be very sharp like that.

It might be that something is loading down the tank circuit, making the Q of the whole circuit too low.

You could try a 10K 30-40W wire wound resistor for the plate choke and see if your tuning's any sharper. The idea is to eliminate the plate choke. The power is not important, just the tuning. Don't use a lot of drive. Try to check these things with a miliammeter in the B+ line, so you are sure only the plate current, not the cathode+grid current, is being looked at. Just be careful the HV, esp. with the meter.