The AM Forum
March 28, 2024, 12:22:40 PM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: 1 ... 4 [5] 6   Go Down
  Print  
Author Topic: 2x 813 amplifier?  (Read 129048 times)
0 Members and 1 Guest are viewing this topic.
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #100 on: November 20, 2009, 12:52:35 PM »

What is the value of your plate tuning cap. I suspect it is too small and the circuit is acting like an L network or a low Q pi network. I think the plate tune should be at least 300 pf but have not run the numbers.
Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #101 on: November 20, 2009, 01:48:56 PM »

I figure at 2500V anf 2400 watts in for 1500 out
Plate 280
load 1281
L= 8uh

I bet that inductor is much bigger than 7.8
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #102 on: November 20, 2009, 01:56:30 PM »

I think the plate cap is 300 pf.
2000 volts in, 300 ma or less, with 60 ma grid current, so its more like 220 ma in, but say 600 watts in, 150 out, 450 watts plate dis for 4 tubes with 125 watt dis.
The plate tuning cap is 7/8 closed.

I shorted a turn and could not get it to resonance...but that might not have been done right.

Brett
 
Logged
N3DRB The Derb
Guest
« Reply #103 on: November 20, 2009, 03:54:05 PM »

that plate tuning cap is the good old national 100 DA... 100pf per section, I have 2 of them here.0

I counted abt 17 turns on that coil.
Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #104 on: November 20, 2009, 03:56:08 PM »

Try putting 500 pf across the loading cap and see what happens. If the power drops down but dip is deeper try shorting the coil or just spread the turns.
Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #105 on: November 20, 2009, 04:34:59 PM »

200 pf is not enough unless the loading is pretty light. The inductor looks kind of large. You will make some real power once you get the tank circuit right.
I have been burned by running the Q down and a number of others have been there also.
Find yourself a 300 pf or dual 150 pf 7000 volt Johnson and mount it stator up. 
Then start lopping off turns and watch the power go up. You may need more loading if it is close to fully meshed now.
Logged
N3DRB The Derb
Guest
« Reply #106 on: November 20, 2009, 05:24:49 PM »

Brett,

I have a big ass Johnson here thats at least 200 pf pr section, and at least 5kv. I'll go take a picture of it and post it, it's yours free if you want it - just paypal me some shipping money - I'm so broke at the moment I cant afford to send it out unless I get it first.  Roll Eyes

lemme get this picture.....standby.... BTW thats a 8 inch ruler for scale...this mofo is large


* P1020257.jpg (206.75 KB, 1280x960 - viewed 1147 times.)
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #107 on: November 20, 2009, 06:06:44 PM »

Brett,

Yes, as Frank mentioned, we can't stress enuff the importance of using a lot of C in the tank for both C1 and C2 for a linear amp.

For years I used breadslicers and compensated by using overly large inductors. When I finally got some vacuum var caps, my amps came to life.

In fact only recently I thought my GG triode ssb amp was running well. Turned out I needed to add another 500pf to C2 and the efficency (and power out)  came up dramatically. I ran out of C even in a vac cap. Many of the C2's out there are 1500pf. Sometimes on 75M you will need more with a big amp. The 813's are of higher impedance, thus less C is required, but I've found even with that class of amp I've always underestimated the tank C values.

Frank's suggestions will get you seeing better efficiency. Keep adding C and playing with the inductance until the power peaks and then drops off and you know you are close.

I find the power peak when tuning resonance is sharp when the Q is right, but it can be made sharper so that the Q is too high. For a linear, the plate current meter doesn't tell you dick when dipping - it's the power meter that tells the real story.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #108 on: November 20, 2009, 06:22:28 PM »

813s are higher impedance but when you have 4 in parallel you go below 2 Kohms in QRO land of class AB. Your dip should be close to max power out usually at a tad more C than dip. When you operate at higher Q the inductor circulating current goes up so it needs to be low resistance.
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #109 on: November 20, 2009, 06:35:46 PM »

Derb,
I will pay more than shipping.
Let me know where to pay pal the money to.

It might not fit, but I can try it.

I think the results I am getting are normal as far as input power and output power go.
Say you had a single 3-500z type tube, that is good for about 150 watts out, no?

I stretched the coil and added 40 pf to the plate tuning cap, no change in the lack of plate current dip.
I DO get a nice power peak.

Brett





Brett,

I have a big ass Johnson here thats at least 200 pf pr section, and at least 5kv. I'll go take a picture of it and post it, it's yours free if you want it - just paypal me some shipping money - I'm so broke at the moment I cant afford to send it out unless I get it first.  Roll Eyes

lemme get this picture.....standby.... BTW thats a 8 inch ruler for scale...this mofo is large
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #110 on: November 20, 2009, 06:57:39 PM »

Forget about carrier power when tuning the amp.

The best way is to use a pulse with a very narrow duty cycle to see the peak power without stressing the amp.. Even a 60wpm CW string of dits is a start. Look at a peak reading power meter and make your adjustments. That's where the amp gets its highest efficiency and the tank circuit should be a good match. There will be a sweet spot in the L/C ratio and tank adjustments where you will see, say 2KW out with a certain drive. Adding more drive just causes saturation... and other L/C combos produce less than 2KW out.  That's how to tune and find the best L/C in the real world.

Then when you set you AM carrier you are assured of having the maximum headroom available for a maximum carrier setting.

Later you can add a little more C2 to load it more heavily for better linearity.

T
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #111 on: November 20, 2009, 08:05:28 PM »

I measured the values:
plate tuning cap, 200 pf at max.
Coil, .006 mh,
loading cap, over 2000 pf.

Normal settings seem to be around 150pf and 1100 pf.

The zl1axb design I used uses a 100 pf plate tuning cap and 1100 pf loading cap.
They dont say about the coil...

Brett
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #112 on: November 20, 2009, 09:59:54 PM »

More tests....
I put a 40 meter coil in, and get the exact same results, 150 watts out, 800 or 900 watts pep, no dip in the plate current at all. I have the flex set to 70% carrier which explains the high pep.

I dont think anything is wrong with the pie net part of the amp.
Into the dummy load, the plate tuning is set to 60%, the loading is at 72%, output power peak is sharp.

Brett
Logged
N3DRB The Derb
Guest
« Reply #113 on: November 20, 2009, 11:36:53 PM »

Brett, on a tetrode amp you wont get much plate current dip when the amp is loaded. it does not act like a triode where no matter how hard its loaded you can still get a good dip. Tetrodes and Pentodes simply dont do that. You get a good dip when the amp is unloaded, and it progressively gets smaller the more you load it.

Heinz & Kaufman recommends that you tune the RF stage of their tantalum plate tubes by the color of the plate, not by using a plate mil meter, for that very reason. you cant do that wih 813's, but basically fool around with the output tank until you get the max blowsmoke you can get out of it, balls to the wall. use a output meter to tune it. H& K recommended a rf ammeter, but we got Bird wattmeters these days. It would be a great idea to send cw dits at 60 wpm instead of making a cb keydown, come on.  Grin

Think power transfer. You want the most blowsmoke out for the least inpoot in.  Grin Lookit the little snippett. You chasing a ghost.

let me know if you still want the cap.

* Tuning yer Pentode.pdf (41.37 KB - downloaded 570 times.)
Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #114 on: November 21, 2009, 11:25:36 AM »

Derb,
The 813s are set up as triodes.
Brett you have to decide on the power you can or want to run. 4 813s should do 1500 watts pep if your power supply is up to it.
Take your plate voltage and divide it by the peak plate current. Then multiply that number by 1.6 for class AB. This is your plate load Z. Assume your exciter will do 100 watts. Then pump 100 watts into the linear and tune it up. 4-813s should handle that drive and also handle making output power. Just watch your input power, output power and the color of the tube plates. This is where you set it up for PEP.
Once you know the peak input power and plate load Z then you have to decide on the plate Q.  12 is a good number. Then it is a simple matter to get component values from a table in most handbooks and a zillion places on line.
Say the table is set up for a Q of 10 and you want 12. just multiply cap values by 1.2 and divide inductor values by 1.2.
At full bore and 2500 volts your plate Z is going to be around 1300 to 1400 ohms. This is pretty low so another thing to look at is the plate coupling cap. I would think 2000 to 4000pf. If your inductor is really 6 uh I bet you are tuning up on 40 meters. It really wants to be around 8 uh.
If you have a signal generator pull the plate caps off and pump a signal into the loading cap with a resistor across the
 plate cap equal to the load Z. Monitor the resistor with a scope.
Logged
N3DRB The Derb
Guest
« Reply #115 on: November 21, 2009, 11:54:25 AM »

Quote
The 813s are set up as triodes.

ack. what I get for not reading the whole post.  Lips sealed
Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #116 on: November 21, 2009, 03:55:51 PM »

And I should finish my coffee before I post. Divide plate voltage by plate current. take that value and divide it by 1.6 to get load Z at class AB.
Brett, a pi network will work over a wide range but as the Z goes down the values get more critical. I'm not sure of the gain of an 813 in GG but 4 of them should make some real power. i would like to know the gain of an 813 in GG if anyone knows but suspect it should be at least 10 dB.
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #117 on: November 21, 2009, 09:40:38 PM »

Well, first off, I had a nice long qso with Eric and Nick today, with plenty of old buzzard transmissions and everything held together.

I also dont have any problem getting lots of PEP out of the amp, at 2500 volts it will do 1100 watts out, at 3000 volts it will do about 1600 or more.

I run it at 2000 volts and about 250 to 300ma on AM. That gives about 125 to 150 watts carrier out.

That gives a plate load of 4166 ohms.
My Bill Orr book says, for 4000 ohms, a plate tuning cap value of 135pf, an inductance of 15.5uh and a load capacitance of 1100pf.
Those cap values are VERY close to what I have dialed in, the coil measures lower, but I dont think the meter I have really measures micro henrys.

It is correct that i cant get much more than 900 watts pep out of it at 2000 volts, but it does not have a tuned input, so output will be lower than it could be...

2000 volts, 300 ma, 600 watts input, 150 watts output, 450 watts plate dissipation, four 813's have 500 watts of plate dissipation.
From what I hear, most guys running amps get 300 watts out of a pair of 3-500z type tubes, so i am in the ballpark.

I have not experimented much with running it at 2500 volts or higher on AM.

Brett
 

Logged
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #118 on: November 22, 2009, 12:00:03 AM »

OK Brett I see. You are setting your load Z based on the impedance at carrier.
When you modulate the plate current will rise but the plate voltage remains the same because there is no modulator. When you modulate the plate Z dives on positive peaks. This tells me you are losing efficiency on positive peaks due to low Q. At a higher Q you may not notice anything different at carrier but I bet the positive peaks improve as long as you have the drive and power supply.
2 3-500Z have 150 watts of emission.  4 - 813s have 200 watts of emission
Blow a little air past them and watch what they will do.
Logged
W1RKW
Contributing
Member
*
Offline Offline

Posts: 4410



« Reply #119 on: November 22, 2009, 08:45:21 AM »

Frank,
When I had the 813 GG amp (1 pair) if I remember correctly I had a gain of around 8 or 9.
Logged

Bob
W1RKW
Home of GORT. A buddy of mine named the 813 rig GORT.
His fear was when I turned it on for the first time life on earth would come to a stand still.
WA1GFZ
Member

Offline Offline

Posts: 11152



« Reply #120 on: November 22, 2009, 10:08:07 AM »

Bob,
I wonder if that was tuned or untuned input?
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #121 on: November 22, 2009, 11:18:41 AM »

I know a guy who had four 813's in GG with air and 4KV on them. He could easily get out 2KW pep.  He ran it like that for years without problems.

As long as you run some air past them, they will do that for you. The voltage will not bother them since they will handle 3KV in plate modulated service = 6KV + at 100% mod.


BTW, I've found most power tetrodes, when in grounded grid with tuned input are good for about 10db gain or so.  My 4X1 Quadzilla rig was no different.  We need specialized triodes like the 8877, 3cx-1200, 3-500Z, etc., to get into the 13db and higher range. But the GG tetrode does OK with IMD due to the cathode negative feedback. I ran some tests recently, especially on Quadzilla (4X1's in GG, now deceased) that showed the IMD stayed above -30 3rd order, which is good. I found very little difference between the FT-1000D and Quadzilla when the amp was driven reasonably. When hit hard close to saturation, then the garbage begins. The 813's should work out the same. The proof is how clean they run in triode connected modulator service which is basically the same config, minus the cathode negative feedback. (But I always run external NFB on modulators tubes anyway.)

T




Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
W1RKW
Contributing
Member
*
Offline Offline

Posts: 4410



« Reply #122 on: November 22, 2009, 01:34:37 PM »

Bob,
I wonder if that was tuned or untuned input?

Frank, I think it was an untuned input but now that I think about it, 90w in yielded about 600w (carrier only) or so out so I think the gain was lower than 8 or 9.  That was 10 or more years ago. Memory is foggy.
Logged

Bob
W1RKW
Home of GORT. A buddy of mine named the 813 rig GORT.
His fear was when I turned it on for the first time life on earth would come to a stand still.
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #123 on: November 23, 2009, 01:37:32 PM »

I will have to try some experiments with PEAK power and different coil inductances, although I dont think the amp lacks, pep wise.
And the way I run it, 150 watts carrier, with 800 watts or more pep seems fine.
On ssb it does at least 1600 watts pep out.
But tests are fun, so I will make some.
I also need to make a better bigger coil out of tubing, purely for looks...

Brett
Logged
K1JJ
Contributing
Member
*
Offline Offline

Posts: 8893


"Let's go kayaking, Tommy!" - Yaz


« Reply #124 on: November 23, 2009, 05:19:23 PM »

I will have to try some experiments with PEAK power and different coil inductances, although I dont think the amp lacks, pep wise.
And the way I run it, 150 watts carrier, with 800 watts or more pep seems fine.
On ssb it does at least 1600 watts pep out.
But tests are fun, so I will make some.
I also need to make a better bigger coil out of tubing, purely for looks...

Brett



Brett, if you can get out 1600w pep on ssb, then you can do it on AM too.  Find a way to add more cooling, and try the class C diode idea I mentioned to increase dead carrier efficiency. Maybe you can get 300W carrier and 1600w pep on AM. That wud be a nice improvement.  I added Dietz lantern chimneys to mine and it certainly helped bring up the plate dis..
Logged

Use an "AM Courtesy Filter" to limit transmit audio bandwidth  +-4.5 KHz, +-6.0 KHz or +-8.0 KHz when needed.  Easily done in DSP.

Wise Words : "I'm as old as I've ever been... and I'm as young as I'll ever be."

There's nothing like an old dog.
Pages: 1 ... 4 [5] 6   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.062 seconds with 19 queries.