[Amps] L-PI amp network

Sat Dec 17 17:15:35 EST 2022

Jim,

Thanks a bunch for your input.

I was close.

I had used .65 for the current and a "K" factor of 2.0 as I'm 99% CW.  
It's been DECADES since I was above 28.1 MHz.  I could have easily 
chosen 1.8.   I also had '9999' for the choke and .1 for suppressors.  I 
also had '5pf' for strays, where you used '3pf'. At least, I wasn't 
totally lost.  I was counting on being able to use GM3SEK's calculator.

Having said that, I followed along with your numbers, and I can ALMOST 
get what you got.  I'm looking for where my entered data is different 
from yours.  At one point, I hit the comma instead of a decimal period.  
Boy, did that mess things up.  Had a heck of a time finding the problem.

Not sure how you got a Q of 19.  I had entered a Q of 12 and that was 
reflected in the C1/C2 info from the get-go.

How much of a big deal is it going to be if I use 1/4 copper tubing?  
It's what I've got.   I've got many feet of .015 x 2.00 inch copper, but 
even if I split it, it's too thing to self support.  Do you have a good 
source for suitable copper strap?

I'm lucky as I have a 14 inch metal lathe I use to wind these coils.  In 
the past, I've used #4 solid copper (3/16 inch) as I had a bunch of it.  
Makes good solid coils, but it work hardens, making changes a PITA.  I 
DO put the lathe in neutral and turn it by hand...

Steve, NR4M

On 12/17/2022 12:55 PM, jim.thom jim.thom at telus.net wrote:
> Date: Tue, 13 Dec 2022 17:36:44 -0500
> From: Steve Bookout <steve at nr4m.com>
> To: amps at contesting.com
> Subject: [Amps] L-PI amp network
>
>
>   I cycled all of the unknown tubes thru
> this amp.? They all were similar, taking 22 to 27 watts of drive for the
> same 1500 watts out;? 550 ma to 600ma @.030 ma grid.
>
> That was the background for my real question for the group.
>
> Based upon GM3SEK's PI/PI-L calculator, I'm working with a plate load
> impedance of ~3000 ohms.? I know from my own experience, and that of
> others, that it's a pain to get the plate inductor 'right' so that it
> all works and tunes, as it should and make power with reasonable
> efficiency.
>
> I have never used an L-Pi, but I basically understand the idea.
>
> IS THERE A RECOMMENDED TOOL OUT THERE, WHICH WILL HELP ME FIGURE OUT MY
> TWO "L's"?
>
> Using GM3SEK's calculator, I see there is a line, 'Lead inductance
> (total from tube to tank, but excluding suppressor)', which I have
> played with.? Adding my 'L' inductance there, say '.6uh', does reduce
> the inductance associated with PI, and it does raise the value of the C1
> capacitor. ? ( I am using a 100 pf /15 KV vacuum variable and right now
> as I write this, I don't remember it's minimum C, but it's in the range
> of what I think you should see; maybe 5 pf?)? Using this, I was able to
> come up with about 15 pf, vs 4 pf in a standard PI configuration.
>
> Suggestions on if this would be an accurate way to figure this out? ?
> Or, another/better way? ? And, what should I be looking for as maybe a
> ratio of the 'L' vs the 'L' in the PI network. ? As I recall, I think
> I'm looking at about 1.7 uh.
>
> I would appreciate any constructive comments, or even precautionary ones.
>
> 73 de Steve, NR4M
>
> ## Steve, I have been using the GM3SEK spreadsheet for a long time now,
> since it 1st came out.  It's what I use to design the L-PI networks.  It
> only works on tubes that DON'T require a suppressor.
>
> ## assume 1.5 kw out..and a tank eff of 60% on 10m.  1500/.60 = 2.5 kw DC
> input.
> 2500w / 4000vdc =  625 ma required plate current.   (28.6 mhz)
>
> ## Use 1.8 for the 'K'.   4000 / (1.8 x .625) =  3555 ohms  plate load Z.
> With an overall network Q of 19 , line 39 (input q = 17.2   output q = 1.8)
> Enter 0  for all the parasitic lps/rps on lines 46+47.
> Enter 10 pf on line 42
> Enter 5 pf on line 43
> Enter 3 pf on line 53
> Enter 50 uh on line 51
> Enter 0 on line 48
> Enter 0 on line 52
>
> Here's what u get.
> C1 = 9.6 pf
> L =  1.26 uh
> C2 =   198.5 pf
>
> An overall  Q of 19 sucks.   11.825 amps of current through the tank
> coil..which is a bunch.
> I would use 3/8" OD tubing...or 1/2" to 5/8" wide strap coil.  With tubing
> flattened at each end, then enough width to punch a hole big enough for a
> 1/4-20 brass screw.   DON'T use  SS  hardware.   It will turn black on
> you..and ditto with plane jane ferrous steel machine screws.
>
> ##  Ok, let's use the L-PI config, which is really a C-L-C-L-C
> We will use the tube's anode to grid C (+ strays)  plus the 1st coil, to
> form a step down L network..... stepped down low enough, that a normal PI
> net can be designed around, with a lot lower loaded q.
>
> Enter .8 uh on line 48.   Lower the overall Q on line 39  to   11.
> Ideally, you want this 1st coil to be a little less than the main coil.
> Yes, u have to keep juggling with the overall q on line 39, each time u
> increment the 1st coil on line 39.   Start with .1 uh..and keep increasing,
> while at the same time decreasing the overall q on line 39.
>
> Ok, here is what we get.   Note, the overall network Q on line 39 is = to
> the (input Q + output Q).
>
> Overall network Q of 11   (28.6 mhz)
>
> C1 =   12.7 pf     ( input Q of  9.5)
> L =   .91 uh
> C2 =   171.6 pf.    (output Q of 1.5)
>
> Note the 1st coil's value (.8uh)  is slightly less than the main coil.
> Again, use  3/8" tubing or  1/2" to 5/8" wide strap coils for BOTH coils.
> Yes, use ur 100 pf vac cap. You won't have any issues, and harmonic
> suppression is typ higher when a vac cap is used.  Air caps have the issue
> with the rotor wiper contact.  I'm assuming ur vac cap goes down to 5 pf.
>
> Harmonic suppression with the L-PI is superb.  With the lower Q,  you won't
> have to constantly be re-tuning  when u qsy.
>
> Jim  VE7RF
> _______________________________________________
> Amps mailing list
> Amps at contesting.com
> http://lists.contesting.com/mailman/listinfo/amps