[Amps] Grid Dipping the Pi network in a new amp? --- ( Itnow Functions)

[Robert Bonner]

Tom,

I agree totally with you and I realize "MAGIC Q" is just flat BS..  12 a
nice average middle of the road number that has good characteristics.
Somebody probably at the league chose to work with, I prefer 13.2 myself
just to be different...  WINK...

The problem with this instance is Chuck has bought two Vacuum variables
already and they are on the ragged fringe of working with the coil. 284pF
measured on his tuner (which should work with <9uH). If you read his test
notes he's having to parallel up a handful of padder caps to get the coil
into the realm of tuning.  After the first round, he was still way up above
the 75 meter band.  His Loader cap is in the range of 1200pF measured at
full position without the padder.

If she don't resonate, she won't work.  So formula numbers always need to be
bumped a little here or there.

Some of my other comments about the long lead over to the 10 meter coil has
already been addressed, as he's going to have to make it work on other bands
after 80.  He's been referencing photos of the Bill McCullough amp from the
late 1960's and mirrored the layout, once he turned the ten meter coil
around.  I am familiar with that amp as I was bidding on it when it sold on
eBay.

My other comment about machining out the cap bracket so the tuner could
ground directly would take the series lead inductance out of the cap.  But
Chuck has mentioned the machine costs were very high so we missed the window
on that setup.  Again cost prohibitive.

Me personally, I have a large stock of 500pF 15KV ceramics I use for tuners
and 1500pF 5KV I use for loaders when I build my 80-10 projects.  I think I
still have 4 or 5 sets left of them.  The 500 will also just make 160 meters
but is close.  Again we play with the coil to get it to work.  I wish I had
a box full of 2000's 3000's and 4000's here and not the 1500's but not so
lucky.

The 1500 on 80 meters is pushing the ragged edge, I think I'm about 4 turns
in on the PLASMATRON (2-8877 amp) when it's on 80.  That's a bit closer than
I'd like to have it.  But I just fudge a tap or two over on the coil and it
brings the thing right in.  If I put it where it was supposed to be, I'd not
have enough cap.

You are misunderstanding my position on his coil Tom...  When I refer to it
as not working, I mean it is past the tunable range of existing
components...  That's all.  I'm looking at the total picture of what he's
done and suggesting slight modifications on his existing coil to bring the
caps into reason.  I think he can get them to work at least the tuner, the
loader will probably still need a little help.  IF he completely threw the Q
out the window, I'm sure he could get it running with the loader he has by
adding a few extra turns of coil.  

Like I offered, I have these EXTRA switch contacts here in the parts bin so
he could add the "Extra" caps on the loader when dropping to CW. As the B&W
has a 6th position that isn't used.

At 284 pF max on the tuner, he'll probably only need to add a couple turns
to the tank coil to get it to run with the existing cap.  I just hate
putting padders on tuner caps.  Here you have these great vacuum units and
then stick a 150pF / 7.5KV cheapo cap across it.  

Chuck is a big Tesla guy (as I am) so winding coils is his game...  Just
little stuff him and I have been talking about.  He's going to have a coil
beefier than an 850A if he builds it like he's been talking to me on the
back channel.  He could also have it silver flashed and it would look great.

SO, Just to clarify and get us both on the same page..  Caps are already
purchased items and might need to be padded, or replaced at great expense.
(Not feasible) coil is pretty easy to modify and inexpensive to do.
Changing voltages is cost prohibitive.

I tried a lot to get the 852 to do what I wanted it too back in the 70's and
gave up. As the caps just got too big.  I answered my issues at the time
with heavily modified 850A's.  Let's face it nobody puts the taps in the
right place as each homebrew construction is different.  To counter that a
guy needs a little more tuning range in the caps.  The problem here is not
enough cap.  If he starts switching caps in and out he'll booger up the amp
more than building a new coil.

SO anyway, that's my thinking on it and let's get it to "WORK" without extra
switching if we can.

BOB

-----Original Message-----
From: Tom W8JI [mailto:w8ji at w8ji.com] 
Sent: Saturday, July 21, 2007 7:08 PM
To: rbonner at qro.com; 'Chuck Curran'; amps at contesting.com
Subject: Re: [Amps] Grid Dipping the Pi network in a new amp? --- ( Itnow
Functions)


> Good testing Chuck.  You confirmed what I had said earlier 
> this week.  The
> maximum plate load impedance for the 852 is around 
> 1400-1500 Ohms and
> responds best down in the 1000 range.

Bob,

After just having dinner with my wife and here sister, I'm 
probably a little too sensitive about drama. A helicopter 
lands down the street, and the drama queens say it landed in 
the front yard. A two inch Pine Vole comes up and eats an 
apple, and I have to cut the whole tree down because it 
attracts giant rats.


It's the same with tank circuits. People write as if there 
is some mystical magical point where the tank quits working 
on the hi-Q side of minimum Q. People are fixated on that 
exaggeration.

Let's look at the real numbers in tank with a Q of 12 at 
1500 ohms with typical vacuum caps and a B&W coil and see 
how much is fact and how much is drama:

3.75MHz
Anode loadline 1500  j0
Anode RF Power 1500W
Anode required C 330.2 pF
L 6.45uH
Load C 1632pF
Loss 2.8%
Phase delay 147.5 degrees
Effective Q 12.0
Inductor current 11.6 amperes
3dB bandwidth (without retuning) .31 MHz

Now let's increase the anode loadline to 3000 ohms and see 
if the tank "stops working"

3.75MHz
Anode loadline 3000  j0
Anode RF Power 1500W
Anode required C 320.2 pF
L 6.45uH
Load C 2400pF
Loss 5.3%
Phase delay 158.5 degrees
Effective Q 23.8
Inductor current 15.8 amperes
3dB bandwidth (without retuning) .16 MHz

It doesn't "quit working". The anode capacitor doesn't 
change, the loading capacitor increases to about 1.4 times 
the original value, the loss increases from ~3 to ~5 
percent, the bandwidth (selectivity) doubles.

Now let's look at the same component quality and a 3000 ohm 
tank at an effective Q of 12.

3.75MHz
Anode loadline 3000  j0
Anode RF Power 1500W
Anode required C 166.5 pF
L 12.1 uH
Load C 992pF
Loss 2.7%
Phase delay 134.3 degrees
Effective Q 12
Inductor current 8.2 amperes
3dB bandwidth (without retuning) .31 MHz

For all the popular drama that "without a Q of 12 the amp 
will stop working", the overall efficiency of the amp might 
change from 65 to 63 %.

The REAL major effects we would notice are only in the value 
of the capacitors required. It doesn't stop working. It 
isn't "only good for 1500 ohms".

If I had an amp as nice looking as the one Chuck built, I'd 
just increase the size of the loading cap and ignore the 2 % 
change in overall efficiency. The loading and tuning caps 
should have at least 50% capacitance headroom anyway in case 
the load isn't perfect.

73 Tom