Gents

On these 3CX-6000's and also the socketless version [ YC-243],
I fool with from time to time, Eimac/ Svetlana state's the grid
diss at 225 Watts CCS. [3CX-3000A7 is also 225 w CCS grid]

Now, since the tube(s) can be operated in either GG OR grid
driven.... it occured to me that in grid driven mode... the grid
should still has a 225 w CCS rating ???

The point here is..... in GG service... you bolt this huge grid
ring to the chassis.... and in the case of the socketless
version.... an even bigger diam grid flange is bolted to the
chassis !

Now... that alone should heatsink the grid ??? Since the grid
is sitting in a vac... the only way for thermal heat to escape is
via the grid flange.

Since the below chassis compartment is pressurized... one might
think the actual grid dissipation can perhaps be increased via
this heatsink action ?? [GG mode]

Am I out to lunch here ???

Later..... Jim VE7RF

--- In ham_amplifiers@..., "pentalab" <jim.thomson@...>
wrote:

Gents

On these 3CX-6000's and also the socketless version [ YC-

243],

I fool with from time to time, Eimac/ Svetlana state's the

grid

diss at 225 Watts CCS. [3CX-3000A7 is also 225 w CCS grid]

Now, since the tube(s) can be operated in either GG OR grid
driven.... it occured to me that in grid driven mode... the grid
should still has a 225 w CCS rating ???

The point here is..... in GG service... you bolt this huge grid
ring to the chassis.... and in the case of the socketless
version.... an even bigger diam grid flange is bolted to the
chassis !

Now... that alone should heatsink the grid ??? Since the

grid

is sitting in a vac... the only way for thermal heat to escape is
via the grid flange.

Since the below chassis compartment is pressurized... one might
think the actual grid dissipation can perhaps be increased

via

this heatsink action ?? [GG mode]

Am I out to lunch here ???

Later..... Jim VE7RF

### Did anybody actually read the above ???

Later.... Jim VE7RF

开云体育

ORIGINAL MESSAGE:

On Thu, 21 Sep 2006 23:41:38 +0200, "Peter Voelpel"
<df3kv@...> wrote:

I would answer if the message was not html, where I can?t quote parts but
have to resend the whole bunch.
I will not allow html on my mailer.
I don?t want to go to the yahoogroup website everytime just for answering
messages.
But I go there periodically and answer then

------------ REPLY FOLLOWS ------------

Slightly OT, but if you use Forte's Agent program you can switch
between HTML display and plain text by pressing the "/" key. Neato.

Bill, W6WRT

Jim,

What you describe may help a little. The problem is the internal connection
from the ring to the grid itself, and it's size. If it's only a piece of
wire, you may not gain anything as the wire would be the bottleneck I would
think. The ring would help a little by being a larger heatsink for the
connection itself, but the grid itself is still wire which it's size and
coating (gold, etc) can only take so much current. The ring would have to
dissapate the heat from the grid via the internal connection(s).

Best,

Will

From: ham_amplifiers@... [mailto:ham_amplifiers@yahoogroups.

com]

On Behalf Of pentalab
Sent: Donnerstag, 21. September 2006 22:12
To: ham_amplifiers@...
Subject: [ham_amplifiers] Re: Can Grid Dissipation in metal GG triodes be
increased ????



--- In ham_amplifiers@ <mailto:ham_amplifiers%40yahoogroups.com>
yahoogroups.com, "pentalab" <jim.thomson@>
wrote:

Gents

On these 3CX-6000's and also the socketless version [ YC-

243],

I fool with from time to time, Eimac/ Svetlana state's the

grid

diss at 225 Watts CCS. [3CX-3000A7 is also 225 w CCS grid]

Now, since the tube(s) can be operated in either GG OR grid
driven.... it occured to me that in grid driven mode... the grid
should still has a 225 w CCS rating ???

The point here is..... in GG service... you bolt this huge grid
ring to the chassis.... and in the case of the socketless
version.... an even bigger diam grid flange is bolted to the
chassis !

Now... that alone should heatsink the grid ??? Since the

grid

is sitting in a vac... the only way for thermal heat to escape is
via the grid flange.

Since the below chassis compartment is pressurized... one might
think the actual grid dissipation can perhaps be increased

via

this heatsink action ?? [GG mode]

Am I out to lunch here ???

Later..... Jim VE7RF

### Did anybody actually read the above ???

Later.... Jim VE7RF

--- In ham_amplifiers@..., "craxd" <craxd1@...> wrote:

Jim,

What you describe may help a little. The problem is the internal

connection

from the ring to the grid itself, and it's size. If it's only a

piece of

wire, you may not gain anything as the wire would be the

bottleneck I would

think. The ring would help a little by being a larger heatsink for

the

connection itself, but the grid itself is still wire which it's

size and

coating (gold, etc) can only take so much current. The ring would

have to

dissapate the heat from the grid via the internal connection(s).

Best,

Will

#### Tnx Will !

### The grid in both a 3000/6000/YC-243 is 225 W CCS... and
resembles a bird cage. The joke is.. you can use it for a dummy
load for your 1000-D .

### The next trick is how to calculate the grid diss ?

### Eimac only has 'specs' on the 6000A7 for GG FM broadcast
service. They rant on about 'class C'... but the specs clearly
show zero bias operation. The 6000/YC-243 have a MU of
200.... and per Reid Brandon at Eimac.... make an excellent
tube for ssb / good imd.

### Eimac can't even tell me what the input Z is in GG.... so we
used W7IUV's slick method. Guestimate.. and build for one
band..say 40m. After its tweaked for flat swr[while driving
the amp] .... leave it alone... and stick the mfj on the input
side... and substitute resistors on the cathode side... till the
swr is flat on the mfj. What ever this resistance is ... is the
tubes's input Z ! Works every time.



### What we see is with 800 w of drive... output is 11,500
watts..... and about 640 ma of grid current. 2.5 A of plate
current @ 6800 vdc under load..... all with a dead cxr.

### Eff is 67% on all bands 160-15m. We tweaked the Pi
net for a Q of 8-10

#### the tuned input is unique too. Instead of the usual 9x
bandswitched 'pi nets'.... we used 2 x broadcast variables ..all
4 x sections strapped in parallel= 72-2080 pf.... and a small
4uh bandswitched coil... made from either 7 or 8 ga solid
copper wire. 17 turns 1.5" ID. [BTW... Multronics still
makes a 4.4 uh roller coil... made from 8 ga wire]

### We used an extra wafer, so we could pad the pair of
broadcast caps on 160m. [4 x 500pf doorknobs per cap]

### We actually installed bird line sections on BOTH sides of
the tuned input.... and with 200 w in... were only getting 160w
out on 20-17-15m... and 195 on 160-40m. Turns out the Q
was TOO high on 20-15m. This nonsense about using a loaded Q
of 5 for a tuned input is just that. I'm talking about a real
loaded Q of 5.... not Eimac's / Rich's method.. based on C1
only. Rich's Q of 2 is in fact 3.3 to 3.6.... which is
about what one wants.

### with a real Q of 5-6... you could actually get 8 ga wire
warm ! 200w = 2 A into 50 ohms.




### We increased the UH on 20-17-15m just barely enough...
so that power out of the tuned input shot up to 190-195 watts.


### Now with 750-800 watts dead cxr..... the tuned input runs
just fine on all bands.... bullet proof. And with vernier's on
both caps.... it's just 'dial up by the numbers' . You get real
flywheel action to boot. In addition... max grid current and
min input swr coincide.

### In actual operation... we pulse tune everything.

later... Jim VE7RF

That is a good point for tubes like the YC-156/179. On tubes like the planar constructed GS-35B, the grid is tied directly to the ring and can move more heat. Unfortunately, the actual construction of the grid itself may actually be the limiting factor on the GS-35.

73, Tony W4ZT


craxd wrote:

Jim,

Grid dissipation is measured by ohms law, grid voltage by grid
current.
Since the grid is at ground, you just measure the catode voltage with
a scope.

For calculation of efficiency you have to deduct the drive power first
in GG, about 70% of the drive power is seen as output power.
Drive power is not lost totally as with GK.

If both capacitors of your input network are the same value you have
no transformation, you will have 50 ohm input from the transceiver
and 50 ohm output to the cathode side.
I would expect the cathode impedance smaller at that power, probably
30-40 Ohms.
I use that tunable pi-network on my 3CX3K amp, but used a smaller
capacitor on the ouput then on the input. My coil is switched as well.
I would like to use ganged capacitors there and ganged with a roller
at the same time, should work with that low circuit Q.

Any idea of the price of the 4?H roller by multronics?


73
Peter

--- In ham_amplifiers@..., "craxd" <craxd1@> wrote:

### The next trick is how to calculate the grid diss ?

### What we see is with 800 w of drive... output is 11,500
watts..... and about 640 ma of grid current. 2.5 A of

plate

current @ 6800 vdc under load..... all with a dead cxr.

### Eff is 67% on all bands 160-15m. We tweaked the Pi
net for a Q of 8-10

BTW... Multronics still
makes a 4.4 uh roller coil... made from 8 ga wire]

--- In ham_amplifiers@..., "Peter Voelpel" <df3kv@...>
wrote:

Jim,

Grid dissipation is measured by ohms law, grid voltage by grid
current.
Since the grid is at ground, you just measure the cathode voltage

with

a scope.

### whoa. Are you saying RF cathode V x DC grid current
= grid diss ????

### That equates to way below 200 w of grid diss.... based
on 45 ohm input Z... and 800w of drive . I can see we
are well within the grid diss ratings.

For calculation of efficiency you have to deduct the drive power

first

in GG, about 70% of the drive power is seen as output power.
Drive power is not lost totally as with GK.

### I never deduct the drive power... since it really enhances
the eff in the end anyway... another benefit of GG. Bottom line
is... "apparent eff" goes up... anode diss goes down a bit.

If both capacitors of your input network are the same value you

have

no transformation, you will have 50 ohm input from the transceiver
and 50 ohm output to the cathode side.

### yes... u can see that effect on GM3SEK's excellent pi/ pi-l
spreadsheets. Funny how every arrl handbook/qst article
using fixed silver mica's always used same values for C1 +
C2.... which never worked.

I would expect the cathode impedance smaller at that power,

probably

30-40 Ohms.

#### it's about 40-45 ohms. The gain is about 11.54 db at
11.5 kw out.

### Funny thing is... with just 100w of drive... power out is 3
kw [ 3CX-6000A7] It appears on all these big tubes... that
gain is higher with low drive levels... then decreases... with
increasing drive levels.

### The 3000A7 is abt 1500w out with just 50 w of drive.

### Tubes like the higher gain YC-156 have aprx 25 ohm input Z.
Gain is almost quasi proportional between plate load Z and input
Z.

I use that tunable pi-network on my 3CX3K amp, but used a smaller
capacitor on the ouput then on the input. My coil is switched as

well.

I would like to use ganged capacitors there and ganged with a

roller

at the same time, should work with that low circuit Q.

### Henry tried that in the 4 K ultra. Trying to get it to
track was [still is] a real bitch. I wouldn't do it. For
160-10m... u need 4000 pf caps. If u pad smaller 2100 pf caps
like I did for 160m... ur tracking will be out to lunch. You
also seem to get anomoly's from one band to the next.

Any idea of the price of the 4?H roller by multronics?

### No clue It appears identical to their 28 uh unit that
was featured in many qst/ handbook articles over the years...
excpet the little 4.4 uh unit is shorter... and uses 8 ga wire
[.128" solid tinned alloy wire]. I believe the 4.4 uh unit also
has the same tapered pitch at the high freq end.

### Multronics is now a division of Cardwell condenser corp.... I
think it's just cardwell condensercorp.com

Later... Jim VE7RF

73
Peter

On Sep 22, 2006, at 1:54 PM, pentalab wrote:

--- In ham_amplifiers@..., "Peter Voelpel" <df3kv@...>
wrote:

Jim,

Grid dissipation is measured by ohms law, grid voltage by grid
current.
Since the grid is at ground, you just measure the cathode voltage

with

a scope.

### whoa. Are you saying RF cathode V x DC grid current
= grid diss ????

That's pretty much what Care and Feeding ... ... says. Avg. DC grid-
current is measured with an ordinary panel meter.

### That equates to way below 200 w of grid diss.... based
on 45 ohm input Z... and 800w of drive . I can see we
are well within the grid diss ratings.

For calculation of efficiency you have to deduct the drive power

first

in GG, about 70% of the drive power is seen as output power.
Drive power is not lost totally as with GK.

### I never deduct the drive power... since it really enhances
the eff in the end anyway... another benefit of GG. Bottom line
is... "apparent eff" goes up... anode diss goes down a bit.

If both capacitors of your input network are the same value you

have

no transformation, you will have 50 ohm input from the transceiver
and 50 ohm output to the cathode side.

### yes... u can see that effect on GM3SEK's excellent pi/ pi-l
spreadsheets. Funny how every arrl handbook/qst article
using fixed silver mica's always used same values for C1 +
C2.... which never worked.

I would expect the cathode impedance smaller at that power,

probably

30-40 Ohms.

#### it's about 40-45 ohms. The gain is about 11.54 db at
11.5 kw out.

### Funny thing is... with just 100w of drive... power out is 3
kw [ 3CX-6000A7] It appears on all these big tubes... that
gain is higher with low drive levels... then decreases... with
increasing drive levels.

### The 3000A7 is abt 1500w out with just 50 w of drive.

### Tubes like the higher gain YC-156 have aprx 25 ohm input Z.
Gain is almost quasi proportional between plate load Z and input
Z.

I use that tunable pi-network on my 3CX3K amp, but used a smaller
capacitor on the ouput then on the input. My coil is switched as

well.

I would like to use ganged capacitors there and ganged with a

roller

at the same time, should work with that low circuit Q.

### Henry tried that in the 4 K ultra. Trying to get it to
track was [still is] a real bitch. I wouldn't do it. For
160-10m... u need 4000 pf caps. If u pad smaller 2100 pf caps
like I did for 160m... ur tracking will be out to lunch. You
also seem to get anomoly's from one band to the next.

Any idea of the price of the 4?H roller by multronics?

### No clue It appears identical to their 28 uh unit that
was featured in many qst/ handbook articles over the years...
excpet the little 4.4 uh unit is shorter... and uses 8 ga wire
[.128" solid tinned alloy wire]. I believe the 4.4 uh unit also
has the same tapered pitch at the high freq end.

### Multronics is now a division of Cardwell condenser corp.... I
think it's just cardwell condensercorp.com

Later... Jim VE7RF

73
Peter

Yahoo! Groups Links

R L MEASURES, AG6K. 805-386-3734
r@...