I went and dug into my memory archives a little more and I also woke up a
little more.
It appears from the text I wrote below that my primary motivation for using
the "L-coil" as a I call it or step down transformer in the anode output was
to force the B&W 850A to work. Actually, that wasn't it at all. It was a
side benefit, but I was ready to roll my own inductor coil.
The primary motivation behind stepping down the impedance was the tune Cap.
The output capacitance of a 4-1000A is on the order of 7 pF. By the time
you add the socket and capacitance of the amplifier housing that all add to
it, you end up with a tube capacitance on the order of say 15 pF. Some
people have commented to me it could be as high as 20 pF.
With a plate load impedance of 5300 Ohms, at 28 MHz, the needed value of
Tune C for a Q of 15 is 15 pF! For a lower Q value, you need even less
capacitance.
Since the output capacitance of the tube needs to be added to the real value
of tune C, that means that tune C would need to be zero pF or less which is
not possible!
This was my primary motivation for using the step down coil. A typical
minimum value of C for a vacuum capacitor is say 10 pF. So that means, the
actually total tune C in the circuit is about 25 pF. With matching 5300
Ohms to 50, a value of 25 pF for tune C would give a Q at 10M of about 26.
A little too high, IMHO.
By stepping down the impedance to 1200 Ohms as I mentioned before, the value
of tune C needed is now in the 40 to 50 pF range. Yes, the step down coil
actually causes the effective output capacitance of the tube to be increased
as well. According to my calcs, it ends up being around 30 pF at 10m. But
now we have a little bit of head room.
Perhaps I have not explained it all that well or phrased it clearly enough.
If anyone has questions, I'll attempt to answer them. I also have the
spreadsheets available in either Excel 97 or earlier format as well. Just
let me know which you may want as they are different files. The
spreadsheets will calculate pi-network values for fixed Q across the ham
bands or at a fixed frequency will show you the values needed for different
values of Q. All the equations are shown as well. Another sheet in the
workbook contains the equations necessary to do series-parallel conversions
and calculate the effect of adding the step down inductor.
73,
Jon
NA9D
on 4/4/01 7:35 AM, Jon Ogden at na9d@mindspring.com wrote:
>
> My 4-1000A runs at an idle plate voltage of about 6000 volts. Under load,
> this voltage is around 5750 Volts with around 600 mA. So that makes the RF
> Plate Load Resistance equal to about 5324 Ohms.
>
> Plugging this number into the equations for a Pi-Net we get the following
> values for a Q=12 (12 was picked because the numbers for a Q of 10 were not
> formatting correctly on the spreadsheet and I don't feel like screwing with
> it) for the tank circuit:
>
> R1 = 5324
> R2 = 50
>
> 3.5 MHz L=20.35
> 7.0 MHz L=10.67
> 14.0 MHz L=5.39
> 21.0 MHz L=3.6
> 28.0 MHz L=2.7
>
> All values of L are in uH.
>
> Now the B&W 850A, has the following inductance values:
>
> 3.5 MHz L= 13.6 uH
> 7.0 MHz L= 6.5 uH
> 14.0 MHz L=1.75 uH
> 21.0 MHz L=1.0 uH
> 28.0 MHz L=0.8 uH
>
> WAY OFF!
>
> Now the B&W spec sheet specifically says that the impedance range it is
> designed for is 2000 to 4000 Ohms.
>
> Let's look at what our L values are for 3000 Ohms plate load and a Q=12:
>
> 3.5 MHz= 12.56
> 7.0 MHz= 6.59
> 14.0 MHz = 3.33
> 21.0 MHz = 2.22
> 28.0 MHz = 1.66
>
> So even in its "proper" impedance range, the 850A still does not have the
> best values on the higher bands. With the 850A's inductance values, the Q
> on 20 M will end up being about 24 to 25 with the 850A (1.75 uH). On 15M,
> the Q ends up being about 28 to 29 (1.0 uH). And on 10M, the Q will end up
> around 26 to 27 (0.8 uH). And that's with the impedance being in the
> "proper" range! So the 850/850A were not good even for the lower impedance
> ranges!
>
> In my case the Q on 10M would have come out to be a whopping 45! That is
> NOT good for proper tank circuit efficiency.
>
> Now, Bob what I did that you may have gotten the idea of something "lower"
> from was to lower my load impedance through the use of a step down
> transformer. It is an inductance placed in series with the DC blocking cap
> before the tune cap. By doing series to parallel conversions, you can
> calculate the math on this. This inductance acts as a simple step down
> transformer and brings the load impedance of 5324 down to a more manageable
> level for more reasonable Q.
>
> By adding 1.1 uH in series as I describe, you end up transforming the plate
> load impedance to about 1287 Ohms at 10m, 2721 Ohms at 15m, 4072 Ohms at
> 20m, 5014 Ohms at 40m and 5265 Ohms at 80m.
>
> It's much easier to match the 1287 Ohms to 50 Ohms at 28 MHz. With the 0.8
> uH that the B&W has, we now have a reasonable Q of about 12 to 13.
>
> Now, my coil that I added may or may not have an inductance of 1.1 uH. I
> had no real good way to measure it. I had to end up removing the 10m coil
> on the 850A and replacing it with just a wide copper strap. So I suspect
> that my inductance value may be a little different.
>
> Thanks for bringing this up. I'd wanted to post my design about the step
> down transformer for these new guys and I'd lost all previous copies of
> stuff I sent out about it.
>
> I can't take credit for this, since someone else on the list suggested it.
>
> One additional nice thing about this transformer was that adding it shifted
> the VHF resonances in the plate circuit WAY away from the 80 MHz or so that
> you get with a 4-1000A. Using a grid dip meter prior to adding the coil,
> the plate circuit had a sharp resonant dip right around 80 MHz. After
> adding the coil, the dip was gone and I could not find it anywhere from 50
> to 120 MHz. My amplifier is stable and the only suppressor network I have
> is a single turn coil and NO suppressor resistor. I took it out as it was
> continually burning up due to the circulating currents at 10m. Some other
> old timers suggested that the 4-1000A could be made stable without it.
> Indeed they were right!
>
> I have mistuned my amplifier into bad loads, operated it on the wrong band
> and regularly run full legal limit on all bands and have had no problems
> with it in over 2 years other than having one of my tubes go gassy.
>
> So Bob, before you go off half-cocked why don't you get your facts straight.
>
> 73,
>
> Jon
> NA9D
>
>
> -------------------------------------
> Jon Ogden
> NA9D (ex: KE9NA)
>
> Member: ARRL, AMSAT, DXCC, NRA
>
> http://www.qsl.net/ke9na
>
> "A life lived in fear is a life half lived."
>
>
> --
> FAQ on WWW: http://www.contesting.com/FAQ/amps
> Submissions: amps@contesting.com
> Administrative requests: amps-REQUEST@contesting.com
> Problems: owner-amps@contesting.com
>
-------------------------------------
Jon Ogden
NA9D (ex: KE9NA)
Member: ARRL, AMSAT, DXCC, NRA
http://www.qsl.net/ke9na
"A life lived in fear is a life half lived."
--
FAQ on WWW: http://www.contesting.com/FAQ/amps
Submissions: amps@contesting.com
Administrative requests: amps-REQUEST@contesting.com
Problems: owner-amps@contesting.com
|