# an RF amplifier with transistor and 807 tube for CW

Discussion in 'Amateur Radio Amplifiers' started by ASHOKDAS81, Apr 17, 2021.

Tags:
1. ### ASHOKDAS81QRZ Member

Thanks to all for the documents and comments. I am now studying these to understand. Will come back again once I have gone through all of these

PU2OZT likes this.
2. ### AC0OBPlatinum SubscriberPlatinum SubscriberQRZ Page

Hi ASH, Jim brings up a good point.

I think your source rig was meant to put out 0.8 Watts and work into a 50 ohm load, where V
rms = SQRT(50ohmsX0.8W) = 6.3 voltsrms or 9 volts peak. That won't be enough to force the 807 into the ClassC conduction regime.

We're biasing the 807 stage at approx. -28 to -30 volts. If my amplifier supplies a peak voltage of 17.5 volts, then the grid voltage would reach about -10 volts on the positive peak which, according to the 807 tube curves at 600 volts DC plate voltage, is the point where the 807's plate current is 100 mA. (I always design tube circuits using RCA, GE, Sylvania, etc., tube data and curves).

If you want to use your QRP rig as the driving source, then you would need to load your output with a 50 ohm 2 Watt non-inductive resistor and a "step-up" transformer with a voltage transformation ratio of at least 1:3 (primary to secondary), yielding an impedance ratio of 0.1:1. ":" means 'to.' That would result in a grid drive voltage of about 27 volts peak.

As always, there are many approaches to attack electronic circuit designs.

Pheel

Last edited: Apr 20, 2021
3. ### W9JEFPlatinum SubscriberPlatinum SubscriberQRZ Page

If the 807 operates Class C, then the grid current should be enough of a load, eliminating the need for a resistor at the output. I would recommend about a 1:10 link-to-main-coil turns ratio, and of course a .001 µfd coupling capacitor feeding the grid, which should be biased to cutoff.

73,
Jim . . . . . . . . . . . . EARTH: LOVE IT OR LEAVE IT

4. ### AC0OBPlatinum SubscriberPlatinum SubscriberQRZ Page

Here is another approach using your QRP rig as the RF source and with a toroidal RF transformer (T130-2 core) which will give you some experience winding coils. See upper left of schematic.

Theory: Your QRP rig feeds a controlled load impedance of 50 ohms. A series resonant circuit is tuned to 7.16 MHz.

The primary of the RF transformer resonates with the 100 pF cap to resonant at 7.16 MHz to provide maximum current through the primary winding. The turns ratio of 3.5 gives us ~ 32Volts peak on the secondary. 20 to 22 gauge enameled wire should suffice for the primary and secondary windings.

At a cathode bias of 32 volts, this will force the 807 to conduct for about 140 degrees of the total 180 degree positive sine wave.

Much to consider and with which to experiment.

Pheel

#### Attached Files:

• ###### 807 AMP CW Cathode Keying V2.pdf
File size:
42.4 KB
Views:
11
Last edited: Apr 20, 2021
5. ### AC0OBPlatinum SubscriberPlatinum SubscriberQRZ Page

Dang it! Forgot to DC decouple the transformer secondary from the grid with a cap, as we will now have grid leak bias well.

I also added a cap across the secondary in case you need to resonate it for more voltage. I would try the circuit without it at first. I haven't determined the self capacitance of the secondary but it can't be more than a couple of pF.

If one needs more drive voltage for the control grid one could also increase the secondary turns to 16 turns.

Pheel

#### Attached Files:

• ###### 807 AMP CW Cathode Keying V2.pdf
File size:
44.1 KB
Views:
18
Last edited: Apr 20, 2021
6. ### ASHOKDAS81QRZ Member

Thanks to all for sharing so many valuable inputs to the idea. AC0OB can you please share me the LTSPICE tube model that you are using.

7. ### ASHOKDAS81QRZ Member

Toroids are very difficult to find in my country. I will go for other approaches that Pheel suggested. For making coils I have 0.75 and 1.25 inch dia PVC tubes. Another hurdle is air variable caps. Those are not available also. I am planning to use high voltage fixed capacitor, but how to calculate the values ?

Ashok

8. ### AC0OBPlatinum SubscriberPlatinum SubscriberQRZ Page

My tube simulation models are all in MatLab.

If you need the .SUBCKT listing for Spice I have a limited amount of those as well.

If you are asking about the schematic symbols for tubes, you will find them in the Select Component Symbol, [Misc] LT directory.

Pheel

K4YNZ likes this.
9. ### ASHOKDAS81QRZ Member

Hi Pheel, thanks for the reply. Yes I have found the tube symbol and successfully attached it to a .SUBCKT file. It looks OK during simulation. I am now playing with it to get some insight into the tube amplifier and understand it.

10. ### AC0OBPlatinum SubscriberPlatinum SubscriberQRZ Page

Marshall Leach has a good explanation on constructing tube models in Spice: Spice Models for Vacuum Tube Amplifiers, Journal of the Audio Society, Vol. 43, No. 3, March 1995;

http://leachlegacy.ece.gatech.edu/papers/tubeamp/tubeamp.pdf and

https://digilander.libero.it/paeng/spice_models_for_vacuum_tubes.htm and

https://www.agsrhichome.bnl.gov/AP/ap_notes/ap_note_97.pdf

For the 807, one could modify this 6146 .SUBCKT model by changing the parameters:

.SUBCKT 6146 A S G K
*
* Calculate contribution to cathode current
*
Eat at 0 VALUE={0.636*ATAN(V(A,K)/10)}
Eme me 0 VALUE={LIMIT{V(A,K),0,10000}/130}
Emu mu 0 VALUE={PWRS(V(G,K),1-(LIMIT{-V(G,K),30,9999}-30)/2000)}
Egs gs 0 VALUE={LIMIT{V(A,K)/10+V(S,K)*4.5+V(mu)*19,0,1E6}}
Egs2 gs2 0 VALUE={PWRS(V(gs),1.5)*16E-6}
Ecath cc 0 VALUE={LIMIT{V(gs2)*V(at),0,V(me)}}
Esd sd 0 VALUE={LIMIT{V(gs2)*V(at)-V(me),0,99999}}
*
* Calculate anode current
*
Ga A K VALUE={V(cc)}
*
* Calculate screen current
*
Escrn sc 0 VALUE={2.5E-4*PWRS(V(S,K),1.5)*V(gs2)*(1.1-V(at))+V(sd)}
Gs S K VALUE={V(sc)}
*
* Grid current
*
Gg G K VALUE={PWR(LIMIT{V(G,K)+1,0,1E6},1.5)*(1.25-V(at))*870E-6}
*
* Capacitances
*
Cg1 G K 13.0p
Cak A K 8.5p
Cg1a G A 0.24p

I think you will find tube modeling and simulation a fun part of any electronic hobby.

Pheel