The AM Forum
March 28, 2024, 11:55:49 AM *
Welcome, Guest. Please login or register.

Login with username, password and session length
 
   Home   Help Calendar Links Staff List Gallery Login Register  
Pages: [1]   Go Down
  Print  
Author Topic: Grid Circuit - Class C Modulated transmitter  (Read 12316 times)
0 Members and 1 Guest are viewing this topic.
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« on: March 04, 2012, 09:46:49 PM »

I have put together a test transmitter for testing modulation ( a learning experience for me mostly and to test Hammond ESE series transformers as modulation transformers) . It will be a preliminary for my dual 813 transmitter.


I plan to run two 807s on 75 meters only with a 400 volt solid plate supply in CCS with everything very conservative.

I have completed the choke input power supply and mounted the two 807 sockets and put together the output pi net and plate choke. I plan to feed the screens with a dropping resistor off the plate supply for easy plate modulation.

What is the best grid circuit for a simple transmitter like this for best audio. I will probably drive the transmitter with a solid state VFO with 2-3 watts drive out. Do I need to make the grid circuit tuned and will I need to neutralize the transmitter. My thoughts are no neutralization and no tuned grid circuit.

Not being a class C tube designer - tell me what is wrong with the following line of thinking:

The CCS ratings for the 807 at 400 volts on the plate and 250 volts on the screen call for grid #1 to be -75 volts at 3.5 mA current - Does this mean that the imput resistance of the 807 is approximately  E/I or 75/.0035 or roughly 21,000 ohms or roughly 10 K for two tubes in parallel.

To match this to 50 ohms from the VFO the Z ratio is approximately   10,000/50 or 200 for a turns ratio of the square root of 200 or - about 14    - so I can wind a transformer with a turns ratio of 14:1.

The drive requirement minimum is rated as .3 watts/tube or .6 watts. At 50 ohms this is  about 5.5 volts which would be boosted to about 77 volts at the tube grid.

Am I off base with my assumptions?

Pat
N4LTA
Logged
K3YA
Contributing
Member
*
Offline Offline

Posts: 134



« Reply #1 on: March 04, 2012, 10:10:47 PM »

You shouldn't need neutralization.

You should have a tuned circuit to feed the grid.  A simple parallel tuned network with a link connected to the driver will provide the necessary transformation from your low impedance solid state driver to the high grid impedance of the 807's.  For good modulation it is important to drive the 807's well into class C with lots of RF drive voltage and plenty of negative bias.  Consider using grid leak to provide at least part of the bias voltage. 
Logged
W7TFO
WTF-OVER in 7 land Dennis
Contributing
Member
*
Offline Offline

Posts: 2521


IN A TRIODE NO ONE CAN HEAR YOUR SCREEN


WWW
« Reply #2 on: March 04, 2012, 10:26:00 PM »

Lots of 807 rigs didn't need neutralization, but it was found best to mount the tube socket depressed about 2" into the chassis to isolate the in from the out.

Check out how the ARC-5 rig mounted the 1625's for the proper setup.

73DG
Logged

Just pacing the Farady cage...
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #3 on: March 04, 2012, 10:31:07 PM »

Yes - I had recessed 1625 sockets on a CW transmitter I built a long time ago. I didn't do it on this one and hopefully I won't have  trouble. I can do it if I have to.

Pat
N4LTA
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #4 on: March 05, 2012, 03:43:41 PM »

I have built a -47 volt zener supply using a 5 watt zener diode. The specs for the tube at 400 volts on the plate in CCS call for 3.5 mA grid current at -75 volts.  -47 volts should hold the tube anode current safely down  if drive is lost I think.

That means I need a grid leak resistor to drop 28 volts at 3.5 mA for each tube.

So I plan to use a  8.2K resistor at each 807 grid - That seem to be going in the right direction?

Use a grid leak on each grid with a seperate choke to the supply for each tube or parallel the grids and use a 3.9K resistor for both?

I plan to use a T106-2 toroid tuned to 3.7 mHz with a small variable cap coupled to the tube with a 100 pf cap and feed the 50 ohms in with a small winding over the secondary using the turns ratio from above (about 14:1.

Am I in the right ballpark?

Pat
N4LTA
Logged
w1vtp
Member

Offline Offline

Posts: 2638



« Reply #5 on: March 05, 2012, 05:33:23 PM »

For your reading pleasure.  Please note that it is a class B amp.  Make changes accordingly

Al

* rcahamtips0702.pdf (1457.32 KB - downloaded 213 times.)
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #6 on: March 05, 2012, 05:52:44 PM »

I am using the following data:

Class C Modulated Amplifier (CCS)
Plate Voltage ................................. 400 Volts
Grid No. 2 Voltage ............................ 250 Volts
Grid No. 2 Dropping Resistor .................. 25K Ohms
Grid No. 1 Voltage ............................ -75 Volts
Peak Grid No. 1 Voltage ....................... 95 Volts
Grid No. 1 Current ............................ 3.5 Ma.
Plate Current (Zero Signal) ................... 80 Ma.
Grid No. 2 Current (Zero Signal) .............. 6 Ma.
Driving Power ................................. 0.3 Watts
Power Output (approx) ......................... 22 Watts


Why is it a class C amp - You can't used plate modulation on a class b amp because it does not have a square law
V/I characteristic can you? Is the data above incorrect?


Pat
N4LTA
Logged
KM1H
Contributing
Member
*
Offline Offline

Posts: 3519



« Reply #7 on: March 05, 2012, 06:09:17 PM »

If you cant drop the sockets then try and find the half shields that were sold by Millen. Otherwise better make provision for neutralization unless the grid is going to be heavily swamped with a resistor.

I havent tried it but the 807W/5933 might be a bit more stable.

Carl
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #8 on: March 05, 2012, 06:13:32 PM »

Carl,

I don't mind swamping the grid circuit as this is more of a test bed. If I need to - I can punch a larger hole and use spacers to drop the socket. I can also try to neutralise also becuase I need a good stable monoband amp.

I still am confused about Class C vs Class B. The data come from Sylvania data for the 807W and also another site on the net.

Pat
N4LTA
Logged
KC2ZFA
Member

Offline Offline

Posts: 441



« Reply #9 on: March 05, 2012, 07:02:51 PM »

the ham tips article describes an 807 push-pull class C amp AND an 807 class B modulator.

If you derive screen voltage from plate voltage then -47V protective bias may not be enough.

Peter
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #10 on: March 05, 2012, 08:01:15 PM »

I was not thinking about the protecting the screen with the protective bias - I need to look at that again. I was looking at the anode current only. I'll review that article. Thanks for the reminder.

Pat
N4LTA
Logged
KM1H
Contributing
Member
*
Offline Offline

Posts: 3519



« Reply #11 on: March 05, 2012, 08:11:27 PM »

Whats confusing about this Pat?

http://tubedata.tigahost.com/tubedata/sheets/127/5/5933.pdf
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #12 on: March 05, 2012, 08:33:27 PM »

Nothing  - That is the same data that I was using for Class C telephony

The comment from Al above about the amp not being class C  - but Class B? - That is what confused me.

Did I or am i reading something wrong?


Pat
Logged
w1vtp
Member

Offline Offline

Posts: 2638



« Reply #13 on: March 05, 2012, 10:54:33 PM »

Nothing  - That is the same data that I was using for Class C telephony

The comment from Al above about the amp not being class C  - but Class B? - That is what confused me.

Did I or am i reading something wrong?


Pat

For some reason my most recent post didn't make it.  That article actually talks about a class C amplifier (the torpedo) and a class B modulator in the classic triode connection.

Hope that clears things up.  The first part of the paper demonstrates the need to decouple the input from the output thus eliminating possible parasitics. My "please note" comment was in error. Again, hope this clears things up

Al
Logged
Rob K2CU
Contributing
Member
*
Offline Offline

Posts: 346


« Reply #14 on: March 06, 2012, 08:22:43 AM »

In this case, the class of operation refers to the conduction angle of the tube.

Simply put:
Class A : Tube conducts (has plate current) for 360 degrees of the signal waveform.
Class B : Tube conducts for 180 Degrees of signal waveform.
Class C : Tube conducts for much less than 180 degrees of waveform.

Class B is the typical mode for push pull type stages, with each tube conducting for the opposite half of the waveform.
cross over distortion is of reduced by running the tubes with slightly more than 180 conduction, aka  AB.

Class C is plate current operation in pulses. The tube is typically biased off with a large negative grid voltage. The RF drive is of large voltage amplitude such that it kicks the type during the positive peaks of grid signal. The tunes circuit in the plate "rings" for the remainder of the cycle. The gird drive often brings the peak voltage on the grid to become positive and not just brought up to zero volts. During the peak voltage period of the tube, grid current will flow as it is positively biased for just an instant. The measure of this current provides a means to ensure enough grid drive is present. Plate modulation adds a lagre AC voltage waveform on top of the DC plate voltage. During positive excursions, there has to be enough grid drive to ensure that the tube is still conducting. Without the extra grid drive, the tube may become starved of current which is manifested by flat topping.

The negative grid bias for class C usually comes from two types of designs. One is simply a separate bias supply. To reduce costs, and when RF drive is always present, self bias is the other method. In self bias, the positive grid current provides the means by which negative voltage is built up on the grid input coupling capacitor. Large voltage swings are provided by the tuned plate circuit of the usually included driver stage.

In your described scenario, you plan to drive the 807 with an solid state exciter. In order to meet the need for the necessary grid drive voltage some type of voltage conversion is required. This is could be done with a broad band transformer. The problem is that you would need a fairly high number of turns in the secondary and self resonances will becomes a problem. The better scheme would be to use a parallel resonant circuit similar to the type of plate circuit you would have with a tube driver. But, instead of a tube plate driving the top of the coil, either tap down on the coil so you have what would look like a resonant auto transformer, or link couple the input to the tuned circuit. The danger with a self biased set up is that with any loss of drive, you will have excessive plate current flow. This can be likely to occur if the driver is not built into the 807 transmitter chassis.

Of course, with any tuned plate tuned grid design you will need to isolate the output from the input as others have described.
Logged
w1vtp
Member

Offline Offline

Posts: 2638



« Reply #15 on: March 06, 2012, 08:46:58 AM »

Rob

Agree. Many use either just enough grid bias to cut off the amplifier stage making up the rest of the bias with grid leak.  Yet another approach would be to include a clamper tube on the screen circuit for the 807 finals. With this approach the need for a fixed bias would be eliminated because with the removal of excitation the screens of the 807s is pulled to ground thus eliminating excessive plate current. 

Another point I might make is that what with two 807s, some consideration should be given to having an adjustable drive control.  This could be done by varying the screen to the driver stage (should "hollow state" be used in the driver). For an idea on this you could refer to a schematic on a Viking 2 or an EICO 720.  This would satisfy your question for having a design that would make for good audio.
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #16 on: March 06, 2012, 09:29:41 AM »

Al,

I now understand. That article is very helpful.

I was planning to use both a bias supply and some grid leak in combo. I chose a fixed bias of -47 volts which should drop the plate current to about 100 ma if drive is lost but that may not be enough. The next easy to get 5 watt zener is 62 volts and that will be what I try first as I have one.

Rob

I was thinking about using a Iron Powder toroid about a 1" size with a resonant winding (instead of BB)tuned with a 100 pF variable cap and coupled to the grids with a 1000pF cap and maybe a swamping resistor in the tuned circuit to lower the Q. I was going to couple the 50 ohm SS drive with a 2 or 3 turn link depending on how many turns were required to resonate the circuit. After reading the article, I planned to put a 1 W 47 ohm carbon comp resistor in both the grids and cathode. Didn't have any in the junk box and they sure are expensive nowdays - $2.50 for Ohmite - but I guess we are lucky they still make them.

BTW - While looking through the Hammond site, I found an interesting choke. They make a 30 H 100 Ma open choke as a replacement for a Vox AC30 - which is much larger than anything in the standard choke line. That would make a modifed Heising choke for a single small tube and allow the use of an audio output transformer as a modulator. I may try a couple in parallel with this rig and a Hammond 125FSE SE transformer. The transformer is large and hefty because it is made for 90 Ma DC current and rated for 20 watts. I think it can do twice that with no DC current and it is multi-tapped.

Thanks again for all the help.

Pat
N4LTA
Logged
KC2ZFA
Member

Offline Offline

Posts: 441



« Reply #17 on: March 06, 2012, 10:11:46 AM »

I was planning to use both a bias supply and some grid leak in combo. I chose a fixed bias of -47 volts which should drop the plate current to about 100 ma if drive is lost but that may not be enough. The next easy to get 5 watt zener is 62 volts and that will be what I try first as I have one.

I've experimented with this on my MOPA. I need -90V to cut off the 807 when its cathode is grounded and I pull the power supply wire feeding the oscillator.

Peter
Logged
KM1H
Contributing
Member
*
Offline Offline

Posts: 3519



« Reply #18 on: March 06, 2012, 10:31:34 AM »

The tube manual specifies -85 at 400V plate for operating bias. Id double that for cut off.

Carl
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #19 on: March 06, 2012, 11:02:13 AM »

I have the bias wired for -125 volts when in standby - which is all that I have as it is wired but I can double that if I have to. I am using a little PCB mount 6 VA transformer running off the 6.3 volts. It has a 115/230 volt primary and I am using a FW rectifier circuit and get about -125 volts with a stiff bleeder. I can re-do the board for -230 volts if need be. I hope -125 will do the trick though.

When transmitting, I plan on -80 volts as a combination of fixed bias and grid leak. I guess I'll first try -62 volts of fixed and a grid leak resistor of about 18/.0035 or  5.1K.

I may put a little degeneration in the cathode - Maybe 22 ohms. Maybe see if I need it first.


Pat
N4LTA
Logged
w1vtp
Member

Offline Offline

Posts: 2638



« Reply #20 on: March 06, 2012, 11:15:02 AM »

One point to recall is that whatever fixed bias you provide you will either need to VR / Zener regulate (preferrably shunt method).  As you start to draw grid current through excitation it will want to charge up any fixed bias supply if it is merely a PS with a capasitive filtering.  Shunt VR regulation will avoid that problem.
Logged
KC2ZFA
Member

Offline Offline

Posts: 441



« Reply #21 on: March 06, 2012, 11:29:55 AM »

I've adopted the schematic below (from elsewhere on AMFone). Bias is all grid leak, and tube is cut off when drive is cut.


* Bias.gif (7.26 KB, 763x398 - viewed 458 times.)
Logged
N4LTA
Contributing
Member
*
Offline Offline

Posts: 1075


« Reply #22 on: March 06, 2012, 12:06:22 PM »

Al

I am planning on a 62 volt 5 watt shunt zener as a VR - when the transmiiter is in standby - I open the zener from ground and the fixed bias rises to -125 (or higher if needed)  - the unregulated supply voltage.

The rugulated volatge passes through a 5.1 K grid leak resistor to generate the additional -18 volts to give me the -80 volts during driven condidtions.

I make go with what KC2FA has shown as that would be an easy adaptation.

Pat
Logged
Pages: [1]   Go Up
  Print  
 
Jump to:  

AMfone - Dedicated to Amplitude Modulation on the Amateur Radio Bands
 AMfone © 2001-2015
Powered by SMF 1.1.21 | SMF © 2015, Simple Machines
Page created in 0.062 seconds with 19 queries.