The AM Forum
March 28, 2024, 04:24:30 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 [2] 3 ... 12   Go Down
  Print  
Author Topic: screen modulation again...  (Read 182952 times)
0 Members and 1 Guest are viewing this topic.
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #25 on: November 21, 2014, 08:52:03 PM »

You can put air on the tubes without it being a loud racket, I do it with a pair as modulators in AB1.
But I would run it at reduced power like I do the plate modulated setup.
No reason to push the tubes, I have rigs that will put out plenty of power without pushing things.

I do have 813's, 4x150's, and 4D32's.
I have an air system socket or two and blower for the 4x150 tubes and could go that way, or a single 813, or just patch the circuit into one or all of the rigs I have already built.

Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #26 on: November 21, 2014, 11:26:01 PM »

Made some tests, bypassed the mod transformer, adjusted the plate voltage to 1000 volts, put a tone into the screen modulator and adjusted things to get a clean sine wave out.
So far so good.
I made adjustments to the grid drive, the screen voltage, and adjusted the carrier level pot on the screen modulator, looking at the waveform on the scope.
I could get 100% negative modulation, and I could adjust things so I could get 150 watts pep out with any carrier power, 5 to 100 watts, and the waveform looked good.
I hooked the audio chain up and had a loud hum, and it sounded a bit stuffy, but not bad at all.
In keying and unkeying the rig, the screen current protection tripped, it had tripped a few times before as its set to 70 ma and the screen modulator draws about that with the voltage at 600 volts going in.
After it tripped and I reset it, something was wrong, the waveform peak no longer looked like a sine wave but a double humped camel. Only the positive part was distorted, I also had some erratic power output and cracklng in the audio, so I suppose one of the devices blew out.

I might have been able to adjust things so it sounded very good if I could get rid of the hum, but I had no chance to  try anything before the blowup.

All the adjustments seemed VERY critical, any change to anything had a lot of effect, loading, any voltage, grid drive, plate voltage, and all the adjustments on the screen modulator.
I do not really like that, but that is likely the way it is with screen modulation.

I put everything back to normal and adjusted the transmitter for plate modulation and it works fine.
 
Logged
KX5JT
Contributing
Member
*
Offline Offline

Posts: 1954


John-O-Phonic


« Reply #27 on: November 21, 2014, 11:57:58 PM »

 This is a helavu interesting thread to follow!
Logged

AMI#1684
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #28 on: November 22, 2014, 04:51:07 PM »

I was working Jerry, W1ZB today and he said to try an isolation transformer on the audio input to clean up the hum.
I will find out what blew out and fix that and try again.
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #29 on: November 22, 2014, 11:16:39 PM »

R2 was open (1 meg trim pot), I replaced it with a fixed 260k resistor.
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #30 on: November 23, 2014, 10:54:17 AM »

Got it working real well today.
I installed the isolation transformer, it did not do anything to correct the hum, so I put an oil filled cap on the B+ feed to the screen modulator and that fixed the hum.
I had hooked it up to the 3x 4D32 rig, and put a tone in, adjusted things so they looked right, and it sounded very good on voice!
40 watts carrier, worked N2LJO who said it sounded fine.
150 watts pep, I can not get as much peak power on the 4D32's, but it might need more adjustment.

1000 volts on the plate, about 100 ma for the 40 watt carrier.

Nothing on the screen modulator gets hot.

I need to try it on the 4x150 deck again, they seem to give a lot more modulation, I guess the screen has more control then it does in a 4D32.
100% modulation in both directions is not hard to do at all, without any modulation of the grid drive.

If you could run the RF deck at about 700 or 800 volts, you could power the screen modulator off the B+ and no other power supply would be needed.
I might try that, running the plate voltage down to 700 volts with the 4x150's and see what happens.

Four 4x150's in a very small RF deck with the screen modulator and a power supply could make for a very small 300 watt carrier rig.


Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #31 on: November 23, 2014, 11:57:26 AM »

Brett

Good!

A suggestion:

If there is hum on the B+ (with respect to the screen modulator's ground), then, by far, the largest effect of that hum on the output of the modulator will be from the base of the first transistor. All of the voltage gain is in the 1st stage, and the audio voltage on the base of the first stage's transistor is directly proportional to hum on the B+

To save space, and the avoid the use of the oil filled capacitor...

Add a 10k ohm resistor between R2 and the base of the first transistor (which, to the extent that it makes a difference, will be compensated by using a value of R2 that is 10k ohms less). Add a 10uF (or larger) capacitor between the junction of these two resistors and ground. The voltage from that junction to ground will be less than 100 volts (measure it before adding the capacitor), so you won't need a very high voltage capacitor.

This will significantly reduce the hum at the output of the modulator, without having to use the oil filled capacitor across the entire B+

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #32 on: November 23, 2014, 03:00:58 PM »

Thanks Stu, I may try that.
I think the issue is that the rf control deck has a bunch of stuff close on the front panel, variac for the screen voltage and big wire wound pots to adjust the grid leak bias and screen dropping.
I think the variac's modulate the voltages through the wire wound pots.
A separate supply would not have that problem.
You want the screen to self modulate, so I can not put a cap at the output of the screen dropping pot if I want to use it for plate modulation.
The control deck also has a not used power supply (with variac) that was for the 3 diode neg cycle limiter, if its keyed, maybe I can use that instead.

I think I want about 600 to 700 volts to cover all the tubes I might use, at about 100 ma.
That is easy to do, and would likely fit on the screen modulator chassis itself, keyed by the plate circuit (step start).
 
Logged
AJ1G
Contributing
Member
*
Offline Offline

Posts: 1283


« Reply #33 on: November 23, 2014, 05:33:54 PM »

I am sure its just adjustments and voltages, but at the moment, it makes old military screen modulated stuff (carbon mic) sound hifi...

Glad you qualified the above by including the "carbon mic".  The screen modulated command sets, when fed line level good quality audio can sound very good.  They do require a bit of tweaking to get them sounding right, very sensitive to final loading, like it on the heavy side.  My T22 modulated by a BC-456 stock dyno powered setup makes about 55W out on CW and about 20 on AM,feed it with line level audio through a decoupling cap into the carbon mic input, or using a military RM-12 remote control unit which allowed coupling field phone circuits into the carbon mic inuts of various sets, even has a VU meter for setting levels.

I swept the frequency response of the setup fed this way, and it looked clean from about 150 Hz to way over 10 kHz.  Feeding it into a dummy load and feeding it with music test audio is very listenable in a local good quality AM receiver.
Logged

Chris, AJ1G
Stonington, CT
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #34 on: November 23, 2014, 09:25:31 PM »

Put the 2x 4x150 rf deck on and tuned everything up, and made all sorts of tests and adjustments.
I think I am a bit limited on my voltage feed to the modulator because it trips the screen current overload protection which trips at about 70ma.
The modulator itself takes more then that if I increase the voltage past about 550 volts.
But it was working very well at about 70 watts carrier out, 250 watts pep, about 1100 volts on the plates, about 200 ma.
Looked at the sdr and the distortion products were way down, it sounded good in the mod monitor, and I made a contact with Larry, W9SX 1000 miles away. He said it sounded good.

Due to the way things are connected, its going to be more of a pain to try it on the 813 rig.

Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #35 on: November 24, 2014, 11:14:33 PM »

I think I will need to do this if I want to use the power supply that is built into the rig.
If I put a 4 uf oil filled cap on the voltage input, the hum is low, its even lower with a 15 uf cap, but there is a trace.

While I am at it, I would like to put some combination of fixed resistors and a pot to fine tune things.
With the standard design, I somehow blew out the R2 trim pot.
Maybe a 10k pot plus some fixed resistors to center it around the total 260K ohms called for.

I also wonder what the effect would be of adding some resistance to the two 10K 50 watt resistors to cut
the current through the system a bit.
As it is, with about 500 volts going in its pulling 70 ma or more.

It really sounds very good to me, at 70 or 80 watts out I get plenty of modulation, to get more power, I would need to increase the voltage into the modulator.
500 is going in, and about 80 is coming out without modulation.
It wants 600 or 700 volts going in with 4x150's.
At 80 watts carrier, the tubes are running at 92F with gentle air flow, 100 watts should be easy to do.

For lower power tubes, you should be able to run the screen modulator off the plate voltage, up to about 1000 volts I suppose.




Brett

Good!

A suggestion:

If there is hum on the B+ (with respect to the screen modulator's ground), then, by far, the largest effect of that hum on the output of the modulator will be from the base of the first transistor. All of the voltage gain is in the 1st stage, and the audio voltage on the base of the first stage's transistor is directly proportional to hum on the B+

To save space, and the avoid the use of the oil filled capacitor...

Add a 10k ohm resistor between R2 and the base of the first transistor (which, to the extent that it makes a difference, will be compensated by using a value of R2 that is 10k ohms less). Add a 10uF (or larger) capacitor between the junction of these two resistors and ground. The voltage from that junction to ground will be less than 100 volts (measure it before adding the capacitor), so you won't need a very high voltage capacitor.

This will significantly reduce the hum at the output of the modulator, without having to use the oil filled capacitor across the entire B+

Stu
Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #36 on: November 24, 2014, 11:50:30 PM »

Brett

Putting the details aside for the moment:

How large does the positive-going screen voltage swing need to be... from carrier level to your desired percentage of peak positive modulation? Is 100V enough? If not, how much?

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #37 on: November 25, 2014, 08:11:07 AM »

Stu,
I do not know, and it likely changes with each tube type.

A pair of 4x150's has 500 watts of plate dissipation, so that would be the best choice.
As it is, I have enough swing for about 80 watts of carrier with about 550 volts going in and 80 out.
I did not write those numbers down so they may be off a little.
1400 volts on the plate I think, and the grid bias also has a big impact, less grid drive equals less bias which equals more plate current and more power output.

I plan on doing some tests using the plate supply as a feed to the modulator as well as the plate of the tubes.
I can turn it down to 700 volts and see what happens, and work my way up to 1000 volts.
It looks like the circuit would work very well with 6146's with 800 or 900 volts on everything....

Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #38 on: November 25, 2014, 09:26:16 AM »

Brett

Okay.

I spent some time thinking about this circuit, and each of its stages in more detail.

I have to retract my earlier statement that the dominant contributor to the circuit's output hum is the power supply hum on the base of the 1st stage transistor. More on that below.

1st stage (common emitter)

A. Reducing the DC collector current

I believe that you could reduce the DC collector current drawn from the power supply by increasing the collector resistor from 10k ohms to 20k ohms. This will reduce the required DC collector current to half its present value. The power dissipated by the resistor, and by the transistor, will also be cut in half. Note: you will have to increase the value of R2 by a factor of about 2... because a reduction in the DC collector current by a factor of 2 must be accompanied by a reduction in the DC base current by a factor of 2. I.e. the transistor has a fixed current gain. You should adjust R2 to produce enough collector current to accommodate the desired positive voltage swing, and not much more. For example, if you need 120V of positive swing to modulate the screen voltage from carrier level to the desired positive peak modulation level, then you should adjust R2 to produce just a little more than 120V of DC voltage drop across the (new) 20k collector resistor. This corresponds to just a little more than 6mA of collector current. The reason that you can increase the value of the collector resistor, and therefore reduce the DC collector current, is that the load... looking into the base of the 2nd stage (the emitter follower)... is significantly higher than 20,000 ohms for all values of screen modulator output voltage. However, you should check to see that there is no distortion at the output of the 1st stage as a result of this change.

Note: The DC current gain of this transistor, under these (low collector current) operating conditions, is around 10.

http://product.ic114.com/PDF/T/TT2062.pdf

Therefore, with 6mA of DC collector current, the base current should be around 6ma/10 = 600uA.

With 600uA of base bias current, and a B+ of 500V, the required value of R2 is 0.883 megohms.

With the existing 10k ohm collector resistor, the collector current should be set to around 12mA, the base current should be around 1.2mA,  and the required value of R2 should be around 0.417 megohms. I suspect that the value you are now using for R2 (260k ohms) is resulting in much too much collector current in the 1st stage. I.e. 19mA instead of 12mA. Maybe even more, if the current gain of the transistor is larger than 10

B. Reducing the hum contributed by the 1st stage

If the power supply (B+) increases by 1% (i.e. the upward part of the peak-to-peak swing, at 120Hz, due to hum), then the base current flowing into the first transistor will increase by 1%. But (this is where I was mistaken in my earlier post), the voltage on the top of the 1st transistor's collector resistor will also increase by 1%. The effects on the voltage at the output of the 1st stage (i.e. at the collector of the first transistor, which is the bottom of the collector resistor), will partially cancel (because the resulting voltage changes are of opposite polarity)... but not completely cancel (because the first effect is smaller than the 2nd). To reduce the hum at the output of the 1st stage, both the collector B+, and the B+ driving R2 (which produces the DC base current) must be filtered. Just filtering the B+ that produces the DC base current will actually make the output hum a little worse!

2nd stage (emitter follower)

A. Reducing the DC collector current

In the general case of an emitter follower, the DC current flowing in the emitter must be large enough to accommodate the required downward swings of the emitter (output) voltage. The emitter current cannot be negative... so the DC emitter current has to be larger than: the required downward voltage swing / the audio frequency emitter impedance.

Note: In this application, the "downward swing" is the difference between the required screen voltage, at carrier, and the (negative) screen voltage required to produce 100% negative modulation of the RF tube. The DC voltage between the emitter and ground should be set equal to the value of this "negative swing".

For example, if the required screen voltage, at carrier, is 75V and if the screen voltage required to produce 100% negative modulation is -20V, then the "negative swing" is 95V. In this example, the base bias on the second transistor should be adjusted to produce 95V between the emitter and ground.

The audio frequency emitter impedance is the parallel combination of the emitter resistor and the impedance looking into the screen of the modulated RF tube. If this combined, parallel audio frequency impedance is at least 5000 ohms (for all values of modulated stage screen voltage below the carrier level), and if you need a 100V negative going screen voltage swing (from carrier level to 100% negative modulation level)... then you will need to have 20mA of DC emitter current (about 20mA of DC collector current) in the 2nd stage. Note, this 20mA of 2nd stage DC collector current includes the carrier level screen current drawn by the modulated RF tube.

You could probably reduce the DC collector current, somewhat, by increasing the value of the emitter resistor to 20k ohms... without any unfavorable effects on the behavior of the circuit.

I suspect that you have set the base bias current on the 2nd stage at too high a value...by using too low a value for R4... and that you, therefore, have too much DC collector current.

B. Reducing the hum contributed by the 2nd stage

The hum on the collector of the 2nd stage does not produce significant levels of hum on the output of the 2nd stage. However, you need to filter the B+ that drives the base bias resistor of the 2nd transistor.

3rd stage (DC offset voltage)

A. Reducing the DC collector current

The collector current of the 3rd stage must be equal to the screen current drawn by the modulated RF tube, at carrier. This current is set by 3rd stage adjustment pot. This current is part of the DC collector current drawn by the 2nd stage (i.e. it does not come directly from the power supply).


Bottom line:

If you change the collector resistor of the 1st stage to 20k ohms, the current drawn, at carrier, by the screen modulator should not be much more than: 5mA (1st stage) + 20mA (2nd stage). Be careful to adjust R2 properly... so that you don't have excessive DC collector current in the 1st stage. Likewise, be careful to adjust R4 properly, so that you don't have excessive second stage DC collector current.

With respect to hum: You need to filter to base bias voltage supplies of the first and second transistors. Each base requires very little bias current (a high value base bias resistor), and is easy to filter (as per my earlier post). You need to filter the collector voltage supply of the 1st stage, but not the 2nd stage.

Stu

 

Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #39 on: November 25, 2014, 05:43:41 PM »

Dashed home from work and tried using the same supply for the plate an screen modulator.
It works well but only puts out about 30 watts at 700 volts.
The 4x150's want a lot more plate voltage to make power...

Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #40 on: November 25, 2014, 09:35:48 PM »

Made more progress, with great results!
I had issues with the rig taking off if I ran things at higher voltages, plate or input to the modulator.
It worked fine at moderate levels, but when I cranked things up, the plate current would take off, and it was because the output of the modulator increased its voltage out to the screen.
As a test, I just stuck a 100k resistor across the output of the modulator to load it down some.

I can crank everything all the way up and nothing bad happens, and I can get lots of power out, 100 watts is easy, with all kinds of modulation and it sounds very good to me.

I also fixed the screen overload protection relay making the trip current adjustable.
This is turning into a very fun project!
Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #41 on: November 25, 2014, 10:20:07 PM »

Brett

Interesting...

I wonder if the 100k resistor made the difference, or whether the process of soldering it in fixed a cold solder joint or loose ground having to do with the screen RF bypass capacitor... or some other faulty connection.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #42 on: November 26, 2014, 08:16:22 AM »

I did nothing to the RF deck.
All I did was to add a pot across the screen current trip resistor and I added some filter caps to the voltage input on the modulator (electrolytic under the chassis).
I still had the issue with the carrier voltage level popping up an extra 100 volts or so, but only if I ran over 1200 volts on the plates of the 4x150a's or if I ran over 600 volts in to the modulator.
I then just stuck the resistor on thinking something is making the last stage increase its bias point.
I had tried various settings of the R4 and R6 pots and voltages on the plate and modulator input and it always did it.
I also tried different bias levels on the RF deck, different grid and plate tuning, but nothing stopped it.
I had thought it was something in the rf deck taking off, but the swr was low and nothing odd shows on the sdr rx. The 100,000 ohm resistor eliminated the problem.
I just picked 100K ohms as something to load/damp the modulator output..

Its only about 100 to 150 volts out to the screens without modulation.


The thing now works so well I am thinking three or four 4x150's with plenty of air would make a very nice 300 watt rig.
 
Logged
KD6VXI
Contributing
Member
*
Offline Offline

Posts: 2648


Making AM GREAT Again!


« Reply #43 on: November 26, 2014, 02:21:35 PM »

I'd be willing to bet the new resistor gave a path for neg screen current to go.

--Shane
KD6VXI
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #44 on: November 26, 2014, 02:41:50 PM »

That is what I thought, as it only happened past a point, but once started stayed.
Carrier power went up 100 watts or so, and turning the voltage to the screen modulator OR the plate voltage down would stop it till next time.

I suppose a light value of resistance to work into will make the modulator cleaner since it will have a more stable load to look into.
Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #45 on: November 26, 2014, 04:19:22 PM »

Brett

You have mentioned that you have a screen current trip circuit.

Can you post a schematic of that circuit? Please indicate where the screen modulator feeds in, and please indicate where the 100k ohm resistor is attached.

I'm assuming that you have a resistor in series with the screen grid, and that you are measuring the voltage across that resistor (both sides of which are at the screen voltage level above ground).

I'm wondering whether, somehow, this screen current trip circuit is causing the screen voltage to lock up in the absence of the added 100k ohm load.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #46 on: November 26, 2014, 07:39:42 PM »

Stu,
The screen current overload relay is built into the rf control deck. Its a relay coil in series with the screen feed to the rf decks. Part of the screen current flows through the relay coil, part through a resistor that sets the trip point.
When excess current flows, the relay gets pulled in and connects the screen voltage to a resistor to ground which holds the relay closed. Another set of relay contacts supplies power to a light in the reset button that breaks the path to the resistor to ground.
Pushing the button resets the overload.
There is a small cap across the relay coil to stop chatter.
I had used a fixed resistor that set the overload to 70 ma, its a fast acting circuit, and to prevent the screen modulator from tripping the overload I wired a pot across the resistor so I can set the trip point from 70 ma to never trip.

The screen overload has nothing to do with the problem I was having.
The 100K resistor goes from the modulator output to ground.

In order to reduce hum, I put two 100 UF 450 volt caps in series across the input voltage with 100K resistors to balance the voltage.
I found that kills some key down power spikes out of the transmitter and increases the positive modulation a lot.

I really like how the modulator is working.
Logged
AB2EZ
Member

Offline Offline

Posts: 1722


"Season's Greetings" looks okay to me...


« Reply #47 on: November 26, 2014, 10:44:31 PM »

Brett

The important thing is that the modulator is now working to your satisfaction.

With the substantial series inductance of the trip circuit relay coil (through which part of the screen current passes), interacting with the large amount of capacitance from collector-to-emitter of the 3rd stage of the screen modulator (effectively in series with the relay coil)... one could get some strange effects (audio frequency oscillations).

I suspect that adding the 100k ohm resistor to ground, at the point between the screen modulator's output and the screen voltage input of the transmitter, was sufficient to damp out those oscillations.

Stu
Logged

Stewart ("Stu") Personick. Pictured: (from The New Yorker) "Season's Greetings" looks OK to me. Let's run it by the legal department
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #48 on: November 26, 2014, 11:27:18 PM »

That was what it was acting like, but it did not show.
The swr into the antenna was low, the spectrum display (100 Khz wide) showed no modulation, nor did the mod monitor.
It sure looked like some sort of DC bias point problem.

But anyway, I really like the way its working now, it sounds really good to me.

That gives another problem, what to hook it up to and where to put it.
It seems to work very well with the 4x150 deck, but that could use more plate voltage, and likely more air and real air system sockets to run at full power.
Too much stuff and not enough rack space.
Logged
N2DTS
Contributing
Member
*
Offline Offline

Posts: 2307


« Reply #49 on: November 27, 2014, 12:41:07 PM »

I used the screen modulator on 40 today, plenty of long transmissions and it worked fine.

I did not like the series electrolytic capacitors I used to filter the B+ input.
I used two new 100 UF 450 volt 105C  caps in series with 100K resistors across each one and found one resistor was getting warm, the other not.
I measured the voltage across each cap, 8 volts across one, 540 volts across the other.
Swapped them out and had 100 volts across one and 450 across the other.
Took them all out and put the oil filled 1500 volt 15 uf cap back on.

Now to figure out what I am going to do with this thing...
Logged
Pages: 1 [2] 3 ... 12   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 0.083 seconds with 19 queries.