New 4-1000A modulated by 4-1000As AM Project - Fabio II Reborn

[KA2DZT]

Not really too sure that the back emf from the reactor gets bled off by the antenna.  Most of it gets absorbed by a higher plate voltage on the finals.  Of course, the antenna does add to the overall load on the tank circuit.  Some of it gets absorbed by the back load on the mod xfmr from the modulators.

The voltage spike from the reactor is a function of the RATE of change in the current through it.  The faster or greater the rate of current change results in higher voltage spikes.  So, when JJ was opening the CTs on the filament xfmrs that dropped the current about as fast as he could do it (nice work Tom ) he was getting the maximum fireworks.

The only logical way to reduce the voltage spike is to increase the time it takes to drop the HV which will drop the current slower.  This is also true for the screen reactor.

So I think, from his comments, JJ is going to try the primary shut-off of the screen and HV supplies.  This will allow the PS filter caps to decay the voltages at a slower rate.  This should result in much less back emf.  Hopefully this will be enough to solve the problems.

We'll all find out after Tom completes the changes.

Fred

[W2VW]

The voltage spike from the reactor is a function of the RATE of change in the current through it.  The faster or greater the rate of current change results in higher voltage spikes.  So, when JJ was opening the CTs on the filament xfmrs that dropped the current about as fast as he could do it (nice work Tom ) he was getting the maximum fireworks.

The only logical way to reduce the voltage spike is to increase the time it takes to drop the HV which will drop the current slower.  This is also true for the screen reactor.

Fred

While this is true there must be a plan B for times something unexpected happens. Loss of drive can sometimes create a fast enough decay to flash things over. I have the Tee shirt.

[AMLOVER]

MY SOLUTION IS TO ACTIVATE/DISACTIVATE RESISTANCES IN PARALLEL WITH THE CHOKES

I keep all the voltages to the final and to the modulator (4 x triodes) all the time ON in order to be very fast on tr/rc. I had the same problem as Tom but not anymore. My solution is in the attached sheme and was a suggestion of Stu that fitted perfectly to my transmitter.
The on-off sequense is very simple and done manually in few milisecs by a simple make before brake rotary switch  4P4T that controls 26.5Vdc.

ON :
1st position = ant relay from rc to tr
2nd position = Rf drive on
and modulator's bias from the cutoff (-200v) to the idle current setting (-100v) by grounding the bleeder pot (only for lengthening tubes' life).
3rd position = reactor's vacuum relay opens R1= 5k/100w and screen's relay opens R2=5k/20w
4th position = audio on

OFF : oposite

It is amazing how quietly turns on and off the transmitter now.
The resistances are involved only when the Tr is off and they are disactivated when it is going to be on.

Stefano

[K1JJ]

MY SOLUTION IS TO ACTIVATE/DISACTIVATE RESISTANCES IN PARALLEL WITH THE CHOKES

Stefano

Hi Stefano,

That seems like a very good solution.  I wish I had heard it before ordering these step-start parts - at least to try it with my existing setup.

I received the vac relays today, so am tempted to use them to add a resistor across the HV Heising reactor and the screen choke as you suggested.   Maybe I can still keep the HV on and ground the fil CTs using this technique.

I see you leave the HV on all the time as I presently do now.   Do you key the fil CT or leave the CTs grounded all the time? I didn't see reference to them in your sequencing.  Maybe you turn the RF final off with a fixed bias to keep it from drawing current. Then the RF drive comes on and still no (or little) plate current.  Does the RF final tube conduct plate current after the screen is turned on or a little current before too?  

Or maybe the RF final's screen voltage stays on all the time and as soon as the RF drive comes on, there is power output to the antenna?  IE, the fixed bias keeps the tube off when there is no RF and the grid leak adds the remaining bias during full drive power.

Then I could key the mod tubes using the reg screen supply keyed at the 120VAC point.

T

[KA2DZT]

The voltage spike from the reactor is a function of the RATE of change in the current through it.  The faster or greater the rate of current change results in higher voltage spikes.  So, when JJ was opening the CTs on the filament xfmrs that dropped the current about as fast as he could do it (nice work Tom ) he was getting the maximum fireworks.

The only logical way to reduce the voltage spike is to increase the time it takes to drop the HV which will drop the current slower.  This is also true for the screen reactor.

Fred

While this is true there must be a plan B for times something unexpected happens. Loss of drive can sometimes create a fast enough decay to flash things over. I have the Tee shirt.

You're right about the drive.  He is using -90V bias which completely cuts off the finals.  That's why I suggested that he use some grid leak bias along with less fixed bias.  If he looses drive the tubes will still conduct some current.

Fred

[K1JJ]

I've been using -90V fixed bias with an additional  -180 V bias generated from 2K of grid leak.  The specs call for about -200V for class C, but I found at -270V the tubes are off the scale for efficiency and no plate color.  The performance appears to be excellent for THD.

I just need to determine the best way to turn the finals on, etc., as mentioned in my last post.  That may be the easiest way to proceed and eliminate the LOUD clacking of step start contactors...

T

[KA2DZT]

Tom,

Looking at Stefano's scheme, it seems that it should also work.  You're right about the relay noise.

I thought you said that without drive there is no current  or the current stops dead.  Wouldn't keying the drive on and off be enough to turn the finals on and off??

For the modulators you could just key the screen voltage on and off, leaving the plate voltage on.

OK FB on the bias, you never mentioned that you were already using some grid leak bias.  I didn't think -90V bias was enough to drive the tubes far enough into cut-off.  So, it's a total of -270V.

Fred

[K1JJ]

Yep, looks like a plan, Fred.

Somewhere back in this long thread I did mention I am running -270 v total fixed/grid leak bias, giving excellent efficiency.

The -90V is probably at about class B cutoff since I have 500V of screen on them.  I chose to make the -90V regulated to avoid RF rectification charging that has been a problem in other rigs.  The unreg fixed bias wud just creep up higher and higher during a transmission, needing more and more drive. Kinda a runaway thang.

So, I'm installing the the relays and resistor dampers now.  I think leaving the HV on with CTs grounded, while using RF drive to activate the RF final -  and screen voltage to key the modulators will work FB.

I may end up keying the RF screen voltage in the pri too, but not sure yet.

Want to add screen safety shutdown relays eventually.

Shud know maybe within a day how it all works out.

T

[AMLOVER]

Tom

I keep all the voltages ON the RF and AF tubes always.

My winter evenings routine goes like this :
The RF tube is a tetrode and the AF tubes are triodes.
So I first turn on the RF blower and modulator funs, after all filaments with CTs permenantly grounded through the Amp meters that are fully protected, in the same time with filaments comes 30% safety bias to the Rf tube (-140v) and cutoff bias to the AF tubes (-200v) because the bias bleeder is ungrounded by a relay at this moment, there is of course a leak resistor for the rest 70% of RF bias (-300v for a total bias of -420v), after I turn on the 5500v RF plate voltage supply (6200v at this moment) with a 3 steps starter, 3 steps of 50ohm resistors in order to help HV capacitors' bank and modulation cap to be peacefully charged up through the modulation transformer (no idle plate current indicated yet), after I turn on the 4000v AF plate voltage from a different supply with an also 3 steps starter, on the end I turn on the 600v RF screen voltage slowly with a variac through screen protection circuit and self mod. choke (at this moment I get a very small indication of plate current about 10-15ma which is the condition with all voltages on and no excitation so no total bias). Now the system is ready for a long ionosphere riding. I leave it in the basement and go to my shack checking it from 2 wireless cameras (one to the Bird43 indication and one to the meters). I do these manually in about 15 min and only once in the begining of the evening. I of course come on the end of the evening to turn it off oposite way.
From the shack I move the rotary switch to the
pos1 and the ant relay moves to transmit,
pos2 and the RF exciter (rice box) turns on (44w) in this point I can see 70ma grid current, 180ma screen current and 1,6A plate current with the same order another relay grounds a 100ohm/200w tapped bleeder in order to set the modulator's bias at
-100v and its idle current at 80ma per tube (total 320ma),
pos3 and the 2 very important relays disactivate the 2 resistors from the chokes letting them free to play their modulating roles,
pos4 and the audio is on

When I turn off first audio goes off, then the relays are shorting the resistors to the chokes and make them relaxed, then the modulator's bias cuts it off and in the same time excitation can be easily off now because chokes are shorted by the resistors and have no energy for sparks and on the end the ant relay moves to receive.

As far you use tetrodes at the modulator keep the cathodes always grounded by filaments CTs and instead of altering the bias for cutting off the modulator, as I do with my triodes, in the sequence position 2 you could turn off the screen voltage from its ac side by a simple 220v/5A relay and you are set.

I was out of country for business and so I had not time for am phone otherwise I would  respond sooner.

Stefano

[W2VW]

I just need to determine the best way to turn the finals on,

T

Put a poster of Marilyn Monroe up on the wall.

[w4bfs]

Stefano has a maul

[K1JJ]

Tom

I keep all the voltages ON the RF and AF tubes always.

My winter evenings routine goes like this :
The RF tube is a tetrode and the AF tubes are triodes.
So I first turn on the RF blower and modulator funs, after all filaments with CTs permenantly grounded through the Amp meters that are fully protected, in the same time with filaments comes 30% safety bias to the Rf tube (-140v) and cutoff bias to the AF tubes (-200v) because the bias bleeder is ungrounded by a relay at this moment, there is of course a leak resistor for the rest 70% of RF bias (-300v for a total bias of -420v), after I turn on the 5500v RF plate voltage supply (6200v at this moment) with a 3 steps starter, 3 steps of 50ohm resistors in order to help HV capacitors' bank and modulation cap to be peacefully charged up through the modulation transformer (no idle plate current indicated yet), after I turn on the 4000v AF plate voltage from a different supply with an also 3 steps starter, on the end I turn on the 600v RF screen voltage slowly with a variac through screen protection circuit and self mod. choke (at this moment I get a very small indication of plate current about 10-15ma which is the condition with all voltages on and no excitation so no total bias). Now the system is ready for a long ionosphere riding. I leave it in the basement and go to my shack checking it from 2 wireless cameras (one to the Bird43 indication and one to the meters). I do these manually in about 15 min and only once in the begining of the evening. I of course come on the end of the evening to turn it off oposite way.
From the shack I move the rotary switch to the
pos1 and the ant relay moves to transmit,
pos2 and the RF exciter (rice box) turns on (44w) in this point I can see 70ma grid current, 180ma screen current and 1,6A plate current with the same order another relay grounds a 100ohm/200w tapped bleeder in order to set the modulator's bias at
-100v and its idle current at 80ma per tube (total 320ma),
pos3 and the 2 very important relays disactivate the 2 resistors from the chokes letting them free to play their modulating roles,
pos4 and the audio is on

When I turn off first audio goes off, then the relays are shorting the resistors to the chokes and make them relaxed, then the modulator's bias cuts it off and in the same time excitation can be easily off now because chokes are shorted by the resistors and have no energy for sparks and on the end the ant relay moves to receive.

As far you use tetrodes at the modulator keep the cathodes always grounded by filaments CTs and instead of altering the bias for cutting off the modulator, as I do with my triodes, in the sequence position 2 you could turn off the screen voltage from its ac side by a simple 220v/5A relay and you are set.

I was out of country for business and so I had not time for am phone otherwise I would  respond sooner.

Stefano

Excellent, Stefano!

I follow your description clearly and will do a very similar set up here.   OK on ALL voltages staying on and the finals controlled by the fixed bias and also RF drive.  Also, the CTs are grounded all the time.

I will use the screen primary to control the modulator keying as you suggested.

The screen relay and resistor are already mounted and now working on the HV Heising relay/ resistor.  That is the key to taming this - the resistive dampers. Now the back-EMF will have somewhere to go and is rock solid.

I will use the contactors to set up an initial step-start that is used only to turn the HV supply on at the beginning of the session and stay on.  Right now I use the HV supply smoothing choke as the inrush softener, but that is really not enough.

The AMfone board should post your solution above in the archives, as it is a well thought out solution and I'll bet many guys building plate modulated rigs could use parts if not all of your suggestions.

BTW, you built  quite a FB transmitter with  5500 vdc  on the plates....   I seem to remember you posting some nice pictures a few years back. Could you post the latest pics of it?

73, and thanks again for the help, OM.

Tom, K1JJ

[K1JJ]

I just need to determine the best way to turn the finals on,

T

Put a poster of Marilyn Monroe up on the wall.

When I turn my rig on it ALWAYS returns the favour.

[KA2DZT]

I just need to determine the best way to turn the finals on,

T

Put a poster of Marilyn Monroe up on the wall.

Dave, I'm not sure how we would ever get by on this forum without your invaluable advice.

Fred

[W2VW]

I just need to determine the best way to turn the finals on,

T

Put a poster of Marilyn Monroe up on the wall.

When I turn my rig on it ALWAYS returns the favour.

Are you British now?

[AMLOVER]

All the thanks to Stu...
He is the creator of this circuit and I just followed his instructions.
I had very serious problems on keying my transmitter, the same like you.
I was very desperade and had almost done all the tricks with sequencing, bypassing, relaying left right. I had also tried to turn off the driver by a potentiometer at its output voltage very slowly like a dimmer but as soon the exciting wattage went from few miliwats to zero I had thunders in the lab.......
Then I asked for help here and Stu cared to check my situation. The result was fantastic. Happy end.......

I'll be very busy for the summer and 90% out of home but I promise you to put many pictures from the new look transmitter as soon I am back for good. It will be a very interesting surprise for you Tom...
Wish you the best.

Stefano

[K1JJ]

OK Stefano -

Glad you got yours running so smoothly.

Yes, Stu has helped me out in a few areas too.  His advice is usually well thought out, as we'd expect from the college level engineering professor that he is.


Well, not wasting any time, here's a few pics of the screen 50 watt 5K resistor and vacuum relay installation. I left in the CT PTT relay as-is in case I need it again.


Notice I have the RF deck on a permanent Lazy Susan  for easy access to all parts.  The cabinet slides off in seconds. Same for the modulator cabinet, though it is on caster wheels instead.

T

[K1JJ]

And, here's the Heising choke 10K 200 watt resistor and vac relay installation. It's in the rear of the rig.

I'm glad it's almost finished cuz things are starting to get crowded...

The cabinets are all built and painted IBM blue.    

Might make a few more mods and fire it up again tomorrow.

As an experiment, I'm going to try it with the new dampers using the original CT keying on both the final and modulator, just to see what happens. This will tell us just how good these dampers work.

T

[K1JJ]

I just need to determine the best way to turn the finals on,

T

Put a poster of Marilyn Monroe up on the wall.

When I turn my rig on it ALWAYS returns the favour.

Are you British now?

You bloody wright, mate!
I apologise for being such an arse, but I practise at it, being off colour, I am.

What are you bloke, a bloody English teacher now?   

Thom

[W2VW]

I just need to determine the best way to turn the finals on,

T

Put a poster of Marilyn Monroe up on the wall.

When I turn my rig on it ALWAYS returns the favour.

Are you British now?

You bloody wright, mate!
I apologise for being such an arse, but I practise at it, being off colour, I am.

What are you bloke, a bloody English teacher now?

Thom

I'm the last guy who should be an English teacher.

[W2VW]

I just need to determine the best way to turn the finals on,

T

Put a poster of Marilyn Monroe up on the wall.

Dave, I'm not sure how we would ever get by on this forum without your invaluable advice.

Fred

You'd figure it out somehow Fred and it would probably involve several transformers.

[K1JJ]

I'm the last guy who should be an English teacher.

English can be corrected by a $10 / hour English major student.  You, my boy, should be writing for David Letterman.  

T

[Steve - K4HX]

Someone needs to write for Letterman. He ceased being funny about 20 years ago.

[K1JJ]

Well, the 60A contactors arrived and I couldn't resist building a HV step start with them. These are BIG mofo's and heavy. Much bigger than I thought.  Only $10 ea on eBay used. 20 watts each to key.  POW!


I ganged the three poles together and this makes a 180 amp  contactor.  I've always had problems with contactors sticking due to arcing, so this should solve it.  I use a 2 pole breaker to feed the contactors, so the 120-0-120 legs are totally disconnected when off.

The delay is about 3/4 seconds  before the second contactor comes on and shorts the 10 ohm 200 watt resistor.  (two 5 ohmers)

I'm using 24vdc contactors -  using 4 ohms and 30,000 ufd for the R/C delay to the second coil.  

Works like a charm on the bench.

The contactors are loud, so I 'm building a cover box that is lined with 1" of styro insulation. This will seal and keep things quiet, I'm hoping.  Notice the 1.5" border around the inside parts to the edge of the plate to allow for the styro..


The HV step start will work on both AM and ssb for the linears. I've always liked the idea of having the HV off when in receive for sanity reasons...


I decided to change a few plans and key the HV and screen supplies.

The new PTT sequence will be:  (With final and modulator fil  CTs always grounded)

1) Ant relay on

2) RF drive on

3) HV step start on

4) Both screen supplies turned on at 120 primary transformers

5) Both RF screen choke and Heising resistor dampers lifted off

6)  Audio on


Unkey reverses the above sequence.

On unkey, I like the idea that the dampers come on immediately and stay on -  and the RF drive stays on until the end.  This should suck out any remaining back EMF.   The HV and screen supplies will ramp down since they are keyed at the primaries.

We'll have to see how it works. Received the 100 watt 1 ohm filament adjustment resistors today and also need to install the screen keying circuits.

T

New HV step-start keying 240 VAC primary:

[w4bfs]

step start with moose contactors .... sounds like social interaction up Timtrons way

I spent a bit of time and designed a digital sequencer .... this mimics a drum switch controller and is easily expandible as long as the control action resembles a push down stack .... many different ways to do this and may already be available off-the-shelf

if anyone is interested, let me know