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

[KA2DZT]

Tom,

Whole set-up looks good.  I have contactors bigger than those for starting motors on some of my machines.

I see you're going with both methods, this should work FB.  If it doesn't, we're back at square one.  Try to keep the timing between the HV and screen supplies close.  You don't want no screen voltage to cut off the tubes before the HV has time to decay.  On start-up it doesn't matter but on shut-off it might.

Fred

PS;  if my lights dim down here in Jersey, I'll know things aren't going too well.

[K1JJ]

It frickin works!!


After some more mods I finally fired up Fabio II. Tested to see if the new sequencing system stops this bad gap arcing when unkeyed.


Here is the final configuration -  when the PTT is hit:  (1 to 5)


1) Ant relay on

2) RF drive on

3) (Three at once) HV step start on, modulator screen supply on, RF final screen supply on. (ALL supplies keyed at AC primary)

4) (Two at once)  Heising choke resistor damper off.  RF final screen choke resistor damper off. (Using vacuum relays)

5)  Transmit audio on

Unkey, the sequence reverses  steps 5 to 1.


I'm still running just 2KV for now. Unlike before, there is no sign of arcing on any of the gaps during key or unkey.  I even tightened the gaps up quite a bit. All of the current meters are well behaved with rock-solid motion. The only minor aberration is that the modulator screen current moves up as it discharges the screen supply after unkey once the HV gets below the screen voltage. No problem. It's only about 50ma total - the tubes are rated at 280ma screen total.  

The HV drops off very quickly as desired.  I think the dampers are really doing an excellent job killing the EMF.  The sequence is maybe 3/4's second long.. not bad.

Next I'll run some more tests and then ramp up the HV to see what happens.

Another problem, but can be fixed:  The HV step start contactor is so loud it makes the rig hit 100% modulation  -  BANG!  It's like Fabio is shooting himself.  I'm going to find a cover for the contactors and line it with 1" of styro insulation.  

I just love the keyed HV. Nice feeling that there is no HV in the shack when unkeyed.


I didn't see any signs of instability in the RF final or modulator.  Think it's gonna turn out to be a FB rig.

Also working on a new 75M 2el wire Yagi at 140' high, facing Washington state WNW. Supported with two 36' long boom bridles, one for each tower to keep it evenly spaced and flat.  We on the east coast have been neglecting the 75M AM guys out west for too long and I want to get some new coast to coast action going soon.  W8, W9, W0, W7, W6, W5, etc.


THANKS to all who have lent their sequencing expertise. It made a big difference.

T

[steve_qix]

Keying the H.V. is much better than leaving it on.  Even with my old tube PWM rig, which used a 10,000V power supply, I keyed it on and off (with a step start).

Much safer, too!

Hopefully, I'll get to hear the rig on the air sometime soon.

Regards,

Steve

[KA2DZT]

I thought you were going to unkey the screens ahead of the HV.  Although, I suggested keeping the timing of the two close.  You have the HV and screens keying and unkeying at the same time.  This shouldn't be a problem as I see it.  The screens will draw more current if the plate voltage is less than the screen voltage.  We had no way to estimate the decay time of the two supplies.  Seems the HV is decaying faster than the screen supply.  I don't think any of this is a problem.  Also, this situation may change when you increase the HV to the normal operating level.

I agree,  no reason to have the HV on when you're on receive.  No need to worry about HV lurking about when you're just in the listening mode.

All you need to do now is to reduce the contactor noise.  I see you're using two of the same contactors to step-start the HV.  The one that shorts out the 10ohm resistor could have been a smaller relay.  There shouldn't be that much current surge on that relay as most of the current is flowing in the resistor.

Anyway, glad you have it working.  I knew it would work FB.  If it didn't, I was out of ideas.  Having increased the time decay of the plate current dropout even a fraction of a second greatly reduce the back emf in the reactor.  Add in the snubbing resistors, it had to work.

Finally, add in what we all have learned with this thread.

Fred

[K1JJ]

Hi Fred,

I think the HV and screen supplies are so slow to ramp up or down that it wouldn't change anything to make them separately keyed.   Maybe lengthening out the sequencing steps would do the most to give them both more time to discharge before the ant relay drops out.   But right now I see no signs of power needing a place to go.  It keys just like a tame linear amplifier...

I have only four steps to work with using this sequencing board, so that was the main reason to group some of the functions together.

In hindsight, I think the resistor dampers on the two chokes were the big breakthrough. I never heard of doing that before and I'll bet it wuda solved the problem even without the HV and screen keying, as well as keeping the CTs keyed.


Steve, I will be looking for a critical report when you hear it. I've already run some THD tests and it looks above average. There are distinct THD sweet spots when running various loading, screen voltage and drive levels - as well as modulator idle bias settings. The modulator seems be cleanest idling at about 200 ma, whereas higher idle gets worse, against conventional thinking.  I can see the distortion go up and down. When optimized, the thing seems to be pretty clean. It will probably get cleaner once I raise the voltage up some.   Fabio I didn't like 2KV very much and got better at higher voltage.  I have about 8dB of NFB dialed in and that seems to help extend the low end down to about 20HZ.  I can see the shark fins, just like the e-rig, so that's a good sign.

Looking forward to getting on 160M and 40M AM too... been a long time and I've made it easy to change bands with pre-sets and coax antennas.

See ya on soon!

T

[w4bfs]

nice going Tom .... sounds like a good solution ... I have finished initial design on a digital sequencer .... can be anything from 2 to 8 off-on-off steps using cheap and readily available LS parts ... this is a clocked design and allows changing step times to allow for relay activation speeds ... this has been fun and remininded of the kind of work was available before NAFTA ...73...John

[PA4WM]

In hindsight, I think the resistor dampers on the two chokes were the big breakthrough. I never heard of doing that before and I'll bet it wuda solved the problem even without the HV and screen keying, as well as keeping the CTs keyed.

Would that trick work with just a mod transformer as well, without a heising reactor?
What about the resistor value? Does that have to be in the final tube impedance range?

Martin

[K1JJ]

Hi John,

If I run into any more problems related to needing more sequencer steps or customized spacing between steps, I will take a look at your new development.  If I had known about it earlier, I wud have used it, but this 4-step unit is hardwired in and working for now...

Martin:   Yes, it wud work FB. There's no reason why the damper technique wouldn't work with a stand-alone mod xfmr. It's all about putting a load across the modulator iron, however it's done to dissipate the back EMF.

Stu suggested 5K which would be the impedance of a typical QRO RF final.  I used 10K simply cuz it's the value I had in 200W.  

Once I ramp up the voltage above 2KV, the real story will be told whether the rig stays glued together.  I bet it will be fine.

T

[Steve - K4HX]

A proven sequencer.

http://www.w2drz.ramcoinc.com/Sequencers.htm

[K1JJ]

A proven sequencer.

http://www.w2drz.ramcoinc.com/Sequencers.htm

That's the exact sequencer I'm using....

BTW, I was up on the tower at 140' for 5 hours today. Some complications. Finally got the first 36' bridle up ready with ropes and pulleys for the new WNW wire Yagi.   Windy day. I'll tell ya, this work is unnerving and I'm beginning to dread going up there.

The rope and pulleys might be the way to go in the future to do most of the antenna changes from the ground...

T  

[steve_qix]

The rope and pulleys might be the way to go in the future to do most of the antenna changes from the ground...

For wire antennas, that's the way to go.  I only have to climb the tower if something goes wrong with a microwave link, or with the pulley itself (which is very rare).  Otherwise, I just lower the antenna and stay on the ground the whole time.  This is particularly useful during the winter months when it's 10 degrees out and windy and where climbing would be VERY uncomfortable or impossible 

[w4bfs]

A proven sequencer.

http://www.w2drz.ramcoinc.com/Sequencers.htm

thanks Steve .... I was not aware this board was available .... the one I designed is similar but different since I never met 2 digital/analog designers who did things the same way...  

[Steve - K4HX]

Lots of ways  to do it. A PIC could be used too and any set of delays and sequences could be programmed.

[KA2DZT]

Tom,

At some point, if you feel like trying it,  I'm wondering how much of a reduction in the back emf was attributed to just the step-start of the supplies.  Meaning, the slower decay of the HV and screen supplies should have greatly reduced the voltage spikes from the two reactors.

You would have to not use (disconnect one lead) the snubbing resistors (just for a one time try) and keyon the rig and keyoff to see how much, if any, arcing may occur.

From the math, the voltage spike is directly related to the rate of change of current in the reactor.  The slower rate of current change caused by the slower decay of plate and screen voltage may be all that is needed to prevent the arcing.

Just wondering,

Fred

[Steve - K4HX]

One of my amps has a snubber circuit on the HV. No arcing on key up/down.

[K1JJ]

Fred,

If I hadn't put so much work into this project I would probably try some experiments and disconnect the dampers, etc.  I'm curious as you are.

But, all of the techniques combined seem to be doing the job and the rig is very quiet and stable during keying. I wouldn't want to risk disabling the dampers. When I was experimenting with sequencing last week I had two instances of fireworks that broke some parts. So I shouldn't tempt fate again and maybe lose a mod xfmr next time...

The next risk is to raise the HV. That in itself is scary.   I'd rather risk parts doing that with all suppression working.

T

[KA2DZT]

Tom,

OK FB,  If there's a chance of damaging parts I wouldn't try the test.  Like the old expression says, quit while you're ahead.

If anything is on the edge, you'll find out when you crank up the voltage.

Fred

[K1JJ]

Fabio II was running FB at 2KV with the new HV step start and damper system.  I then ramped the voltage up to 3KV.  It ran extremely well for about 5 minutes. Beautiful positive peaks.  Then BANG!!!! 

I saw an extended bunch of flames and arcs coming out of the back of the modulator. Please, not the mod xfmr!    There was an acidic smell in the air. 

After 30 minutes of conquer and divide trouble shooting, I isolated the problem to the Heising reactor.  Then I saw it - the new HV damper relay across the reactor was burnt and charcoal on its 24V coil.   It just can't take the abuse. It also took out the wire HV fuses in my HV supply.

Evidently there is a lot of energy still there even with the sequencing.

To make Fred happy, I tried the keying without the HV damper just to see what happens. There was a double arc, bigger than usual at the Heising reactor gaps.   I tried it again and this time shut off the HV supply and let the rig bleed it down before I unkeyed. No arc. Looks like I need to get that HV supply a longer time to bleed off.

Must order a new and bigger HV damper relay and try again.

T

[K1JJ]

Telephoto shot:  Today there were three hawks perched  at 190' for at least 5 hours holding court on 40M. Maybe they are making a nest / shack, I dunno...

The 2nd picture gives perspective.  The hawks are barely visible.

T

[K1JJ]

Here's the finished 36' bridle at 140' ready for the WNW wire Yagi.  Notice the ropes and pulleys ready to connect to the driven el and reflector. (Work from the ground)  

This was a bitch to get up there and rigged by myself.  The other bridle side will reside on the other tower about 50' below the hawks.

I just love wire antennas.


Third shot:   Yaz is showing his age, but still a good guy.  (10 years in August)

T

[Steve - K4HX]

JA short path?

[KA2DZT]

I was worried about the 24V line that close to such HV.  Didn't mention it cause I didn't know the limits of that relay.

You need to let the high voltage decay as slowly as possible.  Remember, it's only the HV filter caps that are discharging through whatever the load resistance is.  The greater the plate current at shutoff, the shorter the decay time.  The greater value of filter caps, the longer decay time.  You're already using big caps (100ufd).

Question,  Did the relay blow while you were testing or did it blow when you keyed off??

[K1JJ]

JA short path?

Nope, this is for the lonely 75M AMers in W9, W0 and W7-land we haven't talked with coast to coast in years...  It will be facing about 290 degrees, Washington state.

Now that you mention it it will make a good JA short path, but that means I'd have to get up before dawn... gag gag.

T

[Steve - K4HX]

Now that's funny!

[K1JJ]

I was worried about the 24V line that close to such HV.  Didn't mention it cause I didn't know the limits of that relay.

You need to let the high voltage decay as slowly as possible.  Remember, it's only the HV filter caps that are discharging through whatever the load resistance is.  The greater the plate current at shutoff, the shorter the decay time.  The greater value of filter caps, the longer decay time.  You're already using big caps (100ufd).

Question,  Did the relay blow while you were testing or did it blow when you keyed off??

It went when transmitting, which leads me to believe that maybe it started a path when unkeyed and then when I ramped up it just blew apart inside to the coil.  I may saw it in half to see what it looks like.   The bottom at the coil is melted but the whole case is intact.

I noticed that the HV damper resistor did get warm but not hot after many keys at 2KV.

Still thinking of what to change. 


T