The AM Forum

Safety Check.....

I'm in the process of heavily modifying some amplifiers here and possibly building up a new 4-1000A PDMed by a 4-1000A AM rig.   So, there's lots of wiring going on in the shack right now.

For some reason, I'm having trouble visualizing the potential electrocution danger of losing a ground [negative lead] from a high voltage supply to an amplifier.

Let's say that chassis ground is the negative lead and the HV supply and amplifier are separate. You run a cable to the amplifier chassis from the 3KV HV supply NEGATIVE point. The positive HV lead goes to the final RF choke, etc. Turn the HV on and then break the NEGATIVE lead. What kind of voltage/current can you expect if you got between the amplifier chassis and this broken negative lead?

What confuses me is that the HV positive lead has a path only through high impedance paths... ie, only through the tube and through the plate bypass .001 capacitor.  So, it would seem the current is limited to microamps and not fatal. But I want to know what others think.  If the tube is 10 megs and the body is 100K, then little voltage across the body. [unless the tube is conducting, but it would stop when the lead was broken]

I do plan to be safe and use a separate  negative ground lead AND a thick braided strap to connect all the amplifier chassis to the negative supply point, so it will be redundant. But, the various rigs are NOT in the same rack as the HV supply - are not bonded together, thus the possibility for a wire break, though remote .

ie, We all take big precautions for the positive HV lead, but how important is the negative?

73,

T

Tom I have seen Zeners used between return and chassis if you need to float the return. This way the voltage is limited to the zener voltage.

Not 'little voltage across the body...'    little current!   If you're lucky enuf that the current is limited to less than what will kill you, you will still get the entire HV kick - maybe enuf to break something just from the muscle reaction.

I always think about Murphy sitting on my shoulder... whatever damaged the ground bond could have also jiggled the B+ to the chassis.

Best to run the B+ in a shielded cable in addition to bonding the frames, chassis, etc.  If there is mischief on the B+ cable it will have some chance of shorting to the shield. I use RG11, rated to 8KV I think.  And no connectors, pigtail the shield off to the chassis.

That's my story & I'm sticking to it.

73, Bill  N2BC

I'm here thinkng of a cathode keyed rig and even though the PA tube is in series with the B+, the key is still hot referenced to ground and many of us have stories of reaching for the key and waking up across the room.

I'm thinking in the scenario you're talking about the full B+ should be across that open.  Maybe I'm not thinking it out clearly, but to me it sounds worrisome.

Ya know, I am thinking of the rigs designed by companies whose HV power supplies had the choke in the negative lead of the CT plate iron...these chokes still had to
handle the Voltage rating, but I believe their current rating was about 30% less....
I always wondered if that really helped the power supply in any way?
73 steve

Hi Steve,

Hmmm...  I was under the impression that a choke in the negative lead still had to handle the same series current, but it did not see the high voltage to ground. Thus, when mounting it on the metal chassis, this kept it's frame insulation risk down.

But, if one were to "float" the choke above chassis, it would not make a difference if it were in the negative lead or positive lead - it would reap the benefits of breakdown protection to ground... but still see full current and have still have the end to end voltage flashover risk as in any config.  I usually mount my chokes on stand-offs, or if real heavy,  just sitting on a block of fiberglass, etc.  Same with mod transformers.

Any other comments on this?

BTW, tnx for the comments on the negative lead HV guys. Yes, the cathode keyed final does demonstrate well a potential risk of the negative lead being off ground and coming between it and the chassis.   I guess the only way to find out is to put a HV probe in the negative power supply lead [bridged with a resistor to simulate the body resistance]  and turn on the HV to see the real truth under various conditions.

T

Tom,
Maybe tube is meg ohms when its cut off, but when its conducting it will
have low impeadance. I think too, it is like a cathode key amplifier, so it will have
near full potenial from chassis to the broken ground.

I once got a nasty shock from running my valliant with an external moduator, using
the B+ off the valliant, I forgot
to ground the modulator chassis to the the valliant chassis. I then was touching the modulator
cassis and reached with my other hand to Plate switch and keyed the vallint. I became the common
ground, and bammm next thing i knew i was seeing stars, fortuanlly my hand released.

Be careful, Chuck

That is correct.

Triple ground everything. It just isn't a good place for research. Just picture Clint Eastwood: "The question you have to ask yourself is do ya feel lucky....".

Tom,
I use RG8 with HN connectors for high voltage. The shields are always terminated to chassis ground just by the fact that the connector is mated.
bone head fool proof. You could do the same thing if you wanted to float the ground. I would run extra safety ground lead back to the PS as another return
myself.
Chokes in the negative lead only see the DC drop and ripple across the choke making it only a couple hunderd volts above chassis ground.


I smell a new PDM rig with the modulator in the right place.....BTW somebody is working on a floating modulator with fiber optic interface.

PL259s suck for high voltage.....

Frank,

OK on the shielded coax and precautions.  I usually cover the whole thing with plastic fexible water pipe for runs outside the cabinet. Right now I'm powering three different amps with one supply. [pair of 3-500Z's, second pair of 3-500Z's and the 8877 for 6M]

I wanted to keep the following questions on the BB so that otheres may learn or get interested too....

On the TUBE PDM rig...  I'm curious why you prefered to use the switching driver FET in the tube's cathode vs: Steve's grid FET on that rig we built in '90?  I blew a lot of FETs up in the cathode, but it seems easier.  Also hoping those switching diodes will last this time too.

I was thinking if I used TRIODES for both final and modulator, I cud use grid leak resistors for both.... just requires a floating fil transformer for the final, regular fil trans for the switching tube and a plate supply -  that's it!!!

Could I use a big string of diodes in the cathode of the modulator for safety bias or wud that muck things up?

Also, where would you mount the PDM coils/ power supply to insure better switching freq suppression?  My HV PS is outboard as you know with about 4' leads. 

T


 

Tom,
I did catdode switching because it was referenced to ground and not a negative supply. Yea we blew a lot of FETs because of the long run back to the supply and bounce on the ground lead. Ground wasn't ground in the shack but down in the basement where the power supply filter was located. Lots of series L.
The FETs were 1 Kv units so if everything was right that should have easily switched the tube off. The damper diodes blowing was due to the same problem.
Poor damping also made life bad for FETs. I have never blown a FET in the V2 and only see about 60 volts peak at the cathode of the 6DQ5s. I'm running 1550 plate volts. Scale that to your voltage and you should have had 100% safety factor in a perfect set up.
Screen voltage can be had by adding a winding to the heater transformer or float a second transformer to make bias and screen. You could get away with a low voltage transformer by using a fil winding as the primary and feed it with the secondary of the final heater secondary but you will have to float the case.
As you learned layout is critical in a pulse system.
I will never forget the Jacobs ladder going up the feed line as Tim and I watched when you went inside and fired up. The spark echo of modulation was quite impressive at QRO.
You can't beat shielded wire for safety. A cool old spec to see AFR122-10
for safety. It has been replaced with newer one repackaged information.
fc

Tom the modulator/damper diodes and filter needs to be close to the power supply.
A switching suppply/modulator needs a low Z source otherwise you are setting up a second tank circuit that as you learned can't be controlled easily at that plate voltage. The damper was ineffective because it dumped energy back into a long inductive lead setting up yet another tank circuit in the B+ lead.
It would be a lot more stable if just audio riding on DC came up from the P.S. / modulator without pulses taveling in everything.
HMMM Tom is dreaming of big dumb triodes again after he learned how to do it at 50 volts. Unstrapable winter coming?

Thanks for the great info, Franz.

Yes, that's a clever way of running the final's screen voltage using the floating fil source as the input to another fil transformer and getting 240AC. I suppose a voltage doubler wud get ya up into the 500VDC zone for 4X1 screens.

This time around I plan to keep the HV supply leads within 3'-4' of the modulator and final. Might stabilize things.

I'm still not sure why I cudn't suck out that 90 khz switching freq more than -60 db ... even using a trap last time.   But this time maybe the closer lay out will help. I plan to use the 110kc? PDM low level board design from the solid state PDM fet rig.

Still trying to decide whether to use the 4X1  or use some ext anus triodes like before. The triode has no soul, wheras the 4X1's wud have viewing appeal.  The 4X1 might be easier to stabilize for band switching plus will be much easier to drive.

I'm gonna start from scratch and design up a new PDM filter system. Maybe I'll go with three coils this time like the solid state PDM rigs use. Gonna use single layer, so the first coil will be about 2'-3' long, 10" diameter again... ;D

I'm leaning towards the 4X1's, but we'll have to sleep on it.

T

Tom I was thinking of a TV style transformer to be used as a bias / screen supply.
Take say the 5 and 6.3 volt winding in series and use it as the primary connected to the floating 7.5 fil source. You couldn't attenuate the switching niose due to circuit layout and distance from the power supply output cap. Dumping damper energy on the B+ lead and ground bounce on the modulator with the long ground lead. The modulator would be a lot happier next to the power supply.
You will need larger wire in the filters if you QRO. #26 that I wound was for the 4-1000A. Remember the inductor doing pull- ups on voice peaks....
sure a lot easier at 50 volts Mr. Tesla

I never thought of doing that!!!  Excellent idea.   Now we will have BOTH the screen and grid protection supply floating for HV off of a single TV type transformer tied to the floating filament transformer.

I plan to wind a filament transformer for 7.5 volts on a large variac core. That usually works well.

The 4X1 looks even better now with the ease of the supplies.

Do you think a string of diodes in the cathode of the PDM tube wud work as safety bias along with a grid leak resistor?  ie, How wud you handle the bias for both the final and pdm tubes?  Grid leak for both?  Do I even NEED protective bias on the class C final since the PDM series tube is there?

I also was thinking of mounting all of the 4X1's in with a large viewing window. I need to put a shiled between the PDM tube and the final to isolate the pulses, right?

T

I understand the need for the modulator to be near the power supply... but talk to me about circuit layout. As I remember, the PDM filters and modulator were in the bottom of the rack with a steel plate separating/shielding them from the final deck above. How would you lay out the new rig???

T

Do you remember Steve's tube pdm rig? He had the PS modulator in one bay and the final in the other.

Take a chassis and mount the modulator tube with the PDM generator and interface circuit inside the chassis. Above the tube do the filter and damper diode stack. You could mount in in a rack near the power supply. I would put a
small oil filled cap near the damper stack to absorb the pulse. The cap would go between B+ and ground. Output of the filter and B+ go th the shack in 2 coax
cables. Your problem was having the modulator tube and final next to each other.
you could do the whole thing up in the shack but the power supply output cap would have to be in the shack near the rig to reduce the high frequency switching current loops.
50 volts is so much easier om and safer...........

OK, Franz, will do.

I'll get everything in the same cabinet, but separated and shielded.  The HV power supply downstairs is a thing of the past, maybe 1993. That was the old "White Room".  My PS is now in the shack next to Mr Ugly.

I guess I'm all set now to design something up except for the questions back in the last post about using protective bias with a diode string in the PDM cathode and if I need it also for the final... Please re-read it..

Tnx -

T

I can't answer bias questions till I know the switch tube and control method.
You could use a high voltage zener across the fet if you switch the cathode.
It would have to be high enough to turn the tube off but low enough to protect the FET.
You could put the modulator on a chassis at the bottom of a rack and the final above it with the filter in the middle. I forgot you moved the pig farm up stairs.
consider a GG 4-1000 A about 20 volts on the cathode will turn it off at 3500VDC plate voltage. 0 bias will
draw about 100 ma. I suppose you will have to put some screen voltage on it to get more current. That will mean it will take more voltage across the fet to turn the tube off. The lower the screen voltage the lower the saturation voltage when the tube is on.
My V2 switches the cathode with g1 at ground and g2 at 60 volts. saturation voltage is about 100 volts across the tube when it is on I seem to remember about 100 volts across the fet to reverse bias the tube. I tie the cathode to the CT of the fil trans and fet Drain yanks both CT and cathode.  A 900 volt fet in the cathode should do it for you.  I would put a noninductive resistor across the fet say 10 K maybe less. Just enough to back bias the tube when the fet is off. Short leads are required. I would put the pdm generator under the chassis near the tube.
I put a 6.8 ohm wire wound resistor at the plate caps of the 6DQ5s to act as a VHF suppressor. see if Steve has and tube pictures on his site to look at??

I like the idea of the 4X1 switch tube in GG. Yes, screen voltage wud be required to get the switching 4X1 into saturation considering the cathode FET cannot add anything below ground.

Simple is good.

Well, looks like you've answered my questions there, OM. Think I will go the 4X1  X 4X1 route.  Already have the tubes, sockets and chimneys.

My existing HV supply is going to be rather low for PDM though. I will be getting about 6KV max across the system which equates to only about 2.5kv across the 4X1. That's OK, cuz I can always switch out the PS filter choke later on once the bugs are worked out.

How wud you calculate the plate impedance to design up the filter?  I was gonna use Bill/GF's filter calculator that we used for the FET E rigs.  Is it as simple as E/I or is there some additional factor used for AM like in Bill Orr's book?

I was thinking of doing up a 3 coil/ 3 capacitor PDM filter this time.

Later -

T

Tom,
plate z is plate z. I remember I did your filter for 2K. Yes start low and work the bugs out. See if you can find a high voltage TEK scope probe. Bill and I never agreed on filters so do what you want. You know his and mine worked. I designed my filter from a table I found years ago in a Don White EMI filter book. It was voltage mode drive and has a larger first inductor. I simulated it and it worked fine. It worked just as well scaled to 2 ohms. Bills works fine also ??
always ask yourself as you lay it out. What is the current return path. Keep it short and you will be fine. Fair amount of lightning tonight heading your way. Front must be pushing this crap out. I want to go to the beach but WX too crappy for a ride down tonight. Can't fit everyone's stuff inside the truck, don't want wet clothes.  Man it has been raing hard all night....all week

Thanks for all the FB info, Frank!   That's real ham radio - helping each other.

I've got the 4X1's, chimneys and sockets laid out on the table now. Trying to figure how to lay it out. I was thinking of making a big viewing window for both the final and modulator... like 12"X12". That wud be slick and is half the fun of running glass tubes.

Later -

T

Cold WX is coming so a little tube heat will be nice.
Good luck and be careful.

That's almost exactly what happened to me with a homebrew rig and a modulator which consisted of a hi-fi amplifier feeding a reversed output transformer.  The external amplifier was not properly grounded, and the reversed output transformer, which was never designed to handle 1000 volts DC on the hi-Z winding, shorted to ground, putting the full HV on the entire amplifier.  The power transformer in the amplifier evidently had better insulation and didn't short out.  With microphone in one hand, when I touched the transmitter to adjust tuning, I had the full 1kv from arm to arm.  Couldn't release it.  The only way I was able to break away was to break the the mic cord loose from the connector.  Lucky it was one of those crappy 50's vintage screw-on high impedance mic connectors where the only thing holding the cord in place was a setscrew, plus a blob of solder on the hot audio line.  If the mic connector had been solidly attached, someone would have found me hanging on the rig the next morning.

I have a HB transmitter, four sections starting from the bottom:

Final HV PS
Mod PS
Mod
RF Deck

They are all in a single Bud cabinet.  The original cord was a two prong.  I replaced it with a three prong and attached the Final HV PS chassis to the ground wire.
The HV supplies use the chassis ground as the common negative path.

How safe is this setup?  There are no ground connections between the sections besides contact with the cabinet.  I don't trust this.  How about running a fat wire down the back connecting all four chassis'?

thanks

Rich

That's what I have on mine Rich. I still haven't gotten around to changing it to a 3-prong plug. Maybe after I change out the brute force cap to a tuned circuit. ;)

Hi All - this has turned into a good PWM discussion !   Lots of good comments.

I'll relate some of my experience with high power tube PWM transmittes.  First, you can put the modulator anywhere - in the next room if you want.  However, as Frank has pointed out, there are a few critical leads which must be kept short.

1) The damper diode is connected directly to the plate of the modulator tube at one end.  The other end must go directly to a capacitor to the local ground of the modulator. You need to also provide a good, low impedance local capacitor between the high voltage and the ground of the modulator, at the modulator.  This local capacitor and the local ground (chassis) of the modulator are the high current points involved in the modulator and damper diode.  You will generate serious ringing without these.

2) You *must* make a provision for analog compensation, or you will have distortion at high negative peaks.  This means you need to provide analog control of your modulator tube.  Either you need to vary the amplitude of the input pulse train with modulation (the drive follows the modulation - higher drive for higher modulator output), and/or vary the screen voltage of the modulator tube.   I have found varying the drive is very important and effective.  The only alternative available is to provide an active pull down (up, in this case) to actively discharge the filter network.  Doing so will compromise the modulator efficiency, but the method does work.  The broadcasters all used analog compensation, and so did I in my rig.  The other problem you will solve with analog compensation is the fact that a tube is an imperfect switch and will exhibit an increasing voltage drop as more current is drawn through the tube.  The analog compensation will linearize the tube at high currents, and will also allow you to achieve 100% negative modulation with much lower distortion.  There is extensive documentation about this on my web site in the "Vacuum Tube PWM Article".

3) Important:  Beware of coupling around the filter - through the ground.  I have found it important to isolate either the RF amplifier ground (impractical in this case) or the power supply / PWM filter ground.  You might think everything is "grounded", and your large rack and all that aluminum will surely never allow any ground currents, but you very likely will have some, and this will reduce the effectiveness of your filter.  In my PWM rig, I actually DC isolated the modulator and PWM filter ground plane (and it was a plane). 

One suggestion is this:  The power supply ground connectes to the modulator (isolated) and PWM filter (isolated) ground point.  This is where everything related to the modulator joins:  The bypass for the damper, the modulator local bypass capacitor and the first PWM filter capacitor.  Then you go through the filter.  Since the filter is large, there should be a ground plane provided for the filter (isolated).  The last capacitor in the filter is located physically inside of the RF amplifier cabinet, and is usually the RF bypass as well.  A ground strap is run from the RF amplifier chassis (which is presumably the actual cabinet/rack ground as well), to the filter ground plane.  The high voltage, of course, runs from the filter output as well.  In doing this, you will reduce considerably the chances of coupling around the filter.

When running high power, an 80dB down spur from your pulse width modulator will be very apparent.

A series resonator at the output of the filter (preferably located very near to the RF amplifier and grounded to it) can be of help, although your 6 element filter should be sufficient.  I have not needed to provide any other circuitry when using a 6 element filter with the corner at 12.5kHz and a 100kHz switching frequency.  The PWM spurs are unmeasurable by other operators who are a short distance from my residence.

Anyway, just a few random thoughts !!! ;)

Regards,

Steve

If HV mistakes and excapades weren't so serious they could be funny.
Can just see it....

Don said, " when I touched the transmitter to adjust tuning, I had the full 1kv from arm to arm.  Couldn't release it.  The only way I was able to break away was to break the the mic cord loose from the connector. "

How did you break the 'mike' cord?  with your teeth?

That was my first "visual" and you gotta admit, it's would make an excellent safety cartoon.  :o

stories like Don's are why I'm spending so much time on my ground system. There will not be a single piece of gear powered by ac in my shack not  connected solidly to the same ground system and going back to the service entrance. I'm torch silver soldering all the split nuts, and acorn nut joints today and pounding down another 2 rods. The one for the service entrance sux. No bonded copper, just galvanized. I dont trust it.

My HB rig had full B+ on the tank coil, plate caps in the open, etc.

I read once some guy in the late 20's swore he would never build a rig with a metal chassis. All his rigs were made with wood.

 

It just happened when I fell backwards.  The only thing that saved me was the crappy mic plug, and the fact that I probably hadn't fully tightened the set screw.  It was one of those old style single wire shielded hi-Z mic connectors that holds in place with a screw ring and uses a piece of bakelite with a blob of solder in the middle to make electrical contact for the audio line.  The mic cord pulled out of the connector and the wire broke loose internally from the blob of solder.

I have a grounding problem at the computer.  RF from my transmitter gets gets picked up and rectified by the amplified computer speakers, and the UPS power strip and the cable modem both generate rubbish that gets into the power lines and causes interference in  my station receiver, even though the shack is separate from the house and located 80 ft away.  I have tried a brute-force a.c. line filter between the mains and the computer power strip, but the rf appears to be  travelling by common mode, and the ground wire simply serves as a by-passes and detours the rf around the filter. I have several isolation transformers, but would most likely have exactly the same problem with one of  those.  One possibility would be to eliminate the ground wire between the computer power strip and power mains, and strap it directly to neutral right at the secondary terminals of the isolation transformer.  But then I lose my lightning protection ground, but perhaps the shielded internet cable, which goes to a separate 8-ft ground rod at the cable entrance, would  serve the purpose.  The computer is located upstairs, so a direct wire leading to an earth ground (including the cable shield) makes a nice Marconi antenna, and would probably make the interference problem worse rather than better.  Anyone have other ideas or experiences to relate?  I suspect this is a very common problem.

PDM Stuff
I parted out a Continental PDM rig that ran three 3-500Z's
I the HV insulated filament iron, etc.if someone could put it to use I will trade it for something more useful. I even have the manual for the ugly thing.
Skip

Wow, would that rig have been SOMETHING on 160 !!!!!  The Continental had some rather unique features.  I can't, off hand, remember what it looked like, but most of the modern broadcast rigs just aren't as attractive as the older ones!

Regards,

Steve

It can't get much uglier than the Gates BC1-T, unless it's that one (CCA?) with the fake wood grain cabinet.

This applies to almost anything.

The operative word is "plastic". Yes. it's all Plastic.  ;D

LOL. A BC rig with a D-104. What a waste.

Same thing you've been drinking. ;D

D-104 on a BC rig. What a waste.

Looks like some good stuff!!

Lotsa good wineries up here in VA too.

is it me or did the price of wine jump up. I have not bought any in a while and was shocked at the prices yesterday. My Uncle Joe has my G.F.'s old press and he tells me he in into making homebrew wine. Maybe I should visit him and get some pointers. $11 for 750 ml of good red stuff is almost as stupid as a $56 Christmas tree.

Oh not to rip off the thread I took a 300 volt defib a couple hours ago been years since I did anything that stupid.

http://www.bumwine.com/ (http://www.bumwine.com/)

prices of bum wines have never been lower.  :D