The AM Forum

Back again!  I scored 20 feet of tinned copper braid and 20 feet of 40KV wire that fits inside like a glove.  Do I have to shield the B+ from the PS to the modulator and from the modulator to the RF deck or is the shield only necessary within the PA enclosure?

I rather prefer shielding it everywhere.  In the very unlikely event that your 40kv wire is damaged, that shield will take the HV to ground before it goes somewhere else.  There's no need to be afraid of high voltage, but it deserves respect, and a belt-and-suspenders approach.

The HV wire does not normally need to be shielded because of RF getting into the power supply, etc. The bypass HV caps and PS filter caps will take care of that.  You could run unshielded HV wire everywhere and the rig would work fine, though shielding it inside the rig is sometimes done.

But as Norm said, the safety issue is what counts.  For outside-the-cabinet runs I prefer to use a grounded shield on the cable to absorb accidental wire pinches to ground and I also put a plastic flexible tubing over the shield as a secondary protector.  

The idea is that you could cut the live cable with an axe and it would blow the breaker.

T

Agree with JJ,  no need to shield your HV wiring anywhere.  Commonly done years ago for what reason, I don't know (probably TVI).  You can run your exposed HV wiring inside some plastic tubing for added insulation.

OK; thanks!   :)

Somewhere in the HV circuit there should be one or more 0.001 uF to 0.01 uF cap about 2XHV to shunt any RF to ground.

That should take care of RF floating around the system.

Phil - AC0OB

I only shield the HV and modulated HV if it comes outside the equipment. Just for safety purposes.
As mentioned there are bypasses around and RF gives no trouble.
The wire is always bypassed when passing through the RF enclosure wall.

A little reminder on HV safety...

On Friday morning I was working on my amp, but didn't give enough time for the B+ to bleed down. I don't know exactly what happened, but I stuck my test lead into the RF deck near the plate capacitor and there was a bright flash and a bang. My pinky came into contact somehow with the plate cap. Luckily, my forearm happened to be touching the chassis, and there is a HV exit burn at that point. You can see it in the picture.

My arm went very tingly for 10 seconds, but I was ok. Everything seems to be healing fine and I still have the sense of touch everywhere.

Jon

I have a pair of 690 pf @ 20 KV for the main HV bleeders.  Schematic calls for 2 x 500.  All HV wiring will be inside a 4 foot Atlas cabinet.  That's Bella Cat.  ;)

Had the same thing happen to me a number of years ago. Exact same situation. PS was turned off but filter caps did not bleed down fast enough. Got the back side of my and and thumb across the filter caps and chassis.  Based on bleed down time and the contact I figured about 1.8kV across the hand.  A flame shot out of my thumb where it contacted the filter cap terminal and had a burn mark below the pinky finger where contact was made with the chassis.  Had to have minor surgery to clean out the burned tissue in the thumb due to infection.  Jon is 100% correct on HV safety.

A few months ago we had a thread that mentioned Grounding Sticks, sometimes called "chicken sticks".  It's an insulated stick, with a hook on the far end connected to an insulated wire that goes to ground, either through a big alligator clip or, better yet, bolted to ground.
When you think everything is off, touch everything in the rig with it, just to make sure.  Most every broadcast transmitter has one built right in.  And if someone calls you a wuss for using one, just rap them with it.  Bleeder resistors can and do open up with age and should never be trusted.  Some of you guys give me the heebee-jeebies.  They only used 2200 volts at Sing Sing, but you can do the job with a lot less.

I always like to hear these kinds of horror stories because I know you guys have successfully learned and will now go on to live long, safe lives around electricity.

The pain and resulting emotional trama that we get by eating bad food (food poison) getting mauled by a dog, hit by a car, falling down the stairs, getting burned on a stove, pulling a muscle in our back, slicing our hand with a knife, etc., etc...   is what it takes to wire the neurons in our brain to evoke a very strong negative emotional response every time we try it again.  And so it is for a bad electric shock.

Without the first-hand personal shock experience, few of us can learn by reading about it of even seeing it. We must experience it first hand to really get mentally programmed. Otherwise our safety procedures may become half-hearted.   This is a pity because some things, like 2KV ++ electrocutions don't give us a second chance. That's the hell of it.

The best thing that can happen to a ham or radio tinkerer is to get a nice healthy jolt early on in his career and live - whatever it takes to wake us up out of our complacency. This is almost a guarantee that we will do everything possible to make sure it doesn't happen again.  

Blessed are those who can learn from others' blunders without experiencing it themselves.

T    (Member of the Belted Bad and Lived Club)

You know, all this for a PW 1500 watt linear.  What's up with that stupid power limit rule anyway?  I talk on 40 meters AM, but I've yet to have a QSO with someone in China.  We need more power.

Jon

Here is my HV Supply for a pair of 811A's. I tried to over engineer this thing a bit in preparation for something larger. This unit is a V doubler with the goal at 1200 V 500 mA. The VA on this xmfr is unknown, weighs in at just shy of 20 pounds. I suspect it is at least 500VA. It is a 120-480V control transformer and with the TR and peak V available from the doubler the transfer from 120 to HV is 10.66:1. I am seeing exactly 10:1. Several of the ideas for handling the HV layout came from folks on the forum. Thanks! Alan

And don't trust the ceramic on bleeder resistors to contain the HV.  Was testing the 1.5 KV PS I built and got the back of my hand close to the resistors while it was operating.  Fortunately, as above, another part of the hand was grounded and the HV did not get up the arm to mess with the heart's electrical system.  A plexiglas shield for the bleeders is being worked on.

Hi Mitch, thanks. Yes! I was a little terse in mentioning what I did in putting this together. I do not have any interlocks, yet... However, it is not clear in photo 1, but there is a plexiglass shield all on the top of the card. As well on the bottom of the card out of view from this photo, there is plexiglass bottom that the card is mounted upon with ceramic spacers. So the HV card is sandwich between some plexiglass plates. Furthermore, the HV wire exists the card and passes to a ceramic standoff, but the wire and the standoff are in cased in a plexiglass sample vial. It's basically a pill container with the top cover secured so there is no HV line exposed. The HV line runs off to a 3/8 auto hose line and is all contained in the hose line which eventually mates with the amp. 

I am glad you are OK.  That white spot. I'd let my doctor look at it. Think of it this way, what if it had been a 3CX3000 or a 4-1000 GG amp? It might have had 5000V waiting there and could have been far worse a burn, like emergency room horror time. I agree that for AM the limit should be higher. I always think about +6dB over the present rule as being reasonable for my taste, but I am too cheap to spend money on anything more than that. 8-)


Those C should be fine for RF bypassing. What are you using for HV bleed?

People criticize me for wasting 90mA in bleeders but I want the plate voltage gone quick when the key is up. I use the permanently attached chicken stick no matter what. If I open the cabinet doors on my TX then the unit is un-keys including the HV supplies via their primary relays. I can bypass that but have to do it intentionally and only done so very rarely.

That is a nice looking cat. Calico are said to be good luck and every one I have ever met was nice tempered and gentle. Never been bitten by a calico.


If it had bleed down time and still had 1800V, did you install a higher current bleeder?


Those are decent size gray resistors next to those caps. Do they make a bunch of heat that will over time ruin things? Its a nice job and layout but I always nitpick my stuff about heat from things, esp. large resistors which I don't like to place on PCBs.

Hi Patrick. I agree. Actually, the bleeder R's are not on the PCB. They along with the meter shunt and series R's are on a terminal card which is raised above the PCB as it sits on standoffs. That one bleeder R is a little close to the jacket of one of the Caps and I was a bit bothered how that worked out. You might note that the top glass top is raised a bit with some black spacers to keep the glass above the resistors and let some air circulate. However, the bleeder system is 100k and at 1200V it's under 15 W.

My suggestion regarding bleeder resistors for HV is to use at least TWO resistors in parallel.
Over rate for wattage and voltage. Substantially if you can. 

The reason is that in some instances - usually after long years of service - the heat can cause failure of a resistor. In this case having 2 resistors reduces the possibility of having NO bleeder to pull the voltage down.

In the case of the Henry 2K-3 amps, the "bleeder" resistors actually act as a type of load or ballast resistor, and pull down the B+ quite a few volts all the time. They get HOT! If they open, the voltage floats up and exceeds the arc over of  quite a few components! These are >100w brown ceramic wirewounds too, not puny sandcast stuff...

Always put a stick or at least a HV meter on everything before putting hands on. I do.

                          _-_-

I scored seven 100K at 50 watt aluminum 3% resistors and will use 4 on the RF supply bleeder; two in parallel in series with two more in parallel for 100K. I have a 3KV meter that requires 8 ma for full scale reading which works out to having about 375K for the dropping resistor.  Three of the 100K plus an 80K with slider for meter adjustment will be paralleled with the bleeder  The whole system will give me roughly 85K total bleeder resistance.

More High Voltage safety suggestions:

I have a large meter (6" X 8") that is dedicated to measuring the main HV in the shack. I use this same HV power supply for all big rigs.   This HV meter is in my face and I watch it like a hawk after I turn the supply off and watch the voltage decay. It is also a great meter to watch during modulation and key up to be sure HV regulation is good.  I have four power supply settings using the primary for various voltage levels. This meter is a double check that I have everything set right.

Another important "idiot light" is a red 15 watt bulb that sits near the HV meter. It is connected to the HV transformer 240V primary to show me when AC is applied across the transformer. The point is, what if the primary was on, and the rectifiers were open due to a problem?   There could be  4KV AC on the transformer secondary with no DC voltage showing on the meter.  The red light lets me know.  Even a Jesus stick across the filter caps would not show this mistake. Of course, we should unplug the AC mains of any power supply before working on it.

In addition to shorting sticks, physical barriers and other precautions mentioned, we want the HV area to be secure enough to allow someone to walk into the room with the HV on, be drunk as a skunk, grope around anywhere he can put his hands and be safe. Same thing for a cat or dog. The room's safety must be well thought out.

T

This is a picture of my 6" X 8"  HV meter. The other meter for tube grid current has a fail safe circuit that shuts down the drive whenever a certain threshold is hit to any amplifier in the shack. Save the grids in my 8877 linear, 4-1000A AM rig, etc. - when I make a drive mistake.

Emphasis mine: Ross Hull, the well known experimenter, was killed when he did not observe Tom's important advise.  He was experimenting with a TV receiver when he groped for the CRT HV switch under the bench and connected with the HV instead.

Good advise Tom,  Al

What a pretty kitty!

http://www.qrz.com/db/w1vtp (http://www.qrz.com/db/w1vtp)    Scroll down to the first row of pics and meet Rosie.

Al

I keep the 220 locked out to the amp when not in use, and I keep the filament/HV locked out as well.

I will be adding a HV meter so that I can be sure the HV bleeds off even when locked out.

Jon

Very good on the locks, Jon.  As the kids grow up, the locks will become even more important. How many dumb and dangerous things have we both done even when warned by our parents?  :-)

Another thing to consider is what would happen if you were on the air and you were suddenly called into the other room for an emergency?  A loud bang, an intruder, a person at the door who was injured,  an accident in the kitchen - whatever. It causes you to run out of the radio room with the HV still on. Could someone walk in there and accidently stick their hand in the soup?   How about a freaked out cat that runs into the radio room for cover?

Another possibility:  What if you got across the juice and were slumped unconscious over a HV power supply? Would the XYL know enough not to touch you - and instead turn the breaker off? Does she know that there could still be HV on you or in the area until it decayed down?  How about cardiopulmonary resuscitation once she is sure the HV is bled off and it's safe to touch you? How about paramedics that rush in? Could they unwillingly become part of the electrocution gang-bang or would they just freeze not wanting to touch you?

The bottom line is that we all make mistakes every day and we need to make rules and procedures to eliminate the major ones. Big mistakes are called bad accidents.   After awhile we finally think out the more common errors until we get to the outlier. An outlier error is the weird stuff we call perfect storms that will happen once in a great while. We cannot predict when they will happen, but we can prepare for them.

If we are successful, our obit will eventually say something about dying of "natural causes."   ;)

T

All my stuff is in locked cabinets with lights and meters, and I have shorting clips.
I clip things to ground because I worry about some odd failure that might key the trans while I was back there.
I have got it a number of times with low voltage, 120vac, the high voltage scares the s%$t out of me.

When my son was young, I always took the keys with me on my key ring.
He was not interested anyway.

A small 20-30A  old style 'fuse box' painted red on the wall near the station door and with an "emergency power cut off" sign above it would be something almost anyone would see, not have to think about, and know how to turn 'off'.

You can only do so much and some people still won't have a clue.

Such an old fusebox might be too old for 'code' but OK as a switch for one or more 24V contactors that control power to the station and any workbenches.  24v is nice because it is low voltage and there are HVAC contactors in large sizes that are inexpensive. Just be sure to buy them that break both sides of the 240V. Some cheapies only break one side.

The whole system could be free:
No one wants the old type fuse boxes any more.
No HVAC technician would -ever- consider using salvaged parts on a job, so any 24V transformers and usable two-pole contactors from worn out furnaces should be free or for pennies.

What would others do or suggest for a power shut off that could be recognized and used by anyone no matter whom?

You guys are making a killer case for MOSFETs.

Heheheheh... very good point. 

As a future pansy station I can envision an Apache SDR rig driving a monster homebrew solid state linear using the latest pre-distortion software for a super clean signal.  Then add in a class E MOSFET rig to keep our fingers in the AM homebrew "efficiency" arena. 

We could still get hurt with that setup, but probably because of getting too lax.

T

My 10,020,000 ohm resistor string to measure B+.  You know what they say... If it can't kill ya, what good is it?

Jon

I think it was Neitzche who said "what doesn't kill us makes us stronger"

check Conan the Barbarian next time it is on ....

A length of glass tubing could be helpful to prevent sagging and add beauty to the installation.

Thread makes me recall the recent event that taught me respect for HV.
Built the PS for my first homebrew.
Had read warnings about working around HV, believed I had a safe solution.
Put the Fluke HV probe on the end of a 3ft fiberglass rod.
Thought that fiberglass was not conductive. Well, possibly it is not, but @ 5kv...
I grabbed the rod to move the probe...
After I moved the probe to the next test point, released the rod.
Noted fingers now numb.........did not feel anything weird while manipulating the probe.
Possibly just the 99 % humidity in the room was enough to let a bit of HV over the fiberglass.
Luck saves another foole.
 

  Back in about 1981 I joined a company that made particle accelerators. One of the older designs ran the "source" at 400,000 volts above ground. The picture attached is from one of those machines that is still in service today.

  It took a great deal of work to make 400KV. First the power supply diodes were prone to breakdown. We used to test these in a special room with a grounded metal ball tied to the rafters swinging past the supply. After a while the power supply would stop dying, so we went to the machine with it. The bullet shaped machine actually was full of electronics. A pair of 4cx250b's were used to excite a cold cathode rf source. The rf was a self excited oscillator somewhere around 2 meters. The white nylon rods all controlled a variac inside for special purposes. The green diagonal rod was the drive shaft for a 400hz 115v generator.

  There used to be a practice of suiting up a person with a rubber suit. That person would squat in the corner and hold a grounded rod. The object was to find all the corona points while in pitch dark as someone else ran up the high voltage. At say 311 Kv, if a corona spot illuminated, the person inside had to hell out to shut it off, and keep his eyes on that exact spot. Usually just a mere burr in the polished aluminum would cause a corona glow. These corona points would pull current, and subtract from the power supply output. The supply was rated at 400KV @ 1 ma...that is 400 watts power.

  One day there was a guy inside that high voltage room when a team of engineers came back from break. Hmmm, why is the door opened one asked? He closed the door, which locked the other guy inside. Then they proceeded to run up the voltage, 100kv, 200kv, 300 kv, then SNAP SNAP, 100KV, 200KV, 300KV, then again SNAP SNAP! The guy inside was the recipient of multiple lighting bolts. He was yelling his head off, but with other equipment running, nobody heard his screams. Finally he got smart and after getting nailed he grabbed hold of that aluminum bullet. With only 1ma of current available, the high voltage would not go beyond about 5KV (dialectric strength of leather sole shoes). Eventually the engineers upset at another failure decided to shut it down, and open the door. The guy inside busted out hollering, and with fists flying...He was all right.

  About a year later at the same company I got nailed by a 40KV 50ma supply. That one could have easily killed me. I think I saw god during that ordeal.

Jim
Wd5JKO

I mounted the resistors vertically and banjo-strung tight. I'm using a regular voltmeter hard wired in with the battery removed. It works pretty good! I can see the voltage drop to zero.
Although, I must admit, I had quite a bit of trepidation sticking my hand back in there again. Without that meter, I wouldn't do it. For those of you who haven't felt a good >1kv shock, they hurt like hell!

I was actually dozing off in the car the other day while the wife was in the store, and I had a daydream of the same sensation in my other arm with the same thing happening. I woke up abruptly with a quick scream. Shell shock or something.

Jon

That's an amazing account. I would not have wanted to be the first guy outside the door. Was the RF source cathode heated by the RF from the 4CX250Bs, like some CFAs are heated by the drive?

Indeed, quite a test wearing the rubber suit in there! OSHA would not like that one.
At our accelerator we are still using Cockroft-Walton DC accelerators for injectors from our ion sources into an RF linac. These monsters run at 690 kV DC, and I remember one was featured on Bill Nye the science guy back in the 1990s. Here is a photo of the ion source room for one at Fermilab near Chicago, and of ours. You would not be allowed to stay inside when it is on, as the capacitance is quite respectable just from the sizes of the components and the multiplier stack. When it sparks, its like a thunder clap.

That's an impressive setup, John. I have to admire you, you get to play with all the big toys, and the particle beams, which are made of all the little toys.