High voltage arcing in amp?

[3022]

Hi guys, I have been looking for an answer to this question for a while now to no avail. I have an old d&a pdx-400 amplifier that has some issues. First off I am only driving it with 2 watts dead carrier, modulating to 12 pep, into a beam that every other amplifier I had worked fine into. Input swr is 1.4:1, output swr on channel 1 is 1.0:1 flat, then it rises into the red past channel 15.

Now the arcing, it will occur on the final loading capacitor inside the amp, In no specific location, it will vary where it arcs at. It will arc as soon as I key the microphone, then stop unless I talk into the mic, then it will actually produce a stong arc and start "sizzling". Almost like a welding arc. It only does it on the high power setting, not the low. The capacitor is clean and no plates are touching.

I found some good info here on this amp from a few users here but not on this topic.. I searched high and low for the answer to this with no solution. Any help would be majorly appreciated

 

[unit_399]

3022 -

How does the unit sound on-the-air on low power? My first thought would be to check the contacts in the antenna relay. If all of the power in the high setting is not getting through the relay and into the load (antenna), it has to dissipate somewhere, and the loading cap would be a good place for that to happen. The antenna connector is also a probable suspect.

If the relay or the connector isn't the problem, I would check everything from the antenna connector back to the loading cap.

- 399

 

[Captain Kilowatt]

3022 -

How does the unit sound on-the-air on low power? My first thought would be to check the contacts in the antenna relay. If all of the power in the high setting is not getting through the relay and into the load (antenna), it has to dissipate somewhere, and the loading cap would be a good place for that to happen. The antenna connector is also a probable suspect.

If the relay or the connector isn't the problem, I would check everything from the antenna connector back to the loading cap.

- 399

Things don't really work that way. If the relay was the weak point then the contacts would arc and burn and usually nothing would happen to the load capacitor.

One thing about an arc, if it happens once it WILL happen again unless everything is cleaned and polished. I know the OP said the load cap was clean and well poloshed but every time it arcs it will leave a small pitted spot that will arc again, and again and again unless it is repaired each time it arcs. There should not be that much voltage on the load cap to make it arc unless there is a piece of dirt stuck between the plates or the plate blocker capacitor is shorted allowing full plate voltage to appear across the tune and load capacitors. Since the tune cap has a larger spacing the load cap will arc first. Check the plate blocker cap and make sure the load cap is clean and no rough edges or traces of carbon left on it.

 

[3022]

Ok heres what I found, plate is clean, not even any carbon deposits on it from where it was arcing??? Input and output so-239 connectors are good with tight fitting connections. My coax is only a few months old and is rg-213 to the antenna, and rg8 to the meter from the amp. Coax all checks out fine.

Now I just read captain kilowatts post, and I flipped the amp over without the case and watched underneath as I keyed, and the middle relay does infact make a greenish/blue spark when I key the radio... What exactly does this mean?

 

[unit_399]

Cap'n K.-

I agree with what you say. I just had trouble understanding why it worked in the low power mode, and why the cap never arced in the same place. Since the antenna switching relay in most of the D&A amps was a DPDT with the 2 contact sets paralleled, I thought if one contact set was bad that this could cause the problem. It was just a shot in the dark.
In any case, I would replace the final load cap, 'cause it can't be in too good a shape after that much arcing.

- 399

 

[3022]

Thanks for the responses, heres what I found... The pdx-400, atleast this one uses 2, 3 pole relays.. On the antenna relay I tried advancing the tank circuit pole a few thousandths, so the antenna would connect before the input circuit, didnt work, still shoots sparks.. Now the 2nd relay is what controls the high and low I think? On the low setting sparks are nowhere to be seen in this amp. on the high setting the middle pole on the 2nd relay will spark along with the cap plates. Maybe this will provide some insight.

 

[dudmuck]

Here is a good article on amplifier arcing:
http://www.eham.net/articles/21381

Some mention there of capacitor plate spacing vs. voltage limit.
And among other things, the various causes of arcing.

 

[3022]

The capacitor is fine. Believe me I read ever single article on arcing I could find, but thanks for that link, it was an interesting read. The capacitor is fine, I looked it over and none of the fins are touching, I tested it with a multi meter, no continuity between plates.. Cleaned it up and polished it to a mirror finish with 3000 grit sand paper, still arcs.

Could this be due to the tank coil....? I bought this and the tank coil looks out of whack, 6 turns, spaced far apart for less inductance I would think?, and soldered or tapped at the last turn to the 90 degree bend. Really want to get this thing on the air, Its just killing me to see it arc and self destruct with ALL brand new 6lq6 tubes and almost ever capacitor and component replaced.

 

[unit248]

I Googled this amplifier and found what I think are some pictures of it from a couple of sources. There was one thing I wanted to suggest, but unfortunately none of the pictures I found helped me figure out of my suggestion would help or not. Anyway, here goes:

In general, arcing means the voltage between the plates of the variable capacitor is so high that it overcomes the resistance of the air between them and current starts to flow. Exactly how much voltage it takes depends on how much space there is between the plates and the level of humidity (moisture in the air will lower the breakdown voltage).

But my question is, is it RF voltage or DC voltage? Ideally, there should be a DC blocking capacitor between the plates of the tubes and the input to the tuning network. This prevents the high voltage DC plate supply for the tubes from being applied to the tuning network. (The capacitor will let the RF pass through, but block DC.) In some of the pictures I found, it looks like there's a couple of ceramic caps connecting the tubes to one of the tuning caps, but in another image it looked like they weren't there.

Just to check, I would get a decent voltmeter and measure the DC voltage across the cap that's arcing. Use a couple of alligator clips to position the voltmeter probes while the amp is powered off -- you don't want to stick your hands in there while it's on. Most digital voltmeters should be able to measure up to 1000 volts. Set it at its highest range, and then key your transmitter up with the amp on and see what it reads. Observe all possible precautions: high voltages are present which can kill you.

I don't think there should be any DC voltage present across the variable cap at all. If there is, then maybe someone removed the DC blocking caps, or else they've gotten old and become shorted over time. Either way, you should probably fix the issue before continuing to use the amp.

If there's no DC voltage present, then the arcing is due to high RF voltages. From what I can tell, this unit was originally designed for 15 meters, and people have modified it for 10/11. It's possible that the modification hasn't been done right, resulting in RF voltages across the tuning caps that exceeds their rating.

-Bill

 

[Captain Kilowatt]

I Googled this amplifier and found what I think are some pictures of it from a couple of sources. There was one thing I wanted to suggest, but unfortunately none of the pictures I found helped me figure out of my suggestion would help or not. Anyway, here goes:

In general, arcing means the voltage between the plates of the variable capacitor is so high that it overcomes the resistance of the air between them and current starts to flow. Exactly how much voltage it takes depends on how much space there is between the plates and the level of humidity (moisture in the air will lower the breakdown voltage).

But my question is, is it RF voltage or DC voltage? Ideally, there should be a DC blocking capacitor between the plates of the tubes and the input to the tuning network. This prevents the high voltage DC plate supply for the tubes from being applied to the tuning network. (The capacitor will let the RF pass through, but block DC.) In some of the pictures I found, it looks like there's a couple of ceramic caps connecting the tubes to one of the tuning caps, but in another image it looked like they weren't there.

Just to check, I would get a decent voltmeter and measure the DC voltage across the cap that's arcing. Use a couple of alligator clips to position the voltmeter probes while the amp is powered off -- you don't want to stick your hands in there while it's on. Most digital voltmeters should be able to measure up to 1000 volts. Set it at its highest range, and then key your transmitter up with the amp on and see what it reads. Observe all possible precautions: high voltages are present which can kill you.

I don't think there should be any DC voltage present across the variable cap at all. If there is, then maybe someone removed the DC blocking caps, or else they've gotten old and become shorted over time. Either way, you should probably fix the issue before continuing to use the amp.

If there's no DC voltage present, then the arcing is due to high RF voltages. From what I can tell, this unit was originally designed for 15 meters, and people have modified it for 10/11. It's possible that the modification hasn't been done right, resulting in RF voltages across the tuning caps that exceeds their rating.

-Bill

NO! Don't do that! Most digital meters are good for 600 volts, some are good to 1000 volts but you have to watch the insulation on the meter leads and clip leads. (story about that later)* Also you do NOT want to key the transmitter/amp and have your test meter connected. The RF will do some nasty things to that meter. It is NOT designed for high power RF. You will most likely fry the meter.If you can key the amp without any RF drive from the transmitter then it should be OK. Just make sure there is NO RF in the tank circuit. Hmmmmmm......typing that make me think of parasitics in the amp.Does the power meter show ANY power out at all from the amp with no input signal and the amp keyed? Something else to check on.

*- A year or so ago I was testing some old plate transformers at the kitchen table using a variac.Now bear in mind I am no stranger to high voltage having worked on gear that had up to 5000 volts on the tubes. In fact 2500 was "low" when it came to high voltage in the broadcast gear I maintained. I was using my trusty Triplet meter and was checking a transformer rated at 6400 volts center tapped. All was good until I reached for the variac knob to turn things down. The variac case was grounded. Everything else was insulated via wooden table, chairs, and floor. I had grabbed the wrong meter leads. I was in a hurry and instead of my 10,000 volt leads I grabbed the Radio Shack leads rated for 600 volts. Two words should now come to mind: leakage currents. Next thing I knew I was on the floor in copious amounts of pain with my wife ready to start CRP. Nope, I'll never make that mistake again. I'm not sure what was worse on me, the shock or the wife.

 

[Toll_Free]

This can be caused by not being tuned correctly.

The PDX-400 was / is a 15 meter TRANSMITTER for CW. It was modified for 10 meter use. It was D & A's answer to the 10/11/6 meter rules that came about in the late 70s / early 80s.

What's surprising is: They didn't get away with the SAME crap that Texas Star is NOW getting away with. Just goes to show you how lenient as a society the US of A has become: What at one time was a blatant violation of rules, and was blatantly done in regards to ONLY being able to get around regulations and resulted in fines..... 14 years later the FCC HELPS another company figure out the EXACT same loophole, and helps them license their products as such.

ANYWHO!!!!

I'd be willing to bet that since that amplifier is > 25 years old, it's worked in it's lifetime. Chances are, you will have a carbon track across the capacitor... They ARE close spaced, and you aren't going to see it unless you KNOW what symptoms to look for.

ALSO, you will find that the amplifier switches in different tank circuits for drivers, etc. depending on power level selected. One of the smarter sweep tube mfg's, they didn't just reduce the drive and throw Q sky high to reduce Pout for the idiots that need "power levels". Buy an amplifier the correct size in the first place, and stop worrying about "lemme flip this switch" which is nothing more than an IMD (SWR) increasing swamping resistor in transistor amps, and a Q increasing element in tubed stuff. MAKE SURE you're looking at the correct cap.

ALSO, you will find that if it's the larger spacing capacitor, you can find resonance at TWO places. Sometimes you will get lucky and be able to resonate your amplifier at a point on the cap on the opposite side, and you will lose the arcing.

Hair Spray can be your friend. I've taken a few breadslicers apart, filed and buffed the carbon tracks, and then LIGHTLY hairsprayed over the area..... Let it all dry, and then reassemble. NOT for the faint of heart, some slicers are BUTTHOLES to assemble!

Also, for your SB220 owners. The TUNE cap can arc. Fix is simple, for 10 meters only operators... Take half the cap out of the equation.... Double the spacing by reducing the plates by half. No more expensive Ctune replacements, and the voltage rating goes WAY up.


--Toll_Free

 

[unit_399]

3022 -

I agree with toll-free that the problem is probably the load cap. D&A varies the plate voltage to get different power levels; 450 volts in low power and 900 volts in high. Since you've got a big investment in new tubes, I think you need to bite the bullet and buy a new load cap. RF parts stocks them. Also what are you using to drive the amp; what radio and how much dead key? D&A lists the maximum deadkey into the amp at 4 watts.

The picture I have of the amp shows the final tank coil with 6 turns spaced about 3/16" apart. The driver tank coil has about 20 turns (it was hard to tell) spaced at about 1/8". SECRET CB Volume 6 page 53 has the conversion from 15 meter transmitter to linear amp. The SECRET CB books are on CB Tricks website. Good Luck and 73s.

- 399

 

[Crusher]

I agree with toll free and 399. I think you stated on lower setting works, but when you turn it on high it arcs. This would lead me to believe the loading cap is not sufficient. What is the plate spacing on the loading cap fully meshed? Also as stated, with sweep tubes, don't need a big carrier from radio and I would,kt use an export to drive it. 3-4 watts carrier max as stated.

 

[Captain Kilowatt]

FWIW even a legal limit ham amplifier has a loading capacitor that uses what is called receiver spacing. IOW the spacing of the plates is tyhe same as thta used in a receiver. The loading cap needs only to handle 220 volts of RF at 1000 watts. I don't think the plate spacing is the problem. Below is a picture of a Kenwood TL-922 amplifier with a pair of 3-500Z's for legal limit. The cap on top is the tuning cap and the cap on the bottom is the loading cap. Note how small the spacing is yet it handles 1500+ watts with ease. I'm not saying the cap is NOT the problem but rather that the spacing of the plates is not the issue. That is unless the capacitor was damaged in some way. I am still suspicious of a sharp edge somewhere. Even something as small as file marks can be enough to cause arcing under the right (or should I say wrong) tuning and loading conditions.

 

[Crusher]

That is true KW, but seems odd that on low it was fine and hi setting it is not. But then this is RF and all sorts of strange things can happen. What about wire from loading cap to relay. Is this a good solid connection. I have seen a transmitter with a bad connection at cap and the rf is trying to find a path. Just a thought.