Eimac air socket for 4-1000A

[KL7OF]

I have built a 4x1 GG amp and the chassis was not deep enough to mount the socket under the chassis so I put it on top..  I'm having some parasitic problems and wonder if having the socket on top makes any difference...I even tapped an additional hole in the nose of the socket to bond that end to the chassis.  I remember reading some posts about the metal socket being prone to gremlins..The guy was adamant that the sockets were no good...I have 2 other HB 4x1 amps with the metal air socket that work fine ..those sockets are under the chassis.. Did I make a mistake ?

[KL7OF]

pictures

[Opcom]

The only gripe I have ever heard on those sockets is that due to the smaller inlet and the 90 degree air path, a higher pressure blower may be needed to force air in at the specified rate.

If they are ticklish, it would probably not be in a GG situation, but in a grid driven situation. I have no socket-caused issues, and use the metal one like that one in a class C plate modulated amp.

Have you tried installing the four little spring clips that ground the tube's metal shell base to the socket? It serves a shielding function.

Have you been sure to remove the paint all under the washers that hold the socket?

Really it should be under the chassis so those big screws on top clamp it like a gasket. Put a box over the underside of that area to enclose the socket and pressure air. But that may not match your rack plan.

Who is the guy that complained about the sockets? Adamant is one thing, stating the reasons for it is another thing. It would be interesting to know more.

[K1JJ]

Nice construction job, Steve!

It all depends how much air comes thru.

What kind of parasitic problems are you having and what freqs?  Try grounding the metal shroud at the bottom as Pat suggested.    Shielding.

I wonder if you recessed the mounting screws on top - if a chimney would fit.  Might as well make the air work for you by hugging the glass envelope on its way to the plate cap.

T

[KL7OF]

Nice construction job, Steve!

It all depends how much air comes thru.

What kind of parasitic problems are you having and what freqs?     I had it going pretty good except for 10 meters .  I put in a KW roller coil and that tuned OK... until it started to spit and arc  The hot end of the coil arcs on any band at any drive level.. The tube socket is mounted to the chassis with 5  10-32 machine screws..bonded very well ...the grid pins are strapped to the chassis with a copper sheet.  I have a piece that mounts to the top of the tube socket that holds the chimney...

[KD6VXI]

 The twin 4-1000 at 8kv I built had a socket problem.  Same socket would oscillate and fry Psupp.


Tried different tubes, same thing.  This was plasdick sockets.

I ended up finding that I could do new grid connections that about halved total length of the connection to ground.  That solved the problem.

How long, total, is the length of ground connection from each grid pin (screen and control).

Also, the shielded area at the base of the tube has to be grounded.  Without doing so is an invitation to disaster.

--Shane

[KL7OF]

The twin 4-1000 at 8kv I built had a socket problem.  Same socket would oscillate and fry Psupp.


Tried different tubes, same thing.  This was plasdick sockets.

I ended up finding that I could do new grid connections that about halved total length of the connection to ground.  That solved the problem.

How long, total, is the length of ground connection from each grid pin (screen and control).

Also, the shielded area at the base of the tube has to be grounded.  Without doing so is an invitation to disaster.

--Shane

the shielded area has to be grounded....I haven't done that.  The pic is of a fit up...All the paint has been removed from the screw and copper contact areas.

[KD6VXI]

Well, I'd say that's about as short as you can get! 😊

The plastic ones have a tab that extends through the bottom..

I ended up running a tan of copper from the top of the socket out the side of the socket. 

Don't see any way you could make those shorter. Just ensure the socket itself is well grounded to the chassis and try some homemade spring clips to the metal maybe....?

--Shane
KD6VXI

[Opcom]

These little clips are often missing when sockets show up.

[K1JJ]

Those clips are slick cuz they hold the tube in place, ground the shield AND hold the chimney tight and centered.  You could flip the amp upside down and everything stays put.  Well, maybe the chimney might slip. Nice.

T

[KL7OF]

I made clips from some springy stock this morning and installed them......I use the lid off a gallon paint can as the base for the chimney...Eimac glass fits right in'    I had this amp going with a bandswitch and a tapped coil.   It would arc when it was way out of tune...and didn't like 10 or 15 meters..As I was working to tame it on 10 and 15, it started arcing the bandswitch...So I took out the 10 coil and strapping..  after that it started arcing the switch on all bands not always but more  often...I was trying different values in the parasitic choke on the plate  and did some bypassing at the cathode...No Joy..Then I got a KW roller coil and put that in...worked first time and really well for a while....Next day arcing started on the hot end of the coil..Low drive low plate voltage doesn't change it..still arcs when you key on all bands..my neon shows that most of the voltage is concentrated in the one end of the coil......is that normal ?  

[Detroit47]

I would totally strip the paint on the bottom where the copper contacts the chassis. Then tin it and sweat solder the copper to the chassis as well as bolt it. It is a little work that will pay dividends.
There is one other thing that will help calm it down. The lead coming off the parasitic suppressor should be flat strap like you have on the plate block. Get rid of the thin wire and run flat strap right up to the parasitic suppressor. I have had to make this change in some commercial amps trying to run Russian 572b's. It seems to calm things down.


John N8QPC

[KL7OF]

roller coil setup 

[K8DI]

What are you driving this with? I’ve been doing some reading on the W8JI pages about amps and he talks about ALC implementations that result in high initial drive that can start arcs regardless of whether you’ve turned the power down.  Something to check...

Ed

[KL7OF]

I decided to put the socket under the chassis...I had another socket and I cut it down to nothing more than the ring to bolt to the chassis and 4 posts to hold the insulator with sockets...I drilled and tapped 4 additional holes in the ring for complete bonding to the chassis. With the socket cut away like this, The airflow really increased, It lowered the tube thus ,shortening some straps, and I found some original clips for the shield grounding/chimney.
   All good stuff but it didn't stop the arc.  I'm driving the amp with a FT 101E ..I have been reading W8JI as well and tried his method for finding the proper inductance for Plate parasitic suppresors..to no avail. 

[Detroit47]

Try something like these they work good on 10 meters. Resistors are 56 ohm 10 watt corning glass with 5kv on plate. Steady as a rock. I have a bunch of the resistors if you want some. I killed the amp for parts. What are the two white wires on the filament choke for? I would love to see a picture of the whole under chassis if possible.

John

[Opcom]

The trouble may be the plate hookup?  One thing I see looks like long straps from the tube to the vac cap then to the pi coil. Also, looks like a "U" shape in one strap after the plate tuning cap? Please do get a pic of the whole thing from the top. I ask because my transmitter had all kinds of annoying instabilities and it was due to the long straps/leads between the tube plate, the blocking caps, and the pi coil, and tuning cap. I was told that these things can make up their own tuned circuits and make ther amp go nuts. At times, large arcs would dance across a few turns of the roller coil, and this sounds similar to what is bothering your amp. Also, is the blocking cap a RF type? Part of my trouble was using a bunch of TV set type doorknobs in parallel.

[w8khk]

Please disregard.  Should not be posting before morning coffee!

[w8khk]

I decided to put the socket under the chassis...I had another socket and I cut it down to nothing more than the ring to bolt to the chassis and 4 posts to hold the insulator with sockets..

Very nice milling job on that aluminum socket!  I have several of those for the 4-400 and the 4-1000.  That technique will make them useful!

[KL7OF]

I decided to put the socket under the chassis...I had another socket and I cut it down to nothing more than the ring to bolt to the chassis and 4 posts to hold the insulator with sockets..

Very nice milling job on that aluminum socket!  I have several of those for the 4-400 and the 4-1000.  That technique will make them useful!

Thanks...I used a side grinder and a cutoff wheel...Quicker than setting up the mill....The airflow is greatly increased ..

[KL7OF]

The trouble may be the plate hookup?  One thing I see looks like long straps from the tube to the vac cap then to the pi coil. Also, looks like a "U" shape in one strap after the plate tuning cap? Please do get a pic of the whole thing from the top. I ask because my transmitter had all kinds of annoying instabilities and it was due to the long straps/leads between the tube plate, the blocking caps, and the pi coil, and tuning cap. I was told that these things can make up their own tuned circuits and make ther amp go nuts. At times, large arcs would dance across a few turns of the roller coil, and this sounds similar to what is bothering your amp. Also, is the blocking cap a RF type? Part of my trouble was using a bunch of TV set type doorknobs in parallel.

I agree and I have shortened all leads as much as I can...several different configs, different blocking caps.  even replaced the load variable vac cap with a breadslicer as It looks like  it may have lost it's vacuum.   The strap loop going to the coil is gone,  (you can't see it in the picture)  I have tried 4 different (values) plate chokes.
  I have made many different parasitic suppresors .  3 different tubes, even tried different drivers..I am running it on an 8KV  Variac controlled power supply.  I haven't gone higher than 6kv...   What is strange is that this amp worked when I first put it together .....75,40, 20, were full output.15 not so much and 10 wouldn't go...  Whilst trying to get 10 and 15 going, It started to act up on all bands...Now  the amp starts to arc at the coil with 100 watts out on all bands..

[KL7OF]

Try something like these they work good on 10 meters. Resistors are 56 ohm 10 watt corning glass with 5kv on plate. Steady as a rock. I have a bunch of the resistors if you want some. I killed the amp for parts. What are the two white wires on the filament choke for? I would love to see a picture of the whole under chassis if possible.

John

Chasing parasitics, I loaded the filament choke with a resistor across each coil..didn't help... that's the white wires..I'm removing them..
the picture of the underside was the earliest configuration.  before modification started.  I'll have to take another of the present underside.