Auto Wave soldering sockets

[dsavitsk]

I am looking for PCB mount sockets with short leads for use with high Gm tubes. Does anyone have any experience with these auto wave sockets from AES? From the picture, the leads look shorter than a standard 9-pin, but it is hard to tell.



More Info for item P-ST9-AWS-C

[kevinkr]

I've had mechanical problems with somewhat similar sockets in the past, just weren't that well made, contacts became intermittent. These could be fine I suppose.

Have you tried a regular PCB socket? You might get away with it even with a very high transconductance type as long as grid stoppers are right there at the socket.

If length is a concern I would use a dual layer board and place the trace to the grid stopper resistor on the top side of the board, shortening the pin length by 0.062" or even more if you also trim the pin(s) short.

To be honest it is something you would have to try as I have not actually tried what I am advocating, but I do use the D3A and 5842 extensively in my own designs with conventional sockets point to point and typically 0.250" of lead length to the grid stopper resistor - and I have had no problems with VHF oscillations so far.

I use your D3A eval boards with cinch pcb sockets of rather substantial height (>0.5") and have never had a problem.

Another thought would be press fit contacts mounted directly into the pcb. A small teflon disk in the center ought to provide sufficient thermal barrier to protect the pcb from tube heat.

[dsavitsk]

Originally Posted by kevinkr

I've had mechanical problems with somewhat similar sockets in the past, just weren't that well made, contacts became intermittent.

Looking at the enlarged picture, it does look like the pins just rest on there with nothing to grab them -- seems like a bad idea.

Originally Posted by kevinkr

Have you tried a regular PCB socket? You might get away with it even with a very high transconductance type as long as grid stoppers are right there at the socket.

Yeah -- this is for 6s45's, and I have found that some of them oscillate no matter what I do, and some of them are fine with just a reasonable amount of care -- I was hoping to maybe salvage some of the ones that were trouble.

Here's what I came up with -- grid and plate stoppers are on the top layer, and cathode stoppers are on the bottom. I used MJ's suggestion of SMDs to get closer to the pins. I also thought I might use teflon sockets as those might (?) be better in terms of parasitics.

I use your D3A eval boards with cinch pcb sockets of rather substantial height (>0.5") and have never had a problem.

Another thought would be press fit contacts mounted directly into the pcb. A small teflon disk in the center ought to provide sufficient thermal barrier to protect the pcb from tube heat.

Interesting idea -- I'll look into it.

[jean-paul]

Hi, from the picture I get the feeling those SMD components won't survive the stress from heating up and cooling down at every power cycle. Or am I too cautious ?

[kevinkr]

Originally Posted by jean-paul

Hi, from the picture I get the feeling those SMD components won't survive the stress from heating up and cooling down at every power cycle. Or am I too cautious ?

In my experience no problem doing what Doug proposes. I have been doing this for a number of years. Perhaps they are a bit close to the pins, I would say a mm of trace would be ok, and I would use 1206 size. As long as the solder does not get hot enough for the parts to fall off, and their dissipation is properly derated for the ambient temperature at the resistor all should be well. (They can be run quite hot without reliability issues.)