[scootsit]

Here's the inside.

 

[Breck]

Out of curiosity, what did you use to affix your boards to the chassis?  Is that some sort of epoxy or putty?  If it works well that would be more convenient than drilling standoff holes.  (specifically I'm looking at the vertical board on the top-right of the chassis in the picture)

 

[scootsit]

The main bit of board is bolted using some metal spacers from ACE. That vertical board is the E12 muting delay. I didn't want there to be any visible screw-holes in the side of the case, so I used an epoxy tamale. I used board-mount tube sockets, and held them in with epoxy, so I figured I'd give it a try. It actually worked pretty well. It dries pretty solid and is rigid as can be. I think I used the Loctite brand one, but I'm sure that any of those epoxy tamales would work about the same. It really dries almost like ceramic, and I was thinking it may have an application for deadening hollow space, too.

 

[scootsit]

 
 
Quote:

The To220 version is slightly more expensive, but allows for heatsinking which is totally worth it IMO.

 
 
Quote:

Told you nikongod liked those TO-220 BUF634's with heat sinks.

 
FYI, it's not really the same as putting a DIP-8 heat sink on top of a chip.

I just found a source for some TO-220 Buf634s, so I'm going to replace the OPA551s. How big of a heatsink do I really need? I noticed that the V- is connected to the tab, given that the case is already grounded, I was debating mounting the Buf634s to the side of the case, but that may be ugly. I only have room for fairly small heatsinks. Radio shack sells tiny little extruded ones, do you think they would be enough?
http://www.radioshack.com/product/index.jsp?productId=2102857&filterName=Brand&filterValue=RadioShack
 
They have a bigger one with fins that is stamped, but I just don't think there will be enough room.

 

[tomb]

In your case (pardon the pun), the case would be ideal, IMHO.  You just need to be sure to use a good heat sink kit with an isolating pad and a shoulder washer.

 

[scootsit]

Quote:

In your case (pardon the pun), the case would be ideal, IMHO.  You just need to be sure to use a good heat sink kit with an isolating pad and a shoulder washer.

I would need an isolating pad? The tab is on the V- pin of the Buf634, and the amp is running with a single rail, wouldn't the V- be the same as ground in that set up? Since my case is grounded, shouldn't that be the same?
 
Thanks!
 
I'm going to try to figure out a way to connect it to the case, I'm not sure it will be possible. It's built on Radio Shack strip board, I find that 90% of the time if I try to reflow stuff on that, the traces come up, so I was thinking of just bending the pins and basically plugging them into the sockets the OPA551s are in. I suppose I could use wire soldered to the leads, going into the socket, but the socket is the kind with the narrow slots, so I'm not sure wire will fit. Any other suggestions? Else, would that heatsink work at all?
 
Ok, last question, is there any performance benefit of throwing a rail splitter in there? I mean, with the buf634's current demands, I'm think a rail splitter could limit the draw. That said, aside from eliminating the output capacitor, is there any other benefit to using a dual rail set-up? EDIT: This would be a nightmare to do, so I'm going to withdraw that question...having the buffer run on two rails and the tube on one would be very challenging, I think.

 

[tomb]

Yes, sounds like you are correct.  I didn't read closely enough and wasn't used to seeing "V-" instead of saying negative ground.  My fault.
 
I don't think a dual rail supply is going to eliminate the output capacitor.  The output capacitor is not there for offset out of the buffer.  It's there because the tubes are offset to 13.5V. There are some designs, I think, where people have tried a dual rail supply with tubes, but my understanding is that things can get squirrelly very quickly.  You can't depend on tubes - especially in the Millett Hybrid - to hold anything more than about a 1/2 volt tolerance on bias.  Even then, that's assuming perfect temperature and they typically vary by as much as 1-2 volts over 2-3 hours as they heat up.  That's way too much offset and allows nothing for safety (a failed tube).  The buffer will pass all of that to the load.  You'd still need the output caps.

 

[scootsit]

Got it, thanks for the info. I think I'm going to give it at try with the heatsinks I linked to above. I have been thinking about it lots, and I cannot for the life of me figure out a way to connect the buffers to the case and get them in the appropriate socket.
 
Tomb, I really appreciate the help.

 

[scootsit]

I just found some 22gauge solid core wire in my closet. I tested it on a socket I had around the house, and it fits like a glove. New plan, mount the Buf634s directly to the case. I'll post pictures.
 
Thanks again, Tom, I really would have a bunch of non-function boards populated with the wrong components and poorly soldered were it not for your help.

 

[scootsit]

I don't think I could mount in the back, so there'd be visible screw holes, which I'm trying to avoid.
 
I'm going to try this stuff: http://www.xoxide.com/phobya-2component-thermalglue.html
 
It seems to be about as conductive as regular silicone grease from radio shack, there's one by Arctic Silver (way more expensive), too. I'm thinking that I will sand the case down a little, and glue the chips to the case. I know that without the screw, it won't be quite as effective, but I'm thinking this is the best option, far better than those tiny heat-sinks.

 

[scootsit]

I've got the buffers in and working and they sound pretty good. There is very little heat, it all seems to be dissipating well. Only problem, when I superglued the sockets back together after removing the center pin, I got a drop or two in one of the pin holes. The tubes have always required a lot of wiggling to get good contact, as a result all of the pins are falling off of the sockets. So, I just ordered some from Tomb, hopefully by this time next week, it will actually be functioning well!

 

[scootsit]

So,
I replaced the bad socket yesterday, resoldered everything, and have a new problem. For whatever reason, there is a popping sound, it sounds almost like straight DC because it's actually causing the driver on my test-phones to reach its full extension. I have tried swapping tubes, and it's still there. It goes away as I turn up the volume, which makes me think it's somehow being introduced at or near to the volume knob. Tomorrow, I'm taking it all apart to search for the source. Before I dive in, anyone have any suggestions of places to look?

 

[scootsit]

Okay, I have some more info. As you can see, the tubes/sockets are mounted onto a top. The popping goes away, more or less when the top is not on the main unit. Also, disconnecting the case ground wire seems to reduce the popping as well. I have a feeling there is something wrong with the ground. Given that the traces lift constantly on radio shack boards, re-checking all of the traces seems like a logical place to start. I did have to take the boards out to put the buf634s in.
 
I'm also sort of thinking that the epoxy I use to hold the tube sockets in may be slightly capacitive, if it were bridging two contacts, that may cause a popping sound, but I'm not sure about that.

 

[scootsit]

It seems to stop when I reconnect the pot to the case. I had it out so that I could test it. I ahve since bolted it back in to the case. It's very weird. At one point, I was measuring 0.6VAC between the case and ground, which made no sense, unless RF can really cause .6V, that was with the case ungrounded. It's really inconsistent, sometimes it happens when I touch the case, other times, when I'm not. It happens with the case grounded, then I'll disconnect the grounding wire and it will stop. Then, any time I think I've got it figured out, it will do it with the opposite, so when it will start with the case ungrounded, too. Touching the tubes sometimes makes it stop. I cleaned the contacts, thinking that they might be dirty, and that seems to have helped marginally. Every resistance and connection inside is as it should be. I resoldered all of the tube socket pins, just in case, and that seemed to help marginally. I'm really at a bit of a loss. If connecting the volume pot to the case really gets rid of it, then I have 0 ideas what caused it.
 
With the volume all the way down, there is also a buzzing. I ahve tried switching tubes, and it always stays in the right channel.

 

[scootsit]

I don't believe it. I disconnected the bandwidth pin on the Buf634, and all of a sudden it stopped. It only was a problem in the right channel. I'm not sure what to do now. What would cause that? I assume it was in oscillation. Why would only one channel oscillate and not the other? What did I do wrong? I disconnected both BW pins, and am noticing that Ti and a few others built theirs with resistors at Pin 1 on the BUF634.