diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Tubelab (http://www.diyaudio.com/forums/tubelab/)
-   -   Tubelab SE - Sockets off PCB? (http://www.diyaudio.com/forums/tubelab/193068-tubelab-se-sockets-off-pcb.html)

dagerman 22nd July 2011 06:52 PM

Tubelab SE - Sockets off PCB?
 
Hi there,
I am about to start assembly of a Tubelab SE board and I am wondering how critical it is that the tube sockets are soldered directly onto the PCB.
The reason for the question is that I'd like to put it in a chassis such that the tube sockets are flush with the top.
Not having deep theoretical knowledge, I still suspect that the signal paths should probably stay as short as possible, but I wonder what the effects are if I would solder extensions onto the board to raise the sockets enough to clear the large electrolytics.
I figure I'd only need about an inch or so.

Any advice/experience if that would have noticeable negative impact on sound quality or other characteristics?

Thanks for your thoughts!
Stefan

nic6paul 22nd July 2011 07:03 PM

Why not mount the components on the bottom?

dagerman 22nd July 2011 07:47 PM

Quote:

Originally Posted by nic6paul (Post 2647084)
Why not mount the components on the bottom?

Thanks Nic.
Yeah, thought about that, too. That's what I did in my Simple SE.
It seems more opportunity to screw it up on the SE due to more silicon... :D
I also had a hard time getting the aluminum top plate for my chassis punched exactly right for the four tube sockets. Having a little more wiggle-room would be helpful. Or just bigger holes.....

I am probably going to do the bottom mount for the tall stuff again, I was just mainly curious to see if extending the sockets off and hence adding track/wire length will have significant undesirable side effects.

Cheers,
Stefan

dnewma04 22nd July 2011 08:53 PM

It should be fine. It's pretty commonly done from what I've seen.

Green77 22nd July 2011 08:57 PM

I once bouild a tubeamp with PCB besides the tubes. Used standard electrical cables to the tube-holder.. sound was perfect!! :)

anatech 22nd July 2011 09:07 PM

Hi Stefan,
All I can say is that sometimes this matters, and some times it doesn't.

If you can build it per the instructions or examples, you are pretty much assured it will work as it was intended. Once you start deviating, you're on your own. I wouldn't extend the connections to any tubes (except rectifiers and "magic eye" tubes) if I were you. This is especially true if you don't have much knowledge and equipment for troubleshooting.

Hi Green77,
Quote:

... sound was perfect!!
That's an awful long way from saying it works properly. It may be doing some things that you are unaware of. Possibly, if i was built properly you may hear that you had problems before. No one can be sure without doing some actual stuff with test equipment. Believe me, I've seen over 35 years off misbehaving equipment where the owner was sure everything was fine. The next comment goes something like this ... "I can't believe I didn't hear a problem for so long, but it's much better now." Life in an audio service shop.

-Chris

Green77 22nd July 2011 11:25 PM

True, but all messure was right and i didn't have any problem..and it sounded greate :)

anatech 23rd July 2011 04:48 AM

Hi Green77,
Well, that's all fine and dandy. The problem as I see it is that your advice to another person isn't based in fact at all. Certainly not in very good odds of success. His build is not familiar to you as far as the layout, so advice like that is not responsible I feel.

I'll assume you checked your wave forms with a 'scope, decent sine wave source and THD analyzer a the very least - right? That is the only way you could possibly confirm that the circuit is in fact operating properly. Just listening really doesn't cut it, although with some practice, trained listening may be able to detect a fault. This does not normally work in my experience with many, many other people.

In the high end audio servicing business, I often had people come in for some "improvement" to their equipment that they've heard or read somewhere. It became much more common place once the internet became popular. In almost all cases, the equipment had problems with it's performance even though the customers indicated that there were no other issues to be concerned about. As is customary, the equipment is tested and some measurements made in a case like this. This is so we can ensure the equipment is not suffering any performance issues, and to also make a performance reference to refer back to. My deal with people in this situation has always been, "let's get this thing running properly before we begin changing things". The idea is to have them hear the equipment working properly at home before messing with it further. In every case the customer has been shocked to find out that the equipment has not been performing well at all for years. They usually are happy enough to leave it alone, depending on the modification. There are times that I'll perform a change for improvement because it will make a positive improvement in performance. Something that follows the guidelines for improving linearity (a real modification). Since I now have performance figures before, after repairing and after changes, I can justify the work for the customer. Things like these are typically discussed beforehand.

Sorry, a technician's viewpoint that is supported by many years of direct experience with audio equipment. Listening for problems normally will only find problems that have already become serious performance problems. BTW, I listen as well as measure. Both methods must agree that changes are in fact positive ones.

-Chris

rknize 25th July 2011 06:30 PM

The 5842s like to oscillate. It's important that the grid stopper is on the tube socket instead of on the PCB if you move the socket off the board. I would do the same to the grid stopper on the output tube, but it is less critical. The output tubes on my TSE are off the board while the input tubes are standing on "socket savers" to get the clearance that I need.

I originally built the PCB with the tall components on the bottom. The heat sinks need to be rotated, which gets a little tricky because these are soldered to the PCB (unlike the SSE). When you do this, they hover over the solder pad of 1 or 2 resistors, so you have to stand-off the heat sink or file a notch into it. The filament regulator is another tricky bit. It can't be mounted to the bottom because it has 4 pins, not 3. I opted to flatten the leads and hang it off the end. This will require a bit of fabrication for a heat sink. Here are some pictures, if you are curious:

Tubelab SE – Pictures – metaruss

dagerman 25th July 2011 07:55 PM

Quote:

Originally Posted by rknize (Post 2649881)
The 5842s like to oscillate. It's important that the grid stopper is on the tube socket instead of on the PCB if you move the socket off the board. I would do the same to the grid stopper on the output tube, but it is less critical. The output tubes on my TSE are off the board while the input tubes are standing on "socket savers" to get the clearance that I need.

I originally built the PCB with the tall components on the bottom. The heat sinks need to be rotated, which gets a little tricky because these are soldered to the PCB (unlike the SSE). When you do this, they hover over the solder pad of 1 or 2 resistors, so you have to stand-off the heat sink or file a notch into it. The filament regulator is another tricky bit. It can't be mounted to the bottom because it has 4 pins, not 3. I opted to flatten the leads and hang it off the end. This will require a bit of fabrication for a heat sink. Here are some pictures, if you are curious:

Tubelab SE Pictures metaruss

Thanks Russ, very helpful!
I think I may put the silicone that will cause mechanical issues when bottom mounted on a little satellite board that will mount to a sizeable heat-sink and run short wires to the solder pad. Those connections should be less critical in terms of wire length as long as I keep them as short as possible. I'll play around with it once I get started; still waiting for parts.
I'll let y'all know how it goes!
Cheers,
Stefan


All times are GMT. The time now is 01:31 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2