Direct wiring instead of a driver board?

[erichayes]

Hi All,

A few remarks regarding grounding . . .

Since pure star grounding is impractical and counterproductive, most grounding schemes will use a combination of star and bus grounding. You want to divide the grounds into like types, such as small signal grounds, power handling grounds (power supply, speaker, etc.), grounds within two inverting stages (this applies more to pre and line amplifiers), and star ground those groups. Then run bus wire (Jones says 16ga minimum; I've always used 10ga for power amps and 12ga for small stuff) between the stars. Make the chassis ground point as close to the input stage as possible.

The safety ground should be separate--if you use a power entry module, it will be self-grounding as long as it can bite into bare metal--and should not be connected to the bus.

[parabellum]

Thanks for the input. I have definitely decided to go with the ST70 as my first project. With this kind of support I feel much more at ease.

Now I'm gonna do more research on the best combination of tubes and transformers and then start ordering parts.

Any opinions on tubes to go with or the differences between hammond and dynaclone transformers?

[mesherm]

Don't rule out Edcor trannies.
http://www.edcorusa.com/products/transf ... index.html

They are a quality transformer at very reasonable prices.
My experiecne with tubes is that JJ KT77s, EH6CA7s, and JJ EL34s sound the best unless you have a source for some original Mullards or the like.
I would suggest also not using the original ST70 driver circuit if you can help it. If you do anyway invest in NOS 7199s, not Sovteks.
Alos don't rule out triode mode.
My personal favorite is ST70 in triode configuration. I have built two, One uses Hammond 1650Gs and KT77s, the other uses Amplimo 3A524s (toroids) and EH6CA7s. One has Shannon's driver and the other uses Roy Mottram's driver. Either one blows a stock ST70 out of its sockets. I loaned the toroid equipped one to a friend at work who I had built a very nice (but stock) ST70 for and now he won't give it the triode amp back...lol

[parabellum]

mesherm, I was going to use diytube driver schematics I got off of this board but directly wire everything instead of using the printed board. This updated driver seems to let you experiment more with the tubes and they are easier to get.

And the reason I mentioned only dynaclone and hammond is because I see schematics for both included in the manual.

Naturally since I'm building a new st70 and not rebuilding an old one i would love to build a best possible one, but I'm just concerned that with limited experience and lack of documentation I'm more liable to mess something up.

[mesherm]

One of the best ways is to try a dry run with just short pieces of wire instead of actual components. You don't have to solder them just bend them in place like a real component. Bolt two 3 or 4 terminal boards (Antique Electronics has loads of different sizes) next to each tube socket and so one or two are the correct B+ voltage. If you use the grounded terminals run a 12 to 16 gauge copper wire through them all but ground one end to the main power supply. The ground bus can be anchored at the ends using the same type of terminal strips and can run at right angles down to the power tubes. Ground only ONE bus end to the power supply, you do not want loops of any kind. Its not really as hard as it sound and if you take your time and do it neatly looks rather impressive.
Proper planning is the key.

[parabellum]

Thats a good idea. I might post a picture or something before I start soldering.

[kheper]

To me at least, it seems like a lot of
fiddling around with alot of components,
bending and soldering them together,
in a point to point fashion, where a
diytube board would only set you back
$50. I do not see the advantage of point
to point over a driver board. Is there
some sonic advantage in this? I would
hate to try and debug a mass of tangled
and soldered leads coming from resistors,
capacitors and even worse transistors, if
the amp does not work out of the
starting block.

[sorenj07]

To me at least, it seems like a lot of
fiddling around with alot of components,
bending and soldering them together,
in a point to point fashion

Trying to do this in a comprehensible and aesthetically pleasing way is a large part of the fun of the physical part of building, at least for me. Plus you can't beat the look of a badass heavy-gauge ground bus wire.

[TomMcNally]

A circuit board is a great way to take advantage of someone else's design and get it done in an hour. Handwiring is more like art you do yourself.

[sorenj07]

hey Tom, this is off topic, but do you think it's worth it to get higher quality pots like the Alps Blue Velvet? I haven't done any A-B but call it a suspicion. I also use multi-turn ETI wirewounds for bias/balance purposes.

[TomMcNally]

I tend to use the Alps (anywhere from $ 14.95 to $ 29.95 depending on where you get 'em) in higher end amps. The little Alpha seem to work pretty well though, I've been grounding the cases lately. I can't say I've heard any sonic difference, but the Alps are smoother physically, but also harder to wire because most of them have pins rather than lugs.

I've used some of the better quality pots with the locking shafts for bias, what type of 10 turn are you using ?

[TomMcNally]

Oh yeah - I have a couple of motorized 100K Alps waiting for a project.
I'm still reluctant to put a microprocessor in an amp ! Although it would
be cool to have it do this ...

1) delayed startup
2) monitor and adjust bias
3) readout parameters on a front panel display
(or Nixie tubes)
4) control these functions via RF or Infrared remote
control the motorized pot
on/off
input switching
mute or attenuate
5) running time meter
6) shut the amp off if no input for 10 minutes

My reluctance is that the amplifier will last forever, but the
CPU and remote will be hopelessly obsolete in 10 years.

[parabellum]

Trying to do this in a comprehensible and aesthetically pleasing way is a large part of the fun of the physical part of building, at least for me. Plus you can't beat the look of a badass heavy-gauge ground bus wire.

A circuit board is a great way to take advantage of someone else's design and get it done in an hour. Handwiring is more like art you do yourself.

Thats exactly right. Like I mentioned in the first post its not a matter of saving 50 bucks, I just feel that would be taking a lot of fun out of it. If I need to make one with he circuit board then I would go thought the process of designing and making one myself.
Maybe one day I'll have enough skill to design a whole amp by myself to my own personal taste. That would be the ultimate.

[parabellum]

Oh yeah - I have a couple of motorized 100K Alps waiting for a project.
I'm still reluctant to put a microprocessor in an amp ! Although it would
be cool to have it do this ...

1) delayed startup
2) monitor and adjust bias
3) readout parameters on a front panel display
(or Nixie tubes)
4) control these functions via RF or Infrared remote
control the motorized pot
on/off
input switching
mute or attenuate
5) running time meter
6) shut the amp off if no input for 10 minutes

My reluctance is that the amplifier will last forever, but the
CPU and remote will be hopelessly obsolete in 10 years.

That would be kinda cool looking with Nixie tubes, but you're right about the CPU and remote becoming obsolete.

[erichayes]

Hi All,

The debate of which construcion technique "sounds" better has been going on ever since circuit boards were invented. PCBs were developed for reliability and cheaper construction costs, both of which they succeed at admirably.

But there's one thing inherent in a circuit board--particularly multisided ones--that can deteriorate sound quality: latent capacitance. This is particularly true of the newer CAD generated boards, where traces run parallel to each other more often than with the older hand drawn layouts.

With single sided boards, the capacitance occurs between adjacent traces, and is relatively benign. Two sided boards also have capacitance between the surfaces. Shannon's 0.1" substrate minimizes the effect, but it's still there. Three sided boards, with an internal ground plane, are the worst for audio frequencies.

Another factor to consider is the material used for the substrate. Glass-epoxy boards are about all you can find now, but they actually sound worse than the old XXXP paper/resin phenolic boards used by Dynaco and others back in the '60s. I've played around with teflon substrates, but they're so hard to work with (and expensive) that I've given up.

A point to point amp that is "sloppily" built can sound superior to the identical amp built with all the wires running parallel to each other and making smart 90° turns for the same reason stated above. This doesn't mean you shouldn't use good wiring practices for your heater and B+ lines. Just allow the signal leads a little more curvature when you can.