The ultimate DIY? A Stax SRM-T2!

[spritzer]

Tru dat but firing up something like this for the first time is a bit scary. In a basic tube amp like the ES-1 the resistors are large enough to take some level of fuck-up before all hell breaks loose.

[Icarium]

If one does it wrong and fires it up will it be like hadron collider styles or like in lawn mower man when all the phones ring at once?

Yeah I feel like the added power and better power supply will lend some of what I feel is conveyed to me as "intensiveness." Mmm!

[spritzer]

Depends on what you do wrong...

[jgazal]

First of all, I am not an electronic engineer so I might be asking something really stupid, but I will try anyway. If the question does not deserve any answer, leave it in blank and I will kindly understand.

I have been following Kevin Gilmore dc coupled designs and this gorgeous dc coupled amplifier and its power supply counterpart are food for thought.

I comprehend that voltage and current come in alternating at 60Hz. But what does it mean to have a fully dc coupled signal path if the power supply needs to handle the 60Hz ripple with capacitors? Aren’t them adding some ripple (or helping to attenuate it) right in the signal path?

I have been reading recently that power line is going to be used to send and receive data. I presume that this modulation will occur at very high frequencies, much higher than the audio spectrum, but, still, I think this will adding be some kind of distortion to audio power supplies and circuits. This new power supply is able to cancel/attenuate that THD better than others?

Well, I presume that bigger transformers doing the current/voltage modifications, diode rectifiers converting AC in DC and those well dimensioned capacitors do help a lot. But, what if power supplies were made with "dynamos"?

I know that this brute force approach would increase costs with moving parts, as well as wasting more energy with heat dissipation, friction and the consequent physical hum. But is this much more inefficient than power supplies with rectifiers? Please imagine cost is no problem (I recognize that diodes shall be cheaper than building dynamos...).

I was thinking that building a dynamo with several phases in series would produce a low ripple dc stream and avoid the power supply capacitors. Then we could call it real fully dc coupled amplifier.

How much of the power line THD in higher frequencies would interfere in the dynamo rotation? I feel that this kind of construction would also reduce the power line THD in the audio circuit.

Last but not least. Is it possible to build a dynamo silent enough for audio purposes?

This might sound economically impracticable, but I just want to know if it is theoretically possible and better.

Best regards,

Jose Luis

[naamanf]

Why would pulsed DC be better than rectified and filtered A/C?

[jgazal]

I thought that adding a couple of 3-phases dynamos in series with slightly inclined stators (therefore slightly delayed phases) could decrease ripple, although not sure if it is feasible with high voltages and other consequences as I am not an electronic engineers.

Just imagining things I do not really know and trying to understand.

Would it be a pulse?

p.s.: picture from wikipedia

[Currawong]

I think anyone building one of these should be required to have a video camera recording at the moment of first power on. No fair blowing stuff up if the rest of us don't get to see it.

[kevin gilmore]

OK so here is the deal so far. I've ordered all the components including the

heatsinks. I'm going to wait till everything is in stock, and a couple more people

have checked the design against the schematic. If you think you are up to that,

give me a PM and i'll point you to all the pieces.

The 4 heatsinks are going to cost roughly $250 total (plus shipping) which

includes the bare heatsink material, drilling and tapping, and powder coating

black. I've purchased enough material for 20 amplifiers.

5 are spoken for already (2 x gilmore, 2 x spritzer, 1 x icarium)

if you are interested, let me know. I'm upfronting the money, so i don't

want any payments now.

The 4 angle aluminum pieces (completely machined) drilled will likely be

$20 each. (you will need 4) Its mostly labor. For stuff like this, i will

pay to use the milling machine.

Working on the prices for the back, front and tops and bottoms now.

The circuit boards are going to be $267 for the board set (one of each) if i buy 10 sets.

Less if i buy 20 or 25 sets. So i need a show of hands as to how many

want board sets. The board is layed out for either an RK50 (me of course)

or the DACT. If a khozmo ever shows up, i'll see if i can add that too.

There will be an additional small mounting bracket to secure the pot to the

amp board. (this will be a few bucks max)

All prices approximate plus shipping. I'm not going to make any money on this,

but i don't want to loose any money either.

So likely about $1100 total for all the chassis pieces plus a circuit board set.

The 3 transformers from SumR are a total of $250.

The rest of the parts including real amphenol military connectors about $1500.

The RK50 you can price for yourself. Expect to pay $1k, maybe more.

Not sure of the price on the DACT.

amp board schematic

http://gilmore.chem.northwestern.edu/t2schem.pdf

amp power supply schematic

http://gilmore.chem.northwestern.edu/t2schempower.pdf

current board layout in a few minutes.

Edited September 5, 2009 by kevin gilmore

[kevin gilmore]

current board layouts

http://gilmore.chem.northwestern.edu/staxt2connexonlyrev0-4-1.jpg

http://gilmore.chem.northwestern.edu/staxt2rev0power-6-1.jpg

I almost forgot.... 4 ounce solid .9999 fine platinum knobs will be available special order. Figure about $11K for the knob.

Solid gold knobs (24k) will also be available. Figure about $5k.

Solid silver knobs (.925) , not coated, and guaranteed to look really cool in a year or two will be about $400

Edited September 5, 2009 by kevin gilmore

[luvdunhill]

current board layouts

http://gilmore.chem.northwestern.edu/staxt2connexonlyrev0-4-1.jpg

http://gilmore.chem.northwestern.edu/staxt2rev0power-6-1.jpg

Looks like the 6DJ8 pinout got fixed

[kevin gilmore]

icarium x 2 sets of heatsinks.

[Icarium]

Titanium knob sir, titanium knob!

[Les_Garten]

Count me in for one of everything.

I'll source the Fire extinguisher on my own!

[cetoole]

Kevin, quick question. Is that meant to be an improved wilson current mirror the servos are driving?

[HeadphoneAddict]

Count me in for one of everything.

I'll source the Fire extinguisher on my own!

Somehow I thought I'd see you in here.

[kevin gilmore]

Kevin, quick question. Is that meant to be an improved wilson current mirror the servos are driving?

yup. That was back when those 2sc3381's were about 50 cents each.

unless you know the right place to buy them from, go and check the bdent

pricing. Way stupid.

Edited September 5, 2009 by kevin gilmore

[HeadphoneAddict]

Dr. Gilmore - I would like to go in for a complete kit, using DACT and plain aluminum knob.

In regards to the $1500 estimate for parts, I know that doesn't including shipping or the DACT, but does that estimate include the case, front and back?

Thanks!

(I guess I need to start selling stuff. RS-1 with APS V3 cable might be first to go. At some point I'll need to hire a builder.)

[deepak]

Dr. Gilmore - I would like to go in for a complete kit, using DACT and plain aluminum knob.

In regards to the $1500 estimate for parts, I know that doesn't including shipping or the DACT, but does that estimate include the case, front and back?

Thanks!

(I guess I need to start selling stuff. RS-1 with APS V3 cable might be first to go. At some point I'll need to hire a builder.)

So likely about $1100 total for all the chassis pieces plus a circuit board set.

The 3 transformers from SumR are a total of $250.

The rest of the parts including real amphenol military connectors about $1500.

.

[Icarium]

Right without volume control or tubes you are looking at nearly 3k in parts for parts + boards + transformers + chassis...

2850 to be exact.

Might need a few RS-1s ;p

[cetoole]

yup. That was back when those 2sc3381's were about 50 cents each.

unless you know the right place to buy them from, go and check the bdent

pricing. Way stupid.

I was actually more wondering if you meant to have the diode connection on the same side, as I was under the impression that a Wilson mirror had it on opposing sides.

[kevin gilmore]

I was actually more wondering if you meant to have the diode connection on the same side, as I was under the impression that a Wilson mirror had it on opposing sides.

That is how it is wired. Just checked the real circuit board pictures to make sure.

Figure on about $4k in parts total including an RK50. If you are creative, you

might be able to save a couple hundred in parts.

[HeadphoneAddict]

Okay, I see. I apologize for my misunderstanding. This one is too rich for my blood. I can't do it. Please count me out.

[spritzer]

...If you are creative, you

might be able to save a couple hundred in parts.

Not with me picking resistors. I wonder how much it would cost with just Riken resistors...

[kevin gilmore]

This is why i'm doing just the pieces for which there will be no argument.

Although the jury is still out on powder coating vs hard anodizing.

or having my friend that owns an audi approved (aluminum) body shop

put a few layers of paint on it. That is how wilson audio gets those

kind of finishes... Spies Hecker black is one of my all times favorites.

[luvdunhill]

I wonder how much it would cost with just Riken resistors...

depends if you understand the circuit enough to be able to rearrange it with a more restrictive resistor value set... or if not, and you decide to try to parallel two values to get the resulting value that's absent from the resistor line.