The ultimate DIY? A Stax SRM-T2!

[Guest sawyers]

updated schematic for the power supply board with all known corrections.

(i think) (its all really minor stuff, missing .1uf caps, wrong resistors for 555 timer)

(labeled diodes, missing 2 diodes)

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

if someone finds more errors in the schematic let me know.

At least the board is known correct.

Hi Kevin

Very nice looking - and good to know that it struts its stuff!

Inconsistencies between board silk screen, which agrees with your stuffing, and the schematic are:

1. Add 300k discharge resistors across C7 and C14

2. Change the values of R8, R9, R23, R35, R36, R48 from 300k to 400k

3. Change the values of R18, R19, R43, R44 from 80 ohms to 40.2 ohms

Inconsistencies in the PSU BOM:

1. 10k 8 needed (not 4)

2. 100k 3 needed (not 4)

3. 1N4007 12 needed. In the BOM there are two lines for this, one for 12 off and a further one for 3 off. Delete the line for 3 off.

Cheers

Craig

[Guest sawyers]

Hi all

I've also gone through the amplifier BOM, and found a couple of minor inconsistencies as compared with the actual board.

1. 300k, 2 needed (not 6)

2. 390k, 8 needed (not 12)

Oh - and there are two R44's, one in the 700V battery (390k) and one and one between the supressor grid of the left hand 6CA7 and anode (220 ohms).

Basically I cross checked the boards (took a photo so I could cross parts out as I checked them), cross referenced it to the schematic parts numbers, then listed out the parts numbers in sequence to ensure I had a sequential list with nothing duplicated or obviously missing. So I *think* it is consistent. It definitely, *definitely* needs cross checking with similar things others on the forum are doing.

I've also broken the BOM into separate sheet for passives, semiconductors and hardware (I haven't checked the hardware) and provided a column for schematic parts cross reference.

BOM is here http://www.tech-enterprise.com/tekstuff/t2parts3cgsedit.xls

Craig

[kevin gilmore]

updated

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

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

(6ca7 resistor now R93) (have no idea how that happened)

generated from the layout software

http://gilmore.chem.northwestern.edu/t2bom.txt

(realize that you have to multiply most stuff by 2 and battery parts x 4)

[kevin gilmore]

Damn Kevin! That needs a glass top, because it's as beautiful inside as out!

I'll get a quote on some top and bottom panels machined out of pyrex and/or synthetic saphire.

You probably won't like the price, but maybe i'll make one set just for fun.

Top for the power supply, and bottom for the amplifer.

[Pars]

I really wanted to use those spacers, because i have seen them used a bunch of times, but

i could never find the part number for those things.

Please post the part number for the spacers. I think its a good idea that everyone use them.

I was out with my wife yesterday and she was looking at beads when I ran across some beads that might work for this called bugle beads. They are available in both glass (czech glass) and in what they refer to as base metal (whatever that is), something like these:

Beadfx - Liquid Metal and Tube Beads

Apparently you can get different lengths and diameters.

[kevin gilmore]

these are pictures of the pins i want.

http://gilmore.chem.northwestern.edu/tagboards.jpg

[Guest sawyers]

these are pictures of the pins i want.

http://gilmore.chem.northwestern.edu/tagboards.jpg

What you mean is these?

http://escol.com.my/Miscellaneous%20parts.html

About quarter way down the page.

Craig

[cetoole]

What is the proposed benefit of the big input CCSs (parallel pairs of 10M90 DMOS devices)?

[kevin gilmore]

1) Noise reduction. Less than 1mv of noise on any of the high voltage lines.

2) output caps charge much slower, no massive input current peak to toast the pass fet.

3) limits output current in case of a tube short. Fuses unnecessary.

For the 2 by 500 volt supplies, the one CCS could have been driven off of the

center tap of the transformer, which currently is optional. Save about 2 watts

that way. Maybe for round 2.

[cetoole]

How much noise reduction do you get before the pass transistor? I would think the reg could get that kind of ripple reduction alone.

Soft start and current limiting is definitely nice, some of the major benefits of shuntregs. Gotta love a supply which will not burn if shorted.

So where does the extra current overhead go?

[kevin gilmore]

So where does the extra current overhead go?

Very little current overhead when running correctly.

Otherwise it goes into heat. Where else would it go.

Unlike some other power supplies out there, this can be

tested with absolutely NO load. Easier for people to

put together.

There is a fully shunt-regulated version of this that

has been on the drawing board for a while. Justin was

going to make an upgraded supply for the BHSE, i do not

know if this is still in the plans. Not cheap. Cooks eggs

for breakfast.

The stability and ultra low noise is the key factor here.

To do a dual tracking version (in this case, quad tracking version)

gets even more expensive due to the 1kv isolation opamps necessary

to get the control signals to the bottom rail.

I could have saved one ref102 by using one for both the +250 and +500.

This is not about saving a few dollars.

Edited March 8, 2010 by kevin gilmore

[guzziguy]

Very little current overhead when running correctly.

Otherwise it goes into heat. Where else would it go.

Into magic MilSpec wire? Isn't that where the extra bass comes from?

[cetoole]

Of course it goes into heat, but where? There is no shunt element, so what is dissipating the extra current that the load isnt using, especially during no load scenarios?

Maybe odd idea, but for dual tracking, how about instead using a differential i/o opamp for the error amplifier, running on something like +/-15v instead of +30v like the current error amp? One would, of course, need to invert the negative supply parts.

[deepak]

Any estimates for the power consumption?

[Guest sawyers]

Of course it goes into heat, but where? There is no shunt element, so what is dissipating the extra current that the load isnt using, especially during no load scenarios?

The current flow is via the series pass transistor, through the amp, and then back. So if the amp is not connected, no current can flow. These actually aren't formally being used as constant current sources as such - they are essentially current limiters with a limit current set by the cathode-gate resistor. In the words of the datasheet:

Applications

Highly stable voltage sources

Current surge limiters

Transient voltage protection

Instantaneously reacting resetable

fuses

Soft start-up circuits

This is totally different to a shunt current regulator, which dissipates maximum power when the current drawn is zero. The principle there is that the current regulator power dissipation plus the amp power dissipation is a constant. They do have significant benefits as compared to a series regulator, apart from (as Kevin says) the frying eggs syndrome. Most shunt current regulators will thermally die if a dummy load is not connected unless they have huge heatsinking.

Craig

[spritzer]

This is not about saving a few dollars.

We should have that engraved on the amp...

Any estimates for the power consumption?

200W+

[Inu]

Hi all

I've also gone through the amplifier BOM, and found a couple of minor inconsistencies as compared with the actual board.

1. 300k, 2 needed (not 6)

2. 390k, 8 needed (not 12)

Oh - and there are two R44's, one in the 700V battery (390k) and one and one between the supressor grid of the left hand 6CA7 and anode (220 ohms).

Basically I cross checked the boards (took a photo so I could cross parts out as I checked them), cross referenced it to the schematic parts numbers, then listed out the parts numbers in sequence to ensure I had a sequential list with nothing duplicated or obviously missing. So I *think* it is consistent. It definitely, *definitely* needs cross checking with similar things others on the forum are doing.

I've also broken the BOM into separate sheet for passives, semiconductors and hardware (I haven't checked the hardware) and provided a column for schematic parts cross reference.

BOM is here http://www.tech-enterprise.com/tekstuff/t2parts3cgsedit.xls

Craig

Craig,

Thank you for the update.

Neutrik Female XLR connectors: 568-NC3FAH-2 > 568-NC3FD-H (or H-B or H-BAG).

I will keep my NC3FAH-2s for other project

[cetoole]

The current flow is via the series pass transistor, through the amp, and then back. So if the amp is not connected, no current can flow. These actually aren't formally being used as constant current sources as such - they are essentially current limiters with a limit current set by the cathode-gate resistor. In the words of the datasheet:

Applications

Highly stable voltage sources

Current surge limiters

Transient voltage protection

Instantaneously reacting resetable

fuses

Soft start-up circuits

This is totally different to a shunt current regulator, which dissipates maximum power when the current drawn is zero. The principle there is that the current regulator power dissipation plus the amp power dissipation is a constant. They do have significant benefits as compared to a series regulator, apart from (as Kevin says) the frying eggs syndrome. Most shunt current regulators will thermally die if a dummy load is not connected unless they have huge heatsinking.

Craig

OK, so it is just a protection device, with no significant benefits to ripple rejection. Thanks, that answers that.

[kevin gilmore]

No, you are WRONG... shorting out the current sources and there is 6db more noise.

I just measured it. But no reason for me to argue.

Dynamic impedance of the current source reduces the noise.

NOTE!!!!!

D20 is labeled wrong on the power supply board. Needs to be flipped.

Schematic is correct. Power supply and time delay circuit now fully

tested.

picture of D20 correctly installed soon.

[kevin gilmore]

OK, here you go

http://gilmore.chem.northwestern.edu/t2powerfix.jpg

http://gilmore.chem.northwestern.edu/t2power.jpg

cetoole...

a naar veizt men nit kayn halbe arbet

Edited March 8, 2010 by kevin gilmore

[Guest sawyers]

NOTE!!!!!

D20 is labeled wrong on the power supply board. Needs to be flipped.

Schematic is correct. Power supply and time delay circuit now fully

tested.

Wow, well spotted. That would've given an interestingly high -12V line!

While thinking of lurking silk screen gremlins, the amp circuit board has R56/57 as 10k on one channel, but 100k on the other. I guess that 10k is the correct value?

Craig

[kevin gilmore]

While thinking of lurking silk screen gremlins, the amp circuit board has R56/57 as 10k on one channel, but 100k on the other. I guess that 10k is the correct value?

Craig

Correct value 10k. Mentioned a few pages back.

And when i tested the low voltage supply and found the -12 was -19 i knew instantly what was wrong.

So far, 2 silk screen errors. Total. Plus the holes on the 3381's a bit tiny.

Plus one zener hole a bit tiny.

Pretty fucking good for a project of this magnitude.

I challenge anyone to do a project of this scope and get it 99.99999999% right the first time.

Edited March 8, 2010 by kevin gilmore

[kevin gilmore]

all 3 transformers mounted.

http://gilmore.chem.northwestern.edu/t2power2.jpg

need slightly longer 1/4-20 nylon bolts

Its starting to get heavy...

[n_maher]

Kevin,

If those are SumR transformers double check the secondaries by hand. On the last trafo he made for me (for the BHse) he mis-marked the center tap on one of the secondaries, which thankfully didn't cook anything.

[Beefy]

That's obscene, Kevin. Certainly puts my meager efforts in their place