goldenreference low voltage power supply

[dsavitsk]

the noise spikes are 120hz (100hz for you) and due completely to the switching noise

of the input diodes. every power supply, even tube rectified ones do this. The idea

is to remove the spike noise, and its not easy. caps across the diodes, etc, just does not

work.

 

Nice design, Kevin. Any thought to changing the first cap to a pi filter?

[Pars]

the noise spikes are 120hz (100hz for you) and due completely to the switching noise

of the input diodes. every power supply, even tube rectified ones do this. The idea

is to remove the spike noise, and its not easy. caps across the diodes, etc, just does not

work.

 

 

Would changing to discrete input diodes, perhaps even Schottkys (31DQ10?) help for this?

[sorenb]

Nice design, Kevin. Any thought to changing the first cap to a pi filter?

isn't at pi filter there to reduce ripple? if the caps doesn't ground the spikes why should the coil resist those?

[kevin gilmore]

everyone wanted tiny.

they got tiny.

i could do singles sic diodes or schottky dioes, board has to get a bit bigger.

and the noise is already 140+ db down. i can't measure that at home, my

led lights over the bench generate 7mv into a single turn of wire. and the switcher

in the burglar alarm, and the high frequency switcher in the whole house UPS...

and the switcher in the variable frequency drive for the AC...

and at work, just barely can i measure.

the stuff that does -160 to -170db seems few and far between.

And in the stupid price category.

Edited July 19, 2015 by kevin gilmore

[JimL]

kevin gilmore on 19 Jul 2015 said: 

"...by the way, so far, I can't even get close to these numbers with a shunt regulated supply, the noise rips right thru the main current source."

 

I wonder if using a cascode current source would help?  Just a thought.

Edited July 19, 2015 by JimL

[kevin gilmore]

cascode current sources at 1 amp, not so easy

parts are not available

[luvdunhill]

the noise spikes are 120hz (100hz for you) and due completely to the switching noise

of the input diodes. every power supply, even tube rectified ones do this. The idea

is to remove the spike noise, and its not easy. caps across the diodes, etc, just does not

work.

I find a C || R+C across the secondary and a common-mode choke after the first cap seems to be about as good as you can do. The filter needs to be tuned to the transformer though.

[PretentiousFood]

cascode current sources at 1 amp, not so easy

parts are not available

 

Not with depletion mode parts, but with enhancement mode devices it works quite well. Here's something I'm using for a phono supply at 200mA or so, but it works well at higher currents with a different biasing scheme. You could use a mosfet cascode CCS (like Gary Pimm's) and feedback to the shunt device for better noise rejection, too.

 

[Helium]

Kevin, is this thread all devoted to goldenreference4.zip?

What about goldenreference and goldenreference3? Is goldenreference4 direct replacement for those?

And you opted to get rid of TO-3 transistors?

[kevin gilmore]

all are the same design, the first one has on board to3 transistors.

 

you can always use the goldenreference4 and mount the to3 transistors on

an external heatsink.

[mwl168]

On 7/16/2015 at 7:17 PM, kevin gilmore said:

its a 10 volt reference, the feedback resistors are 1.5k which is a divide by 2.

So if you change the top 1.5k resistor you can get any voltage you want.

750 ohms makes it 15v

 

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

 

change R8,R9

For my own benefit and possibly others, can someone confirm or correct the following:

The regulated output voltage is:

for V+: ((R8 + R7) / R7) x 10 (reference voltage of D5 - LT1021-10)

for V-: ((R9 + R10) / R10) x 10 (reference voltage of D7 - LT1021-10)

On the goldenreference4 board silkscreen, R7, R8, R9 R10 are all 1.5K. R7 and R8 are connected and R9 and R10 are connected. R8 and R9 are the 1.5K resistors closer to the DN2540 on each side of the board.   

Thanks!

 

Edited February 7, 2017 by mwl168

[kevin gilmore]

yep that is the idea

over the weekend i replaced the 1.5k resistors with 2k original vishay resistors rated at .01% and 5 ppm

voltages much more accurate.

version 5 of the goldenreference board has a place for 1M pots that go in parallel with these resistors for

those that just absolutely need to adjust to 6 places

(actually those pots may influence overall thermal stability, cermet pots required)

[Pars]

On 7/19/2015 at 3:10 PM, kevin gilmore said:

everyone wanted tiny.

they got tiny.

i could do singles sic diodes or schottky dioes, board has to get a bit bigger.

For the Gilmore dynamic V2 I built (on Justin's PCB), I did some adapter boards and used 31DQ10 Shottky diodes in place of the monolithic bridges (which I think are the same bridges used here).

 

These use a Mill-Max 90 degree pin and plug right into the monolithic bridge footprint. I would think these could be adapted to most any discrete rectifier diode assuming the room was there to accommodate them.

Edited February 27, 2018 by Pars

[UFN]

Nice idea, I may want to copy that 

 

//UFN

[b0bb]

Parts connexion sells them.

http://www.partsconnexion.com/kits_sub_assemblies.html

[insanity]

I have a few questions concerning the capacitors on the GR LV board, if someone is kind enough to answer

1. What are the recommended voltages of the tantalum caps on the GR LV? Is it the same for the 10uf and 47uf?

2. Whats the voltage rating for the 47pf disc cap?

3. Is this the correct 4.7uf 63v cap? http://www.mouser.ch/ProductDetail/WIMA/MKS2C044701M00JSSD/?qs=ISKKXg98%2b1mdri3wReBbVQ%3d%3d

4. Are these the correct 0.01uf caps? http://www.mouser.ch/ProductDetail/WIMA/MKS2C021001A00JI00/?qs=%2fha2pyFaduj8D%2b74lAuj%2fVfU0ISRQQyVkMGVwoGjK1pT08hCW9OSJ62gFby%2fdbTu

 

Thanks for helping

 

 

 

[kevin gilmore]

1)  35V  581-TAP476M020CCS  don't have a link for the 10uf

2) does not matter I used these 810-FK28C0G1H470J

3) I used these  505-MKS2-4.7/50/10 which are in stock and 50V, more than enough

4) I used these  505-MKS2.01/63/10

 

Edited August 29, 2015 by kevin gilmore

[Pars]

As far as transformer for these things, if I wanted to build one up to power a multiamp, and configured for +/- 20V output, would a 2 x 23VAC transformer be too high? Just thinking of stuff I have laying around without buying a new trafo. If I did use this, obviously there is going to be some voltage leftover to dump and this will cause some heat... anything I would need to be aware of here?

[insanity]

1)  35V  581-TAP476M020CCS  don't have a link for the 10uf

2) does not matter I used these 810-FK28C0G1H470J

3) I used these  505-MKS2-4.7/50/10 which are in stock and 50V, more than enough

4) I used these  505-MKS2.01/63/10

 

Thanks Kevin.

[kevin gilmore]

2 X 22 volts should be just fine, still need some heatsink on the pass transistors

does not seem to be a standard value though

[kevin gilmore]

I have posted the goldenreference6 board, has options for center tap transformer, and dual windings

for next time

[Pars]

I find a C || R+C across the secondary and a common-mode choke after the first cap seems to be about as good as you can do. The filter needs to be tuned to the transformer though.

I've tried this a couple of times (R+C across the secondary), but have never tried to measure it. 330 ohms and 0.022uf. See John Swenson's posts here:

http://www.diyhifi.org/forums/viewtopic.php?f=15&t=1957&hilit=transformer+ringing

[mwl168]

Does it matter what forward voltage LED to use so long as D2/D8 and D1/D9 match? 

[kevin gilmore]

nope

[nopants]

For KGSSHV duty does this even need heatsinks? How much heat is the package itself good for?