Looks like your Iset Rdrop is getting smaller the lower your Vin-Vout dif. Now you run circa 70mA per rail. Not bad. Is your mains at nominal value if you measure it with DVM on max AC scale? Also how much your row of 3 leds drop?
Thanks Tea-bag, it sound good to my ear. Now start to do something different, DIY some equipment see how far can my system go.
My main now AC is 238VAC, after rectification 15V. Row of 3 leds drop = 5.39V and 5.36V.
sourcing 2sk170 fets
Hi,
I'm looking to construct a B1 myself, but Mouser shows the fets on backorder for 11 weeks? does anyone know of a local (US) source for the 2sk170's?
also, I have 10K noble and 25K alps black beauty pots in my parts bin. which would be better for this application?
thx
Hi,
I'm looking to construct a B1 myself, but Mouser shows the fets on backorder for 11 weeks? does anyone know of a local (US) source for the 2sk170's?
also, I have 10K noble and 25K alps black beauty pots in my parts bin. which would be better for this application?
thx
You might try Blues over at the swap meet...
http://www.diyaudio.com/forums/swap-meet/161366-matched-toshiba-fets-sale.html
I aquired my matched and unmatched FET's from spencer here on the forums, but he's in Hong Kong. No problems with delivery though, I'm a satisfied customer.
The alps would be easier on the sources I guess, recommended value is 15-25K. I use the LDR attenuator myself, sound is fantastic.
http://www.diyaudio.com/forums/swap-meet/161366-matched-toshiba-fets-sale.html
I aquired my matched and unmatched FET's from spencer here on the forums, but he's in Hong Kong. No problems with delivery though, I'm a satisfied customer.
The alps would be easier on the sources I guess, recommended value is 15-25K. I use the LDR attenuator myself, sound is fantastic.
Hi Sugden.
The voltage across the current setting resistors should not change. There is slight drift with temperature.
I think your 12Vac transformer is not giving sufficient voltage before the regulator and that's why you see a lower volts drop across the triple 68r, cf. the double 68r.
Do you have a 15+15Vac transformer to try?
The voltage across the current setting resistors should not change. There is slight drift with temperature.
I think your 12Vac transformer is not giving sufficient voltage before the regulator and that's why you see a lower volts drop across the triple 68r, cf. the double 68r.
Do you have a 15+15Vac transformer to try?
Current measurements for my DB1 shunts:
http://www.diyaudio.com/forums/pass...symmetrical-psu-b1-buffer-62.html#post1998407
also see salas post right after mine.....
Fran
http://www.diyaudio.com/forums/pass...symmetrical-psu-b1-buffer-62.html#post1998407
also see salas post right after mine.....
Fran
Tweak to test
For all of you tweak inclined DCB1 owners, I have a an idea to test. If you happen to have a pair 0.1-1uF axial MKT or MKP that can fit upright replacing the 100uF 5X LEDs filter caps (the pair behind the transformer connector), give it a try and evaluate subjectively. Due to the node's impedance of the LEDs Vref, there will be no averse effect with the regs very low frequency gain & phase. Its a trade off between much noise filtering or little filtering and adequate or better cap distortion. Since the LEDs are good enough for self noise, and we have line level audio here, lets see what it may subjectively matter more to you. Don't destroy the 100uF caps, you may want them back or not. Let us know your evaluations.
For all of you tweak inclined DCB1 owners, I have a an idea to test. If you happen to have a pair 0.1-1uF axial MKT or MKP that can fit upright replacing the 100uF 5X LEDs filter caps (the pair behind the transformer connector), give it a try and evaluate subjectively. Due to the node's impedance of the LEDs Vref, there will be no averse effect with the regs very low frequency gain & phase. Its a trade off between much noise filtering or little filtering and adequate or better cap distortion. Since the LEDs are good enough for self noise, and we have line level audio here, lets see what it may subjectively matter more to you. Don't destroy the 100uF caps, you may want them back or not. Let us know your evaluations.
The bigger axial film cap it can fit upright, the more the filtering too. Positive suggestion then. But maybe better if smaller but MKP. V1.2s are proving excellent on phono for now, a friend here evaluates a double on his 57dB MC, he is chuffed head to head with V1.0s. Line level is less demanding, but Bill in Sydney is chuffed on op amps too. He will let us know with them on DCB1 point to point soon. I will integrate special LED ref V1.2s only if I can find a way to relatively simply nail the B1's Achille's heel manifesting in any package guise first, namely its tendency to up THD with level. Although pleasant (N.P. knows his stuff VERY well), its still what a so simple 2 Jfet circuit has to trade off. Not objectionable as a max number, but still a non linearity trend that has to do with 2SK170 Ciss.
I'm talking about tiny radial caps p 5 mm that fit nicely in the pcb.... MKPs AFAIK aren't available bigger than 0.33 uF I'm afraid.
Axial MKPs are generally 250V.... that means quite big stuff.
he is chuffed head to head with V1.0s Sorry, I don't undesterstand: means he's ultra happy with V1.2 over V1.0?
Axial MKPs are generally 250V.... that means quite big stuff.
he is chuffed head to head with V1.0s Sorry, I don't undesterstand: means he's ultra happy with V1.2 over V1.0?
Last edited:
I do not have any 15+15Vac transformer, but I connect DCB1 to my PS audio power plant which I can get adjust to 250VAC main power supply. Now voltage after rectification are +16.15V & -16.22V, volt at 3 parallel 68R are 1.547V & 1.666V. Look like not much different to the CCS. These all can do untill now, if necessary I will buy one 15+15VAC transformer to test it.
I'm talking about tiny radial caps p 5 mm that fit nicely in the pcb.... MKPs AFAIK aren't available bigger than 0.33 uF I'm afraid.
Axial MKPs are generally 250V.... that means quite big stuff.
he is chuffed head to head with V1.0s Sorry, I don't undesterstand: means he's ultra happy with V1.2 over V1.0?
Whatever fits.
Yes, chuffed means very happy, well satisfied.
I have already finished my DCB1 and put it in a case. For that matter I used an old telecom equipment chassis. Old stuff has no plastics and is more robust than what is made today I will post some pictures as soon as I get a camera.
It works and sounds wonderful. I managed to have a very close match between leds (and I had the luck of finding matched pairs of jfets in my bag as close as 0.02% between them).
Here are the measurements, done with inputs shorted to ground and waited 2 hours after turning it on. 100-150h of burn-in already:
DC offsets: -0.86mv and -1.84mv
AC output: 0.01mv (less than DVM noise floor)
Tx voltage: 13.48vac
After rectification:17.4vdc
Positive rail: 10.748
Negative rail: -10.041
PSupply ripple: 0.00mv
Vdrop68//68: 2.456v (72mA) and 2.216v (65mA)
I'm thinking in rising the CCS to 200 or 250mA. I've done my calculations, and maybe I'm wrong, but I've found that you dont need 1W res. If you are passing 2.456v along 34 ohms, that is 72mA.
P=V·I, so 2.456 x 0.072 = 0.176W. Wich is less than the standard quarter-watt resistor. So you don't even need to paralled them, neither making them 0.5W.
Am I wrong in something? Probably sure because you guys have thought of it before. As we say, don't try to teach your father how to make children
I will post some pictures as soon as I get a camera.It works and sounds wonderful. I managed to have a very close match between leds (and I had the luck of finding matched pairs of jfets in my bag as close as 0.02% between them).
Here are the measurements, done with inputs shorted to ground and waited 2 hours after turning it on. 100-150h of burn-in already:
DC offsets: -0.86mv and -1.84mv
AC output: 0.01mv (less than DVM noise floor)
Tx voltage: 13.48vac
After rectification:17.4vdc
Positive rail: 10.748
Negative rail: -10.041
PSupply ripple: 0.00mv
Vdrop68//68: 2.456v (72mA) and 2.216v (65mA)
I'm thinking in rising the CCS to 200 or 250mA. I've done my calculations, and maybe I'm wrong, but I've found that you dont need 1W res. If you are passing 2.456v along 34 ohms, that is 72mA.
P=V·I, so 2.456 x 0.072 = 0.176W. Wich is less than the standard quarter-watt resistor. So you don't even need to paralled them, neither making them 0.5W.
Am I wrong in something? Probably sure because you guys have thought of it before. As we say, don't try to teach your father how to make children
Thanks Salas, now I hook it up to my system. It make my system sound more lively, continue burn-in before my volume pot coming. Will report back the sound. Appreciate for all the advices.