With a 317 preceding the super reg, when driving some current into headphones would the super reg performance degrade to that of the 317?
Would that be correct? As current has to come though the 317 to get to the super reg pass transistor. And with the lowered voltage across the pass transistor, as recommended in the article, there is less voltage overhead for the super reg to make use of before the 317's performance dominates?
Perhaps best then to not use the pre-reg for headphones?
My +5V use just now is for a Dragonfly Red DAC I have; to make it's supply not USB bus fed, which they say improves it sound. Whilst this device is intended to drive headphones I don't use it for that, I directly drive a stereo power amp inside which I've mounted this DAC so as to do away with the analogue interconnects, yielding a lot better clarity and immediacy, and the DAC is apparently very efficient drawing very little current, so might that be OK with a 317 pre reg?
Would that be correct? As current has to come though the 317 to get to the super reg pass transistor. And with the lowered voltage across the pass transistor, as recommended in the article, there is less voltage overhead for the super reg to make use of before the 317's performance dominates?
Perhaps best then to not use the pre-reg for headphones?
My +5V use just now is for a Dragonfly Red DAC I have; to make it's supply not USB bus fed, which they say improves it sound. Whilst this device is intended to drive headphones I don't use it for that, I directly drive a stereo power amp inside which I've mounted this DAC so as to do away with the analogue interconnects, yielding a lot better clarity and immediacy, and the DAC is apparently very efficient drawing very little current, so might that be OK with a 317 pre reg?
The superreg is very good at isolating the load from any ripple on the reg input. You can use a prereg to lower that ripple but it is unlikely that you see or hear any difference at the output of the superreg.
Of course knowing you spend money and effort to implement a prereg will most likely let you conclude it sounds sooo much better
Jan
Of course knowing you spend money and effort to implement a prereg will most likely let you conclude it sounds sooo much better
Jan
Hi Jan, The weather here has been wet and windy, so finally have some time to start testing my superregs.
My power source is a 460mA 15v-0-15v torroidal transformer, powering L7815 and L7915 regulators on a cheap ps kit. The output from the ps is 15v and -14.89v.
I have connected the +ve and -ve outputs from the ps to the internal pins on the super regulators, and the common ground to the outer connectors. I have installed sockets at U1 and U2. Without a dummy load or jumpers at the sense outputs, I get the following readings with dvm:
U1
Pin 5= 14.96v
Pin 6= 14.35v
U2
Pin4= -14.25v
Pin5= -14.69v
At this stage, both LEDs are not lighting up - which I suppose is no surprise considering both are connected to Q2 and Q4 which depend on the Return connections to be made.
Any advice or warnings before I proceed?
My power source is a 460mA 15v-0-15v torroidal transformer, powering L7815 and L7915 regulators on a cheap ps kit. The output from the ps is 15v and -14.89v.
I have connected the +ve and -ve outputs from the ps to the internal pins on the super regulators, and the common ground to the outer connectors. I have installed sockets at U1 and U2. Without a dummy load or jumpers at the sense outputs, I get the following readings with dvm:
U1
Pin 5= 14.96v
Pin 6= 14.35v
U2
Pin4= -14.25v
Pin5= -14.69v
At this stage, both LEDs are not lighting up - which I suppose is no surprise considering both are connected to Q2 and Q4 which depend on the Return connections to be made.
Any advice or warnings before I proceed?
Well you really do need to connect those jumpers to make it work. But be aware that when you feed them 15V you won't get 15V out.
They probably need at least a volt headroom, probably more to be on the safe side.
So if you want 15V out you need to feed them 17V. Or, leave the 15V in and set the output to 13V. If you are feeding modern opamps in a preamp or filter, 13V is just fine.
Jan
They probably need at least a volt headroom, probably more to be on the safe side.
So if you want 15V out you need to feed them 17V. Or, leave the 15V in and set the output to 13V. If you are feeding modern opamps in a preamp or filter, 13V is just fine.
Jan
Thanks for the quick response Jan. I am still up here - 11.30pm in Perth Western Australia. I have connected jumpers across pins 1 and 2 and pins 3 snd 4 on both regs. and have the following to report:
On U1:
Pin Voltage
1 0v
2 6.68v
3 7.13v
4 0v
5 15.04v
6 14.08v
7 0.01
U2.
1 -0v
2 -6.67v
3 -7.75v
4 -13.97v
5 -0v
6 -14.87v
7 -0.0v
8 -0v
At the output I am getting 14.8v and -13.97v
Both LEDs are NOT lighting up.
On U1:
Pin Voltage
1 0v
2 6.68v
3 7.13v
4 0v
5 15.04v
6 14.08v
7 0.01
U2.
1 -0v
2 -6.67v
3 -7.75v
4 -13.97v
5 -0v
6 -14.87v
7 -0.0v
8 -0v
At the output I am getting 14.8v and -13.97v
Both LEDs are NOT lighting up.
You mean the LED in series with the opamp output? Of course they do not light up. Consider what I said before, and look at the voltage across the LED.
Jan
Pin 2 and 3 should have the same voltage within a few millivolts.
I'll suppose you use a LM329 as voltage reference?
Values for R6, R7, 1k+1K? If yes, you should get 14 V out but you must have 17 V in. What is you supply voltage?
Edit: Do you really use JRC4556? This is dual opamp, can't be used for this.
I'll suppose you use a LM329 as voltage reference?
Values for R6, R7, 1k+1K? If yes, you should get 14 V out but you must have 17 V in. What is you supply voltage?
Edit: Do you really use JRC4556? This is dual opamp, can't be used for this.
Last edited:
Yes, as per BOM: LM329 is voltage reference and R6, R7 are 1k + 1K.
No, I just stuck the JRC4556 for test purposes, I do have the recommended AD825 but haven't sourced the adaptor.
The supply voltage is 15v-0-15v. I will need to get 18v
You could just check by temporarily paralleling the top 1k with another 1k; that should give you just above 10V. Just to check correct operation, if you are impatient. ;-)
Jan
Hi Jan, is there any issue in using your positive supply board with nelsons B1 design ? He uses a voltage divider with Two 10K resistors to pull down the gate of jfet? Just wondering if I need to use the 15000u cap in B1 as it may cause the regulator to oscillate ?
http://www.firstwatt.com/pdf/prod_b1_man.pdf
C1 C2 and C3 and 1 ohm 3 watt resistor could be skipped in attached schematic in page 2?
http://www.firstwatt.com/pdf/prod_b1_man.pdf
C1 C2 and C3 and 1 ohm 3 watt resistor could be skipped in attached schematic in page 2?
There should be no problem at all. In the B1 you should leave out both the 1 ohm R1 and the 1 15000uF cap. The superreg looks pretty much like a battery - a battery better than a real one ;-)
C2 also becomes rather useless, but maybe leave C3 for good measure, possible down to 100nF.
Jan
Last edited:
Andrew did you do the sums?
For a very clean power supply like the superreg, this is a prime example of thoughtless overkill, only serving to waste space, waste money and increasing potential RF pickup. Bad.
Jan
Last edited: