@fab, so was USSPA originally designed around the AMB backplane? Would you recommend this as a good way forward? I was going to use Salas UltraBIB, but this looks tempting as it makes for a really neat finish.
https://fabaudio.online/usspa-pre-amplifier
https://fabaudio.online/usspa-pre-amplifier
R1, R2 and C1 input cap - are these needed? Could they be bypassed and supply straight into the MOSFETs provided that the power supply is of sufficient quality and wired using short (say 5cm) wires?
Also, interested in why C1 is on the negative, not positive, side... Just interested as I'm not electronically trained.
...this question also partly relates to the input cap on USSA as well, would a high quality film cap on C2 still be beneficial even if you used very high quality electrolytic caps on your power supply (e.g. Nichicon Gold or Mundorf). OR can you get away with / negate the effect of cheaper electrolytics (e.g. CDE 380LX) if you use a premium film cap on the input of USSA or USSPA? Would be interested in experiences with this.
Also, interested in why C1 is on the negative, not positive, side... Just interested as I'm not electronically trained.
...this question also partly relates to the input cap on USSA as well, would a high quality film cap on C2 still be beneficial even if you used very high quality electrolytic caps on your power supply (e.g. Nichicon Gold or Mundorf). OR can you get away with / negate the effect of cheaper electrolytics (e.g. CDE 380LX) if you use a premium film cap on the input of USSA or USSPA? Would be interested in experiences with this.
I express my opinion based on my experimentation. I removed C1 because I don't like input filters. For C2 I used and highly recommend Elna Cerafine or alternatively Silmic II
R1 for input is needed in case the preamp may be connected to a source output without defined impedance to ground.
R2 ( along with source output impedance) and C1 form a noise frequency input filter. C1 value is chosen for a 10 kohms log potentiometer as source input of preamp. 100k ohms pot should reduce this value by 10 times, but 10 times lower than specified value is the input capacitance of the jfet thus C1 is not needed in that case. However, you can still reduce the original C1 value or not install it if you do not experience problems in your whole setup. I prefer to recommend a more general no issue value to start with…
For C2 and C3 on-board PSU filter caps, you can use other cap types than polymer one but the recommended parallel 10uf cap C4/C5 value and type is optimised for that specific 560uf polymer cap for an efficient frequency wide-band low impedance….
PSU caps on-board allows to minimise effect of inductance wiring and even 5cm is a long distance….
Fb
R2 ( along with source output impedance) and C1 form a noise frequency input filter. C1 value is chosen for a 10 kohms log potentiometer as source input of preamp. 100k ohms pot should reduce this value by 10 times, but 10 times lower than specified value is the input capacitance of the jfet thus C1 is not needed in that case. However, you can still reduce the original C1 value or not install it if you do not experience problems in your whole setup. I prefer to recommend a more general no issue value to start with…
For C2 and C3 on-board PSU filter caps, you can use other cap types than polymer one but the recommended parallel 10uf cap C4/C5 value and type is optimised for that specific 560uf polymer cap for an efficient frequency wide-band low impedance….
PSU caps on-board allows to minimise effect of inductance wiring and even 5cm is a long distance….
Fb
Last edited:
psu not so big deal. lt1963 symetry u can buy in ebay is ok. I prefer that for most of my project.
tried alot.. 1963 is simple. and sounds good. with the highest staging. R core sounds best to me but a good toroid is good as well.
cap coupling the higher u go the more roll off and better bass. 3.3uf to 4.7 is what I mostly prefer. if u want to go a bit crazy. parafeed it with 3.3uf. with a good line output transformer. 600:600 ohm. it's magic. and even better use a good tvc. feedback use carbon . or if u plan to use LOT. try tantalum resistors. juicy juicy
I only tried LOT it brings the sound to more density texture. analog naturalness. color it whichever u want. use litz wire too or neotech. Kle rca if u want to go crazier each contribute more special sound. otherwise use canare cables. it's the best cheap ones.
power cable use Belden instrument. 1030 1032. I tried many its a safe choice.
. have fun
tried alot.. 1963 is simple. and sounds good. with the highest staging. R core sounds best to me but a good toroid is good as well.
cap coupling the higher u go the more roll off and better bass. 3.3uf to 4.7 is what I mostly prefer. if u want to go a bit crazy. parafeed it with 3.3uf. with a good line output transformer. 600:600 ohm. it's magic. and even better use a good tvc. feedback use carbon . or if u plan to use LOT. try tantalum resistors. juicy juicy
I only tried LOT it brings the sound to more density texture. analog naturalness. color it whichever u want. use litz wire too or neotech. Kle rca if u want to go crazier each contribute more special sound. otherwise use canare cables. it's the best cheap ones.
power cable use Belden instrument. 1030 1032. I tried many its a safe choice.
. have fun
I’m considering using one of Ian Canadas UcPure ultra cap power supplies set up for 15v output for my USSPA build.
The ideal bias for USSPA is stated as being 50ma, is this a good indication of its typical draw? UcPure can deal with larger current requirements easily, but if USSPA-1 needed closer to 1A then it’d only last about an hour.
Is an ultra cap supply like this a good idea for volume control as well, with voltage dropping over time and needing continual volume adjustment?
The ideal bias for USSPA is stated as being 50ma, is this a good indication of its typical draw? UcPure can deal with larger current requirements easily, but if USSPA-1 needed closer to 1A then it’d only last about an hour.
Is an ultra cap supply like this a good idea for volume control as well, with voltage dropping over time and needing continual volume adjustment?
Hi Stretchneck
50ma is minimum output stage bias current. Ideal bias can be between 50 to 100ma. 70ma is a good setting. You also need to dd inoutbstage current of about 8 ma depending on your jfet. And this is for one channel. 1A PSU is recommended to have minimum ripple. Output voltage should be fixed and not reduce over time.
Fab
50ma is minimum output stage bias current. Ideal bias can be between 50 to 100ma. 70ma is a good setting. You also need to dd inoutbstage current of about 8 ma depending on your jfet. And this is for one channel. 1A PSU is recommended to have minimum ripple. Output voltage should be fixed and not reduce over time.
Fab
Thanks Fab
Narrowed my choice down to either:
1. Jan Didden Superreg, 1A max though is this sufficent?
2. Similar to what @fabiodipietro27 made, silicon carbide diodes, 4-pole Mundorf AG + 6800uf and 4-pole claritycap TC4 100Uf, belleson or NewClassD regulators (able to do >1.5 to 3A).
Any preferences?
Also, USSA has 1.8vrms input sensitivity, so best to configure USSPA for unity gain and run it as a buffer? My DAC gives out 2.5v already.
Narrowed my choice down to either:
1. Jan Didden Superreg, 1A max though is this sufficent?
2. Similar to what @fabiodipietro27 made, silicon carbide diodes, 4-pole Mundorf AG + 6800uf and 4-pole claritycap TC4 100Uf, belleson or NewClassD regulators (able to do >1.5 to 3A).
Any preferences?
Also, USSA has 1.8vrms input sensitivity, so best to configure USSPA for unity gain and run it as a buffer? My DAC gives out 2.5v already.
Hi Stretchneck
For power supply I have not experimented several different ones. AnthonyA did and some other members. The important thing is that it needs to be very good quality in such a preamplifier design.
As for running USSPA as unity gain it sure can be done but have in mind that this simple design is sensitive to such changes for sound. Could be better or worse in your system… you would need to try it out. But, having said that, a gain of 2 gives only more margin to cope with lower level recordings….
Fab
For power supply I have not experimented several different ones. AnthonyA did and some other members. The important thing is that it needs to be very good quality in such a preamplifier design.
As for running USSPA as unity gain it sure can be done but have in mind that this simple design is sensitive to such changes for sound. Could be better or worse in your system… you would need to try it out. But, having said that, a gain of 2 gives only more margin to cope with lower level recordings….
Fab
Hi baggerbole
with 16 ohms you will need only a low voltage signal at output to get the power needed. Ideally, you would use a small gain of 2 (6 db) for the preamp and lower the resistor R17 to something like 15 ohms or so for example.
As for the bias current, it depends on the heatsink size. With the one shown in the picture of the manual you can go to 80ma while keeping a reasonable temperature on the heatsink for the transistors. This would give you about 200mW in class A which is probably enough for headphones Listening if you want to keep your hearing intact….
Fab
with 16 ohms you will need only a low voltage signal at output to get the power needed. Ideally, you would use a small gain of 2 (6 db) for the preamp and lower the resistor R17 to something like 15 ohms or so for example.
As for the bias current, it depends on the heatsink size. With the one shown in the picture of the manual you can go to 80ma while keeping a reasonable temperature on the heatsink for the transistors. This would give you about 200mW in class A which is probably enough for headphones Listening if you want to keep your hearing intact….
Fab
Last edited:
I'm a bit perplexed. I believe that to power the USSPA with a +15/-15V dual of these supercaps and Ian's UcPure rigs you'll need two.