excellent, thats what you need especially for the input stage. I found that this matching and the input bias adjustment have the biggest impact on sound, probably bigger than the choice of RIAA components (opinions vary....) and output buffer (impact depending on cable length and line stage input impedance).
IMO Riaa components are determinant for sounstage and "reality" effect.... slight différences between channels introduce fase problems and the image is not so coherent. I payed much more attention in my third build and now soundstage is much wider than in my previous builds.
The effects coming from riaa filter shortcomings are surely different from what you get fidling with input stage.
Input stage is determinant to final low noise and definition.
One must be very carefull with every part in this complex preamp.
yes, I agree. My comment was referring to the type of components. Whatever you choose for the RIAA section, the values must be as close as possible between the channels - measuring and then measuring again and fiddling until you get it REALLY close.... the last 0.1% make a huge difference.....
Obviously you speak from experience, I am not sure anybody has built more Paradise's than you
, not even me
Obviously you speak from experience, I am not sure anybody has built more Paradise's than you
, not even meParadise modding
Thinking of it, here is my list of possible improvements I was brainstorming with myself. Are they all significant and worthwhile? Maybe not. But we all are not here to produce a cost-effective piece of industrial equipment, this is supposed to be wild , crazy fun
Interested in what you guys think......
Symmetrical output – Not straightforward in this circuit. But with an inverting buffer (gain = -1) in parallel to the existing buffer this could be done. Advantage would be if the rest of the chain is symmetrical.
Relais on board for switching the input impedance – Nice to have, allows a cleaner build. Not everybody wants that flexibility, though.
Balance adjustment +/-2dB in 0.25dB steps – Most cartridges have a slight channel imbalance, and that way that could be taken care of straight away. Misaligning the cartridge in the tonearm may swamp that effect, though, so this is mostly for the precision fanatics. I am trying this now in another build, lets see.....
Different settings for the RIAA components – Interesting for those who have older records. Can easily be done with the Paradise as it is today, using the „Ext RIAA“ Connector.
Volume adjustment – would allow for connecting it straight to a power amplifier and have the shortest possible signal path. Guess thats another nice-to-have.
Input stage heatsink – many have done this (or similar), and that has a positive impact on drift and offset. Paper or cardboard is good enough to avoid airflow around the input stage transistors. Better would be a little piece of aluminum with holes where the TO92 packages would fit, mechanically fixed to the PCB. (Yes I am an engineer.) Another way would be to implement the input stage on a separate PCB and put it in silicone oil or similar.... Cooling it down with a Peltier element doesnt help much, for a significant reduction in noise it would have to go down to 60°K absolute temp (Joachim calculated this once, I believe).
Headphone output – This should of course be in conjunction with the volume adjustment. Can't get any shorter.....
RC combination in parallel to the input electrolytics – This is a more esoteric adjustment which I determined from simulating the input stage with its parasitic inductances as present in the current PCB. Ttied it in my build, didnt make a difference. But I may be able to not hear it.
Mute switch – that comes in handy when changing records. Its not as easy as shorting the output, though, as this will disturb the servo. More thought required.....
Demagnetizer circuit – that would be an add-on (and I know these can be bought separately), but I have seen at least one phono stage that has it built-in. I am using another trick – shorting the output of the cartridge and playing a few records helps a lot.
Servo with 3-pole compensation – That will speed it up significantly, and at the same time have less impact on the music signal. Calculated but not tried yet.
Switchmode power supply – This has been shown to reduce noise coming from the grid quite significantly. It would replace the pre-reg, and still sit in a separate case, away from the Paradise.
Battery supply – Alternative to the switchmode power supply.
Vibration isolation – From all reports I have heard, the Paradise builds are not microphonic. But it doesnt hurt to mount it hanging in springs, with a heavy stone or copper platter, in a closed case, so that no noise from the outside can get in (of course, implementing proper electrical shielding too....)
Thinking of it, here is my list of possible improvements I was brainstorming with myself. Are they all significant and worthwhile? Maybe not. But we all are not here to produce a cost-effective piece of industrial equipment, this is supposed to be wild , crazy fun
Interested in what you guys think......
Symmetrical output – Not straightforward in this circuit. But with an inverting buffer (gain = -1) in parallel to the existing buffer this could be done. Advantage would be if the rest of the chain is symmetrical.
Relais on board for switching the input impedance – Nice to have, allows a cleaner build. Not everybody wants that flexibility, though.
Balance adjustment +/-2dB in 0.25dB steps – Most cartridges have a slight channel imbalance, and that way that could be taken care of straight away. Misaligning the cartridge in the tonearm may swamp that effect, though, so this is mostly for the precision fanatics. I am trying this now in another build, lets see.....
Different settings for the RIAA components – Interesting for those who have older records. Can easily be done with the Paradise as it is today, using the „Ext RIAA“ Connector.
Volume adjustment – would allow for connecting it straight to a power amplifier and have the shortest possible signal path. Guess thats another nice-to-have.
Input stage heatsink – many have done this (or similar), and that has a positive impact on drift and offset. Paper or cardboard is good enough to avoid airflow around the input stage transistors. Better would be a little piece of aluminum with holes where the TO92 packages would fit, mechanically fixed to the PCB. (Yes I am an engineer.) Another way would be to implement the input stage on a separate PCB and put it in silicone oil or similar.... Cooling it down with a Peltier element doesnt help much, for a significant reduction in noise it would have to go down to 60°K absolute temp (Joachim calculated this once, I believe).
Headphone output – This should of course be in conjunction with the volume adjustment. Can't get any shorter.....
RC combination in parallel to the input electrolytics – This is a more esoteric adjustment which I determined from simulating the input stage with its parasitic inductances as present in the current PCB. Ttied it in my build, didnt make a difference. But I may be able to not hear it.
Mute switch – that comes in handy when changing records. Its not as easy as shorting the output, though, as this will disturb the servo. More thought required.....
Demagnetizer circuit – that would be an add-on (and I know these can be bought separately), but I have seen at least one phono stage that has it built-in. I am using another trick – shorting the output of the cartridge and playing a few records helps a lot.
Servo with 3-pole compensation – That will speed it up significantly, and at the same time have less impact on the music signal. Calculated but not tried yet.
Switchmode power supply – This has been shown to reduce noise coming from the grid quite significantly. It would replace the pre-reg, and still sit in a separate case, away from the Paradise.
Battery supply – Alternative to the switchmode power supply.
Vibration isolation – From all reports I have heard, the Paradise builds are not microphonic. But it doesnt hurt to mount it hanging in springs, with a heavy stone or copper platter, in a closed case, so that no noise from the outside can get in (of course, implementing proper electrical shielding too....)
I used some kind of heatsink on the input stage, but also on the mirrors. It is more a thermocoupling than a heatsink. I put toghether all the transistors in a stage, thightly tied with a copper strip and silicon paste. In my build in each mirror stage there is a transistor warmer than the others.
Now, and after changing all the transistors with Diotecs, I have finaly good and more stable voltage measurements. If anybody wants some good matched transistors, just let me know.
Now, and after changing all the transistors with Diotecs, I have finaly good and more stable voltage measurements. If anybody wants some good matched transistors, just let me know.
high voltage over ext.RIAA
Hi,
finally I have my paradise up and running with paradise pre-reg.
But I have high voltage across ext. RIAA and 18V across output in one amp. The other is ok.
The reference voltages measure all the same as in the hesener-schema.
Where should I start looking? I tried lowering R43 a/b to 75 kOhm without much effect. Should I try other transistors in the input stage, new matching?
Thanks
Stephan
Hi,
finally I have my paradise up and running with paradise pre-reg.
But I have high voltage across ext. RIAA and 18V across output in one amp. The other is ok.
The reference voltages measure all the same as in the hesener-schema.
Where should I start looking? I tried lowering R43 a/b to 75 kOhm without much effect. Should I try other transistors in the input stage, new matching?
Thanks
Stephan
Hello Peter,
that doesn't seem to be the problem. Voltage regulator seem to work perfect. I get +18V on audio out of the amplifier and 16V across ext. RIAA.
Thanks
Stephan
VPlus is +18,3 V as the regulator is cold
VMinus is -18,3 V
Input is 0v with 10 Ohm
I just checked all the soldering, maybe one was suspicious, but it hasn't changed anything.
Now we are getting there
That is just what I found during my extensive research.... I have built more than 20 riaa amps with different topologies and always find myself fiddling with the last 10pF caps needed to finetune the curve.... It might seem strange but a very small difference is enough to go from "edgy" to "magistral" soundstage.