That's a really big bill, but it must have bought a very fine dinner. What was for dinner?
Hi,
While the circuit you show does remove crossover distortion it is not suited to driving low impedances.
What I have attached can drive 15V peak into 100 Ohm while leaving the Op-Amp essentially unloaded and offering under 0.2 Ohm open loop output impedance.
Note, due to the limitations of the software used to create the drawing I have to show the LM317 as a NPN-Transistor T2 (base = Adj, emitter = Out, collector = In). The boosted rails for the output buffer where needed to allow maximum level from the circuit. If a few lower clipping levels are acceptable the rails can be shared.
This circuit originated in the 1980's as line driver for use with a 1:1 output transformer in studio electronics, for use with a special german "Audio" Op-Amp (the east german B761 which in turn was a copy of the TFK TAA761) that had an Open-Collector Output.
I do also seem to remember that we loaded the "Comp" Pin of the Op-Amp to ground and assymetric supply rails, but memory gets hazy now and I have not seen the original schematics in well over two decades. The original easily drove well over +20dBU into 200 Ohm at very low THD.
If sufficient heatsinking is employed the quiescent current can be increased from the 150mA present and even as much as 15V into 16 Ohm can be handled at 1A Iq (yes, that is 30W dissipation per channel).
If using a genuine 5534 this buffer may be attached to pin 5 instead of the output.
If using modern Op-Amp's with rail-to-rail output or diamond buffer output the bipolar transistor should be replaced with a low capacitance (e.g. IRF610) Mosfet and the polarity should be flipped (N-Channel parts are still better than P-Channel ones)
Ciao T
Crossover distortion?
I guess this is worth a look:
How about that thing?
....
Edit: I guess that my question for you is, should we swap or should we bias the op-amp?
While the circuit you show does remove crossover distortion it is not suited to driving low impedances.
What I have attached can drive 15V peak into 100 Ohm while leaving the Op-Amp essentially unloaded and offering under 0.2 Ohm open loop output impedance.
Note, due to the limitations of the software used to create the drawing I have to show the LM317 as a NPN-Transistor T2 (base = Adj, emitter = Out, collector = In). The boosted rails for the output buffer where needed to allow maximum level from the circuit. If a few lower clipping levels are acceptable the rails can be shared.
This circuit originated in the 1980's as line driver for use with a 1:1 output transformer in studio electronics, for use with a special german "Audio" Op-Amp (the east german B761 which in turn was a copy of the TFK TAA761) that had an Open-Collector Output.
I do also seem to remember that we loaded the "Comp" Pin of the Op-Amp to ground and assymetric supply rails, but memory gets hazy now and I have not seen the original schematics in well over two decades. The original easily drove well over +20dBU into 200 Ohm at very low THD.
If sufficient heatsinking is employed the quiescent current can be increased from the 150mA present and even as much as 15V into 16 Ohm can be handled at 1A Iq (yes, that is 30W dissipation per channel).
If using a genuine 5534 this buffer may be attached to pin 5 instead of the output.
If using modern Op-Amp's with rail-to-rail output or diamond buffer output the bipolar transistor should be replaced with a low capacitance (e.g. IRF610) Mosfet and the polarity should be flipped (N-Channel parts are still better than P-Channel ones)
Ciao T
Attachments
Hi,
While the question "why not discrete" is valid, often the discrete circuitry becomes way more complex than the Op-Amp plus Buffer and performs less well.
For example my original 1980's B761 & SE Output Buffer replaced a by far more complex circuit with by far more parts and which needed a larger, airgapped output transformer with more windings that performed by far less well objectively and subjectively. No doubt, the discrete circuit could also have been tuned to give better performance, but from a manufacturing viewpoint it was a better choice to go integrated.
Speaking of the low impedance Baxandall Volume and Tone Control, if one was not slavishly committed to absolutely bodging a NE553X into any application, by whatever means and hook and crook, one could have looked beyond the rim of ones plate for better suited Op-Amp's.
For ONE example one could have used a pair of AD815 per channel. One as inverter for the balanced circuit, one in the volume control, one (without extra buffers) in the Tone Control and one as buffer preceding the tone control.
The AD815 will happily drive 10V peak into 25 Ohm with 0.005% THD and no rise in THD up to 30KHz and it will manage 10V peak into 100 Ohm with < 0.0005% THD at 20KHz...
Ciao T
While the question "why not discrete" is valid, often the discrete circuitry becomes way more complex than the Op-Amp plus Buffer and performs less well.
For example my original 1980's B761 & SE Output Buffer replaced a by far more complex circuit with by far more parts and which needed a larger, airgapped output transformer with more windings that performed by far less well objectively and subjectively. No doubt, the discrete circuit could also have been tuned to give better performance, but from a manufacturing viewpoint it was a better choice to go integrated.
Speaking of the low impedance Baxandall Volume and Tone Control, if one was not slavishly committed to absolutely bodging a NE553X into any application, by whatever means and hook and crook, one could have looked beyond the rim of ones plate for better suited Op-Amp's.
For ONE example one could have used a pair of AD815 per channel. One as inverter for the balanced circuit, one in the volume control, one (without extra buffers) in the Tone Control and one as buffer preceding the tone control.
The AD815 will happily drive 10V peak into 25 Ohm with 0.005% THD and no rise in THD up to 30KHz and it will manage 10V peak into 100 Ohm with < 0.0005% THD at 20KHz...
Ciao T
Hi forr,
This is a fair question. I have addressed concerns about the phenomena of distortion in the Baxandall volume control when set to low levels, but have not addressed the relative magnitude of the problem. Since we are talking about current-induced distortion when the output signal voltage is almost zero, I don't think we can draw many conclusions pro or con from the distortion created by the 5532 when driving a high voltage into a 600-ohm load. For one thing, the high signal voltage greatly dilutes the THD created in the case I descibed. For another, the high signal voltage case does not let us separate out voltage-induced and current-induced distortions.
If one were to really wave their hands and suggest that worst case all of the 0.00018% distortion voltage would also appear due to signal current when the output of the op amp were driving a load with essentially no signal, and one were to say that the baxandall control was set to -40dB with respect to a nominal signal level, then MAYBE one could argue that the relative THD % could be nearly 60dB higher at 0.18%, but this is an extreme stretch.
The best way to do it, in my opinion, is to operate the line stage at nominal signal level output of, say, 1V, then reduce the volume control to -40dB, obtaining a signal level of 10mV. Then look at THD of that signal at both 1kHz and 20kHz.
Cheers,
Bob
More questions for Mr Self
Dear Mr Self,
Hope you are still keeping an eye on this thread. Firstly, thank you for sharing your expertise and most recent design innovations via the article. I've learnt so much about audio design from your succinct and enjoyably written articles and books over the years.
Comparing this design to the Precision Pre '96, I have some further questions along the lines of Mooly's post earlier that I'm hoping you might be able to answer. I've recently ordered a PCB for the Precision Pre 2.0 courtesy of the Signal Transfer company (superb PCB design, and no regrets having seen that the new design doesn't have those brilliant variable freq tone controls) but before I commence building it the new design raises a few questions that I would be grateful for your advice on:
1. Component Choice
There was considerable emphasis on avoiding the use of exotic components and obtaining high performance through good design in your '96 article. The new design seems to have compromised this to some extent. Specifically:
(a) the specified pots, at >$20 each and long lead times (if available at all to hobbyists) add up to some serious bom cost. Granted at 1k they have to handle much higher power than usual so there may not be much choice when going so far down the low-Z design path.
(b) the 100p feedback capacitors are all specified as axial polystyrene, which cost about $3 each in small qtys. Did you find it necessary to use these to achieve the specified performance? Having read C. Bateman's capacitor distortion articles I'm wondering if COG ceramic or polypropylene (both <$0.50) could be substituted without affecting performance. (COG Ceramic is listed in the BOM for the '96 preamp). Or was the choice for less technical reasons - eg avoiding audiophile ceramic cap phobia
(c) as previously mentioned, the coupling caps are mostly not required in this design, but those that are are now bipolar instead of polarised electros. I guess this is mainly because they're at the inputs and outputs where they are connecting to external equipment with potentially unknown polarity DC offset. But again referring to the Bateman articles he asks why some designers use polarised electros in the signal path, having measured large distortions when the signal voltage across the cap is significant. But even for applications where it isn't, he still suggests the use of bipolar caps. I know the values used in the '96 preamp (220uf) should ensure the signal voltage across is insignifcant, however I still wonder if using bipolar caps instead in the '96 design (after all there are quite a few) has the potential to measureably reduce the distortion generated to any extent? Has your stance on this changed any recently?
2. Resistor values in AGC
The '96 design used 680R in the inverting input of the shunt feedback stage, which as stated reduced noise slightly compared with the '83 incarnation which used 1k. The new design is back up to 1k - was this a deliberate choice to minimise distortion by reducing the load driven by the preceeding buffer? I would certainly think an increase in distortion due to loading would have more potential to be audible than a tiny increase or reduction in noise at this level, and wonder if I should use 1k instead (adjusting the other R to keep the gain the same) in the '96 design?
3. 5532 manufacturer
I know you've already answered this to some extent (and admit your active crossover book is the only one I don't own) but since the cheapest source of non-TI 5532s here seems to be JRC/NJM from RS Components, how did these stack up compared to the others? I don't think there has been any specific mention of this, and the On-Semi (which I know was good from the Elektor 5532 opamplifier Pt2 article) is about twice the cost. Is it worth the extra or is the NJM just as good? Also the NJM is not available in the selected low noise "A" suffix. Is this likely to make much difference in this application.
Obviously I'd much rather find all this out for myself but unfortunately don't have access to any high-res audio test gear at present, so can only hope you might have time to indulge me Apologies for long post...
Dear Mr Self,
Hope you are still keeping an eye on this thread. Firstly, thank you for sharing your expertise and most recent design innovations via the article. I've learnt so much about audio design from your succinct and enjoyably written articles and books over the years.
Comparing this design to the Precision Pre '96, I have some further questions along the lines of Mooly's post earlier that I'm hoping you might be able to answer. I've recently ordered a PCB for the Precision Pre 2.0 courtesy of the Signal Transfer company (superb PCB design, and no regrets having seen that the new design doesn't have those brilliant variable freq tone controls) but before I commence building it the new design raises a few questions that I would be grateful for your advice on:
1. Component Choice
There was considerable emphasis on avoiding the use of exotic components and obtaining high performance through good design in your '96 article. The new design seems to have compromised this to some extent. Specifically:
(a) the specified pots, at >$20 each and long lead times (if available at all to hobbyists) add up to some serious bom cost. Granted at 1k they have to handle much higher power than usual so there may not be much choice when going so far down the low-Z design path.
(b) the 100p feedback capacitors are all specified as axial polystyrene, which cost about $3 each in small qtys. Did you find it necessary to use these to achieve the specified performance? Having read C. Bateman's capacitor distortion articles I'm wondering if COG ceramic or polypropylene (both <$0.50) could be substituted without affecting performance. (COG Ceramic is listed in the BOM for the '96 preamp). Or was the choice for less technical reasons - eg avoiding audiophile ceramic cap phobia
(c) as previously mentioned, the coupling caps are mostly not required in this design, but those that are are now bipolar instead of polarised electros. I guess this is mainly because they're at the inputs and outputs where they are connecting to external equipment with potentially unknown polarity DC offset. But again referring to the Bateman articles he asks why some designers use polarised electros in the signal path, having measured large distortions when the signal voltage across the cap is significant. But even for applications where it isn't, he still suggests the use of bipolar caps. I know the values used in the '96 preamp (220uf) should ensure the signal voltage across is insignifcant, however I still wonder if using bipolar caps instead in the '96 design (after all there are quite a few) has the potential to measureably reduce the distortion generated to any extent? Has your stance on this changed any recently?
2. Resistor values in AGC
The '96 design used 680R in the inverting input of the shunt feedback stage, which as stated reduced noise slightly compared with the '83 incarnation which used 1k. The new design is back up to 1k - was this a deliberate choice to minimise distortion by reducing the load driven by the preceeding buffer? I would certainly think an increase in distortion due to loading would have more potential to be audible than a tiny increase or reduction in noise at this level, and wonder if I should use 1k instead (adjusting the other R to keep the gain the same) in the '96 design?
3. 5532 manufacturer
I know you've already answered this to some extent (and admit your active crossover book is the only one I don't own) but since the cheapest source of non-TI 5532s here seems to be JRC/NJM from RS Components, how did these stack up compared to the others? I don't think there has been any specific mention of this, and the On-Semi (which I know was good from the Elektor 5532 opamplifier Pt2 article) is about twice the cost. Is it worth the extra or is the NJM just as good? Also the NJM is not available in the selected low noise "A" suffix. Is this likely to make much difference in this application.
Obviously I'd much rather find all this out for myself but unfortunately don't have access to any high-res audio test gear at present, so can only hope you might have time to indulge me
Apologies for long post...I had the same thought. Before I committed to the entire build I only ordered one pot - just to be sure I was ordering the correct item. It had been back-ordered since April and showed up last week along with a freight bill for $20 from Farnell UK. Including tax, just the single pot cost me $44! Well, it's the right pot, correct size and all and fits. I ordered the other 3 that I need, just this week, because I want to finish the pre-amp and see how well it performs. Delivery date for the other 3 is September 19th
Say what!Oh well, it is what it is. I had to order some other parts for my next project, Doug's MM/MC pre-amp from Newark along with the 3 1k pots. I've instructed that the back-ordered items (from Farnell, UK) be held up until they can be fufilled and mailed as a single shipment. This will help justify the freight for the 3 additional pots.
Just for the 1k pots alone I will have spent over $120 dollars! I feel your pain owdeo. This will be my first Doug Self designed pre-amp so I'm looking forward to completing it. It will just take some waiting! I would never have imagined stereo pots costing >$20. This hobby is getting expensive lately.
Ps. I just finished up soldering everything else on the pre-amp board tonight sans the pots. It looks great. Hopefully the pots will show up sooner, rather than later. All the other back-ordered items have an earlier availability date.
Pss. It too bad that a full kit of parts, including the PCB cannot be made available for a few extra dollars. I would gladly pay a little more knowing I'm getting everything I need in one bag - the exact components with correct physical size, type and other rating. It's a full days work just to look-up and order the parts for these projects.
Ouch redjr, that's rough having to pay shipping on each back order. I think we may be luckier in Australia in that the full UK Farnell catalogue seems to be available via Element14 here with free shipping, the only drawback being parts sourced from their UK warehouse take 6-7 days to arrive here if in stock over there. Not too sure what happens with that kind of back order though, perhaps they still would charge additional shipping.
Agreed, as hobbyists we aren't nearly as concerned with component cost as manufacturers have to be, but the bom cost can still get out of hand, and long lead times can take the momentum out of the project...
I expect there was not much choice in specifying those pots in this design, but I'm very much hoping to hear about the reason for specifying expensive caps etc.
I'm actually not so keen on kits, though this may be just from my experience with local kit suppliers than only ever include cheap and nasty passives of unknown origin. One of the reasons I like to build my own gear is that I have total freedom to choose the quality of every component (that's not to say I want to use the most expensive parts) while controlling the cost. For example, I managed to reduce my shopping bill for the Precision Pre '96 BOM by more than $100 by carefully comparing prices between Farnell and RS for each part and sharing the order between them. True it took ages but I guess that's part of the fun of the hobby for me - keeps me off the streets anyway
Look forward to hearing how your preamp sounds. One thing is certain - despite the long wait and high cost you'll end up with a preamp that has state-of-the-art measured performance.
A thought occurs, since you have one pot, could you use it for the volume control, keep the tone controls bypassed, and temporarily use fixed resistors for the other pots (eg a 510R resistor across each pot terminal position)? At least maybe you could get it going before September in straight-through mode anyway.
Agreed, as hobbyists we aren't nearly as concerned with component cost as manufacturers have to be, but the bom cost can still get out of hand, and long lead times can take the momentum out of the project...
I expect there was not much choice in specifying those pots in this design, but I'm very much hoping to hear about the reason for specifying expensive caps etc.
I'm actually not so keen on kits, though this may be just from my experience with local kit suppliers than only ever include cheap and nasty passives of unknown origin. One of the reasons I like to build my own gear is that I have total freedom to choose the quality of every component (that's not to say I want to use the most expensive parts) while controlling the cost. For example, I managed to reduce my shopping bill for the Precision Pre '96 BOM by more than $100 by carefully comparing prices between Farnell and RS for each part and sharing the order between them. True it took ages but I guess that's part of the fun of the hobby for me - keeps me off the streets anyway
Look forward to hearing how your preamp sounds. One thing is certain - despite the long wait and high cost you'll end up with a preamp that has state-of-the-art measured performance.
A thought occurs, since you have one pot, could you use it for the volume control, keep the tone controls bypassed, and temporarily use fixed resistors for the other pots (eg a 510R resistor across each pot terminal position)? At least maybe you could get it going before September in straight-through mode anyway.
Yeah, I'm easily over $200 on this project already. All the back-ordered stuff from Newark/element14 always seems to come from Farnell, UK. And they seem to send each piece by airmail and charge my CC!!
I may try firing up the pre before the other pots arrive. That will at least allow for a smoke test and see if it's functioning electrically. I'm pretty meticulous when it comes to putting these 'kits' together, but without the help of the guided Heathkit instructions of yester-year I've been use to, I double check all values before committing the solder. Just getting all the parts together is painful enough - and expensive.
Anyway, I'm anxious to hear what it sounds like once I get it up and running. I don't have any hi-end equipment so this may all be in vain. But I figured you don't want your pre-amp introducing any more noise into the audio stream than possible.
All the back-ordered stuff from Newark/element14 always seems to come from Farnell, UK. And they seem to send each piece by airmail and charge my CC!!I may try firing up the pre before the other pots arrive. That will at least allow for a smoke test and see if it's functioning electrically. I'm pretty meticulous when it comes to putting these 'kits' together, but without the help of the guided Heathkit instructions of yester-year I've been use to, I double check all values before committing the solder.
Just getting all the parts together is painful enough - and expensive. Anyway, I'm anxious to hear what it sounds like once I get it up and running. I don't have any hi-end equipment so this may all be in vain. But I figured you don't want your pre-amp introducing any more noise into the audio stream than possible.
Last edited:
http://www.diyaudio.com/forums/atta...new-doug-self-pre-amp-design-classabuffer.gif
Why the LM317 CCS? This is where a discrete solution is certainly better: wide bandwidth, lower noise, cheaper.
Why the LM317 CCS? This is where a discrete solution is certainly better: wide bandwidth, lower noise, cheaper.
As Nick Fury said in Iron Man 2 "I'll be keeping my eye on you"
I have actually done an imoroved version of the variable-freq tone control since I designed the preamp. Not sure when it might see the light of day...
You have to understand how Elektor works. They chose some of the parts, without any consultation with me. My prototype used parts from Omeg, which were wholly satisfactory and much cheaper.
COG ceramics were used in the prototype and worked fine.(b) the 100p feedback capacitors are all specified as axial polystyrene, which cost about $3 each in small qtys. Did you find it necessary to use these to achieve the specified performance? Having read C. Bateman's capacitor distortion articles I'm wondering if COG ceramic or polypropylene (both <$0.50) could be substituted without affecting performance. (COG Ceramic is listed in the BOM for the '96 preamp). Or was the choice for less technical reasons - eg avoiding audiophile ceramic cap phobia
Using bipolar caps does not reduce capacitor distortion and often increases it. They are used at the inputs and outputs because they may be faced with unpredictable voltages from faulty and ill-conceived equipment without DC-blocking. The important thing is to keep the signal voltage across the capacitor low, say below 80 mVrms
You surely must have noticed that there are four parallel sections to the volume control. The effective value is therefore 1000/4 = 250 Ohm
There is little difference so long as you avoid the Texas version.
Hope that helps.
Bateman's "Capacitor Sound" series measurements show less distortion with bipolar Al electros - different measurement conditions?
"Conclusions.
Having measured a considerable number of aluminium electrolytics using test voltages from 0.1 volt to 3 volt, with and without
bias, a single Bi-polar type produced lower distortion than larger, more expensive, specialist polar capacitors..
Much better results were obtained by connecting two double capacitance value Bi-polar electrolytics in series. Using 1 volt or
smaller test voltages and no bias, distortions for a double Bi-polar and the metallised PET assembly were similar."
Last edited:
Thank you very much for the advice Mr Self. All understood and as expected - has indeed helped to remove my doubts. A couple of things do still bug me though:
I understand that with no signal voltage across a capacitor there is no potential to produce distortion, however your advice still seems slightly at odds with Bateman's in that for eg in Capacitor Sounds Part 6: at 100Hz and 0.1V across a Rubycon 100uf 25V he measured -102.8dB second harmonic (not far above 80mV). Bipolar electros weren't tested until the 0.3V level and I believe the implication is that they do not produce any measureable distortion below this level. In the conclusion he advises that to avoid capacitor distortion the economic solution is to use two bipolars in series for any large cap in the audio signal path provided signal and DC bias voltages are low.
However you're saying that bipolars often produce more distortion - are you able to elaborate or suggest possible reasons for why your experience differs to the results in the article? I'm really hoping you can clarify this futher given the potential for capacitors to affect sound quality is such an ingrained topic for audiophiles at least...
I did of course notice the paralleled AGC sections, but I was wondering why you've used 1k instead of 680R in each section, which would effectievly give even lower noise @ 170R vs 250R?
Many thanks.
I understand that with no signal voltage across a capacitor there is no potential to produce distortion, however your advice still seems slightly at odds with Bateman's in that for eg in Capacitor Sounds Part 6: at 100Hz and 0.1V across a Rubycon 100uf 25V he measured -102.8dB second harmonic (not far above 80mV). Bipolar electros weren't tested until the 0.3V level and I believe the implication is that they do not produce any measureable distortion below this level. In the conclusion he advises that to avoid capacitor distortion the economic solution is to use two bipolars in series for any large cap in the audio signal path provided signal and DC bias voltages are low.
However you're saying that bipolars often produce more distortion - are you able to elaborate or suggest possible reasons for why your experience differs to the results in the article? I'm really hoping you can clarify this futher given the potential for capacitors to affect sound quality is such an ingrained topic for audiophiles at least...
I did of course notice the paralleled AGC sections, but I was wondering why you've used 1k instead of 680R in each section, which would effectievly give even lower noise @ 170R vs 250R?
Many thanks.
Can someone recommend a suitable transformer for the PSU necessary to drive this new pre-amp? To achieve the desired and recommended voltages of +- 17.6VDC, a tranny with secondary outputs of 12.5 - 13.0 would be required. After rectification you would get 17 - 18VDC. I'm assuming that range of voltages would be 'ok' for the pre-amp. Yes? What is not clear, is the amount of current the PSU must supply to the pre-amp. Would this tranny provide enough juice?
Also, does anyone know of an existing PSU pcb kit that would include the necessary rectification, regulation, and filter caps suitable for this pre-amp? Would one of the Rod Elliott design PSU boards work? Are there any exotic PSU designs required for this pre-amp supply?
Also, does anyone know of an existing PSU pcb kit that would include the necessary rectification, regulation, and filter caps suitable for this pre-amp? Would one of the Rod Elliott design PSU boards work? Are there any exotic PSU designs required for this pre-amp supply?
Not as easy a question as it appears.
Well there are 18 dual opamps. Lets assume worse case and also say they "are all" 5532's. Worst case is around 16ma quiescent per package so that gives a requirement of nearly 300ma worst case. In practice it will be considerably lower probably nearer 180ma at a guess but that can not be guaranteed.
Then there is the additional loading when driving a signal around all those low impedances at maximum level. It all adds up.
Maybe Doug has some figures and could advise on that.
As to the supply voltage. What you are describing is an unregulated PSU. The preamp should be fed from a pair of regulators arranged to give the required voltage which means they need a bit of headroom to work with. Typical values would be say -/+ 25 volts DC to feed the regs. That implies an 18-0-18 transformer.
It's a classic PSU set up. The regs will need heatsinking.
If you use a lower DC voltage (the regs and transformer) then the maximum available output from the preamp will be reduced and I suspect absolute levels of distortion may rise slightly.
Last edited:
Thanks Mooly. Yes, I was initially thinking unregulated, but now realize it would need regs to maintain the proper voltages. As for the current demands, I have no idea what would be required, beyond what you quoted, and I guess a worse case scenario should be used. In that case, the larger the tranny the better (in terms of current delivery). I just don't want to buy one yet that's 'extreme' overkill for the project(s).
However, after re-reading the article again last night, and with the newest addition of the MC/MM board to Doug's project, it sounds like this is going to be a multi-pcb project. Therefore, until all the sub-modules are identified, and their respective voltage/current requirements known, building a suitable PSU now would be premature. I have a 4 month wait on the 1k pots anyway, so time is on my side.
However, after re-reading the article again last night, and with the newest addition of the MC/MM board to Doug's project, it sounds like this is going to be a multi-pcb project. Therefore, until all the sub-modules are identified, and their respective voltage/current requirements known, building a suitable PSU now would be premature. I have a 4 month wait on the 1k pots anyway, so time is on my side.