Salas and everybody,
The delay is for the audio part's output, not for the rail, because its a DC coupled circuit that bumps a bit during switch off mainly. Skip it. The Hypno as a reg only does not utilize it either.
Alright, the relay is out. I thought you guys are trying to avoid some instability at startup of the regulator itself. The S/J can be a bit bouncy at startup, for instance, using a timer with a few seconds delay doesn't sound too unreasonable to me. If there's no transitory aspects to worry about, it's out.
How long it will be from load? You could utilize remote sensing.
Are you talking about a 'current sense' type of deal (Jung terminology)? Can you tell me more about that, especially in the context of your shunt reg? I mean, I get the theory (the Sjostrom uses this too), but I both wonder how effective it is (it sounds a bit overkill), and if it applies to this design. Excellent suggestion, I can't wait for more discussion on this.
The length of the supply wires are in the range of 3-5inch. I have everything in a HTPC enclosure.
For best cap distortion that is, use NP caps across the Vref and output.
For instance, across the 5 LEDs?
Mini chokes can filter where the CCS stops still, they are elegant, cheap
I have some right now, in the units of uHs range. They isolate between the various stages of the DAC (clock, USB receiver, DAC, etc.).
Thank you all.
aR
The delay is for the audio part's output, not for the rail, because its a DC coupled circuit that bumps a bit during switch off mainly. Skip it. The Hypno as a reg only does not utilize it either.
Alright, the relay is out. I thought you guys are trying to avoid some instability at startup of the regulator itself. The S/J can be a bit bouncy at startup, for instance, using a timer with a few seconds delay doesn't sound too unreasonable to me. If there's no transitory aspects to worry about, it's out.
How long it will be from load? You could utilize remote sensing.
Are you talking about a 'current sense' type of deal (Jung terminology)? Can you tell me more about that, especially in the context of your shunt reg? I mean, I get the theory (the Sjostrom uses this too), but I both wonder how effective it is (it sounds a bit overkill), and if it applies to this design. Excellent suggestion, I can't wait for more discussion on this.
The length of the supply wires are in the range of 3-5inch. I have everything in a HTPC enclosure.
For best cap distortion that is, use NP caps across the Vref and output.
For instance, across the 5 LEDs?
Mini chokes can filter where the CCS stops still, they are elegant, cheap
I have some right now, in the units of uHs range. They isolate between the various stages of the DAC (clock, USB receiver, DAC, etc.).
Thank you all.
aR
Second Issue PCB (no timer delay, sense outputs)
Hi all,
Here's my second take on the single rail shunt PS PCB. The relay is out, and per Salas' suggestion, I put in 'current sense' outputs ('remote sensing'), intended to provide compensation for losses in cables from regulator to load.
Again - please comment.
Thank you,
Radu.
Hi all,
Here's my second take on the single rail shunt PS PCB. The relay is out, and per Salas' suggestion, I put in 'current sense' outputs ('remote sensing'), intended to provide compensation for losses in cables from regulator to load.
Again - please comment.
Thank you,
Radu.
That one will play better. The way you arranged your components, gnd is bus bar like, you will get no problems from there. Rs is a strong asset too. On a glimpse it looks correct but don't take my word on it. Review the connections many times against my schematic before you etch it. What kind of software is that BTW? I like the ''construction blueprint'' style presentation.
Great - thanks man. This is full-featured AutoCAD. It's gotta look blueprint like, right?... Only 3D design software / BIM gets beyond this, I think (like Revit).
The overlapped (graphically) parts are due to using 'blocks' (CAD lingua) for various parts I like and use (for instance, 5mm F lytics 16mm dia, such as 100uF cerafines). I wasted a lot of time creating these blocks, but you only do them once. They'll be good for use further. They overlap because the board wants to be customizable for 5/7.5mm lytics, for instance. Same for chokes, they can be through-hole discrete parts, or jumbo/core/coil bobbins.
Tried to keep all heat sources as far from each other as I could. Rset - which is R1 on the schematic, right? - is as far from the HS as possible while keeping the traces short. I will have questions on adjusting this after assembly.
Regarding reviewing the connections - I found a mistake on the first issue which is now corrected (couplah JFET legs wrongly connected kind of thing, easy to have your eyes go divergent after some hours of focus). I plan to take a very close look at all traces and compare with the schematic before I'll call anything final.
Have you guys had luck with some particular PCB creation method? I'm also thinking artwork on the back of the PCB, silkscreen stuff. I've seen some guides, but I don't know if toner transfer does color correctly. I've only used photo-presentisized boards this far myself, but they're rather approximate as far as final product is concerned. Am looking at toner transfer, and some local suppliers are of interest (Express PCB, Pulsar, my own homebrewing machine).
Let me know. This could get somewhere.
aR
The overlapped (graphically) parts are due to using 'blocks' (CAD lingua) for various parts I like and use (for instance, 5mm F lytics 16mm dia, such as 100uF cerafines). I wasted a lot of time creating these blocks, but you only do them once. They'll be good for use further. They overlap because the board wants to be customizable for 5/7.5mm lytics, for instance. Same for chokes, they can be through-hole discrete parts, or jumbo/core/coil bobbins.
Tried to keep all heat sources as far from each other as I could. Rset - which is R1 on the schematic, right? - is as far from the HS as possible while keeping the traces short. I will have questions on adjusting this after assembly.
Regarding reviewing the connections - I found a mistake on the first issue which is now corrected (couplah JFET legs wrongly connected kind of thing, easy to have your eyes go divergent after some hours of focus). I plan to take a very close look at all traces and compare with the schematic before I'll call anything final.
Have you guys had luck with some particular PCB creation method? I'm also thinking artwork on the back of the PCB, silkscreen stuff. I've seen some guides, but I don't know if toner transfer does color correctly. I've only used photo-presentisized boards this far myself, but they're rather approximate as far as final product is concerned. Am looking at toner transfer, and some local suppliers are of interest (Express PCB, Pulsar, my own homebrewing machine).
Let me know. This could get somewhere.
aR
Please to be sure could you clarify me, if I understand well I have to put two resistors 33R or 22R in parallel for R1, right?
A photo:
So with 1 PCB I can get +-24V?
An externally hosted image should be here but it was not working when we last tested it.
So with 1 PCB I can get +-24V?
coupla questions
Howdy everybody,
I got a couple of questions I'd appreciate some input on.
1. Does anybody have info on a TO-247AC heatsink that would be under 2.7C/W? The largest (individual) TO-247 HS I could find is the Aavid Thermalloy 6400BG (which is the 2.7C/W I have in mind). This would allow, say, about 18W dissipation in my application (pretty warm environment)... This is for the single rail 'pseudo-hypnotize' I've been working on. The idea is to make the board as versatile as possible. 2.7C/W doesn't sound particularly versatile to me. Ah yes - if possible, easy to source in the US.
2. Has anybody else looked at the PCB? Salas was telling me he's a P2P kinda guy - are there some experienced PCB designers here that could give me some feedback? Salas - maybe at other threads, but sharing interest for the simplistic shunt regulator?...
Best,
Radu.
Howdy everybody,
I got a couple of questions I'd appreciate some input on.
1. Does anybody have info on a TO-247AC heatsink that would be under 2.7C/W? The largest (individual) TO-247 HS I could find is the Aavid Thermalloy 6400BG (which is the 2.7C/W I have in mind). This would allow, say, about 18W dissipation in my application (pretty warm environment)... This is for the single rail 'pseudo-hypnotize' I've been working on. The idea is to make the board as versatile as possible. 2.7C/W doesn't sound particularly versatile to me. Ah yes - if possible, easy to source in the US.
2. Has anybody else looked at the PCB? Salas was telling me he's a P2P kinda guy - are there some experienced PCB designers here that could give me some feedback? Salas - maybe at other threads, but sharing interest for the simplistic shunt regulator?...
Best,
Radu.
Oooops, looking for RCA connector for RIAA I get these nice caps suitables for C5 Shunt reg. PSU
BG WK Power Tank 220uF/200V, wow, can be suitables?
An externally hosted image should be here but it was not working when we last tested it.
BG WK Power Tank 220uF/200V, wow, can be suitables?
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.