Thanks Jan, that's what I hoped this one would do as I've found some regulators appear to have very little 'insulation' in this regard. Whilst most systems will be regulated apart from the amplifiers output stage, the mains supply changes are not coming just from there. Plug any (so far in my limited experience) item of fully regulated Hi-Fi into the wall socket then into the super spur and hear the vast difference. A discrete regulator I have now on the headphone driver was on the old CD player DAC chip / analogue stage; it's the three transistor item used by Philips and others in their better CD Player analogue stages. It came with some Black Gates before, in and after it, I added a bank of Nichicons best 1000uF caps before it and tried various transformers. Everything altered the eventual loud speaker sound buy a large amount, the current drawn was only about 100mA for +ve and a lot less for -ve and the scope showed no ripple at all after the regulator. Even getting rid of the mains leads then the IEC connectors made a large difference when I was able to fit a plug to the primaries wires and plug into the spur. With an AV pre-amp / processor (no power amps just line outs) I had, just moving it 3 feet up the chain from a trailing block lead to the very thick lead it was plugged into, that came direct from the spur, improved it's sound. It's not just about better earthing as some items don't have an earth wire.
When I was having a read though some of the Super Teddy Reg threads the other year, there were comments that it did result in no longer a need for any fancy elaborate components prior to it, then as I was about to buy the parts, Toshiba discontinued the specified 'through hole' FETs and I didn't find alternatives and anyway the need went away for a while.
Which Jung /Didden super reg is the one tested in the Linear Audio Super Regulator Article?
I have a VALab 4395 DAC here with six 'Gold Ref Regs' in, all using the LME49710. Can any of you tell me how these compare in sound to the Jung / Didden one talked about in this thread, or the one in the Linear Audio graphs?
Also, if I have this 2000 Jung / Didden reg up very close to the load, will the many film caps across the op-amps, etc, then have a negative effect on the regulator?
Also, are these new boards only available from the USA? I presume they are made there or in China and shipped to the USA DIY Audio shop keeper? So none available in the EC?
Thanks.
Last edited:
I might like to use just one of these, components set appropriately for 0 to 30V, spanning across both rails of the fully balanced headphone amp, because it's 'balanced', ie, two parallel amplifiers, no earth reference. The current path through the load is strait from one voltage rail to the other, the output capacitors one plate, through the load, to it's other plate. Supply effectively ± 15V.
Can you chaps suggest any reason why would not work?
Will I need a more beefy out put cap? The amp currently has polymer caps across it, and black gates, and tiny 10uF multilayer polyester caps on the pins of the output chips, so a very low impedance, maybe 10mOhms, will these need to be removed or will the regulator be OK?
Thanks.
Can you chaps suggest any reason why would not work?
Will I need a more beefy out put cap? The amp currently has polymer caps across it, and black gates, and tiny 10uF multilayer polyester caps on the pins of the output chips, so a very low impedance, maybe 10mOhms, will these need to be removed or will the regulator be OK?
Thanks.
Vin, I think you missed this: http://www.diyaudio.com/forums/diyaudio-store/247281-super-regulator-8.html#post3756870
Jan
Jan
Dag Jan,
Is it possible to use this regulator in separate box, feeding a DAC via a quite long 30/40 cm cable? I've got an Antelope DAC, and this box needs 18V/1A DC. I would like to replace the supplied stock SMPS with a linear one. And is it possible connect the SReg sensing input to the end of this cable (connector on the DAC side) without making things even worse due to the also long sensing wires, picking up noise?
Eddy
Is it possible to use this regulator in separate box, feeding a DAC via a quite long 30/40 cm cable? I've got an Antelope DAC, and this box needs 18V/1A DC. I would like to replace the supplied stock SMPS with a linear one. And is it possible connect the SReg sensing input to the end of this cable (connector on the DAC side) without making things even worse due to the also long sensing wires, picking up noise?
Eddy
I think one super regulator will create lower output impedance also up in frequency and lower noise.
Dag Eddy,
Yes you can do that, be sure to use a screenend cable for the sense. Screen connect to ground at the reg PCB. It also is a good idea to connect the output sense line, coming from the load, at the PCB via something like 100 ohms, and put something like 100pF from the point where the 100 ohms connect to the PCB to ground. With such a long cable the inductance can cause some stablility issues and with the hf filter that is cured.
I'm thinking to put this option on the next batch of the PCBs as well.
Jan
Thanks Jan, it shall be a nice winter project. So, circuit will be like attached picture? Can you advice me for the buffer cap? Is 5000uF enough? This gives 2V ripple @ 1A. If my math is correct.. For the transfo AC voltage I get: (18v DC output + 2v regdrop+2v ripple+1.5v rectifier) / 1.41 = about 17V AC?
Eddy
An externally hosted image should be here but it was not working when we last tested it.
You only need the 100 ohms + 100pF in the load sense line, not in the gnd line.
The transfo values look OK, but don't forget that the line voltage can drop as much as 10% during a day.
To be safe I'd go with a 20VAC xformer, preferably center-tapped so you only lose one diode drop.
Jan
Thanks Jan,
I need 1A ouput current, with a hfe of 40 for the series transistor, 25mA is needed to get this output current. Is it ok to increase the current source from now about 8mA (I guess) to 25mA by reducing the 249R resistor? The opamp must sink 25mA but seems to be within spec. Or is it better to use a darlington power transistor for Q1?
Eddy
I need 1A ouput current, with a hfe of 40 for the series transistor, 25mA is needed to get this output current. Is it ok to increase the current source from now about 8mA (I guess) to 25mA by reducing the 249R resistor? The opamp must sink 25mA but seems to be within spec. Or is it better to use a darlington power transistor for Q1?
Eddy
I would just look for a pass device with higher Hfe - 40 seems quite low. Is that a minimum spec or have you measured it?
Jan
Sorry Ian, missed this as the notification mail ended up in my spam folder.
I would just use one reg, the pos one, set to 30V. Always better than having two in series, plus you have the neg one for another project or as a gift
Jan
It was the minimum stated in the D44H11 datasheet table. But checked it again, this hfe 40 is at max current trough the transistor. At 1A it is still about 200, so no problem!
It was the Jung Regulator on the boards Jan designed with the LM317/337 pre-regulators. AD825AR was the error amplifier.