Super Regulator

[jan.didden]

This is correct - I have forwarded the files for the latest version to diyaudio but they have not implemented them yet.
Will ping them again.

The new files are downloadable from the bottom of this page: The Superregs | Linear Audio NL

Jan

 

[albertNL]

Hi, Jan.

I had some trouble unpacking the zipx file (my Windows laptop broke down yesterday) and cannot open some of the files. Not problem I think, as the main file
"Superregs for your signal level projects V2.2a.pdf"
is fine. I will re-try the other files when I get my laptop fixed.

Thanks,

Albert

 

[CeeVee]

..you need winzip 12 or later....present version is 19 i think.

Winrar will not do.

 

[peranders]

First time for this icon

Just wondering, do you prefer zipx of some reason? It's not standard even on Windows 10 and at not at Mac either (on Maverick at least, haven't checked on Yosemite).

 

[jan.didden]

Just wondering, do you prefer zipx of some reason? It's not standard even on Windows 10 and at not at Mac either (on Maverick at least, haven't checked on Yosemite).

I really don't now, I just used the zip command in the Word context menu. It appears that there are TWO versions, I must figure out what selects which.

Jan

 

[pchw]

You can download 7-zip, it can open zipx.

 

[yilingcheng]

does the LED D1 suppose to glow?

or is it just for some form of protection...

another question:

if i apply 12.6 volts dc on J1, i'm getting between 11.27 <--> 11.34 DC on J2(1)
than i get 6.59 DC on J2(2) and 6.59 DC on J2(3) to GND J2(4)

seems J2(1) is not regulated where J2(2) and J2(3) are regulated...

i'm just a beginner and i don't understand why so many output Vout and Vout sense

is the output to use is J2(2) + to J2(4) - ?

(i'm just trying to replace a ULNPS-HC4)

thanks

Martin Gagnon

 

[Triodethom]

I really don't now, I just used the zip command in the Word context menu. It appears that there are TWO versions, I must figure out what selects which.

Jan

winzip also seams to work.

 

[yilingcheng]

oh i just read post #253

i need to sum Vout + Vout sense together, same with return with return sense

my bad

Martin

 

[yilingcheng]

finally got it to work nicely with an adjustable pre regulator in the form of a LT1083 from 5 to 12 volts dc,

not so much luck with a passive bridge rectification (may be i have to try bigger cap) the bridge is too noisy.

i change the D1 twice, it has 1 volt across it but i won't light up...

i will be building a tube pre in the future and it needs -24 volt dc for some form of drainage...

and i will use one on every tube for heather (4 x 6.3 volts 600 mA)

pretty cool regulator can't wait to test it on something else than a router...

Martin

 

[johngalt47]

Are D2 and D7 just 6.8 volt zeners?

 

[yilingcheng]

yes, check post #235

you adjust the zener to your voltage need...

 

[B&W_arthur]

This is correct - I have forwarded the files for the latest version to diyaudio but they have not implemented them yet.
Will ping them again.

The new files are downloadable from the bottom of this page: The Superregs | Linear Audio NL

Jan

In the Version2.2, the C9, D5 and R5 are connected differently, compared with from the previous version.

Any benefits of such modification?

 

[peranders]

Seems to be a mistake? I assume that the schematic is identical to the pcb layout.

 

[HiFiNutNut]

Do you need to load the regulator with some current to have reasonable output impedance? If so, what is the minimum load current to ensure reasonable output impedance?

 

[jackinnj]

Do you need to load the regulator with some current to have reasonable output impedance? If so, what is the minimum load current to ensure reasonable output impedance?

Zout is linearly and inversely related to Iout.

 

[HiFiNutNut]

Zout is linearly and inversely related to Iout.

Jackinnj,

Thanks. That is what I expected. What current was used in your published measurement?

Let us assume that the output voltage is 15V, the regulator using 1k for R7 and 1174k for R6 induces 6.8mA current when there is no load. If the load consists of a couple of opamps, the opamps may draw 12mA most of the time. That may not be enough to make the Supereg having super low output impedance.

Should we put some dummy load at the output or lower the R7 and R6 in order to lower the output impedance?

Do you have a graph showing output impedance vs frequency vs current load? That would be great.

Regards,
Bill

 

[HiFiNutNut]

Seems to be a mistake? I assume that the schematic is identical to the pcb layout.

On the PCB, C9 is connected to ground, not ground sense. R5 is connected to Vout, not Vsense. So the PCB is identical to the schematic.

If you suggest that C9 should be connected to ground sense and R5 connected to Vsense, that is a different story.

 

[peranders]

The original idea was to have the whole reference and feedback circuits floating and to be connected to SENSE.

 

[jackinnj]

Thanks. That is what I expected. What current was used in your published measurement?

As Walt configured the original regulators for 13.7V -- the load was 100R||120uF.

The opamp in the SR is an important part of the discussion -- the faster the better. But the AD797 will give you heartburn. LME49710 worked fine, but LT1028 didn't.

One of the things which a poorly designed regulated power supply will do is put harmonics back onto the power supply rails -- Walt's SR doesn't have this problem.