The simplistic Salas low voltage shunt regulator - Page 55

Salas · 17th June 2009, 07:54 AM

If you want to avoid the transformer, try this.

Tham · 17th June 2009, 08:00 AM

Certainly will build one when I receive 2SK170BL that I just ordered. Will a lower/higher Cout make a difference?

Salas · 17th June 2009, 08:36 AM

Its a system thing. If you drive 100k line input, 1uF is OK. If you drive 50k, 2u2 is OK. 3u3 will do well in to 20k. There is a below arm-cart resonance roll off rate that I want to preserve in combination with the interstage coupling. Very low coupling just promotes mechanical feedback if you have large speakers.

Tham · 17th June 2009, 10:10 AM

Thanks Salas for the reply. My line stage is a B1 with a simplified LDR volume control...will try BG 4.7uF at the RIAA output and see if it sounds better.

AndrewT · 17th June 2009, 03:00 PM

remote sensing is the equivalent to extending the PCB all the way to the amplifier.

What have we been saying about instability and it's possible causes?
Did PCB layout get suggested as a possible cause of instability?

Now if we all use remote sensing, what proportion of us are going to come back and say "my SQ has gone all screechy" or "the treble has been exaggerated" or "I'm hearing a lot of sibilance that wasn't there before".

Keep the sensing on board.
Keep the output VERY CLOSE to the amplifier power input. B
A back to back pair of PCBs with a 2mm links soldered between the respective holes.

If you can guarantee that all remote sensing implementations will have no stability concerns then go ahead.

Doing other wise will seriously jeopardise the regulator/amplifier performance.

AndrewT · 17th June 2009, 03:05 PM

is post537 an attempt to add remote sensing to both the +ve supply lead and to the -ve supply lead?

iko · 17th June 2009, 03:24 PM

I don't think I can guarantee any stability with remote sensing or otherwise. To be safe, it's best to have the regulator as close to the load as possible, and connected with thick wire. Do your best to minimize the resistance between Out and Load.

So, for now, maybe we should ignore all the remote sensing stuff. For those that really want to risk or for some reason have no option but leave some distance between the regulator and load, would anyone be interested in exploring the issue? Any expert opinions on the attached circuit? Note that I don't have much experience with this remote sensing stuff.

AndrewT · 17th June 2009, 03:44 PM

Hi Iko,
start by adding various combinations of inductance and capacitance to your sense leads, then simulate.
These reactances will have a lot more effect than a couple of milliohms of resistance in either or both of the sense leads.

BTW,
the Dz to R6 route is part of the sense lead circuit.
The circuit is a measuring bridge. The top and bottom of the bridge must tap into the sense leads and these sense leads must go to the sense locations.

iko · 17th June 2009, 04:14 PM

Andrew, thanks! See attached. Do you think it's sloppy enough to resemble a careless implementation? I'm trying to deliberately do this so we can look not at best case scenarios first. Feel free to suggest different parts values possible closer to reality. The attached circuit is both stable and with performance as good as without all the sensing stuff. In simulation, of course.

BTW, initially I have tried to include the DZ-J2-R6 block in the sense circuit but lost 20dB in output impedance. I don't have an explanation for that.

AndrewT · 17th June 2009, 05:23 PM

connect the extra side of the measuring bridge and then reduce the parasitics to almost zero. The performance should then come back.
If the parasitics are reduced to zero the performance should be identical.

Add some inductance between the CCS and the main shunt element, both +ve and -ve legs.
Add inductance to every wire connection.

Add some capacitance across the load.
Add some capacitance between parts that are very close together.

Set all the capacitances and inductances initially to 1pF and 1nH.

Thread Tools	Search this Thread
Show Printable Version Email this Page	Search this Thread: Advanced Search

New To Site?	Need Help?
Register to Participate Search Privacy Statement Contact Us	Frequently Asked Questions Did you forget your password? Mark Forums Read


	Please consider donating to help us continue to serve you. Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving

17th June 2009, 08:00 AM	#542
Tham diyAudio Member Join Date: Jun 2006	Certainly will build one when I receive 2SK170BL that I just ordered. Will a lower/higher Cout make a difference?

17th June 2009, 08:36 AM	#543
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	Its a system thing. If you drive 100k line input, 1uF is OK. If you drive 50k, 2u2 is OK. 3u3 will do well in to 20k. There is a below arm-cart resonance roll off rate that I want to preserve in combination with the interstage coupling. Very low coupling just promotes mechanical feedback if you have large speakers.

17th June 2009, 10:10 AM	#544
Tham diyAudio Member Join Date: Jun 2006	Thanks Salas for the reply. My line stage is a B1 with a simplified LDR volume control...will try BG 4.7uF at the RIAA output and see if it sounds better.

17th June 2009, 03:00 PM	#545
AndrewT diyAudio Member Join Date: Jul 2004 Location: Scottish Borders	remote sensing is the equivalent to extending the PCB all the way to the amplifier. What have we been saying about instability and it's possible causes? Did PCB layout get suggested as a possible cause of instability? Now if we all use remote sensing, what proportion of us are going to come back and say "my SQ has gone all screechy" or "the treble has been exaggerated" or "I'm hearing a lot of sibilance that wasn't there before". Keep the sensing on board. Keep the output VERY CLOSE to the amplifier power input. B A back to back pair of PCBs with a 2mm links soldered between the respective holes. If you can guarantee that all remote sensing implementations will have no stability concerns then go ahead. Doing other wise will seriously jeopardise the regulator/amplifier performance.

17th June 2009, 03:05 PM	#546
AndrewT diyAudio Member Join Date: Jul 2004 Location: Scottish Borders	is post537 an attempt to add remote sensing to both the +ve supply lead and to the -ve supply lead?

17th June 2009, 03:44 PM	#548
AndrewT diyAudio Member Join Date: Jul 2004 Location: Scottish Borders	Hi Iko, start by adding various combinations of inductance and capacitance to your sense leads, then simulate. These reactances will have a lot more effect than a couple of milliohms of resistance in either or both of the sense leads. BTW, the Dz to R6 route is part of the sense lead circuit. The circuit is a measuring bridge. The top and bottom of the bridge must tap into the sense leads and these sense leads must go to the sense locations.

17th June 2009, 05:23 PM	#550
AndrewT diyAudio Member Join Date: Jul 2004 Location: Scottish Borders	connect the extra side of the measuring bridge and then reduce the parasitics to almost zero. The performance should then come back. If the parasitics are reduced to zero the performance should be identical. Add some inductance between the CCS and the main shunt element, both +ve and -ve legs. Add inductance to every wire connection. Add some capacitance across the load. Add some capacitance between parts that are very close together. Set all the capacitances and inductances initially to 1pF and 1nH.