The simplistic Salas low voltage shunt regulator - Page 53

Salas · 16th June 2009, 07:54 PM

Sensing returns are used (syn08 and others) with success. Its valid. Can you show it graphically for those in need of long leads?

iko · 16th June 2009, 08:17 PM

Have they (others and syn08) discussed the changes when sense is routed this way?

BTW, I made a mistake above, saying that 20cm of cable would be 2mOhms, when it is actually 4mOhms.

Here is the output impedance of v1 as I always simulate it, without a resistance between OUT and Load.

A resistance of 5 milliohms, sense components connected as usually shown in the schematic.

And sense components connected directly to the load.

I think people in general may not realize what a big difference this can make. I didn't until I actually simulated it.

Salas · 16th June 2009, 08:30 PM

OK with the difference graphs, but I actually meant schematically of how to practically connect.

iko · 16th June 2009, 08:56 PM

Somebody with more graphical talent please fix this.

Salas · 16th June 2009, 09:12 PM

Is this OK?

iko · 16th June 2009, 09:31 PM

Not really: source of Q4 does not connect directly to drain of jfet J3 but it connects directly only to drain of Q1 and OUT; otherwise it is ok. I have to run now, but I'll come back later and try a proper drawing, sorry about the misunderstanding.

Salas · 16th June 2009, 09:51 PM

Better?

iko · 16th June 2009, 10:13 PM

Perfect! Thanks. The simulation shows that output impedance is almost not affected at all even by quite a resistance between OUT and LOAD, where as before even 200uOhms would have an effect.

In theory I knew that sense and main current paths should be separate, but didn't think that it would make such a difference in this case. I will test it in practice and see if anything can be observed empirically.

Salas · 16th June 2009, 10:16 PM

You use long leads so you will probably hear better bass, differentiation and depth. At least that happened when I got it from test phase long cable down to 5cm 1mm2 solid core next to the audio circuit.

iko · 16th June 2009, 10:54 PM

You mean that a shorter cable/distance between regulator output and next stage/load improves bass, differentiation, and depth?

OK, so I just tested my theory with a load of 80mA DC + 30mA AC.

Case 1: all sensing components connected directly to drain of Q1 and source of Q4. 20cm cable connecting source of Q4 to a 80mA DC 30mA AC load.

a) scope showed about 200uV ripple on source of Q4
b) scope showed about 2mV ripple on load

Case 2: disconnected all sensing components from source of Q4 and connected them directly to the load with a separate cable of about 20cm.

a) scope showed about 200uV ripple on load
b) scope showed about 2mV ripple on source of Q4

We're talking about a 20cm wire, not even very thin, soldered on both sides! It would be good if someone else double checked this.

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

16th June 2009, 07:54 PM	#521
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	Sensing returns are used (syn08 and others) with success. Its valid. Can you show it graphically for those in need of long leads?

16th June 2009, 08:17 PM	#522
iko diyAudio Moderator Join Date: May 2008 Location: Toronto	Have they (others and syn08) discussed the changes when sense is routed this way? BTW, I made a mistake above, saying that 20cm of cable would be 2mOhms, when it is actually 4mOhms. Here is the output impedance of v1 as I always simulate it, without a resistance between OUT and Load. A resistance of 5 milliohms, sense components connected as usually shown in the schematic. And sense components connected directly to the load. I think people in general may not realize what a big difference this can make. I didn't until I actually simulated it.

16th June 2009, 08:30 PM	#523
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	OK with the difference graphs, but I actually meant schematically of how to practically connect.

16th June 2009, 08:56 PM	#524
iko diyAudio Moderator Join Date: May 2008 Location: Toronto	Somebody with more graphical talent please fix this.

16th June 2009, 09:31 PM	#526
iko diyAudio Moderator Join Date: May 2008 Location: Toronto	Not really: source of Q4 does not connect directly to drain of jfet J3 but it connects directly only to drain of Q1 and OUT; otherwise it is ok. I have to run now, but I'll come back later and try a proper drawing, sorry about the misunderstanding.

16th June 2009, 10:13 PM	#528
iko diyAudio Moderator Join Date: May 2008 Location: Toronto	Perfect! Thanks. The simulation shows that output impedance is almost not affected at all even by quite a resistance between OUT and LOAD, where as before even 200uOhms would have an effect. In theory I knew that sense and main current paths should be separate, but didn't think that it would make such a difference in this case. I will test it in practice and see if anything can be observed empirically.

16th June 2009, 10:16 PM	#529
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	You use long leads so you will probably hear better bass, differentiation and depth. At least that happened when I got it from test phase long cable down to 5cm 1mm2 solid core next to the audio circuit.

16th June 2009, 10:54 PM	#530
iko diyAudio Moderator Join Date: May 2008 Location: Toronto	You mean that a shorter cable/distance between regulator output and next stage/load improves bass, differentiation, and depth? OK, so I just tested my theory with a load of 80mA DC + 30mA AC. Case 1: all sensing components connected directly to drain of Q1 and source of Q4. 20cm cable connecting source of Q4 to a 80mA DC 30mA AC load. a) scope showed about 200uV ripple on source of Q4 b) scope showed about 2mV ripple on load Case 2: disconnected all sensing components from source of Q4 and connected them directly to the load with a separate cable of about 20cm. a) scope showed about 200uV ripple on load b) scope showed about 2mV ripple on source of Q4 We're talking about a 20cm wire, not even very thin, soldered on both sides! It would be good if someone else double checked this.