Simplistic NJFET RIAA - Page 699

RCruz · 20th November 2011, 10:04 PM

I am using 30kr for R Vref so I use 10uF for Vref cap. Use the best you can get because it determines spatiality very much, as well as tone.

This cap can be heard

I like AmpOhm, I find Obbligato Gold much better, maybe a Supreme might be even better.

kstlfido · 20th November 2011, 10:42 PM

Ricard0-

It's an IRC RN60C 121 ohm, 1%. Metal film. Also on resistor- MEA and C6910T2.

You have a preferred value and type?

kstlfido · 21st November 2011, 12:37 AM

CCS mosfet gate stopper (R2), that is. 121 ohm. Metal film RN60C. Is carbon comp better?

Salas · 21st November 2011, 01:36 AM

Yes theoretically if not a noisy sample. I haven't got oscillations in this CCT with MF though.

RCruz · 21st November 2011, 06:35 PM

It works ok with MF or CF but I keep having startup issues with some of my builds.

I found that there is an inductive interaction that sometimes happens, so the shunt only starts if I touch the connection between R2 and Q2.

It is not common and I already built 14 V12R without any issues. Once it happens I need to redo R2 placement until I get it working perfectly.

I am not sure what causes this behaviour and initially I believed it was related with resistor type or value but now I am not sure.

Once I experimented with 140r instead of 120r and it did not solve the problem. maybe someone experimented with lower values (80r ?)

I am very interested in your layout that is quite different from mine so I would like to know if you had any similar issues.

Salas · 21st November 2011, 07:52 PM

Its because the cascode is referenced to the regulated output voltage. That gives it superior PSRR but if there is a slight sync delay it can block the procedure. Well worth it to retain as a feature design wise because it works in the end.

RCruz · 08:53 PM

Would you please explain (in human language)...

Cascode.. what parts are involved in this cascode ? (Q3 Q4) ?

Salas · 21st November 2011, 09:06 PM

Q3,Q4, associated bias resistors and LED. If the output won't catch up fast due to layout parasitics, it won't bias up the cascode to ignite the CCS well. That is why even a non optimum layout can kick start with a probe touch. Also some transistors can have bit higher capacitance than the next sample.

kstlfido · 22nd November 2011, 05:27 AM

I Salas (et al)-

I wired up my 1.2R shunt and fed it power. Loaded output with 1kohm, as you instructed.

Results-

Input- 44.0vdc

Output-
Varied from 37vdc to 42vdc in 1 v steps.
Always consumed 188.3 to 188.6mA (after a little settling time).

IRFP9240 and load R were warm to the touch. IRF9610 was lukewarm.

At 43.5v, current starts to drop.
At 44.0v, 113.9mA
At 44.6v, current bottoms out at 56.9mA.

Scope observations-

Tek 20 mHz scope. I am limited as I only have 100:1 probe, so max sensitivity is only 500mA/division.

No sign of oscillations at the output, across load resistor at any frequency.

So, is 500mA/div sensitive enough?

I'm traveling for the US holiday this week, will be back next week to continue working on the RIAA. Pics of shunt so far, below. Temporary power wiring and sinks.

Cheers! -Kent

AndrewT · 22nd November 2011, 10:33 AM

Quote:

Originally Posted by kstlfido

Input- 44.0vdc

Output-
Varied from 37vdc to 42vdc in 1 v steps.
Always consumed 188.3 to 188.6mA (after a little settling time).

IRFP9240 and load R were warm to the touch. IRF9610 was lukewarm.

At 43.5v, current starts to drop.
At 44.0v, 113.9mA
At 44.6v, current bottoms out at 56.9mA.

check the voltage across the current setting resistor in the CCS.
When Vin-Vout >~5V, you should find that the resistor Vdrop is substantially constant.

When Vin-Vout <~5V you will find that resistor Vdrop is falling rapidly as Vout closes in towards Vin.
The Shunt is reducing it's shunt current to try to maintain Vout. Effectively your expensive CCS+Shunt regulator is not being allowed to do it's job properly.
The earlier Salas Regs had a recommendation of Vin-Vout >7Vdc for all operational conditions.
What is the recommendation for V1.2?

20th November 2011, 10:04 PM	#6981
RCruz diyAudio Member Join Date: Sep 2006 Location: Saxon Blog Entries: 1	I am using 30kr for R Vref so I use 10uF for Vref cap. Use the best you can get because it determines spatiality very much, as well as tone. This cap can be heard I like AmpOhm, I find Obbligato Gold much better, maybe a Supreme might be even better. __________________ RC

21st November 2011, 06:35 PM	#6985
RCruz diyAudio Member Join Date: Sep 2006 Location: Saxon Blog Entries: 1	It works ok with MF or CF but I keep having startup issues with some of my builds. I found that there is an inductive interaction that sometimes happens, so the shunt only starts if I touch the connection between R2 and Q2. It is not common and I already built 14 V12R without any issues. Once it happens I need to redo R2 placement until I get it working perfectly. I am not sure what causes this behaviour and initially I believed it was related with resistor type or value but now I am not sure. Once I experimented with 140r instead of 120r and it did not solve the problem. maybe someone experimented with lower values (80r ?) I am very interested in your layout that is quite different from mine so I would like to know if you had any similar issues. __________________ RC

21st November 2011, 08:46 PM	#6987
RCruz diyAudio Member Join Date: Sep 2006 Location: Saxon Blog Entries: 1	Would you please explain (in human language)... Cascode.. what parts are involved in this cascode ? (Q3 Q4) ? __________________ RC Last edited by RCruz; 21st November 2011 at 08:53 PM.

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

20th November 2011, 10:42 PM	#6982
kstlfido diyAudio Member Join Date: Mar 2006 Location: Berkeley, CA	Ricard0- It's an IRC RN60C 121 ohm, 1%. Metal film. Also on resistor- MEA and C6910T2. You have a preferred value and type?

21st November 2011, 12:37 AM	#6983
kstlfido diyAudio Member Join Date: Mar 2006 Location: Berkeley, CA	CCS mosfet gate stopper (R2), that is. 121 ohm. Metal film RN60C. Is carbon comp better?

21st November 2011, 01:36 AM	#6984
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	Yes theoretically if not a noisy sample. I haven't got oscillations in this CCT with MF though.

21st November 2011, 07:52 PM	#6986
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	Its because the cascode is referenced to the regulated output voltage. That gives it superior PSRR but if there is a slight sync delay it can block the procedure. Well worth it to retain as a feature design wise because it works in the end.

21st November 2011, 09:06 PM	#6988
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	Q3,Q4, associated bias resistors and LED. If the output won't catch up fast due to layout parasitics, it won't bias up the cascode to ignite the CCS well. That is why even a non optimum layout can kick start with a probe touch. Also some transistors can have bit higher capacitance than the next sample.