Simplistic NJFET RIAA - Page 97

iko · 8th February 2009, 02:59 AM

This one simulated nicely. I need to look at it more closely to check that the simulation isn't BS, and I'll probably build a prototype for comparison, if sanity checks hold.

Salas · 8th February 2009, 03:04 AM

Nice one. We will call it ''Salas Hybrid - Iko Push''

iko · 8th February 2009, 03:13 AM

We'll see if it works in reality

but it sure looks promising.

The other thing I was playing with, but didn't get far, was meant to replace the 20 something V zener with some other reference. The higher voltage zeners are, apparently, more noisy than the lower voltage ones. Something based on the lm369 or some other low noise buried zener reference. We'd probably end up with a super nice solution. Not that we need anything better than what we've got already, but hey, what the hey.

Salas · 8th February 2009, 03:16 AM

The real goal is to really feel subjectively some more resolution and dynamics. If not, we will spare the additions. We dig the bottom now really...

Devil will find work for idle hands to do

Salas · 8th February 2009, 03:27 AM

Quote:

Originally posted by ikoflexer
We'll see if it works in reality but it sure looks promising.

The other thing I was playing with, but didn't get far, was meant to replace the 20 something V zener with some other reference. The higher voltage zeners are, apparently, more noisy than the lower voltage ones. Something based on the lm369 or some other low noise buried zener reference. We'd probably end up with a super nice solution. Not that we need anything better than what we've got already, but hey, what the hey.

Remember to match J2, J3 when prototyping. Also the Zener is well bypassed and CCSed so its self noise is pushed down in practice. Don't make it more elaborate there IMO. What we did by now is to idealize the npn driver bjt which bares logic.

iko · 8th February 2009, 04:49 AM

By the time I saw your reply I already built it. I didn't pay attention this time to matching j2 j3, just wanted to put something together quickly.

Attached is a trace. The bottom trace is the input to the regulator, a 100 mV p-p voltage, the top one is the regulator output, shown on 5mV/div on the scope, the tightest so far I got from any version regulator I tried. Don't know how it sounds.

I've replaced Q1 with a 2n6387 (darlington

) because q1 was getting hot and I didn't want to search around too much.

I'm sure this circuit can be done even better than my aerial construction and probably the results will be even better.

We might have hit a small jackpot here. If only I had one of those cards to run some real psrr measurements on it... would be neat. I think I might get this, what do you think:
http://cgi.ebay.ca/ws/eBayISAPI.dll?...link:middle:ca

iko · 8th February 2009, 05:02 AM

and forgot the attachment...

Salas · 8th February 2009, 11:16 AM

The card is good to get. To tell you the truth I haven't measured any 100Hz ripple spike on FFT with simpler shunt versions on the phono output as well, always buried in the grass, but the buffer to the V MOSFET can not hurt. Where is your measuring point? At dummy load or on phono's B+ including cable? How much current on load if with dummy? The key is the sound. To really hear something better. Try with headphones, you may discern and tell us. Only let it run enough before critical listening (2 days powering), I know that it smooths out in a day or two when from new, by experience. To preserve lower impedance gains, the output cap must be very low ESR (bypass with 0.1uF film), and the twisted cabling to phono B+ point and GND point, speaker grade thick. Congratulations for your results! Refreshing.

Salas · 8th February 2009, 11:53 AM

P.S. Good to know that this shunt won't oscillate even with a Darlington driver. Many PSUs are quite neurotic to manipulate. This one, even basically being my first Vbe CCS one with a buffer addition, seems rock solid.

RCruz · 8th February 2009, 01:56 PM

Hi Salas

Just finished measuring the jfets.

I arranged them in lots with values from 10.9 mA to 8.4 mA.

I also have smaller values like 6.4 6.5 6.7 6.8 (without pairs)

What avlues should I use for eac place in the RIAA ?

(Also matched the BC550C using the DMM hFE)

Ricardo

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

8th February 2009, 03:04 AM	#962
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	Nice one. We will call it ''Salas Hybrid - Iko Push''

8th February 2009, 03:13 AM	#963
iko diyAudio Moderator Join Date: May 2008 Location: Toronto	We'll see if it works in reality but it sure looks promising. The other thing I was playing with, but didn't get far, was meant to replace the 20 something V zener with some other reference. The higher voltage zeners are, apparently, more noisy than the lower voltage ones. Something based on the lm369 or some other low noise buried zener reference. We'd probably end up with a super nice solution. Not that we need anything better than what we've got already, but hey, what the hey.

8th February 2009, 03:16 AM	#964
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	The real goal is to really feel subjectively some more resolution and dynamics. If not, we will spare the additions. We dig the bottom now really... Devil will find work for idle hands to do

8th February 2009, 04:49 AM	#966
iko diyAudio Moderator Join Date: May 2008 Location: Toronto	By the time I saw your reply I already built it. I didn't pay attention this time to matching j2 j3, just wanted to put something together quickly. Attached is a trace. The bottom trace is the input to the regulator, a 100 mV p-p voltage, the top one is the regulator output, shown on 5mV/div on the scope, the tightest so far I got from any version regulator I tried. Don't know how it sounds. I've replaced Q1 with a 2n6387 (darlington ) because q1 was getting hot and I didn't want to search around too much. I'm sure this circuit can be done even better than my aerial construction and probably the results will be even better. We might have hit a small jackpot here. If only I had one of those cards to run some real psrr measurements on it... would be neat. I think I might get this, what do you think: http://cgi.ebay.ca/ws/eBayISAPI.dll?...link:middle:ca

8th February 2009, 11:16 AM	#968
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	The card is good to get. To tell you the truth I haven't measured any 100Hz ripple spike on FFT with simpler shunt versions on the phono output as well, always buried in the grass, but the buffer to the V MOSFET can not hurt. Where is your measuring point? At dummy load or on phono's B+ including cable? How much current on load if with dummy? The key is the sound. To really hear something better. Try with headphones, you may discern and tell us. Only let it run enough before critical listening (2 days powering), I know that it smooths out in a day or two when from new, by experience. To preserve lower impedance gains, the output cap must be very low ESR (bypass with 0.1uF film), and the twisted cabling to phono B+ point and GND point, speaker grade thick. Congratulations for your results! Refreshing.

8th February 2009, 11:53 AM	#969
Salas diyAudio Chief Moderator Join Date: Oct 2002 Location: Athens-Greece	P.S. Good to know that this shunt won't oscillate even with a Darlington driver. Many PSUs are quite neurotic to manipulate. This one, even basically being my first Vbe CCS one with a buffer addition, seems rock solid.