Simplistic NJFET RIAA

Jump to Latest
merlin el mago

merlin el mago

Member
Joined 2009
Paid Member
Yes the tone has changed a lot of, now is more full, rich & warm less analytical, more deep & three-dimensional like with more planes, not sound flat sound more like live.

Yes the tracking is near perfect (you know the perfection doesn't exist).

Yes it's a Lingo driven the Linn.

Some photos just for fun:

Linto

An externally hosted image should be here but it was not working when we last tested it.


Setup Tracking Force 2,5g +-0.3g, as you can see at the photo does not coincide

An externally hosted image should be here but it was not working when we last tested it.


The setup was do it with this precision arm load meter & a the ArcProctractor specific for the Ekos:

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


LCD display 4 digits
Tare full capacity
Max 5g
d=0.002g

I also calculated the different load due to the difference between the height of the vinyl LP the height of the arm load meter:D.
 
Last edited:
Salas

Salas

diyAudio Chief Moderator
Joined 2002
Paid Member
When you will have the test record and you will run the tracking test modulation grooves, try to set anti skating by ear. Sit exactly in the middle of your speakers at 2.5m and set it for when the mistracking distortion starts so the crackling noise will come out from both speakers equal at the same time. Also you can use a couple of pins where your load resistor is, so you can easily change between values and types. Like Shinkoh or bulk foil 150R-180R-220R. There must be a subjectively ''right there'' after all work in the phono has been done, considering VTA, VTF, Load, TT support.
 
RCruz

RCruz

Member
Joined 2006
Paid Member
merlin el mago said:
@Salas
Many thanks, tomorrow I will start the Rs:)

@Ricardo
Yes the rectifiers are doubles, I don't compare but possible do a new PS for the regs. & phono RIAA with them to compare against Sikorel, but mine opinion is that are better the Sikorels, I will let you know.

I added the silver mica caps & the resistors to the RIAA with exact values as per your suggestion + 2pF to reach 16nF.

Subjective impressions: 1st is the bass control & now can follow him easily, the mids are rich & full, the highs are sweets, in general seems the music less lost like more tight or focused, also seems more soundstage & more separation between channels, the stereo effect is more marked. Very, very good job Ricardo:).

If I extract the 2pF what happens with the sound?
Click to expand...

I know that I can hear a 20pF difference but I believe ir will be very difficult to hear the differences using the 2pF or not... you let us know... in theory it should give a little more trebble enfasis without the 2p.

Can you post the final values you chose for the riaa circuit so i can feed them in my spreadsheet... just for comparison purposes :)
 
merlin el mago

merlin el mago

Member
Joined 2009
Paid Member
LEFT
R1 3K3
R3 47K//1M4 (several resistors to reach the most aprox. value)
R4 6K8
R5 1M13
C3 15.6nF6 // 330pF // 68pF // 2pF
C2 46.3nF // 470pF (several caps to reach the most aprox. value)


RIGHT
R1 3K3
R3 46K8 // 1M9 (several resistors to reach the most aprox. value)
R4 6K79
R5 998K
C3 15.6nF // 330pF // 68pF // 2pF
C2 46.4nF // 450pF (several caps to reach the most aprox. value)
 
RCruz

RCruz

Member
Joined 2006
Paid Member
Thank you so much. You must keep in mind that we are only adjusting Salas initial values due to the differences encountered when measuring the real components.

If we could reach the exact values Salas provided in the schematic we would certainly get a more even result.

In your case it is a compromise because the resistors are not paired correctly :)

I have one build that I use as reference where I could pair everything and the values are precicely acording initial Salas schematic... He made one specifically for my benz ACE cart and it still beats all others. (The optimum gain obtained in this build is also critical for the overall result)

Now you must listen for a while and focus on the VTA and loading. Both are much more critical than all the work done so far.
 
Last edited:
merlin el mago

merlin el mago

Member
Joined 2009
Paid Member
RCruz said:
If we could reach the exact values Salas provided in the schematic we would certainly get a more even result.
Click to expand...

As per your advice using Kabusa RIAA calculator with exact values as per Salas schematic gives:

R1 Kabusa = 3.3000 (R1 Salas) + 47.000 (R3 Salas) // 1.000.000 (R5 Salas) = 47.891 input R1 Kabusa value

R4 Salas = 6k963 instead 6k8 diff. +163R or -2%
C2 Salas = 45.644nF instead 47nF diff. 1.36nF or +3%
C3 Salas = 15.661nF instead 16nF diff. 0.34nF or +2%

Or I'm losing something or can't see a real matching between the ideal RIAA & our RIAA values.
 
merlin el mago

merlin el mago

Member
Joined 2009
Paid Member
Ricardo, I am thinking to buy a Vishay "nude" for R4 6K963 with exact value to Michael Percy, also to buy several Shinkohs 47k to Pcx to measure & get the 47K needefor Right channel & use mine Caddocks MK132 1Meg for R5.

Now I know how much important is to have a well matched RIAA, do you do the same if was for you?

I am really interested in your opinion, thanks.
 
RCruz

RCruz

Member
Joined 2006
Paid Member
merlin el mago said:
As per your advice using Kabusa RIAA calculator with exact values as per Salas schematic gives:

R1 Kabusa = 3.3000 (R1 Salas) + 47.000 (R3 Salas) // 1.000.000 (R5 Salas) = 47.891 input R1 Kabusa value

R4 Salas = 6k963 instead 6k8 diff. +163R or -2%
C2 Salas = 45.644nF instead 47nF diff. 1.36nF or +3%
C3 Salas = 15.661nF instead 16nF diff. 0.34nF or +2%

Or I'm losing something or can't see a real matching between the ideal RIAA & our RIAA values.
Click to expand...

I remenber Salas sugested you to // 2M over 47kr.
Doing so it will match almost perfectly.

Also I believe he does simulate riaa behaviour including all the parts (not only the eq part)
 
RCruz

RCruz

Member
Joined 2006
Paid Member
merlin el mago said:
Ricardo, I am thinking to buy a Vishay "nude" for R4 6K963 with exact value to Michael Percy, also to buy several Shinkohs 47k to Pcx to measure & get the 47K needefor Right channel & use mine Caddocks MK132 1Meg for R5.

Now I know how much important is to have a well matched RIAA, do you do the same if was for you?

I am really interested in your opinion, thanks.
Click to expand...

IMO you should pair all components as closely as possible. This will get you a true stereo image. After that you must retune.

If you change R4 from 6k8 to 6k963, you will change the presentation in the mids and will need to change all the other values.

I consider a best practice to determine the resistors values, and place the caps a little below spec so you can finetune just by placing more small caps //
 
RCruz

RCruz

Member
Joined 2006
Paid Member
Very good but I would stick to 2M or 1M8 (even closer if you get the exact 6k8)
IMO a difference of 20ohm in the 6k8 can easily be heard.

For cap tuning always take in consideration that C2/C3 should be 2.916.

Thank you again for your input.... It prooves my calculator works correctly.
 
merlin el mago

merlin el mago

Member
Joined 2009
Paid Member
Last edited:
merlin el mago

merlin el mago

Member
Joined 2009
Paid Member
RCruz said:
Very good but I would stick to 2M or 1M8 (even closer if you get the exact 6k8)
Click to expand...

Kabusa Riaa calculator
R= 3.300+45.804 (47k//1Meg8) // 1.000.000 = 46.805 input Kabusa R1
Kabusa R1 46.805 - R4 6.805 - C2 46nF723 - C3 16nF024

C2/C3 = 2.92

Your suggestion is very good because maintain the R4 value 6K8 & the linearity also is very good.

So I need to match the following values that I don't have:

LEFT
R3 add 1Meg8
R5 change 1Meg13 for 1Meg
C2 start with 46nF3 & reach adding more caps till 46nF723 max.
C3 start with 15nF6 & reach adding more caps till 16nF024 max.

RIGHT
R3 change 46K8 to 47K& add 1Meg8
R4 change 6K79 to 6K8
R5 change 998K to 1Meg
C2 start with 46nF4 & reach adding more caps till 46nF723 max.
C3 start with 15nF6 & reach adding more caps till 16nF024 max.

That's right?
 
You must log in or register to reply here.
Top Bottom