Simplistic NJFET RIAA

Jump to Latest
Salas

Salas

diyAudio Chief Moderator
Joined 2002
Paid Member
AndrewT said:
Salas,
I have said this a few times over the past few months.
Can you provide corroboration?
or have you discovered this for yourself and your own measurements?
Click to expand...

I don't have an RF spectrum analyzer, but I have read it in radio station guides too and its logical to also apply for grounding a shield that could be collecting some interference. An example read is:

"(1) The ground wire should be as short as possible, preferably much shorter than a quarter-wavelength long on the highest frequency band operated."

Station RF Ground
 
nezbleu

nezbleu

Member
Joined 2009
Salas said:
You are supposed to have traded some extra gain for some extra local feedback so its rather logical to have heard "something". Did you see some easier thermal settling too? It takes an order of magnitude Rs change to really see such but anyway.
Click to expand...

I am currently listening to a lovely Prestige compilation album. First song was John Coltrane "Traneing In". This album always sounds good, but tonight I feel like it is a little cleaner, a little more solid.

Of course, when making changes like this one cannot make A/B comparisons. I do not have so much hubris as to assume that I can hear something other than what I expected or hoped to hear.
 
A

AndrewT

R.I.P.
Joined 2004
Salas said:
I don't have an RF spectrum analyzer, but I have read it in radio station guides too and its logical to also apply for grounding a shield that could be collecting some interference. An example read is:

"(1) The ground wire should be as short as possible, preferably much shorter than a quarter-wavelength long on the highest frequency band operated."

Station RF Ground
Click to expand...
thanks.
1/4 wavelength of 1GHz is 75mm.

much shorter than this would maybe be 40mm?
So a 40mm tail from screen to chassis would give good transfer of 1GHz interference signal from Mains input to the enclosing Chassis.

The total length from Screen to Chassis must include the hidden length inside the transformer insulation and the fixing length at the chassis.

This forces us to turn the toroid upside down with the tails at the Chassis face, not hanging up in the air as is usually done.
 
Last edited:
Salas

Salas

diyAudio Chief Moderator
Joined 2002
Paid Member
nezbleu said:
I am currently listening to a lovely Prestige compilation album. First song was John Coltrane "Traneing In". This album always sounds good, but tonight I feel like it is a little cleaner, a little more solid.

Of course, when making changes like this one cannot make A/B comparisons. I do not have so much hubris as to assume that I can hear something other than what I expected or hoped to hear.
Click to expand...

That's what I meant to say. i.e. its normally expected to detect some little subjective change after reconstructing the gain and local feedback by a few dB. Without controls never 100% sure but not technically unfounded in this case.
 
Salas

Salas

diyAudio Chief Moderator
Joined 2002
Paid Member
Not a problem in our case, I could not find random frequencies on the phono output with the 200MHz oscilloscope. The RIAA HF filtering and no loop feedback certainly help. At least in this PCB layout. Gate stoppers would be a major noise contributor for MC otherwise.
 
A

andyr

Member
Joined 2003
Salas said:
Not a problem in our case, I could not find random frequencies on the phono output with the 200MHz oscilloscope. The RIAA HF filtering and no loop feedback certainly help. At least in this PCB layout. Gate stoppers would be a major noise contributor for MC otherwise.
Click to expand...

Great to hear, Salas. Thanks. :) As you say, Gate stoppers would introduce noise.


Regards,

Andy
 
A

AndrewT

R.I.P.
Joined 2004
Not just limited to 1Mhz on the Mains supply.
It's all the frequencies that could enter via the mains cable and gets past the mains filter that many don't fit because they believe the filter screws with the sound quality.

Many interference sources exceed 1GHz and some are getting to near 4GHz.

If a builder specifies a screened transformer, there is no point in ruining it's potential for interferance attenuation, by using a long tail from screen to Chassis.

If the circuit does not suffer from interference pick up, then delete the mains input filter and delete the screen from the mains transformer.
 
Last edited:
Salas

Salas

diyAudio Chief Moderator
Joined 2002
Paid Member
Its supposed to drain what a screen could collect between primary and secondary so it better drain it more effectively for highest frequencies possible. Since we learned something from RF guys lets do it properly for its own part in the puzzle, why not. Without becoming fanatic about it either since we did not suffer special susceptibility clues in FSP and we don't normally have Agilent 3GHZ k$ analyzers in the average audio oriented home labs to confirm that stuff in detail.
 
S

stajo

Member
Joined 2011
Paid Member
As challenger of the title "get stuck in preparations" I set up to prepare my workbench with built in isolation transformers just for the sake of elliminating external disturbance pickups when doing small signal measurments like in riaas. The trannys I had was old (as Andrew was part of helping me discover) so I am now schooled in weber, tesla, turns ratios and all of the relating and there is copper wire all over my apartment.

And the workbench stands in the middle of the room, surrounded by tools.:p
 
R

rickmcinnis

Member
Joined 2006
Stajo,

No need to explain yourself.

How about this for travails. My first attempt went very wrong and I am not sure where. After a few semiconductor substitutions I gave up and decided to start over.

The saintly "Tea Bag" sold me some new FETs but the 369s are not the same as the first so I had to order new resistors and since I never know when to stop I ordered these from TEXAS COMPONENTS who then mailed them to our old address even though it was made plain to them where to ship. It will take three weeks for them to make it to me (1 1/2 weeks to go).

Michael lent me his prototype FOLDED and I blew up one channel of it. I wanted to try it with my autoformer attenuators and when I placed the boards in the box I connected one channel to the power backwards. I have yet to recover from the depression resulting from realizing I can be that stupid. I know I am stupid but I hate to be proved THAT stupid.

So you are not the only straggler and I bet you have not ruined anything in the interim.

Take care,
 
A

AndrewT

R.I.P.
Joined 2004
stajo said:
I agree. Then its not only the length but also the resistance so the cable should be thick to effectively drain away the hf pollution.
Click to expand...
resistance is so low that it is insignificant at frequencies above 50/60Hz.

It's cable impedance from Screen to Chassis that is very significant.
Inductance is your enemy here.
You must strive for minimum inductance and that requires very short connection length from Screen to Chassis.
 
S

stajo

Member
Joined 2011
Paid Member
Well Rickmcinnis its a hobby so one can just dig into what ever interesting problem that comes up on the table:D

While I was talking about it I was looking into the resistor component lists and I wonder what you have to say about some choices here.

My build is going to be for 2,5 mV Dynavector 10X5.

For R1, main cartload resistors I have 47k VARs. I am still thinking about what values to choose on Rx1-4. I have 6,8k VAR and also 18, 47 and 100k VARs

I also have VARs for R4 2,2k, R14 47k, R5 and R10 6,8k.

For R2 68R I have some Welwyn RC55D68RBB 0,1% 25 ppm 0,25W https://www1.elfa.se/data1/wwwroot/assets/datasheets/RC-series_eng_tds.pdf

For R6, R7 and R9 I have some IRC CAR5D1MBB 0,1% 25 ppm 0,25W https://www1.elfa.se/data1/wwwroot/assets/datasheets/qaCAR_data_e.pdf

For R8 47R I have either Welwyn RC55D47RBB 0,1% 25 ppm 0,25W https://www1.elfa.se/data1/wwwroot/assets/datasheets/RC-series_eng_tds.pdf

Or

Welwyn RC65Y-47RBI 0,1% 15 ppm 0,5W https://www1.elfa.se/data1/wwwroot/assets/datasheets/RC55_RC65_RC70_eng_tds.pdf

For R11 1,5k I have 3 choices:
Akane ohm RNM12TB 1K5 B E PB-FREE 0,1% 25 ppm 0,5W https://www1.elfa.se/data1/wwwroot/assets/datasheets/jjRN-Scan_data_e.pdf

Or

Arcol MRA0207 1K5 B 15PPM TA 0,1% 15 ppm 0,25W https://www1.elfa.se/data1/wwwroot/assets/datasheets/MRA_eng_tds.pdf

Or

Welwyn RC65Y-1K5BI 0,1% 15 ppm 0,5W https://www1.elfa.se/data1/wwwroot/assets/datasheets/RC55_RC65_RC70_eng_tds.pdf

For R12 120 R I have Arcol MRA0207 121R B 15PPM TA 0,1% 15 ppm 0,25W https://www1.elfa.se/data1/wwwroot/assets/datasheets/MRA_eng_tds.pdf

For R15 3,3k I have Akane ohm RNM12TB 3K3 B E PB-FREE 0,1% 25 ppm 0,5W https://www1.elfa.se/data1/wwwroot/assets/datasheets/jjRN-Scan_data_e.pdf

Or

Welwyn RC65Y-3K3BI 0,1% 15 ppm 0,5W https://www1.elfa.se/data1/wwwroot/assets/datasheets/RC55_RC65_RC70_eng_tds.pdf

Some of the datasheet links are the same since its different values in same series but for your convenience…

What do you guys think about my choices for resistors?
 
You must log in or register to reply here.
Top Bottom