Simplistic NJFET RIAA

[AndrewT]

Originally Posted by Joshua_G

It can also be due to improper azimuth adjustment.

Could you explain why azimuth affects the respective output levels of the two channels of a stereo cartridge?

[Salas]

Many factors can be suspicious but the audible hiss difference points to different electronic gain to check out first. Its easy to have different transconductance in one Jfet, and there is no feedback loop to set the gain certainly.

[Joshua_G]

Originally Posted by AndrewT

Could you explain why azimuth affects the respective output levels of the two channels of a stereo cartridge?

The balance between the L/R channels is done by adjusting the azimuth of the cartridge, or head-shell.

[RCruz]

Originally Posted by AndrewT

Those channel differences are ~1.8dB to 2.1dB.
Your cartridge balance may have something to do with this.
I suggest you swap the cartridge channels first and check again before making any adjustments.

Reversing the inputs, with a 300Hz test tone I measured the following values:
Tone volume.........Right.....Left
+12dB.................0.61V.....0.68V
+14dB.................0.85V.....0.93V
+16dB.................1.05V.....1.13V

In this case I calculated a 0.9 ~0.6dB difference between channels.

Playing with the numbers I found a cart imbalance of:

12dB test tone --- L/R cart imbalance of 0.4dB
14dB test tone --- L/R cart imbalance of 0.5dB
16dB test tone --- L/R cart imbalance of 0.7dB

Subtracting these values from Andrew´s calc, I get an average 1.3dB difference due to the riaa gain only.

It is not very much and nobody would tell just by listening but I will do the trimmer mod on second stage and verify if the differences of 1.3dB in gain are responsible for the differences in hiss.

[RCruz]

Originally Posted by Salas

That is 2nd stage. You got it OK. If you want to play with gain you vary R2 slightly. That one drops more on R1 and ups gain or vice versa. Because the signal level is slight there, THD remains about the same. I don't recommend playing with 2nd stage for gain because it influences the total gain less drastically but changes THD by moving the bias point of the Q5 buffer. You may try optimize the distortion profile to your ears there, 10-11V Q2D-GND is where it measures best in general. Whatever you may like, stays matched between channels on second stage. First stage is much more suitable to balance out any total gain differences between channels. Takes output measurement capability when steady under mV signal is present on input somehow.

Now I am confused

[Salas]

That's an old post, I changed my mind when I factored in noise contribution if upping first stage Rs or to choke the cascode transistor for voltage if needing go down. I match channels by varying the second stage using 1:200 attenuator, 1kHz & FFT. The channels don't start by more than 0.5dB difference in my experience when the JFETs are just IDSS matched. It still holds that varying the second stage enough, changes the THD also. So if a JFET is too strong or too weak better think of spotting and changing that to a better matching one and then trim accurately. Its not so easy with just a cart and a test disc, but in your case its good enough since the differences are substantial. 1.3dB is not good. Get it down to 0.5dB if you can. Its broad range and the stereo will snap in better. Hiss can be coming from low level circa 5MHz oscillation if something is not perfect with the regulator if it is 1.2R and double mono so to chance on the bad channel, or the wild JFET can be noisier. You need an Oscope to check the rails are tame. K170s range from 0.9nVrtHz to 1.2-1.3nVrtHz and can do that slight hiss difference if in the head-amp (1st stage) along with high gain. Try substituting a proper IDSS one in the hissy channel first.

[Joao@AltheaMusica]

Originally Posted by RCruz

Reversing the inputs, with a 300Hz test tone I measured the following values:
Tone volume.........Right.....Left
+12dB.................0.61V.....0.68V
+14dB.................0.85V.....0.93V
+16dB.................1.05V.....1.13V

In this case I calculated a 0.9 ~0.6dB difference between channels.

Playing with the numbers I found a cart imbalance of:

12dB test tone --- L/R cart imbalance of 0.4dB
14dB test tone --- L/R cart imbalance of 0.5dB
16dB test tone --- L/R cart imbalance of 0.7dB

Subtracting these values from Andrew´s calc, I get an average 1.3dB difference due to the riaa gain only.

It is not very much and nobody would tell just by listening but I will do the trimmer mod on second stage and verify if the differences of 1.3dB in gain are responsible for the differences in hiss.

Ricardo

how do you measure? With inverse RIAA at the input? Output impedance of your signal generator and which signal voltage in relation do 0dbV?

This is, in my opinion as an engineer, important to know that you can compare other measurements. In the audio industry you are dealing with relations and given test scenarios.

Musical regards
Joao

[AndrewT]

Originally Posted by Joao@AltheaMusica

which signal voltage in relation do 0dbV?

This is, in my opinion as an engineer,

he is comparing the 0.61V to the 0.68V and etc.
In post6636, 0.58V to 0.71V and etc.

[RCruz]

Originally Posted by Salas

1.3dB is not good. Get it down to 0.5dB if you can. Its broad range and the stereo will snap in better. Hiss can be coming from low level circa 5MHz oscillation if something is not perfect with the regulator if it is 1.2R and double mono so to chance on the bad channel, or the wild JFET can be noisier. You need an Oscope to check the rails are tame. K170s range from 0.9nVrtHz to 1.2-1.3nVrtHz and can do that slight hiss difference if in the head-amp (1st stage) along with high gain. Try substituting a proper IDSS one in the hissy channel first.

Can Q3 (BC550C) contibute in any way to the differences in gain and hiss ?
I had no problems before I did my last mod that consisted in changing the 15n cap by a FT-1 and also replacing R3 resistor bypass, so I may have overheated Q3.

Hi João
I am using a test record that produces 300Hz at +12dB +14dB and +16dB.

That is why I managed to determine cart imbalance and subtract the value from overall output differences. I am using a AC voltmeter.

[Salas]

The BJT can contribute in hiss only if it has a problem. Its normal contribution is uncorrelated. Ground the FT-1 can.