Need to build JC 2 preamp

R8, R9, R10 are not present in Johns schematic, but my original modules obviously have a second pot for offset adjustment so I "borrowed" this from the ML-2 power amplifier which uses the same topology as input stage.

Thanks for the info, gk7. So, it was a bit of guesswork.. Did John "blessed" it?

Please post it, I´m curious to see it.

It's a ML-5 record amplifier. PM me with you email - I'll send you some more.
 

Attachments

  • ml-5-rec.gif
    ml-5-rec.gif
    133.2 KB · Views: 771
You make it too complex. Don't use the first gain block. Come in directly to the gain pot. That is how we do it.

Thank you Mr. John Curl,

You are right - the simpler schematic is better. But in my case I have tube out CD Player which is anode out and the output impedance is around 1-1.2k which is too high to drive 10k gain pot. That is the reason to put the first gain block. Is there any other technique to avoid the use of the first module? In my case, what would you do?

Thank you in advance

Ivo
 
Thank you Mr. John Curl,

You are right - the simpler schematic is better. But in my case I have tube out CD Player which is anode out and the output impedance is around 1-1.2k which is too high to drive 10k gain pot. That is the reason to put the first gain block. Is there any other technique to avoid the use of the first module? In my case, what would you do?

Thank you in advance

Ivo

It would be best to just use an open loop dual fet follower.
located at the tube output.
 
Hi Paolo,

the ferrite beads were already originally there, a photo with the old resistors:
l1000109.jpg

I did not experience oscillations but only used the modules in the ML-1 with
a gain of about 10 (20 dB). They are not unity gain stable, how much feedback
you can apply (lower gain) without stability problems I don´t know, I would
estimate less than a gain of 5 is too low.
 
Thank you Georg. Actually I didn't change the feedback from the original design values, I have used 22k from output to inverting input and 2k2 from inverting input to ground.
Anyhow, the presence of the ferrite beads in the original design makes me think they can be useful.

Basically the preamplifier sound very well but it is missing detail.

I have made some measurements using ARTA and a E-MU 0404 USB.
The first image is the loopback measurement that represents the noise/thd of the audio card, the second one is the spectrum of the output signal of the preamplifier, the last one is the frequency responce of the preamplifier.
As you can see there is a lot of garbage in the spectrum measurement of the preamplifier that makes me thinking that there could be a problem of oscillations. The frequency responce is enough extended but shows also some noise at upper frequencies.

Ciao
Paolo
 

Attachments

  • loopback.jpg
    loopback.jpg
    94.5 KB · Views: 652
  • spectrum.jpg
    spectrum.jpg
    98.6 KB · Views: 634
  • frequency responce.jpg
    frequency responce.jpg
    89.6 KB · Views: 619
  • CIMG0335r.JPG
    CIMG0335r.JPG
    166.5 KB · Views: 641
  • CIMG0336r.JPG
    CIMG0336r.JPG
    146.9 KB · Views: 644
I will try to repeat the measurements with ARTA but this
has to wait until I´m back from my holiday (beginning of September).
All I can contribute for the moment is an old measurement, it shows
about the same amount of 3rd you have, somewhat higher 2nd and not much else.
 

Attachments

  • line_right_1kHz.jpg
    line_right_1kHz.jpg
    60.3 KB · Views: 324
Member
Joined 2005
Paid Member
Borbely cascoded his line amp output stages when they were fets ( as did Pass), but both removed the output cascode becuase of a percieved loss of dynamics. However Erno indicated that his Japanese distributor and clients preferred the ouput cascode when jfets are used for the output stage. ( Borbely's Super buffer 2)