B1 with Korg Triode

Every day a little step .... time to find a housing with a window!

preview
 
The Korgie lights are cute. They look like blue green lightening bugs. My chassis has an open section on top to see them. I have ordered a lab glass beaker cover to go over them as a dust cover.

I do wonder how much the flouorescent filament contributes to the sound, since cathode chemistry and configuration were dark arts with DHT triodes. The Korg seems to do quite well in this regard.
 
I made a test bench with the PCB from diyaudiostore.
I think the sound is different from the original because I selected parts in Japan,but It's sound is good. Thank you to Nelson and the community.


I did two tests.


Reduce microphonic noise with a magnet.
The result was similar to that of post# 1745.The potential at T7, 8 changes, and the sound changes accordingly.
I do not adopt this method, for fear of the side effects that magnets give to electronic devices.


Fairchild J113 (supplied with pcb) vs Toshiba 2SK170
I placed the socket at the Q1, Q2, R1 position of the PCB and changed the parts.2sk170 are GR grade, matched. The potential of T7, 8 is adjusted each time.
Of course, I know that 2sk170 is more expensive and this is not a strict test.
The result: Both sound good and the differences are quite small.
The 2sk170 may sound slightly solid, but it is probably my assumption.
I choose J113 and keep 2sk170 for the next project.
 

Attachments

  • Nutube_1.jpg
    Nutube_1.jpg
    819.6 KB · Views: 959
  • Nutube_2.jpg
    Nutube_2.jpg
    981.8 KB · Views: 955
The one and only
Joined 2001
Paid Member
J113's are all over the map, especially after OnSemi took them over.

25% or so are in the 6 to 8 mA range. Most of the rest are 20+. This is by
split by lot codes, and so my "raw" inventory is sorted that way and when I
want the lower Idss, I go to those lot codes.

It is unknown which group is before/after the ownership transition, but
I have found plenty of uses in both categories.
 
Linearity

Nelson, you know there are only a handful of truly linear low plate/drain/collector devices on the planet. For small signals, the Toshiba JFETS have long been the best. For large signals you have your Semisouth parts and the Sony 2SK82 and 2SJ28. The IRF240 is probably another one IF you bias it high enough.

On the tube side you have a variety of miniatures (12AX7, 12AU7, 6DJ8, 7DJ8), and two biggies (the 300B and the EL34 in Ultralinear grid connection). IMHO, EVERY great amplifier in existence is built around these parts. Everything else requires a lot of monkey work to make halfway linear and ends up compromised by part count alone.

My question - only Linear Systems has tried to resurrect a couple of these devices - the Toshiba JFETS. I know that the MOSIS project has supported the fabrication of custom IC's for many years (see:MOSIS Integrated Circuit Fabrication Service). Why hasn't a comparable provider stepped up to make custom linear amplification devices ? It would seem to me they could print money if they did.
 
answer to Nelson Pass #2035

I fully agree! JFet-prices can make a big difference. :rolleyes:


Because of this the LX-Mini-crossover as well as the B1 NUTUBE preamp is a wonderful possibility to build a nice sounding machine using the much cheaper J113 JFet (or perhaps using a similar N-Ch JFet).

Many persons on this planet (or in this diyaudio - universe) will appreciate this chance. I do.



Thank you! :D


Greets
Dirk
 
The one and only
Joined 2001
Paid Member
They seems to only specify a min. Idss for J111, J112 and J113.
J113 has a min. Idss = 2.0 mA.…..but what can that be used for when most has 20+?
If it was typical values then the "number" J11x was the "grade".

For many purposes you can use all these parts - the 20+ Idss parts
operated at lower current figures, for example 7 to 10 mA, will give
comparable distortion performance in the buffers for crossovers and such.

Also, the high Idss parts are quite good as constant current sources -
they only require the appropriate Source resistance.
 
Ok.....I need to study JFETs a bit more to get a full understanding of why and how to select the optimum grade for different purpose. As far as I remember the BA3-pre the bias of the JFETs is approx. the Idss. My guess is that it is best to have Idss close to the max. current needed to get lowest noise…...but it is just a guess so far.....there are many curves in the data sheet for JFETs that I need to understand better.