Jfet Cascoded BOZ

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi all!
I have received some 2sk170BL and built this preamp circuit using Nelsons articles about Cascode operation/ZV9 and of course BOZ...the preamp did actually work after some hours trimmings...
it shows good square waves and linearity figures on the scope, and sound is clear, and bass sound impressive...
this lead me to think that I could now show it to you as I would really like to have your advices concerning any trimmings that could be done in order to have it work in the good spot for the Jfet...I , for sure have made several beginner mistakes , as it is one of my first trial as building something from scratch, but I would be glad if someone could help me point them out:)
here is the schematic:
 

Attachments

  • jcboz.jpg
    jcboz.jpg
    72.9 KB · Views: 1,412
By the way, the nodes voltages that are on the schematic were from 2 diffrent trimmings, the ones in the triangles were from a first bias setting, then the others are the actual ones that are now set...
here are some picts of the work bench
 

Attachments

  • jcboz2.jpg
    jcboz2.jpg
    68.9 KB · Views: 1,012
a close-up showing the regulated PSU, it delivers 71V, but as there are diffrent switches acccross the Zeners, it can be ajusted to 60v/71V/82V and even 94V, so it's versatile enough for many preamps projects...(it is a copycat of Nelsons BOZ PSU except that the TIP 29 is replaced by a TIP 41C...:)
Vince
 

Attachments

  • jcboz1.jpg
    jcboz1.jpg
    52 KB · Views: 935
bobsinclar, Nice :D I would like to give advice but I'm no expert either. However, Let me stimulate some thinking in a few areas... Don't you think that 50K "volume ctrl" is a problem for source impeadance of the following stage? I would evaluate using a pot at the input instead...
And, what's the 330k and the 4.75k at the input for? And, how about this? I believe that 10u at the input could be a higher quality cap of lower value ($$$) with the 100k input R...
What do you think? Thanks :smash: :smash: :smash:
 
Hi Flg!
actually you pointed the 50K and it is right that I could use a lower pot, but I have no 5K...so I used it...but I will try to find a 25K I may have one lying around...

I put it at the end of the preamp, because Nelson did it on BOZ and I thought it was just right keeping that...but I will try it in front, and we'll see what happens, do you think I can keep it as 50K if I put it in front?

actually the 330K is a key part of the schematic, remove it, and the amp doesn't work, the sine waves are cut flat on the upper side, I don't know exactly why, but it happened that when I put it in place the sine wave become perfect, and the IRF 610 actually become conducting the way it should to make a voltage drop in the CCS resistor(6K), I guess this resistor helps the Gate of the Jfet to be referenced at an other voltage than a virtual ground...the truth: I don't know :confused:

as for the 4.75k at the input, you are right I wasn's sure either that it was necessary to use it...mayve 200Ohm to avoyd parasitic oscillations would be enough...

about the 10uf cap at the input, do you mean that I could have used a 1uf? and then that would have cost less? actually the whole schematic is built with trash components, and these are electrolytics...I for sure have good audio caps of nearly every values...I will try them, but I would like to set right all the concept before...
Thanks for your help Flg, I will try to modify the pot right now!
Vince
 
Well, NASCAR is getting ready to start and I'm trying to get my L Loaded ZV9 proto set up to watch racing and play with it at the same time. What a combination ha?
Yes, I think a 50k pot would be good at the input. It will serve as the input load risistance and the hi Z of the JFET probably won't mind at all.
330k? I'm still thinking about it,,, but, I think you can remove it. You probably need a little tweak somewhere else???
I have used JFETs many times without any input Rs or oscilations...
The 10u... Well, I think without the 330k and the 4.75k, you don't need it at all! But it's a good idea sometimes, depending on what your source is. Having said that, it needs to be sized based on the resistance after it for low frequency rolloff. If you only have that 100k you could probably go down to 2.2u or so. Then you could spend the same money on an expensive "audiophile approved" type. I've used Mundorf "M-Cap Supreme" in places like that but they are $$$.
 
Well I'd be ok to remove the 330k if only I manage to make it work without! but looks like a key part...
I already changed the pot value to 4.7k, but I 'll let it at the output for a time, looks like nothing has been changes, except a little less gain(what is actually desirable as it had too much!):)
actually my question, would be concerning the IRF 610 VGS, if you look at the voltages on my schematic, it seems that it is only 9.5V-6.8V so 2.7V or so...this means that the fet isn't running correctly?? any idea?
Vince
 
Magura said:



How far are you with your LZV9 ?

Magura :)


Hmmm. Well, Ive had it running with the sig gen and the scope mostly with unmatched parts. I've listened to it on some cheapy spk'rs a little. but it had to low of an output cap for descent low freqs. Let me show you a pic of what it's set-up like for todays playing around...
The 120mH is a Hammond 195T5 5A 100mH Choke (measures 120mH @ 3.5A)
I'm going to look at the Open Loop Gain and Freq Response today and then add the R1 & R2 and see what I can get for Closed Loop response and gain...
 

Attachments

  • sej.png
    sej.png
    22.5 KB · Views: 989
This was my first attempt to try at that...using the cascode could enable to use it just in the right spot for that application...
any idea? in fact I would like help to point out what voltages would be desirable accross the schematic in order to say "the preamp looks ok..."
thanks!
Vince
 
After some trimmings and changes in the circuit here is what I get..looks like voltages looks ok, but maybe it would be better to run the Jfet with less voltage? also it's vgs is near -0v ...is that normal?
actually the preamp works, sound is good...

but I am surprised that the upper fet isn't getting hot, I wanted to make something hot:D , it drives abour 7mA, should it run hot, so that the linearity and distortion figures would be better?

anyway, here is the actual schematic with all voltage nodes
Cheers!
Vince:)
 

Attachments

  • cboz3.jpg
    cboz3.jpg
    76.8 KB · Views: 895
Hello,
here is the new schematic, I paralleled 4 Jfet driven at about 7mA each and then the IRF610 has to give about 28mA , I also raised the PSU voltage up to 92V , and loaded the fet with a 1.2K resistor, so that the voltage at his drain is 60V...if I am correct it now dissipates 0.028 A x54 V so about 1.4W...(if I am correct?)
at this point il seems to sound better than the other version, but maybe voltages could be ajusted a better way...:confused:

I will build the BOZ, so i'll have something to compare to...
any input is appreciated:)
Vince
 

Attachments

  • jcboz3.jpg
    jcboz3.jpg
    70.9 KB · Views: 815
Does anyone have expirience how low should be Vds of 2sk170 for "triode" caracteristics mode?

No practical experience but from what I was reading at various places (Borbely articles amongst others !) the region is well below 1 Volt and probably not very practical, so you should use the K170 in the pentode region. It should still be pretty good, though.

If you want to use a JFET in the triode region you should try the LU1014, then you may have 2 - 3 Volts.
 
Dear Nelson,
Thanks so much for that good advise,I am actually trynig much trimmings on it, but the problem is that I only have my ears to listen and judge...would the only way to determine this sweet spot be using a distortion analyzer? or could a Dyer with just a scope find an arrangement to look at something specific on the scope for that trimming?I actually only have a scope and a LF generator(sine and square waves)
Best regards
Vince
 
The one and only
Joined 2001
Paid Member
You can trust your ears, but in my experience any kind of
distortion analyzer will get you there quicker. Consider the
number of permutations of Source resistor, gain, bias, and
cascode voltage, and imagine listening to them all. It's a lot
easier to find the best measured values and start your listening
there.

:cool:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.