LSK pre - BAF 2013

Measuring voltage drop across R2/R4 will reveal the current through JFETs and measuring it over R3/R5 will show the current through Q6/Q7 (minus about 1mA for LEDs). Lowering the values of R5/R3 will increase the current through Q6/Q7.
If DC offset doesn't fluctuate, connect the preamp to power amp, turn them both on and check for DC at amp's output - if everything is fine connect the speakers and put some music on...
You can experiment with feedback values (R7/R6) - 1k5/330R or 470R instead of 680R/150R. That means less load for Q6/Q7 (and slightly higher Zout) but might sound a bit better.
Setting P1 after every change is mandatory.
 
7 mA through JFETs is great. It depends on JFETs' Idss and Source degeneration and not on current mirror. 10-15 mA is what we want through BJTs and you got them there. All is well as it is.
If you lower the R2/R4 it won't change the current through JFETs, it will just lower the current through BJTs and you don't want that.
 
Last edited:
How are the preamps in this thread compare with bellow :
http://www.diyaudio.com/forums/pass-labs/271926-f5-headamp.html#post4269989
http://www.diyaudio.com/forums/pass-labs/146310-bf862-preamp-35.html#post4426944

How does pass ba-3 like preamp, build with complementary jfets , compare with the bjt (easy to get and cheaper) version above ?
And how does bf862 based preamp above (simple b1 like buffer following single bf862 gain stage) compare with these pass ba-3 like preamps ..
Also looking at many of the amps preamps like ba-3 , F5, le monster etc , all have similar front side of complementary n and p channel jfets .. and from ba-3 description adjusting pot between the complimentary jfets adjusts hormonic content with dominant 2nd or 3rd hormonic distortion ... Is it because of this feature it is being used in front of many amps and preamp?
 
Last edited:
It’s hard to answer those questions because no one but maybe Juma has compared these preamps.

The ba-3 is basically the f5 with no feedback...it has a lot of gain and uses p3 to match the jfets and set the harmonic profile. P3 will also compensate somewhat for mismatched jfets. That’s why the jfets don’t need to be tightly matched although it’s better to have a matched set to start with.

I have built the f5 head amp as well but i didn’t enjoy it as a preamp and the dc fluctuation scared me for preamp use.

Juma’s current mirror preamp has far less current drive and gain. And uses feedback. You could implement p3 adjustment here as well (you could do it with any amp or preamp using jfets on input this way). If you don’t need current (amplifier has high input impedance) it could be argued that simpler/lower powered devices such as in jumas preamp will sound better. But the topology is different...jumas preamp is closer to Pass UGS and preamps like the blowtorch I believe.

As for sound, I don’t know. I have the ba3 but I used the jfets from to make jumas preamp. So there won’t be side by side comparisons. Currently Iam using Salas dcg3 which is by all accounts excellent.
 
Here’s some progress.

It works. Cap coupled for now although offset seems very stable. But Iam being cautious. Boards do not bolted down until I figure out if I need the caps.

Iam going to listen for a while and see what I think.

THANK YOU JUMA!
 

Attachments

  • 31C95F4E-0E3B-4E51-9F62-F3156B77DEE0.jpeg
    31C95F4E-0E3B-4E51-9F62-F3156B77DEE0.jpeg
    667.5 KB · Views: 687
  • 287271A7-23CB-4773-8F3A-8BC7FB7B86FC.jpeg
    287271A7-23CB-4773-8F3A-8BC7FB7B86FC.jpeg
    703.9 KB · Views: 670
I have to say...this preamp is excellent. Really good. Among the best I have heard. Vocals are silky smooth and there is an openness yet smoothness to the sound. Transparency and depth is top notch. Extremely lacking in any sort of grainy sound. Tonal balance is good too. Open with wide soundstage but not bright. Delicate and Extremely fast as well. And this is all with a perhaps unneeded output cap which is not the best quality. Also, something that is very important to me and this preamp gets right is the sense of “space”...it’s not too forward or distant and vocals sit in the right place. Preamps are “space” controls to me...tonality is less important to me than where they place the pieces although I believe this preamp is well balanced tonally.

My only complaint is it’s slightly loose at the very bottom...bass is not quite as tight as I would like. I have other preamps with better control but they do not sound as open or lifelike. Perhaps the Silmic caps on the rails will break in and this will improve.

I wasn’t expecting it to be this good.

Well done Juma. I can see why you like this topology and use it in your cubie amps.
 
I've had the opposite experience with Silmics - they start really bassy and then calm down within a couple hours. Though, this was for coupling duty, and not on the rails.

I have had the preamp on for 24 hours or so...I almost always use silmics on the rails and usually they open up a lot in the first couple of days. The PSU of this preamp (Nichicon KG followed by KZ) and regulator have been used for a couple years before so they are broken in.

Honestly, the bass is not bad. It's actually better than some other preamps I have heard. I could totally live with it. But I also have the Salas DCG3 which I have been using before Juma's preamp. It has drill-like bass that is deep and powerful. It's kind of the opposite flavor...Also excellent. Probably my two favorite preamps I've tried. But I haven't tried them all.
 
... I wasn’t expecting it to be this good...
I'm glad you like it. :)

About the bass response, I like it clean and natural, the way I experience it in live acoustic performances. If your preferences are somewhat different or you are a fan of electronic music - see post #521 (3rd paragraph), check also the value of your output caps (if the Zin of your amp is 10k caps should be 10u at least, 1u for 100k and so on...). For tougher loads (very long cables, headphones, very low Zin amps), Emitter/Source follower can be added at the output. BJTs at the input, instead of JFETs tend to change sound in the direction you described (drill-like bass). Power supply/decoupling plays a role too. Volume pot (10-20k) should be at the preamp's input.There's a lot of room for experimenting, have fun with it... :cheers:
 
Power amp is very light load (Pass F5T/100K), output caps are 8.2UF. I am sure a follower on the output would be interesting but another stage might ruin what I have now...especially since the power amp load is so light. But definitely room for experimentation there.

I will play with the feedback resistor values like you have suggested...but really it might be perfect as-is. I was listening last night and it was really live sounding. So open and clean.
 
Member
Joined 2000
Paid Member
Hey, have a question about post number 1.

In the board trace the parts placement can be seen.
Is there a trick to getting the parts placement to go away or does the PCB image need to be edited to remove the parts placement to get a clean PCB transfer?

Thanks,

Vince
 
Member
Joined 2000
Paid Member
Like that. Was it automated or did you edit the traces?

I keep thinking there's a printer function where only pure black and white will print.
I see these type of PCB traces all the time and was just wondering if it's done with the software that made the PCB, or a 3rd party such myself could automate the removal of the parts placement.

Make sense?
 
Last edited:
I edited it in photoshop. Took out the red and filled in the traces. There is really no automated way to do it unless you start from a "Layered" file. If you printed only black the red will cut the traces. The red does not "overprint" over the black. It just cuts through it. I am a graphic designer...so finally we can talk about something I know lots about on this forum. Ha.

If you use the software that created this file I am sure you can turn off and on the parts you want.

The easy way would be just to tell adobe acrobat or whatever printer program you use to only print pure black and then fill in the traces with a sharpie.

In other news, I'd consider Juma's Current Mirror version instead...seriously.
 
LSKpre-CM+FB.gif
I'm glad you like it. :)

About the bass response, I like it clean and natural, the way I experience it in live acoustic performances. If your preferences are somewhat different or you are a fan of electronic music - see post #521 (3rd paragraph), check also the value of your output caps (if the Zin of your amp is 10k caps should be 10u at least, 1u for 100k and so on...). For tougher loads (very long cables, headphones, very low Zin amps), Emitter/Source follower can be added at the output. BJTs at the input, instead of JFETs tend to change sound in the direction you described (drill-like bass). Power supply/decoupling plays a role too. Volume pot (10-20k) should be at the preamp's input.There's a lot of room for experimenting, have fun with it... :cheers:

I did in fact add more feedback...520R/150R was as close as I could get to what you suggested. The sound is a little bit more lively now. Seems like a little less gain too? Not sure which I prefer, the original has a smoothness and silkiness I like. But I will have to listen for a while and see.

Also, while the circuit seems DC stable in a temp controlled situation it does fluctuate a bit (20mv or so) with temp changes. It probably doesn't help that my phono amp with shut regs is on top of it. I will leave the cps for now.

Still impressed. Thanks Juma.
 
Last edited:
Yes, R7/R6 ratio defines the gain and the total value of those resistors sets the load for BJT transconductance stage. Test and find out what you like the best.
I didn't bother much with small DC fluctuations at the output - I used coupling caps (Silmic at first, WIMA MKP 10 later). Thermal coupling and tight matching will help (closing the preamp's case goes without saying). DC servo can be used too.