Krill - The little amp that might...

Status
Not open for further replies.
Jump to Latest
#23
Very interesting output stage, I just started to have a look! Thanks for sharing!

May I ask you where the base current for Q8 and Q11 comes from?

A link or some comments on how the biasing part works would also be much appreciated.

As I'm currently playing with sliding bias schemes, I would also be very interested in improvements for this nice, little circuit. Appears very promising!

Have fun, Hannes
 
#25
juma said:
Current flows as a consequence of potentials difference between two points in elec. circuit; and there is a difference in potentials of 0.8V between those two transistors' bases. 😉
Click to expand...
That is correct. As shown in my posted schematic it is actually .85V. This is adjustable with R12 and sets the base voltages on the drivers and outputs which sets their idle current. I have shown this at a minimum. Increasing the bias will reduce distortion and increase bandwidth - up to a point. Beyond this distortion null, the distortion increases and bandwidth is reduced.

That is why I recommended using a distortion meter to set bias. If you do not have access to a distortion meter then set the idle current between 25 and 100mA. This will depend to a large extent on the transistors you use.
 
#26
h_a said:
Very interesting output stage, I just started to have a look! Thanks for sharing!

May I ask you where the base current for Q8 and Q11 comes from?

A link or some comments on how the biasing part works would also be much appreciated.

As I'm currently playing with sliding bias schemes, I would also be very interested in improvements for this nice, little circuit. Appears very promising!

Have fun, Hannes
Click to expand...


Your first question has been answered. I have no link to provide because to the best of my knowledge this bias scheme is mine. I have been unable to find anything that resembles it in the 25 years since I first designed it. How it works? That is what makes it interesting and I will keep that to myself for now. There are plenty of people here that can figure that out. I want to see what they say.

Improvements will be coming. For now, I wanted a project that was simple and inexpensive so the less skilled (or rich) could benefit also.
 
#27
Steve Dunlap said:
... this bias scheme is mine. I have been unable to find anything that resembles it in the 25 years since I first designed it. How it works? That is what makes it interesting and I will keep that to myself for now. There are plenty of people here that can figure that out. I want to see what they say...
Click to expand...

Well, the intrinsic property of PN junction (diode in this case) to change very linearly its' internal resistance as a function of temperature is well known and widely used.
Here we have, as a relative novelty, use of R12 to make this effect "milder" - R12 takes its' share of current from D6 and D7 and influences voltage drop on D5.
How effective this sub-circuit is in controlling the Iq of output stage cannot be precisely said without real-life circuit measurement, which I'm sure has been done extensively by Steve and adopted as empirically proven method.
My only concern is practicality of mounting those diodes on heatsink - it might turn off some DIYers; otherwise, great circuit Steve!
🙂
 
#28
Current flows as a consequence of potentials difference between two points in elec. circuit;
Click to expand...

Hi Juma, Hi Steve,

my precise question was where the current comes from. I do see that the diodes build a potential difference, but the bases are otherwise not connected to anything.

And as long as current can't be generated out of nothing it has to come from the collector currents which at least for me is a novelty as it's flowing into the opposite direction for at least one transistor.

As this does not sound very realistic, I asked 😉

I hope that at some point Steve will unveil the workings of the bias scheme, it would be a pity to loose interest of the people that just don't have the time or understanding to figure it out themselves.

Have fun, Hannes

By the way I'm still interested in a reply for my initial question 😀
 
#29
I had a tiny PCB made just for the diodes. It mounts with with 1 screw and a 3 wire ribbon cable runs back to the main board where VR12 is located. In early prototypes I simply soldered the diode string together and stuck them to the heat sink with epoxy.

You are correct about the thermal tracking with the diodes. You could accomplish the tracking also by mounting Q7, Q8, Q10 and Q11 on the heat sink.

The voltage drop between the bases of Q8 and Q11 set the base voltage of the outputs and thus the idle current. R12 is to set that voltage where desired.
 
#32
h_a said:
By the way I'm still interested in a reply for my initial question
Click to expand...

Sorry, I thought juma answered that. The base current for each transistor flows in the base of the other. All in one direction, and I trust, the correct one.

As this does not sound very realistic, I asked
Click to expand...

Actually, it sounds great!

I hope that at some point Steve will unveil the workings of the bias scheme, it would be a pity to loose interest of the people that just don't have the time or understanding to figure it out themselves.
Click to expand...

The base current in the bias pair is equal in both transistors at all times. I is not, however constant. Changes in the emitter current cause changes in the base current. When one transistor conducts less, the other conducts less. If for example Q8 conducts less because the voltage on the emitter of Q7 goes more positive (causing more current to flow in the base of Q13) then Q11 conducts less causing a greater voltage drop from emitter to collector. This makes more current available to the base of Q15 causing greater conduction to maintain an on state.

I am afraid I will be unable to describe it more simply.
 
#36
Would be great if I had the parts and time to build that fast, Steve 😉

No, I just took a break yesterday and added a variant of your outputstage to my project amp. I used MicroCap (there's a free evaluation version available:

http://www.spectrum-soft.com/demoform.shtm

)

Shows nicely that the outputs do not shut off independantly of output current.

When I have some more time I want to rework it a bit more, since I want to use either a complementary feedback pair or a single emitter-follower (with driver) as output. IMHO that makes sense as I want to use sustained beta bjts (high beta parts), so no need for a darlington.

If the results continue to be that well I'm going to build an amp in the next couple of months (pcb layout nearly finished, except outputstage).

Have fun, Hannes
 
#37
h_a:

I tend to forget that not everyone has a full workshop at home.:xeye:

I am glad you like what you see so far. I will help in any way I can with your changes and construction tips.

Hopefully there will be more interest in this design eventually. So far the participation has been minimal.

Steve
 
#38
Hello,

To me, this circuit could never work properly in Micro-cap. I have Micro-cap 9. I would be very interested in having the file you used yourself and that worked like a charm!
Maybe you could post it here or send it to me at : maziere_jeanyves@yahoo.fr

It would be a blast! I spent a lot of time on it and now I am frustrated I couldn't make it work !

Thanks
 
#39
darian:

I have not been using MicroCap. The MicroCap results were posted by h_a.

This circuit has proven to be very difficult to simulate on many sim programs. I don't pretend to know why. If h_a got good results, then his sim program would seem to work. Perhaps he will send you the info you requested.

Fortunately I conceived and built the first several without the benefit of sim software. I tried 3 different programs before I found one that would show this circuit to work. I believe an earlier version of MicroCap was on of the ones that did not work, but that was about 7 years ago. I'm sure the program has changed substansually since then.

Others have also reported sim results that simply do not match what can be measured in a working unit. I really can't say why.

Steve
 
Status
Not open for further replies.
Top Bottom