DoSuSy - LTspiced project. Comments?

Status
Not open for further replies.
Jump to Latest
#3
Hi, All!

Probably some additional comments are needed (many "Views", too few "Replies").

Background:
I have not DIY-ed during rather long period, about 8-10 years. A week ago
I didn't know anything about spice modelling at all.

The aim:
Cheap but musical amplifier. Well, anybody wants it, I see 🙂

Will be appriciated:
I'll be very happy to see your comments concerning

- possible schematics drawbacks,
- possible realisation caveats and traps.

There are rarher simple difficulties, like a selection of bjt model for
current sources in preamp-driver stage (well, do you remember? - only
a week ago I have discovered spice modelling. There are too many
parts around... Own spice models library is so small...). And there
are more serious problems, for example, frequency correction.

Schematics:
It is very simple. I'll be glad to answer your questions. Probably,
I must say about schematics part which is right hand of output stage.
The aims of this part is:

- make both current source currents equal (there is feedback from output
stage to differential stage, you see?),
- set (this is class A) these current sources current value.

With hope to get your feedback,
Andrew
 
#4
Hi Andrew

I really think that your effort deserves more feedback.
It looks like a very interesting but complicated (to me 🙂) approach.
It’s a bit annoying that we can’t understand the Russian text at your site.
Perhaps you could do a little effort and translate the important parts in an English page.
What about moving this thread to the Passlabs forum?
After all it’s about Nelson’s SuSy if I’m correct.

There you go, a few ideas to get the ball rolling.

/Hugo 🙂
 
#6
Hugo,

OK, I'll try to translate to english my little page. The problem is I'm
not familiar well with schematics terms, so, please, do not be angry
with my translation 🙂

Is it right way to move to Passlabs forum via mentioning this thread
and giving a link to this new page (and mentionig here about new
created thread)?

Andrew
 
#9
Congratulations, Anli. You have made a power amp that is interesting to even experienced designers.
I might point out a few things: First, I understand the common mode latchup problem. I made an amp with the same input stage (with transistors only) back in 1969. The first thing that it did was latch up! I, too, had to put in a common mode servo to fix the problem.
A year later, I tried to feedback to the emitters(sources) rather than the gates, and the common mode latching went away. Most of my designs, since then, have been with emitter or source feedback.
If you think that adding source resistors (10ohms or so) to your input devices might limit you, perhaps you should find a higher Gm fet, such as the Fairchild IRF 9510, rather than the 9610 to make up the difference.
 
#10
John,

Saying "latchup" - do you mean a process just after turning supplier on?
(I warned about my poor knowledge of terms 🙂 If it's so - OK, I'll
try to analyze first second behaviour.

And I'll try 9510/510 instead of 9610/610.

As for sources resistors... Well, preamp own gain is about 35 times only
(it works with rather high current and low load resitors - to recharged
output transitors gates quickly). So, at any point, I see a problem in
introducing such resitors. I need some time to recognize and answer
more defintely 🙂

Andrew
 
#12
#14
altor said:
Good job, Andrew, but when you'll be satisfied with the real sound from this ? 🙂

We'll be glad to hear good news!
Click to expand...


Alex, unfortunately, I don't think my own realisation will happen at near future.
There are some non-technical difficulties (well, they are to find parts and money for
them before 🙂 So, I hope there are DIY-ers with home laboratories "under
hot run", and they can try the project probe laying out during few hours.

Of course, some experience is needed here. For example, just now the
schematics contains drawback (I'll eliminate it with new version): there are
current sources in the preamp-driver stage. The sources use transistors with
a rather small max power dissipation. I'm going to replace them with more
power darlingtons.

After all, my goal is to share the development 🙂 Welcome!

Andrew
 
#16
Andrew,
I have a circuiit similar to yours that I will throw into the mix at some point. Not the same, mind you, I like your circuit as well. I think I've got enough parts on hand that I can build yours, compare it to mine, and report back. I'll try to inventory the stuff I've got on hand and see where matters stand sometime in the next few days.

Grey
 
#18
I see that you've just joined diy, so you may not be aware...any time estimate I give on how long it will be before I get to something must be multiplied by at least a factor of two or three, maybe even ten. As an example, I just got my car back from the body shop where it spent the last five weeks undergoing repairs after a kid ran a stop light and tried to remove the front end without unbolting it first. So far, so good. The problem is that the car isn't right yet, so I'll have to waste yet more time on getting my car re-repaired. I'd really rather use that time doing electronics or writing another story, but that's the way life is.
Sometimes the only way I get things done is to stay up late at night, but after about a week of getting three to four hours of sleep per night, I start getting loopy.
On the other hand, this general topology has intrigued me for quite some time. I first sent Nelson a schematic of this sort of thing about two or three years ago. Since then, I've continued to putter with it from time to time, most recently in August of this year. Your schematic is the same--yet different enough from mine to spark my interest again. I'm pretty certain that your idea will work, and that mine works...the question is will they work together. If so, great. If not, then perhaps we'll have two topologies that are cousins. Should be interesting.

Grey
 
#19
Grey,

I'm agree with you. "2-3" is rather optimistic factor 🙂

The problem I have mentioned above seems to be serious: I have got a
situation when the amplifier remains almost "off" after turning PS on. I have
resolve the problem, but I don't like the resolution way. Now I'm trying
to find something more elegant.

Andrew
 
#20
version 0.20 is published

Hi,

The preamplifier-driver and the current regulation part were redesigned.
I have fulfilled many experiments to make the model to be closer to reality.
In particular, I have used:

- different spice models (with different output threshold voltage) for the same
mosfets (yes, I have got even harmonics 🙂,
- power supplier with 10000uF filter capacitor,
- simulation of "first second" behaviour (just after turning supplier on).

1. Preamplifier-driver. Mosfet dif-pairs were loaded with AC/DC
repeaters/invertors. Such design eliminates "first second" problems.

2. Current regulation part. The dif-pair was replaced with two transistors used
with a common base. Now, the regulation loop gain is much smaller, and it is
possible to avoid any integration chains.

3. Output pulsation. To eliminate a penitrating of power supplier pulsations
to output:

- the current regulation part must be symmetrical (it is),
- all integration capacitors (in preamp current sources too!) must be
removed (they are).

With 10000uF (!) power supplier capacitor I have got output pulsations
less than ~1mV (with single 50V supplier and 5A output stage current).
It is important as one of the project goal is to design a cheap amplifier.

The project page is updated. New schematics is here:

http://freedns.ru/andrew/dosusy/DoSuSy-0.20.pdf

I hope, these changes are a big step. And I'm waiting for your comments!

Andrew
 
Status
Not open for further replies.
Top Bottom