Power amp under development

Jump to Latest
quasi

quasi

Member
Joined 2004
sites.google.com/site/quasisdiyaudiosite/
jethari said:


Quasi
This is very good. But what about current values for the different power supply versions shown in your selection table? Is there a way we can derive the values for the particular power supply that I'm going to use?

Cheers
Hari
Click to expand...

The current flows shown are the same for all rail voltages. What will vary are the voltages across some of the components (not all) depending on the power supply. For example the voltage across T4 = negative rail minus the voltage across R6 and R7. The voltage across R7 is always 0.6v and if R6 is 18k then its voltage is always 49v. Therefore with 75 volt rails the voltage across T4 is 25.4 volts.

The same applies elsewhere in the cct.

Cheers
Q
 
J

jethari

Member
Joined 2002
quasi said:


The current flows shown are the same for all rail voltages. What will vary are the voltages across some of the components (not all) depending on the power supply. For example the voltage across T4 = negative rail minus the voltage across R6 and R7. The voltage across R7 is always 0.6v and if R6 is 18k then its voltage is always 49v. Therefore with 75 volt rails the voltage across T4 is 25.4 volts.

The same applies elsewhere in the cct.

Cheers
Q
Click to expand...

Quasi
That sure clears up my doubts. Thanks a ton. I'll be posting the latest version of my pcb design shortly. I've already started collecting the components for the amp. Should be commencing assembly sooooon!!
Cheers
Hari
 
M

marus

Member
Joined 2006
new test

NEW TEST :

----------------
Stage 1 :

One Nmos350 , +/- 40V DC, [ R6=2.7K and R17=22K ]

Currents :
* [ 1.33 mA / 1.33 mA / 2.7 mA ] @imput
* [ 13.4 mA ] @ T6-T7
* [ 16.7 mA ] @ T9-T10
* [ 20-26 mA] @ bias/pair

Big Heatsink : cold - unloaded / little warm - 50W
Driver Heatsink : a very little warm ( with fan )

Result : sounds great, no distorsion, output 50W (i think)
---------------
Stage 2 :

Exactly the same Nmos350 , +/- 78V DC, [ R6=18K and R17=33K ]

Currents :
* [ 1.43 mA / 1.44 mA / 2.98 mA ] @imput
* [ 15.1 mA ] @ T6-T7
* [ 16.4 mA ] @ T9-T10
* [ 19-27 mA] @ bias/pair

Big Heatsink : cold - unloaded / near cold - 1W
Driver Heatsink : warm ( with fan )

Result : distorsion, can't get 1W ...
---------------

I leave [ R6=18K and R17=33K ] and put +/- 40V : the distorsion appears at a little higher imput level
When I rise the DC rails the distorsion rise and the level drops.

This don't make sense ... won't follow the laws of phisics.
I think it's a black hole in my area ....
 
keypunch

keypunch

Member
Joined 2004
Hi All,

Could you folks explain to me how the currents remain the same with different rail voltages in an amplifier circuit? I would play with a SPICE program if I could do so with Linux to be more specific in my question. I have searched several times over the last year to try to find a SPICE program I can use with Linux with no success so far. The main reason for doing a circuit simulation was for me to understand the currents and voltages of the quasi NMOS with different rail voltages. The other reason was to give me voltage drops at all points as a reference for build validation purposes and diagnostic purposes should I encounter a problem.


Regards,

John L. Males
Willowdale, Ontario
Canada
08 September 2007 23:18
08 September 2007 23:42 Typo correction. jlm
Official Quasi Thread Researcher
 
quasi

quasi

Member
Joined 2004
sites.google.com/site/quasisdiyaudiosite/
keypunch said:
Hi All,

Could you folks explain to me how the currents remain the same with different rail voltages in an amplifier circuit? I would play with SPICE program if I could do so with Linux to be more specific in my question. I have searched several times over the last year to try to find a SPICE program I can use with Linux with no success so far. The main reason for doing a circuit simulation was for me to understand the currents and voltages of the quasi NMOS with different rail voltages. The other reason was to give me voltage drops at all points as a reference for build validation purposes and diagnostic purposes should I encounter a problem.


Regards,

John L. Males
Willowdale, Ontario
Canada
08 September 2007 23:18
Official Quasi Thread Researcher
Click to expand...


Hi John,

There are 4 main current flows in the amplifier and these are generally measured when the amp is idling.

The first is the current through the first stage and it is derived like this. D1 & D2 set a reference voltage of about 1.2 volts with respect to the negative rail. This sets a fixed voltage of 0.6 volts on the emitter of T4 and across R7. No matter what value the negative rail is this voltage will always remain constant because it is the junction voltage of T4. Therefore 0.6 volts / 220 ohms gives us a constant current of 2.7mA. This current is shared equally by T1 & T5.

The current source for the second stage works in a similar way. The voltage set by D1 & D2 sets a voltage on the emitter of T7 of 0.6v and across R15 (47R) creates a constant current of about 12.7mA.

So what's the big benefit of these 2 constant current sources. The main benefit (apart from huge gain) is immunity to rapid changes in the voltage rails including immunity to power supply ripple. This helps make the sensitive first and second stages hum free.

The 3rd stage is different. The current during idle is set to create about 4.0 volts across R20 & R22. Whilst the setting of VR2 may change depending on the rail voltage the current through the 3rd stage will always be 4v / 220R or 18mA. This voltage also appears as the gate voltage on the output FETs and this sets the output bias current (the 4th main current flow).

Cheers
Q
 
M

marus

Member
Joined 2006
Samuel Jayaraj said:
Marus, reduce the value of R6 to10K and reduce the value of R4 and R8 in steps, around 360E or 330E should get rid of your distortion problem. Let R17 remain at 33K equalling R3.
Click to expand...

I try, and it's the same s**t ! :bawling: :bawling: :bawling:


jethari said:
Marus
Are T1 to T5 exactly as per Quasi's schematic or have you used any alternates? One of them may be breaking down at the higher voltage.
Hari
Click to expand...

No, I use exactly what Quasi said.


It's anyone who made Nmos350 with 75-80V rails and it works ?
What schematic do you use ? I use the latest on Quasi site.
 
M

marus

Member
Joined 2006
Hi Quasi,

Yes, it's stereo setup. The other board do the same distorsion on +/- 80V. :( ( and both works great at +/- 40V )

Between the stage that works great and the stage with distorsion it's just 3 things changed:
1. power rails : 40V to 80V
2. R6 : 2.7K to 18K
3. R17 : 22K to 33K

If i make it exactly like yours how come yours works and mine doesn't ? :whazzat:

Do you use [ R6=18K and R17=33K ] on your +/- 73V ?
 
M

marus

Member
Joined 2006
This is how it looks on +/- 75V :

4k9y3us.jpg


5xpuujb.jpg


522mvqe.jpg
 
M

marus

Member
Joined 2006
to be or not to be

Hi Samuel, Aandy

This were taken with R6=18K and R17=33K.

Samuel, i tried what you said with 360R and 10K and it's just a little better.
I don't understand how "Over 100 (maybe 200) Nmos350's have been built " with 18K and 33K and can't be built one more ! :rofl: :rofl: :rofl:
I don't want to change anything. I make it like Quasi's i want it to work like Quasi's :violin:
 
J

jethari

Member
Joined 2002
Marus
Like Quasi commented a few posts ago, there's got to be a fault somewhere. Here are some possible causes:
1. One of the transistors is breaking down at the higher voltage.
2. It could also be the caps C5,C6, C9 and C10. What voltage rating are you using for these caps. One or more than one may be breaking down at voltages greater than 70V.
3. You should also check the voltage rating of the electrolytic caps C11 and C13. They should be rated at atleast 1.5 times the working voltage.
Go on, dont give up, Quasi's design will work!!!
Cheers and all the very best.
Hari
 
You must log in or register to reply here.
Top Bottom