Another quasi-complementary design

Status
Not open for further replies.
Jump to Latest
#541
What are voltages across All them 0.33 Ohm resistors
Now?
R32 and R33 and R34 and R35 etc.

And what is DC voltage across two speaker Output Terminals?

Do all this measurings without any music (Volume zero)
but with some Old Loudspeaker or 8 Ohm Dummy Load connected.
Use never your best speaker for amplifier testing
----------------------

Across Output should be as close as possible to 0.0 VDC (DC-Offset)
Adjust with VR1.

Across R32 should be like 0.020V - 0.050V (~60-150 mA bias).
Same for R33, R34, R35
Adjust with VR2.
----------------------
If should be 0.0 Volt across any of your R32,R33,R34,R35 it might be your MOSFETs are not Matched.
Both NPN MOSFET should be close matched with eachother.
Both PNP MOSFET should be a matched pair, too.
 
#542
lineup said:
What are voltages across All them 0.33 Ohm resistors
Now?
R32 and R33 and R34 and R35 etc.

And what is DC voltage across two speaker Output Terminals?

Do all this measurings without any music (Volume zero)
but with some Old Loudspeaker or 8 Ohm Dummy Load connected.
Use never your best speaker for amplifier testing
----------------------

Across Output should be as close as possible to 0.0 VDC (DC-Offset)
Adjust with VR1.

Across R32 should be like 0.020V - 0.050V (~60-150 mA bias).
Same for R33, R34, R35
Adjust with VR2.
----------------------
If should be 0.0 Volt across any of your R32,R33,R34,R35 it might be your MOSFETs are not Matched.
Both NPN MOSFET should be close matched with eachother.
Both PNP MOSFET should be a matched pair, too.
Click to expand...

hi
lineup

my amp is running fine with 78-0-78 92 volts dc with 3 pairs
biasing of 1.2 volt across 10 ohm resistor
voltage across r32=8.2mv , r34= 7.6 mv , r36=32mv

r33,35,37=0 volts

ravs
 
#543
hi quasi

your amp is running fine with 3 pairs rails of 78-0-78 92 volt
sound is clean and fine. i think that was all matter of biasing no fault.
what if i increase the biasing current from 1.20 volts to 1.30-35 volts across 10 ohm resistor .voltage across my source resistors

r32=8.2mv
r34=7.6mv
r36=32mv

r33 , 35 , 37=0 volts



are all my mosfets working

thanks in advance
ravs
 
#548
Hi,
the output node has 5 connections into it.
from the upper transistors, from the lower transistors, from the speaker output, from the Zobel and from the NFB tapping point.

The speaker must be disconnected to carry out the bias and offset adjustments.
The NFB should have a many kohm load so cannot carry much current.
The Zobel has a blocking capacitor that prevents DC current flow.
If the upper transistors show current is flowing into the node then it must be coming out through the lower transistors. There is nowhere else for it to go.
So check the speaker is disconnected and that you have a shorting link in the input socket.
Check or replace the Zobel capacitor.
Check the zero volts drop resistors.
The current must be flowing out somewhere.
Current goes around in loops!!!!!!!!
 
#549
AndrewT said:
Hi,
the output node has 5 connections into it.
from the upper transistors, from the lower transistors, from the speaker output, from the Zobel and from the NFB tapping point.

The speaker must be disconnected to carry out the bias and offset adjustments.
The NFB should have a many kohm load so cannot carry much current.
The Zobel has a blocking capacitor that prevents DC current flow.
If the upper transistors show current is flowing into the node then it must be coming out through the lower transistors. There is nowhere else for it to go.
So check the speaker is disconnected and that you have a shorting link in the input socket.
Check or replace the Zobel capacitor.
Check the zero volts drop resistors.
The current must be flowing out somewhere.
Current goes around in loops!!!!!!!!
Click to expand...


hi andrewt

i found what i was doing wrong actually i was connecting my meter leads to only one side of source resistor.i thought i was connecting across source resistor.here are my voltages across the source resistors with biasing of 1.2 volts across 10 ohn rail resistors

r32=11.3 mv
r34=8.6 mv
r38=38.4mv

r33=59 mv
r35=10.2mv
r37=4.4mv

here is the image

http://i33.tinypic.com/e7m8u9.jpg

regards
 
#552
AndrewT said:
there is still something wrong.
The total mV on one side should add up to the same total mV on the other side.
Look also at how unmatched the currents are.
Each FET is running at a different temperature.
Click to expand...

Nearly true but not quite. The negative rail output stage also carries most of the current going through T8 and R19 (a little goes through R20 and R21 etc).

This current is about 10mA so the total of voltages across the negative rail output resistors will be a little higher.

Note, none of this current goes through the speaker because the output is held at 0v.

Apart from the mismatches (that aren't worth worrying about) the amp seems to be working normally.

Cheers
Q
 
#553
quasi said:


Nearly true but not quite.
Click to expand...
no,
the mV are indicators of the currents passing through the output FETs.
The upper FETs MUST pass the same current as the lower FETs.
The total mV of the Upper FETs MUST match the total mV of the lower FETs.
The speaker must be disconnected.

the output node has 5 connections into it.
from the upper transistors, from the lower transistors, from the speaker output, from the Zobel and from the NFB tapping point.

The speaker must be disconnected to carry out the bias and offset adjustments.
The NFB should have a many kohm load so cannot carry much current.
The Zobel has a blocking capacitor that prevents DC current flow.
If the upper transistors show current is flowing into the node then it must be coming out through the lower transistors. There is nowhere else for it to go.
So check the speaker is disconnected and that you have a shorting link in the input socket.
Check or replace the Zobel capacitor.
Check the zero volts drop resistors.
The current must be flowing out somewhere
Click to expand...

The minimum Vre=4.4mV, the maximum Vre=59mV.
A ratio of 13.4:1 when 1.5:1 would be a more sensible maximum.
That is close to a disaster.
 
#554
AndrewT said:

no,
the mV are indicators of the currents passing through the output FETs.
The upper FETs MUST pass the same current as the lower FETs.
The total mV of the Upper FETs MUST match the total mV of the lower FETs.
The speaker must be disconnected.

The minimum Vre=4.4mV, the maximum Vre=59mV.
A ratio of 13.4:1 when 1.5:1 would be a more sensible maximum.
That is close to a disaster.
Click to expand...


Hi Andrew,

The current through R19 must go somewhere. With the speaker disconnected and the output floating at 0v the only place this current can go is through the negative rail output stage devices. This is the only DC path available with a voltage differential. This is not true of course for the Nmos series where the currents through the positive and negative rail output stages are equal.

Try to have another look at the schematic and follow the current path of the second stage ccs.

It is for this reason that the construction guide for Actrk amp specifies the bias current is set by measuring the voltage across the setup resistor on the positive rail.

How well output FETs are matched are a matter of opinion after a certain point and this is mainly an issue when the amp is idling. At any sort of power the FETs will all be conducting a similar amount. But it's up to individual constructors how they feel about this.


Cheers
Quasi
 
#559
Most likely not, but they are sold here..

Imported from china for
20$ apiece , sold for 50-60$, Antec then pays N.J tax + fed tax
so we can finance welfare state, bailouts , and endless wars.

Same for mouser , import malaysian and taiwanese trannies
for $.01, sell for $.15, pays immigrant labor in texas to pick
and pack order, taxes paid to state and fed.
OS
 
Status
Not open for further replies.
Top Bottom