Crossover distortion

[darkfenriz]

hi lumanauw
i must have missed something cause i don't understand it at all


current through 68 ohm seems to be always 2*drop on diode/68, say 25mA
and thus constant-not driving q5

could you explain it a little more??

 

[lumanauw]

OK, now try to omit the 2diode drops, there's only 68R. Does this works?

Since the output transistors will give enormous current, even 1A will give 68V drop to the 68R resistor. So, the purpose of the 2diode is to limit the drop at bigger current pass.

 

[lumanauw]

If I only got small voltage change (like 0,020V) and want to use it to trigger something, how to do it with discrete component?

 

[cunningham]

A comparator??

 

[lumanauw]

If possible, not using comparator or opamp.
I want to sense the voltage in 0.22ohm RE drop, so it can be used to maintained never-off condition, but having 10mA steady flow.

 

[Mr Evil]

A long tailed pair?

 

[lumanauw]

Interesting possibility. Just how to put differential to sense RE (0.22ohm for 10mA) between Transistor's emitor (in EF) and output line (to feel the voltage at 0.22ohm with 10mA)?

 

[fab]

Interesting possibility. Just how to put differential to sense RE (0.22ohm for 10mA) between Transistor's emitor (in EF) and output line (to feel the voltage at 0.22ohm with 10mA)?

Are you thinking of something like that?

 

[AKSA]

Daniel,

You must use two LTPs; NPN for positive rail, PNP for negative rail.

Both are fed from CCS; NPN LTP from the negative rail, PNP from the positive rail.

Right side bases of the LTP connect to the OP Device side of each emitter sense resistor (0R22), and left side bases connect to a voltage reference, a resistor of a few ohms driven from secondary CCSs supported from each rail.


You then use a large resistor on the correcting transistor's collector to kick start the driver base at each side.

This is all conjecture. I've not built anything like this. It might work.......

Have you considered Blomley's 1970 design? It should be somewhere on the web, and uses small signal transistors to divide the signal for each output half, whilst ensuring the output devices are always on, but in Class B.

Cheers,

Hugh

Cheers,

Hugh

 

[lumanauw]

Hi, FAB,

Yes, something like that, but for classB, so the drop in 0R22 is not always constant, but only to provide non-turnoff.

Hi, AKSA,

Nice to see you again
Thanks for the tip. I will think from there.
Do you know where the Bloomey amp is here (URL)?

 

[thanh]

is xover synonym for crossover?
with 1k5 load in input stage , there are no smooth at low volume level
Can you understand?
I think the reason is crossover .I can't increase iq at output stage because i don't like heat

 

[thanh]

oohhh! a interesting thread ! I look forward to hear from you

 

[lumanauw]

The power amp by Peter Blomley is talked everywhere, but I cannot see the sch. Does anyone knows where to see it?

 

[thanh]

The power amp by Peter Blomley is talked everywhere, but I cannot see the sch. Does anyone knows where to see it?

yes ,i'm too. Lumanaw! What's time is it ?
Now forum is sad because there are not much member

 

[Graham Maynard]

Hi lumanauw and thanh,

Here is the 'new' Blomley circuit of over 30 years ago.

Hope it gets through okay, my Internet connection is presently as bad as the weather.

Cheers ............ Graham.

 

[PMA]

I also do have his series of articles - one of my great inspirations (signal splitter)

 

[lumanauw]

Hi, GM,

Thanks for the drawing. It doesnt use differential, so it cannot maintain DC offset (output using cap)?

You have seen more schematic than I have all this time Do you have other SCH that doesnt turn-off in the final stage (in class AB, offcourse). This Blomley is hard to understand, and using cap in final output.

GM, the issue that you mentioned, the reactive speaker kicks back to the amp, is it got back to the amp via feedback (directly to differential) or ramping from final stage to VAS to differential (since if we use bipolars, emitors/collectors are always connected to base)? Is this also happens in pure classA (with feedback)?

 

[thanh]

thank graham!
he didn't use ltp and used cfb , cfb-ef configuration
i don't know what the function of tr6 tr3
carlos! this is a good design

 

[Steven]

Hans Hartsuiker once published another Blomley type amplifier with an opamp input iirc. It was in the Wireless World say 5-10 years ago, maybe more.
I heard that one in 1984; it was actually very good. He published it much later.
I should have a copy somewhere....

Steven

 

[PMA]

Hans Hartsuiker once published another Blomley type amplifier with an opamp input iirc. It was in the Wireless World say 5-10 years ago, maybe more.
I heard that one in 1984; it was actually very good. He published it much later.
I should have a copy somewhere....

Steven

There is one similar by Ian Hegglun, Electronics World, March 1995. The same author, September 1995.