this should be easy

[Mook Johnson]

I'm looking for a sine wave driver that can deliver 2 volts p-p 30Khz sine
wave into a 3 ohm load. +/- 4.5V rails are available and sine wave will be
the input. Distortion needs to be reasonable but not ultra tight. >1%
maybe.

I have a design using complementary NPN and PNP biased to class AB (audio
amplifier) with a opamp completing the loop and providing feedback to
minimize the residual xover distortion.

Thermal runaway is a high concern. I'm using large 2 ohm emitter resistors
on the output and diode biasing to compensate for temperature affects.
Anything else I should be looking at? both xisters will be mounted to a 6
layer PC board with lots of copper underneath.

This amp will be running above 125C (ambient) so the audio amp ICS with the
thermal cutoff will be of no use to me.

Just looking for different opinions..

thanks.


P.S. Class D is also out due to noise issues in the immediate areas.

 

[Ban]

I'm looking for a sine wave driver that can deliver 2 volts p-p 30Khz
sine wave into a 3 ohm load. +/- 4.5V rails are available and sine
wave will be the input. Distortion needs to be reasonable but not
ultra tight. >1% maybe.

I have a design using complementary NPN and PNP biased to class AB
(audio amplifier) with a opamp completing the loop and providing
feedback to minimize the residual xover distortion.

Thermal runaway is a high concern. I'm using large 2 ohm emitter
resistors on the output and diode biasing to compensate for
temperature affects. Anything else I should be looking at? both
xisters will be mounted to a 6 layer PC board with lots of copper
underneath.
This amp will be running above 125C (ambient) so the audio amp ICS
with the thermal cutoff will be of no use to me.

Just looking for different opinions..

thanks.


P.S. Class D is also out due to noise issues in the immediate areas.

This is probably not so easy because of the +125°C operation limit. If you
want something stable and with adjustable current limit, I recommend the
OPA547. With 1.5W dissipation and moderate heat sinking the die temperature
will be still below 160°, where it shuts down. Not too cheap tho, but
certfied up to 125° operation temp.

 

[Mook Johnson]

Ambient temperature may go higher than 160C 125 is the minimun temrpature of
oparation.

the thrustof my question is if anyone knows of an audio type IC tat doesn't
have thermal shudoff or it can be disabled?

If a discrete design must be used, what would be the most stable biasing
scheme for the most reliable operation overwide temperature swings? (hence
my concern about thermal runaway)

In my business I know for a fact that garden variety transistors, opamps and
other components will work for several hours at 200C if they are well
selected. Thats all i need.

 

[Jim Thompson]

Hi,

You have not said what this whole thing is for. From the description
you have given, it could be an electric cigar lighter. You get better
directions if you let people know just where you are going.

RS-232 to TTL??? You have something against 1488 and 1489 ?

...Jim Thompson

 

[Luhan Monat]

I'm looking for a sine wave driver that can deliver 2 volts p-p 30Khz sine
wave into a 3 ohm load. +/- 4.5V rails are available and sine wave will be
the input. Distortion needs to be reasonable but not ultra tight. >1%
maybe.

I have a design using complementary NPN and PNP biased to class AB (audio
amplifier) with a opamp completing the loop and providing feedback to
minimize the residual xover distortion.

Thermal runaway is a high concern. I'm using large 2 ohm emitter resistors
on the output and diode biasing to compensate for temperature affects.
Anything else I should be looking at? both xisters will be mounted to a 6
layer PC board with lots of copper underneath.

This amp will be running above 125C (ambient) so the audio amp ICS with the
thermal cutoff will be of no use to me.

Just looking for different opinions..

thanks.


P.S. Class D is also out due to noise issues in the immediate areas.

Hi,

You have not said what this whole thing is for. From the description
you have given, it could be an electric cigar lighter. You get better
directions if you let people know just where you are going.

 

[Mike Fields]

RS-232 to TTL??? You have something against 1488 and 1489 ?

...Jim Thompson
--

OK, sometimes I have trouble "reading between the lines", but how
did we get from a 30khz amp at 120 deg C to RS-232 ??????????

mikey

 

[Winfield Hill]

Mook Johnson wrote...

I'm looking for a sine wave driver that can deliver 2 volts p-p
30kHz sine wave into a 3 ohm load. +/- 4.5V rails are available ...
This amp will be running above 125C (ambient) so the audio amp ICS
with the thermal cutoff will be of no use to me.

That's 330mA delivered from a 4.5V supply with a 3.5V rail-to-output
drop allowed, right? What's wrong with unbiased NPN and PNP followers
inside an opamp feedback loop? There won't be any thermal runaway,
and < 1% distortion shouldn't be any problem with high opamp f_T.

 

[Jim Thompson]

OK, sometimes I have trouble "reading between the lines", but how
did we get from a 30khz amp at 120 deg C to RS-232 ??????????

mikey

You didn't check out Luhan Monat's website ?

...Jim Thompson

 

[Luhan Monat]

RS-232 to TTL??? You have something against 1488 and 1489 ?

...Jim Thompson

Yo,

Those are fine for putting in a design itself. This unit was for quick
hookups to bench prototypes so each one does not need rs232 just to try
out an idea. Also, it provides for 'network' type mixdown to a single
ttl with 'echo suppression'.

All of this could be done with the chips you suggest, or, as most of my
other later designs did, just a single pnp transistor for transmit, and
a cmos inverter gate for receive.

 

[Mark Jones]

I'm looking for a sine wave driver that can deliver 2 volts p-p 30Khz sine
wave into a 3 ohm load. +/- 4.5V rails are available and sine wave will be
the input. Distortion needs to be reasonable but not ultra tight. >1%
maybe.

I have a design using complementary NPN and PNP biased to class AB (audio
amplifier) with a opamp completing the loop and providing feedback to
minimize the residual xover distortion.

Thermal runaway is a high concern. I'm using large 2 ohm emitter resistors
on the output and diode biasing to compensate for temperature affects.
Anything else I should be looking at? both xisters will be mounted to a 6
layer PC board with lots of copper underneath.

This amp will be running above 125C (ambient) so the audio amp ICS with the
thermal cutoff will be of no use to me.

Just looking for different opinions..

thanks.


P.S. Class D is also out due to noise issues in the immediate areas.

One thing. Consider this when using multi-layer boards:
http://www.geocities.com/CapeCanaveral/Lab/9643/TraceWidth.htm

 

[Ban]

You have not said what this whole thing is for. From the description
you have given, it could be an electric cigar lighter. You get better
directions if you let people know just where you are going.

I once encountered a similar problem, but with 180°C temp spec. Impossible
to do reliably with silicon, I had to assemble these brittle GaAs chips on a
hybrid. It was supposed to be for "Telecommunications".
By chance I found out later that this circuit was mounted above the gates
of those F16 bunkers and was used to arm the atomic weapons in the last
possible moment. The afterburner was heating up that area, that is where
that spec came from. Well telecommunication was not exactly wrong...

 

[Ken Smith]

That's 330mA delivered from a 4.5V supply with a 3.5V rail-to-output
drop allowed, right? What's wrong with unbiased NPN and PNP followers
inside an opamp feedback loop? There won't be any thermal runaway,
and < 1% distortion shouldn't be any problem with high opamp f_T.

.... and highish frequency transistors. At these currents that should be
no problem. This is one application where you don't want the TIP35/36
pair.

 

[Tony Williams]

That's 330mA delivered from a 4.5V supply with a 3.5V
rail-to-output drop allowed, right? What's wrong with unbiased
NPN and PNP followers inside an opamp feedback loop? There
won't be any thermal runaway, and < 1% distortion shouldn't be
any problem with high opamp f_T.

Maybe reduce the dissipation in the Tr's with a resistor in
series with each collector. Up to 10 ohms looks possible.

 

[Tam/WB2TT]

Mook Johnson wrote...

That's 330mA delivered from a 4.5V supply with a 3.5V rail-to-output
drop allowed, right? What's wrong with unbiased NPN and PNP followers
inside an opamp feedback loop? There won't be any thermal runaway,
and < 1% distortion shouldn't be any problem with high opamp f_T.

This sounds reasonable to me. The class B operation will also cut down on
dissipation. I would look for semiconductors intended for automobile
operation. It gets pretty hot under the hood. There would also be some
advantage to transformer coupling; that would let you match the 2V P-P to
the +/- 4.5 V supply. Lastly, for a single frequency amplifier, why not
tuned class C (or class B) ?

Tam