Hello everyone!
I'm trying to build an amplifier based on the integrated circuit TDA2050.
Datasheet informs the TDA2050 uses approximately one volt for operation and the rest is delivered to the load, for example:
If the power supply is +-22V, TDA2050 consumes one volt and we have +-21V peak to peak over the load.
I built datasheet circuit and when I measure the peak voltage over the load I find 18V, in other words, I have 4V over TDA2050.
Is that correct? There is how to achieve the specified in the datasheet?
I thank everyone who can help me.
I'm trying to build an amplifier based on the integrated circuit TDA2050.
Datasheet informs the TDA2050 uses approximately one volt for operation and the rest is delivered to the load, for example:
If the power supply is +-22V, TDA2050 consumes one volt and we have +-21V peak to peak over the load.
I built datasheet circuit and when I measure the peak voltage over the load I find 18V, in other words, I have 4V over TDA2050.
Is that correct? There is how to achieve the specified in the datasheet?
I thank everyone who can help me.
If you measure 4 volts across the chip then it is what it is. Load impedance will be a big factor to.
Are you simultaneously using a dual channel scope to observe peak output voltage and the supply ? The spec quotes a 10% distortion figure at max output to, so you may not be reaching those levels if you simply take the onset of visible distortion as the criteria.
Also try using a tone burst of just a few cycles to get a more accurate impression.
Are you simultaneously using a dual channel scope to observe peak output voltage and the supply ? The spec quotes a 10% distortion figure at max output to, so you may not be reaching those levels if you simply take the onset of visible distortion as the criteria.
Also try using a tone burst of just a few cycles to get a more accurate impression.
Hi,
At full power a +/- 22V unregulated will droop, for certain. It will
also ripple. If max output is +/- 18V, the specification implies the
lowest voltage of of the droop + ripple is +/- 19V, not unreasonable.
Not a problem as music is not sine waves, it spends 80% of the time
under 20% of the peak value, and an amplifier under clipping with
music is hardly ever drawing more than 50% of sine wave power.
rgds, sreten.
A 4:1 duty cycle test signal is a better reflection of reality.
At full power a +/- 22V unregulated will droop, for certain. It will
also ripple. If max output is +/- 18V, the specification implies the
lowest voltage of of the droop + ripple is +/- 19V, not unreasonable.
Not a problem as music is not sine waves, it spends 80% of the time
under 20% of the peak value, and an amplifier under clipping with
music is hardly ever drawing more than 50% of sine wave power.
rgds, sreten.
A 4:1 duty cycle test signal is a better reflection of reality.
Last edited:
So you can connect the second channel of the scope to the supply (scope DC coupled) and observe and measure the difference in voltage between the peak of your output voltage waveform and the supply voltage at that point in time.
Are you see 4 volts differential now ?
sreten
Thanks for the answer.
Molly
When connecting one of the scope channels to the power supply and the other to the load, with music, the positive peaks of the music relative to the negative peaks of the ripple of the source, the difference between them reaches a volt, in other words, the drop voltage across the TDA2050 comes to what is specified in the datasheet.
With sine wave I observe between the positive peak of the load and the negative peak of the ripple of the power supply 5V difference.
Thanks for the answer.
Molly
When connecting one of the scope channels to the power supply and the other to the load, with music, the positive peaks of the music relative to the negative peaks of the ripple of the source, the difference between them reaches a volt, in other words, the drop voltage across the TDA2050 comes to what is specified in the datasheet.
With sine wave I observe between the positive peak of the load and the negative peak of the ripple of the power supply 5V difference.
That sounds correct then. I mentioned using a toneburst type of test signal above although the official test procedure for the TDA2050 is this:
A.1 - MUSIC POWER CONCEPT MUSIC POWER is (according to the IEC clauses n.268-3 of Jan 83) the maximum power which the amplifier is capable of producing across the rated load resistance (regardless of non linearity) 1 sec after the application of a sinusoidal input signal of frequency 1 KHz. According to this definition our method of measurement comprises the following steps:
- Set the voltage supply at the maximum operating value; - Apply a input signal in the form of a 1KHz tone burst of 1 sec duration: the repetition period of the signal pulses is 60 sec; - The output voltage is measured 1 sec from the start of the pulse; - Increase the input voltage until the output signal shows a THD=10%; - The music power is then V2out /RL, where Vout is the output voltage measured in the condition of point 4 and RL is the rated load impedance;
The target of this method is to avoid excessive dissipation in the amplifier.
A.2 - INSTANTANEOUS POWER Another power measurement (MAXIMUM INSTANTANEOUS OUTPUT POWER) was proposed by IEC in 1988 (IEC publication 268-3 subclause 19.A). We give here only a brief extract of the concept, and a circuit useful for the measurement. The supply voltage is set at the maximum operating value. The test signal consists of a sinusoidal signal whose frequency is 20 Hz, to which are added alternate positive and negative pulses of 50 µs duration and 500 Hz repetition rate. The amplitude of the 20 Hz signal is chosen to drive the amplifier to its voltage clipping limits, while the amplitude of the pulses takes the amplifier alternately into its current-overload limits.
A.1 - MUSIC POWER CONCEPT MUSIC POWER is (according to the IEC clauses n.268-3 of Jan 83) the maximum power which the amplifier is capable of producing across the rated load resistance (regardless of non linearity) 1 sec after the application of a sinusoidal input signal of frequency 1 KHz. According to this definition our method of measurement comprises the following steps:
- Set the voltage supply at the maximum operating value; - Apply a input signal in the form of a 1KHz tone burst of 1 sec duration: the repetition period of the signal pulses is 60 sec; - The output voltage is measured 1 sec from the start of the pulse; - Increase the input voltage until the output signal shows a THD=10%; - The music power is then V2out /RL, where Vout is the output voltage measured in the condition of point 4 and RL is the rated load impedance;
The target of this method is to avoid excessive dissipation in the amplifier.
A.2 - INSTANTANEOUS POWER Another power measurement (MAXIMUM INSTANTANEOUS OUTPUT POWER) was proposed by IEC in 1988 (IEC publication 268-3 subclause 19.A). We give here only a brief extract of the concept, and a circuit useful for the measurement. The supply voltage is set at the maximum operating value. The test signal consists of a sinusoidal signal whose frequency is 20 Hz, to which are added alternate positive and negative pulses of 50 µs duration and 500 Hz repetition rate. The amplitude of the 20 Hz signal is chosen to drive the amplifier to its voltage clipping limits, while the amplitude of the pulses takes the amplifier alternately into its current-overload limits.
You may use a single channel scope also, and in a way it's more accurate:
1) scope output, rise voltage until it *just* clips top and bottom
Notice exactly where those peaks sit onscreen.
2) changing nothing, unclip probe and clip to +V or -V rails ...you'll see that rail voltage is a little beyond what you just marked,the difference is voltage drop across output transistors, whether discrete or encapsulated inside a chip.
Notice this new value, also that it shows ripple.
3) now stop driving the chipamp, by lowering volume, etc.
Rail voltage will rise to unloaded value, also ripple will basically disappear ... that's the power supply drop.
As you see, with a relatively simple test you can find all you need.
1) scope output, rise voltage until it *just* clips top and bottom
Notice exactly where those peaks sit onscreen.
2) changing nothing, unclip probe and clip to +V or -V rails ...you'll see that rail voltage is a little beyond what you just marked,the difference is voltage drop across output transistors, whether discrete or encapsulated inside a chip.
Notice this new value, also that it shows ripple.
3) now stop driving the chipamp, by lowering volume, etc.
Rail voltage will rise to unloaded value, also ripple will basically disappear ... that's the power supply drop.
As you see, with a relatively simple test you can find all you need.
My thinking starts with
"This is what I have. What can I do to investigate performance?"
I use comparison most of the time.
I compare one with another. If I see a measurement difference then I can be pretty sure that difference exists.
Then I look for a way to explain what I have just measured.
Here's a very good example of how COMPARISON works to improve our testing technique.
The Hamon Divider.
How to Build a Hamon Resistor Divider Network
I cannot measure better than whatever my best DMM says in the spec sheet plus or minus how far out of specification it has drifted since it's last calibration check (years ago)
With a Hamon divider I can compare to get difference measurements that approach 1ppm (0.0001%)
Similarly using a "Wheatstone Bridge" to compare REF to many DUTs
Learn to use the tools you have.
JMH has just explained how to use the scope to COMPARE voltages for two different operation modes and to compare voltages at two different parts of the circuit. The absolute voltage accuracy one can get from a scope is not good. But the comparison will show a real difference if that diff is just big enough.
And AC coupling a DC voltage allows one to EXPAND the scale of what one can see during the comparison.
"This is what I have. What can I do to investigate performance?"
I use comparison most of the time.
I compare one with another. If I see a measurement difference then I can be pretty sure that difference exists.
Then I look for a way to explain what I have just measured.
Here's a very good example of how COMPARISON works to improve our testing technique.
The Hamon Divider.
How to Build a Hamon Resistor Divider Network
I cannot measure better than whatever my best DMM says in the spec sheet plus or minus how far out of specification it has drifted since it's last calibration check (years ago)
With a Hamon divider I can compare to get difference measurements that approach 1ppm (0.0001%)
Similarly using a "Wheatstone Bridge" to compare REF to many DUTs
Learn to use the tools you have.
JMH has just explained how to use the scope to COMPARE voltages for two different operation modes and to compare voltages at two different parts of the circuit. The absolute voltage accuracy one can get from a scope is not good. But the comparison will show a real difference if that diff is just big enough.
And AC coupling a DC voltage allows one to EXPAND the scale of what one can see during the comparison.
Last edited:
That is a fantastically high quality post.
There is all of the details with what exactly to do about them.
He's got it right; so, do please try to use it.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.