fine-tuning a transistor amplifier

Thread Starter

mike_canada

Joined Feb 21, 2020
239
amp1.png
amp2.png

The last amp I made (as per the calculating wattage post) worked very well. Then again, I could blame my speakers and caps for lack of volume and bass. My caps in that design were limited to 470uF because that's the stock I had. I also had to make my source audio volume at about 95% to hear anything (13 bars out of 16 lit on my old blackberry Q10 which was where the audio came from). Then I had to use the airplane equalizer setting to get a volume boost but then again part of the sounds were junky because of the boost.

So now I want to avoid ramping up the volume like that and convert a low-ish volume input to a nice amplified output that someone outdoors can hear from a good 50-100 feet away. My other design allowed for about 20 feet distance (with airplane equalizer) or about 10 feet without equalizer.

This time, I intend to ramp the voltage up to 14.4 with 2 6800mAh 7.2V batteries in series. I was wondering, am I getting the best output with my component choices or should I change some (on the input side)?

And will my 4.5W speakers effectively handle this new circuit? the equation in the other thread suggests the wattage usage is about 3.8.

And how far do I need to go with the coupling capacitor values? Do I need anything higher than 4700uF? I want to conserve money, but the higher values generally cost more.

And if possible, I want to use as many common resistor values as possible without negatively affecting the circuit.

The input source in the test is as follows: DC Offset: 0V, Amplitude: 20mV, Frequency: 10Hz
 

BobTPH

Joined Jun 5, 2013
5,810
What is the speaker you are using? If it is the one you showed in the other thread it will not produce any bass no matter what capacitor you use, and it cannot handle 1W. Cheap speakers are often way exaggerated in their power ratings.

And what do you mean by hearing it at 100 feet? Do you mean determine an alarm is sounding, or listening to music at normal level? The former might take 1W and the latter might take hundreds of watts. Outdoor sound is difficult.
 

MrSalts

Joined Apr 2, 2020
2,386
If I am calculating right, you'll have about 2.7 to 3 watts of output through your 8-ohm speaker (at most) with a class AB amplifier and 14.4v power supply. Not enough to make anything audible unless your living on a remote pond with no traffic and a wind-still morning. You need more (higher) voltage for your power supply and/or lower impedance speakers. Possibly a bridge design (essentially two amplifiers 180° out of phase fo drive each input to the speaker).
 
Last edited:

MrChips

Joined Oct 2, 2009
27,196
This circuit does not need fine tuning. It needs to be sent to the waste bin.
You are just slapping in components with no consideration whatsoever as to the function of each component.

R7 and R8 in the emitter leg of the transistor has a function and that is to provide negative feedback. Make R7 and R8 about 1Ω to provide some negative feedback.
Yes, they will reduce the maximum output power. I have already said, if you want more power then you need to rethink the design of the output stage.

C3 is the DC blocking capacitor to the 8Ω load. Without it the load will be subjected to Vs/2 DC voltage. C3 is a high pass filter and will determine the low frequency cutoff. Use any online RC high pass filter calculator to determine your cutoff frequency. C3 can be 1000-4700μF. There is no need to go any higher for your application.

Now look at C6. This again is a high pass filter and the same calculation applies. Your input resistance is 56kΩ.
C6 can be 0.1-1μF

And again with C1. Your input resistance is R11 in parallel with the base-emitter junction of Q2. Hence the input resistance is much lower than R11. Choose 10μF for C1.

I have no idea what C2 does. The whole purpose of R1 and R2 can be made better.

As for amplifier gain, you don't have enough gain. Add a single transistor preamp stage to your circuit for more gain.
If Q4 is supposed to be your preamp then there are ways to improve it.
 

BobTPH

Joined Jun 5, 2013
5,810
I think your best bet is to buy a 20 or so watt class D amp board. These produce 4 times the power at the same voitage, and are more than twice as efficient as the class AB you are playing with, which means your batteries last 2 to 3 times as long.

Is you application busking by any chance? If so, you likely have lots if noise to overcome, which will take even more power.
 

Ian0

Joined Aug 7, 2020
6,326
I think your best bet is to buy a 20 or so watt class D amp board. These produce 4 times the power at the same voitage, and are more than twice as efficient as the class AB you are playing with, which means your batteries last 2 to 3 times as long.

Is you application busking by any chance? If so, you likely have lots if noise to overcome, which will take even more power.
They produce the SAME output power for the SAME power supply voltage, unless you are comparing a bridge-tied-load Class D with a single-ended Class AB.
And with the efficiency of a Class AB being around 78.5%, doubling that efficiency will be tricky. (78.5% is the theoretical maximum, the actual is more than likely >50%)
 

Ian0

Joined Aug 7, 2020
6,326
View attachment 275000
View attachment 275001

The last amp I made (as per the calculating wattage post) worked very well. Then again, I could blame my speakers and caps for lack of volume and bass. My caps in that design were limited to 470uF because that's the stock I had. I also had to make my source audio volume at about 95% to hear anything (13 bars out of 16 lit on my old blackberry Q10 which was where the audio came from). Then I had to use the airplane equalizer setting to get a volume boost but then again part of the sounds were junky because of the boost.

So now I want to avoid ramping up the volume like that and convert a low-ish volume input to a nice amplified output that someone outdoors can hear from a good 50-100 feet away. My other design allowed for about 20 feet distance (with airplane equalizer) or about 10 feet without equalizer.

This time, I intend to ramp the voltage up to 14.4 with 2 6800mAh 7.2V batteries in series. I was wondering, am I getting the best output with my component choices or should I change some (on the input side)?

And will my 4.5W speakers effectively handle this new circuit? the equation in the other thread suggests the wattage usage is about 3.8.

And how far do I need to go with the coupling capacitor values? Do I need anything higher than 4700uF? I want to conserve money, but the higher values generally cost more.

And if possible, I want to use as many common resistor values as possible without negatively affecting the circuit.

The input source in the test is as follows: DC Offset: 0V, Amplitude: 20mV, Frequency: 10Hz
Probably best to start reading some books on the subject.
Douglas Self's is probably the most compreshensive.
http://www.douglas-self.com/ampins/books/apad.htm

First thing to do would be to connect your batteries as a positive and negative supply and eliminate the output coupling capacitor. Then change the input to a long-tailed pair to reduce the distortion. With a few years experimentation you should be able to get the distortion below 10 parts per million.

With a source of 20mV you will probably also need a preamp, as most power amplifiers are designed for an input voltage of 775mV
 

BobTPH

Joined Jun 5, 2013
5,810
They produce the SAME output power for the SAME power supply voltage, unless you are comparing a bridge-tied-load Class D with a single-ended Class AB.
And with the efficiency of a Class AB being around 78.5%, doubling that efficiency will be tricky. (78.5% is the theoretical maximum, the actual is more than likely >50%)
If course I an talking about full bridge class D amps, which all of the cheap boards you can buy are.

Class AB amps are particularly inefficient when not at full power, which is nearly all the time, whereas class D are efficient over the entire range, so I think double the efficiency is not out of line.
 

BobTPH

Joined Jun 5, 2013
5,810
This article explains what I mean. When listening to music at a volume where the highest levels are just below clipping, the average power is only at about 10% of the max. And at this level, class D is much more efficient than class AB.

https://www.analog.com/en/analog-dialogue/articles/class-d-audio-amplifiers.html
The differences in power dissipation and efficiency widen at moderate power levels. This is important for audio, because long-term average levels for loud music are much lower (by factors of five to 20, depending on the type of music) than the instantaneous peak levels, which approach PLOAD max. Thus, for audio amplifiers, [PLOAD = 0.1 × PLOAD max] is a reasonable average power level at which to evaluate PDISS. At this level, the Class D output-stage dissipation is nine times less than Class B, and 107 times less than Class A.

For an audio amplifier with 10-W PLOAD max, an average PLOAD of 1 W can be considered a realistic listening level. Under this condition, 282 mW is dissipated inside the Class D output stage, vs. 2.53 W for Class B and 30.2 W for Class A. In this case, the Class D efficiency is reduced to 78%—from 90% at higher power. But even 78% is much better than the Class B and Class A efficiencies—28% and 3%, respectively
 

Ian0

Joined Aug 7, 2020
6,326
It's all down to whether the TS wants to build a better amplifier, or to understand how to design a better amplifier.
As he said "fine tuning", I expect that he wants to learn how to build an improved Class-AB amplifier, not start from scratch with a Class-D module.

Interesting to note that the low-power efficiency of class-D has improved a lot. Originally, the output at idle was a 400kHz squarewave, and switching losses and output inductor losses dominated to such an extent that a Class-AB had lower losses. I think the introduction of three-level modulation made a huge improvement.
 
Last edited:

DickCappels

Joined Aug 21, 2008
9,347
One solution: Works fine with a single +12 power supply. +14V would be even better. Full bridge configuration doubles your P-P voltage to the load.
1661782640021.png
 

MrChips

Joined Oct 2, 2009
27,196
To the TS, if you want more power output here are things you can do:

1) increase the supply voltage
2) lower the impedance of the loudspeaker
3) use a pair of bridged amplifiers
4) change the design of the amplifier
5) all of the above
 

LowQCab

Joined Nov 6, 2012
2,513
Why has no one asked what the end result is supposed to be ?,
so that a viable solution might be arrived at ?,
or so that the TS might be disabused of the idea, if it is not feasible in the first place ?

"" The input source in the test is as follows: DC Offset: 0V, Amplitude: 20mV, Frequency: 10Hz ""

?? 10-Hertz ?? outside !!! with a tiny speaker !!! and ~5-Watts !?!?!?
What planet is this ?, I must be lost or something.

No intended goal has been outlined as of yet.
.
.
.
 

Audioguru again

Joined Oct 21, 2019
5,211
Mike, your amplifier is very distorted and has a output before clipping of only 1.8W.
Capacitor values should be calculated, not guessed. Your enormous input capacitor of 4700uf will take almost 10 minutes to charge. The terrible speakers (shriekers) you selected produce no bass so an amplifier input -3dB of 80Hz is
(1/2 pi) 56k= 80Hz x 56k= 0.036uF which is almost 39nF.
The output capacitor value of 47000uf is crazy. 330uF will produce an output of -3dB at 60Hz.

I fixed the biasing of your input transistor so its collector is nearer to ground, and removed the negative feedback attenuation of the extra 56k resistor. Then the distortion is very low.

MrChips, C2 bootstraps the collector load of the second transistor which increases its positive output swing, the positive feedback increases the effective value of its load which increases the transistor's voltage gain. Then the amplifier has a higher output power before clipping and has less distortion.
 

Attachments

Thread Starter

mike_canada

Joined Feb 21, 2020
239
I have to admit I get faster responses here than I do from local people. Anyways....

...if you want more power output here are things you can do:

1) increase the supply voltage
2) lower the impedance of the loudspeaker
3) use a pair of bridged amplifiers
4) change the design of the amplifier
5) all of the above
Ok, so it seems my best bet for now given the parts on hand and my budget is to increase voltage. I'll go from 7.2 to 14.4V

Mike, your amplifier is very distorted and has a output before clipping of only 1.8W.
Capacitor values should be calculated, not guessed. Your enormous input capacitor of 4700uf will take almost 10 minutes to charge.
But what if I want to import the bass sounds from the music source?
If the capacitor value is too small then I'm cutting out the bass. Come to think of it, if someone uses a capacitor value of under 50nF as a coupling capacitor between amplifier stages, I'll have a comical talk with that person.

The terrible speakers (shriekers) you selected produce no bass so an amplifier input -3dB of 80Hz is
(1/2 pi) 56k= 80Hz x 56k= 0.036uF which is almost 39nF.
Well sadly they're the best i got and upgrading the speakers to even qualify for better sound is $15+ a piece.
 

Thread Starter

mike_canada

Joined Feb 21, 2020
239
wave.png

So I also did more research and started playing with another transistor amplifier and as I was playing with this one the waveforms seemed to look normal and gain is largely influenced by values of C1 and R7. I forgot what website I got this circuit from but it claims gain is calculated as R9/R11.

Just out of curiosity, with a 14-volt battery supply, when I look at the waveform, is the biggest and best output at 6.4V peak to peak (like from +6.4V to -6.4V and back)? I say 6.4 instead of 7 because I assume a transistor will use 0.6V for processing the signal.

I included the LTSpice circuit for one to play with if need be.
 

Attachments

Ian0

Joined Aug 7, 2020
6,326
If you don't want to lose the bass, use a ±7V supply instead of a 14V supply, so you can eliminate the output capacitor.
Are you sure it's the amplifier that is limiting the bass, not the loudspeaker?
 

Audioguru again

Joined Oct 21, 2019
5,211
Mike, again you did not calculate the value of C1 so it is cutting the bass that your tiny speaker cannot produce.
22uF feeding the 100 ohms of R7 cuts 73Hz to half the output power of middle and high audio frequencies.
Use 220uF then the half power (-3dB) frequency is 7.3Hz then audio frequencies above about 22Hz will not be cut.

You use only one slow diode for D1 instead of two fast diodes. Then at low levels there is crossover distortion that sounds like buzzing:
 

Attachments

Thread Starter

mike_canada

Joined Feb 21, 2020
239
Mike, again you did not calculate the value of C1 so it is cutting the bass that your tiny speaker cannot produce.
22uF feeding the 100 ohms of R7 cuts 73Hz to half the output power of middle and high audio frequencies.
Use 220uF then the half power (-3dB) frequency is 7.3Hz then audio frequencies above about 22Hz will not be cut.
The other day when I presented high value capacitors, you told me they're too high. So you're saying the 100 ohms + 22uF capacitor make up a high-pass filter?

You use only one slow diode for D1 instead of two fast diodes. Then at low levels there is crossover distortion...
Wouldn't the 1N914's or 1N4148's more likely to blow up compared to the 1N4007's? I mean to get max voltage peak-to-peak I probably will be using more current.

If you don't want to lose the bass, use a ±7V supply instead of a 14V supply, so you can eliminate the output capacitor.
capacitor that connects to the speaker? Then I'm asking to use a bit too much current and possibly transistors that will go as hot as an oven goes when a thanksgiving turkey is 100% done.

Are you sure it's the amplifier that is limiting the bass, not the loudspeaker?
It is probably both but my budget is a bit too terrible to change the speakers.
 
Top