# Audio amplifier design, basic questions.

#### Coollestersmooth

Joined Dec 18, 2014
72
So im looking at designing a transistor based amplifier, type AB. and have a few very basic questions.

Im aiming for pushing 10 watts through a four ohm load speaker from a 30V supply.
That being said what sort of current/voltage should i be aiming for at the output.

from what i can gather

30V-2V(allowing for some voltage drops) = 28 Vpp

peak voltage = 14V

Power = V^2 / 2R= 98V/8 ohms = 12 Watts not entirely sure where the 2R comes from in this but have seen it on a few calculations?

current = 14V/4 = 3.5A or would i need to use 8ohms here again giving 14/8= 1.75A?

i also have some questions about the input stage.
i was planning on having it run from the line out jack of an ipod or similar.
when measured for a sine wave output this is around 0.475Vrms or 0.662Vrms.
Measured using a multimeter (very high impedance).
my input current is going to based on the input impedance im assuming? i was planning on putting a 10k pot at the input for volume control, how would i account for this variable resistance? should i just ignore it and and/or just use its lowest resistance value in series with the resistance of the initial transistor stage?

Any help would be greatly appreciated.

Will

#### MikeML

Joined Oct 2, 2009
5,444
You need to convert the expected peak-to-peak voltage across the speaker load to RMS voltage before computing the power.

#### MrChips

Joined Oct 2, 2009
31,087
That's where the 2 comes in.

#### MikeML

Joined Oct 2, 2009
5,444

#### Coollestersmooth

Joined Dec 18, 2014
72
figured that might be the case. wasn't sure if voltage output from a dc lab style psu would already given in rms? if so me calling it Vpp is slightly confusing.
any ideas on where the 8 ohms is coming from?

#### MrChips

Joined Oct 2, 2009
31,087

#### Coollestersmooth

Joined Dec 18, 2014
72
sorry, hadn't seen those posts. so convert to rms and i can use 4 ohms in that calculation?

#### MrChips

Joined Oct 2, 2009
31,087
RMS of VDC is VDC

RMS of Vpp sinewave is (Vpp/2)/√2 = Vp/√2

(Vp/√2)^2 is (Vp^2)/2

That is where you get the 2R.

#### MikeML

Joined Oct 2, 2009
5,444
Audio power = ((Erms)^2)/Rs

#### Coollestersmooth

Joined Dec 18, 2014
72
thank you, realised as soon as i posted that rms of a dc voltage is obvious going to be the dv voltage, derp.
so im about on track for my voltage/current at the output.
now i just need to start working out the sort of current voltage gains i want at each stage.
for the input i'm just going to treat the pot at 0 ohms to give maximum possible current at the input stage.

very impressed at how quick everyones responses have been and helpful you all are, so thank you all!
seems like a great community.

no doubt i shall be back with more questions once i get into the actual design stages.

many thanks,
Will

#### MrChips

Joined Oct 2, 2009
31,087
If you put a 10kΩ pot as volume control at the input then the input impedance is 10kΩ (assuming the input impedance of your input stage is greater than that).
Don't worry too much about the input current. Focus on the voltage gain required.

#### Coollestersmooth

Joined Dec 18, 2014
72
Ok, thank you, will do.

#### Coollestersmooth

Joined Dec 18, 2014
72
So im trying to design a 10 watt, type ab transistor amplifier and am having some difficulty in selecting transistors for the input pair.

I'm looking to get 10 watts off a 30V supply, and need a total gain of around 21 (26dB odd)
I'm aiming to get around unity gain from the input pair and letting the VAS do most of the legwork, i may up this to say 2x from the input pair if it helps my cause.

I'm having trouble selecting input transistors that will operate off a line in current at the base,
the current values for line in are dismally small
0.475V/10k ohm input res = teeny tiny current.
All the datasheets i can find for transistors that work in my kind of voltage range require higher bias currents at the base just to switch on.

i have a feeling i may just be being silly and need to piggyback my input signal to a small dc bias voltage using a potential divider off my voltage rails, but i'm really not sure, i'm kinda new at this...

any help would be great
many thanks, Will

#### MrChips

Joined Oct 2, 2009
31,087
Why not stick to one thread?

I'll merge the two.

#### MrChips

Joined Oct 2, 2009
31,087
I may suggest a FET input with MPF102.

#12

#### Jony130

Joined Feb 17, 2009
5,539
But why you want amplifier the input current ??

#### Coollestersmooth

Joined Dec 18, 2014
72
apologies, wasn't too sure on whether anyone would still be looking at previously answered questions, i shall stick to one in future.

I was wanting to make a purely BJT based amp as this is more of a learning exercise than anything else.

Unless i have completely misunderstood how the input stage works i was applying my input voltage signal to the base of one of the input paired transistors, however from what i can gather from most datasheets they require a minimum current to be applied, which my input signal simply doesn't provide. i may be misreading the data sheets, like i say, im pretty new at this ^^

#### Jony130

Joined Feb 17, 2009
5,539
You are completely wrong. You never herd about "biasing circuit" and why we use them ?
Input current is not equal to Vin/Rin, and we don't want to amplifier any current, in fact we want Iin = 0A . All we want from the input stage is to amplifier the voltage but first we must bias the transistor into the linear region.

#### MrChips

Joined Oct 2, 2009
31,087
BJT are current devices and have low input resistances. An FET is a voltage device and since the input resistance is very high it requires very little current to be turned on.

#### flat5

Joined Nov 13, 2008
403