How much current gain do I need for a guitar amp build?

Damien83

Member
2015-07-20 2:20 pm
Hey Peeps

Please forgive me if I make any mistakes here as I am a newb.
Wondering if someone could please help me though -

I'm building a guitar amp from scratch from discrete components, trying to opt away from op-amps (excuse the pun hehe) and do it the good old fashioned way to learn as much as possible along the way and to say that "look mum I did this all by myself with no hands" ..well with hands but..you know what I mean.

I've got my voltage gain sorted out.

I worked out that my pickup produces 100mv of voltage and if I want clean undistorted sound up to the rails on my +/- 37 volt power supply and so I need an amplifier with a voltage gain of 370. Is that correct?

Now I just need the current gain ...I have no idea what kind of power my pickups are producing and have no way of measuring it, I have a multimeter with a.c voltage reading but no oscilloscope or anything. I could possibly put a resistor across the pickup and measure voltage drop but..nah, silly idea...I'm begging someone will please just give me the answer.


So I have a push pull stage in darlington pair configuration for my power side of the amp (which I have built on a breadboard) and have calculated the total current gain to be 200 (for one side NPN or PNP) ...does that mean that it has actually current gain of 400? I got the 200 by multiplying the HFE of the first NPN transistor by the HFE of the second transistor.

And I can work backwards from there and work out the current gain of each stage but that doesn't really help as I need to know what I need to amplify in order to be able to build the stages for it.

Also wondering - for power calculation purposes...


Hypothetically if I have a class a common emitter stage with a transistor with a HFE of say 50 and I run 50 volts as the power supply....
will the current gain remain at 50 or does this HFE value only apply if the transistor is run in common collector mode? (as a current amplifier not a voltage amplifier) ?

I am also wondering....is the current output of the small signal important... IE -
If I run an op amp that produces 32ma but a voltage of +/- 37 does that mean that to calculate my power I use ohms law to calculate Volts X amp's = W? (Bearing in mind the root mean squared thing with amps, can't remember what it's called but that 1.4 number that we have to divide by)

I guess my question here is...will the small signal current gain have much effect on the final product (speaker) or is it more about voltage? (as is the case with mosfets being voltage controlled valves as opposed to b.j.t's being current controlled.

Any help appreciated - thanks
 

Attachments

  • 004.jpg
    004.jpg
    608.4 KB · Views: 66
  • Final.jpg
    Final.jpg
    89.2 KB · Views: 66

thoglette

Member
2008-12-20 5:59 am
You're about to discover a whole world of pain and disappointment. A single 2N3055/2955 pair will likely die the moment they meet a real world speaker. And that's before I look at the loop stability of what you're proposing.

The venerable ETI-413 utilised a pair on 40V rails (see here) and had a reasonable reputation for staying alive.

+/-25 would be a happier point for a single pair - but only with adequate heatsinks (not as shown in your photo)

p.s. I went down this route many decades ago. There's no need to repeat all my mistakes!
pps- you're going to want some "tone" which a properly built 2N3055 diff amp WILL NOT provide.
Have a look at some of these
http://www.runoffgroove.com/articles.html
http://www.duncanamps.com/technical/technical.html
http://www.classictubeamps.com/Tonestack.html
 
Last edited:
Hi,

Its a toneless total disaster area waiting to happen / go bang.
Get the service manual of a decent budget amplifier and copy it.

FWIW you can buy a decent used guitar amplifier for far less
(and hear it first) than the cost of sourcing the parts and the
hassle of building it to be a robust reliable workhorse.

Even buying new will far cheaper in the long run, and FWIW
the DSP options in budget amplifiers cannot be homebuilt.

FWIW current gain is pretty irrelevant to the gain structure.
Voltage gain matters, and high input impedances mean
loads of possible current gain in each voltage gain stage.
Except for the speaker the output is not usually heavily
loaded, so the possible current gain is not exercised.

Checkout the awesome FENDER MUSTANG II V.2 .

rgds, sreten.
 
Here's the Marshall 8008, similar to your concept but 3 extra small transistors Marshall Valvestate 8008 Service Manual free download,schematics,datasheets,eeprom bins,pcb,repair info for test equipment and electronics
The op amp doesn't have to go to the rail voltage. the first and second transistor pairs just produce more and more current. The output won't go to the rail voltage if you want music instead of square waves.
That 470 uf cap you have on the output of your input amp is daft. You need a +-15v supply for the op amp and first transistor pair. Many audio and instrument type toroid transformers or package switcher supplies will have that.
One pair 2n3055/mj2955 can produce about 16w into an 8 ohm speaker with massive heatsinks. From On semi they cost almost the same as mj802/mj4502 whish are a lot tougher in the current spec. The MJL4302/MJL4281 cost less and are a lot easier to mount with one screw hole. However counterfeit 2n3055/mj2955 are very cheap and maybe good for 100 ma max. For 4 ohm speakers definitely the transistors with more soa are required. People in the Pacific frequently use 2SA1943/2SC5200 output transistors which are cheaper there, but not here in the US.
For drivers look at MJE15028/29 or 15030/31 which have lot more soa than TIP41c/42/. I wouldn't use no suffix TIP41/42 for that rail voltage.
 
Last edited:
A Tip41/2 has a typical hfe (current gain) of 70 and 2N3055/2955 20. So 20 x 70 = 1400. That means with 1 drive current of 5mA the output current produced will be about 5 X 1400mA = 7A. That is how to work out the minimum requirements. 2N3055/2955 selected on test I have had hfe of 120 and the Tip types 190. The data sheet describes exactly what to expect.

Without heatsinks, the transistors will quickly overheat and fail.
 
Last edited: