This is my first time using a transistor as a simple amplifier, just wanted to expand my horizons wider than Op amps.... and I wanted to make sure I was biasing it correctly.  Is a slight amount of distortion normal in this design?



Feel free to copy it and whatnot, it's nothing exciting or original at all.

-Colin

Looks okay to me. To minimize distortion, check the collector with the voltmeter, and adjust one or both resistors until you get 1/2 supply voltage at that point. This allows maximum output swing before clipping.

-Joe

Quote from: Joe DavissonLooks okay to me. To minimize distortion, check the collector with the voltmeter, and adjust one or both resistors until you get 1/2 supply voltage at that point. This allows maximum output swing before clipping.

-Joe

Thanks for the tip, Joe.  Also-I'm noticing a significant amount of high end loss, is this normal for this sort of thing?

-Colin

yeah like joe said adjust the collector resistor so the collector is about half dc voltage of the power supply.  my quick calculations yield 680 ohms not 2.2k id suggest changing that or just using a trim if you want to account for device differences.  you could also leave that 2.2k and change the base resistor to 270k.  i just think the bias is a little off as is, my calculations are showing the collector at 2.7Vdc.  but these are just my calculations which could be wrong.

also is the output electrolytic cap backwards?  i dont know if it matters but i usually do it the other way because the output should have close to 0Vdc and the collector end should be close to 4.5Vdc.  not sure about that though.

Smaller value in & out caps will reduce bass, add treble.

SOCKETS

RDV

It's bassy probably because the input impedance is really low. Under 5k. Massive tone sucking. However, the circuit will amplify about as much as it possibly can nonetheless. Probably causing some distortion. Some of that gain is eaten up by the low input impedance, some by negative feedback via the 100k resistor. It will bias well enough pretty consistently, though it may be insufficient depending on your needs.

turn round your output cap, or it`s gonna spit...

Following on from Tim's comments, you can divide the 100k feedback resistor into two (a couple of 51ks should do) and decouple the junction with a suitably large cap (1u or larger). This would up the input impedance to around 51k and still run the amp at high gain.  

Ok, not brilliant (though a lot better!), but you can up the input impedance by using a higher gain trannie (allows for a larger feedback resistor).

gez, are you sure that splitting & decoupling the feedback resistor will increase input impedance? because, it is supplying negative feedback after all. (why isn't there an emoticom for 'totally mystified')

When you split the feedback R it removes the AC feedback but keeps the DC feedback (which makes the bias more stable).  You actually quite a large cap to ensure the AC feedback is removed over the whole audio range.

The input impedance isn't 51k, it's 51k in parallel with the input impedance of the transistor stage which is somewhat lower- in the 2.2k region.

The low input Z problem start by choosing such a low Rc value.  Perhaps kick this up to 10k or so.

Further improvements involve adding an emitter resistor and trading input Z vs. gain.  A non feedback biasing scheme would help too.

Boof, you pipped me to the post!  I went out shopping this morning and remembered than the resistance seen at the base is still low (as you said, an emitter resistor would help here) and is in parallel.  Must stop making early morning posts!  :oops:

Paul, as Boof mentioned, the feedback resistor still provides the requisite DC bias but AC feedback is decoupled.  An input signal will 'see' the 51k in parallel with the resistance seen at the base.  The other 51k in the divider is now in parallel with the collector resistor, but (can't see the schematic while I'm typing this) I think it was just over 2k so there's not much reduction in gain.

Quote from: BoofheadYou actually quite a large cap to ensure the AC feedback is removed over the whole audio range.

Agreed, I was implying that nothing smaller than 1u should be chosen.  For guitar (but probably not bass) 1u would work ok.

Interesting, thanks everyone for the suggestions.  I think I might stick to op amps for amplification and transistors for switching, though

Where is the calculator that gave you those values, downweverything  (or the formulas if you were doing it yourself)?

-Colin

Quote from: anonymousexperimentalistI think I might stick to op amps for amplification and transistors for switching, though

It's worth spending some time with them as they're a handy tool for the designer.  

Other biasing techniques can provide higher input impedance, such as bootstrapping in conjunction with divider bias (as Boof mentioned).  Using higher gain trannies/darlingtons and a emitter resistor would help with the design you've posted - this WOULD make decoupling AC feedback worthwhile (but not as is).

Trannies are useful in follower applications - can provide a fair amount of current and tend to have a lower output impedance than FETs used in the same way.