Hi,

I'm currently working on a project involving direct coupled transistor amplifier. My circuit currently consists of a differential pair, level shifter, and output gain stage.

I'm now looking into methods of increasing the output voltage swing of the circuit, as well as improving the input impedance.

I was wondering if anyone would be able to point me in the right direct, as to what methods would help me with this?

Thanks.

 

Can you post the schismatics of what you have don so far.
And in general look at op amp rail to rail datasheet
http://www.analog.com/static/imported-files/data_sheets/AD8031_8032.pdf

 

Can you post the schismatics of what you have don so far.

http://img512.imageshack.us/img512/9797/transistoramplifier.png

 

Looks like a very promising start. For some pretty good ideas, take a look at the simplified schematic of bipolar op-amps.

Your level shifter will severely limit your dynamic range.

Using common emitter output transistors (PNP for pulling toward plus and npn for pulling toward minus) will get you a little bit larger dynamic range on the output.

 

Looks pretty neat for studying such amps.

Have you considered active loads for the differential pair? This will get you a substantial gain increase. Even increasing R3 and R4 would help.

Another possibility is to interpose a second (PNP) differential pair, fed differentially form the first.

Keep us posted.

 

1. Get rid of R7. It limits output swing & is not required because Q5 is a constant current source.
2. Replace Q1 and Q2 with FETs for higher input Z.
3. Replace Q4 with a PNP (emitter to V+) for more gain, and more output swing.
4. The output is across R9, not R10. Get rid of R10 while you're at it.
5. The output will have some crossover distortion. This can be improved on by moving the Q7 & Q8 bases apart by 2 diode drops.

Good Luck, have fun,
Ifixit

 

Thanks for the help. I'll let you know how i get on.

 

Some additional things in addition to the comments above:

As mentioned above, if you replace R3 and R4 with an active load, you will double the transconductance of the differential stage. If you do stick with using load resistors, however, you don't need R3...it looks nice on paper but it doesn't really balance anything.

I often see a small capacitor put across the zener diode in applications as they can be noisy. Alternatively, one could use the current mirror transistor Q6 to derive the current source references for the whole amplifier by just changing emitter resistors, this is often done in ICs.

The circuit may work ok in simulation, but when constructed with discrete components differential pairs and current sources may oscillate at RF due to package parasitic capacitances. If this happens, you should try putting small resistors (~100 ohms) in the emitter legs of Q1 and Q2, and maybe Q5 and Q6.

As configured right now, if the output is at the junction of R10 and R11 and is asked to source or sink more than about 60 mA, the drive to the respective output transistor will be shut down. You should be certain that is what you want based on what loads you expect to drive and the output devices you are using. Either way, you won't be able to drive the negative output swing of the amplifier any lower than a diode drop above Vee; often ICs have a current source connected to the output that can sink some current to make the output go at least to a saturation voltage above Vee on the negative swing.


Good luck!

 

Hi again,

I've tried adding a PNP active load to the differentail input, as shown in the attatchment. I'm gettting the output shown in the other attatchment for an AC input of 0.002V. However, for a 0.001V AC input, i'm getting a constant 10V output.

I'm guessing this is down to a design fault, but i've tried various methods in trying to correct this but can't seem to solve the problem.

Any further help would be appreciated.

Thanks.

 

I haven't found the "stuck to 10 volts" answer yet, but I've never seen better advice. These are some smart people! I suggest you do what you've been told. Delete R7 and R10. Delete R6 and D1 then connect the collector of Q6 to the base of Q3. That allows deleting R5. If you want the additional 50 microamps through Q3 Change R8 to 3.9k (3.77k to be exact). The current in Q5 can be reduced by putting a resistor from its emitter to -10, but that doesn't seem necessary. Put 2 diodes between the bases of Q7 and Q8 and drive them with a connection from the emitter of Q4.

Do these one step at a time, resimming to see if one of them causes a problem.

Then there is the chance that I'm completely wrong. These are just suggestions.

 

1. Your opamp circuit has very high gain so the output will be impossible to stabilize without positive feedback to the inverting input with external resistors.
2. Read up on "off the shelf" opamps regarding feedback to control gain.

Good Luck,
Ifixit