Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Discrete Op-Amp Design

Status
Not open for further replies.

garu525

Newbie level 4
Joined
Feb 14, 2016
Messages
6
Helped
1
Reputation
2
Reaction score
1
Trophy points
3
Activity points
56
I have a few questions regarding my project. I'm a CompE major student so my analog skills are rusty. Thanks in advance, all my questions are in bold.

Design an amplifier which has the following capabilities:

1. Two inputs one plus , one negative
2. Capable of driving a 50 K ohm load to +/- 10 V Vout at 10v with RL of 50K?
3. Frequency response DC - 10MHz + 2 db. Not sure about this, Gain x BW?
4. The input impedance must be at least 1 Megohm
5. The output impedance must be less than 50 ohms
6. Only discrete devices (BJT, FET, TUBES) may be used.
7. This amplifier will be packaged in such a way that it does not exceed more than 3 inches on any side.
8. Temperature range: 25 -- 100oC Just sum the temperatures from datasheet?
9. The gain must be between 1000 - 1500
 

Hi,

To 2) ...and drive -10V with a load of 50k to GND
To 3) gain flatness from DC to 10MHz: -0dB ... +2dB
To 8) Suming makes no sense. It's just the range of ambient temperature your opamp should be opareabele (within given specifications).

Klaus
 
8. Temperature range: 25 -- 100oC Just sum the temperatures from datasheet?[/I]
Basically this means that the opamp must be able to operate in that temperature range. Normally the hottest part of the circuit - usually, but not always, the internal junction of a transistor - must not exceed its maximum operating temp when running under load.

It's a complex issue in practice. When power is dissipated in a transistor, its internal (junction) temperature rises above the ambient temp. The increase in temp is determined by the amount of power and the junction-to-ambient thermal resistance.

Example: a typical low-power transistor may have a j-a thermal resistance (theta) of 200°C/W and a max junction temp of 150°C. If the max ambient temp is 100°C, then the maximum allowable rise in junction temp is 50°C. This means that, with a theta of 200°C/W, it should not be allowed to dissipate more than 0.25W.

- - - Updated - - -

Note: The condition of maximum power dissipation is not necessarily at maximum load.
 
To 3) gain flatness from DC to 10MHz: -0dB ... +2dB
Sure? I fear the "Frequency response" specification is rather unclear. We would expect a GBW specification for an OP or a closed loop BW specification for an amplifier with fixed gain. It's not even clear if the amplifier is originally specified as OP-Amp, the term doesn't appear in the quoted specification.
 

From 1. this requires a differential input circuit.
2. and 5. this means that you really need a + and - power supply. Because of this requirements and its small size valves ( tubes) are ruled out unless you have +- 200V supplies. Also that the supplies must be more then 10V, +- 15V would seem a good figure.
3. I think this should read +- 2dB at 10 MHZ. Else the frequency at which the increase in gain occurs should have been stated. This means that buffers will be required in between amplifier stages.
4. Means a FET or a darlington JBT must be the input stage.
5. Emitter follower would OK for this.
6. done!
7. With transistors there would be no problem with the size.
8. With an upper temperature range of 100 degs C and a maximum junction temp. of 150 degs C the dissippation of devices must be checked and changes in operating points, due to leakage current variation (use DC feedback?).
Frank
 
Thanks for all your replies. With regards to the project assigned to me, it is "Discrete Op-Amp" so I need to make an amp that has 2 inputs (+/-) with the specs listed on top. I'm starting to have some ideas on how to start the circuit but this part still confuses me:

3. Frequency response DC - 10MHz + 2 db. (to gain flatness from DC to 10MHz: -0dB ... +2dB)

What "to gain flatness" means? Does it mean that as I increase the frequency from 0 to 10MHz the Vo should still be flat?

About the temperature, I've read somewhere to glue the transistors back-to-back. I was thinking of gluing tiny heatsinks.

Thanks again all,

- - - Updated - - -

3. I think this should read +- 2dB at 10 MHZ. Else the frequency at which the increase in gain occurs should have been stated. This means that buffers will be required in between amplifier stages.

You're right it's 10MHz +-2Db, although my professor said in reality it would be around 17MHz +-3Db. Yes, I'm planning on using a voltage buffer on every stage (my professor said it would take at least 3 stages because of Miller effect, etc...).

8. With an upper temperature range of 100 degs C and a maximum junction temp. of 150 degs C the dissippation of devices must be checked and changes in operating points, due to leakage current variation (use DC feedback?).
Frank

Please elaborate, maybe an example of DC feedback to avoid current leakage.

Thank you
 

Hi,

Sure? I fear the "Frequency response" specification is rather unclear. We would expect a GBW specification for an OP or a closed loop BW specification for an amplifier with fixed gain. It's not even clear if the amplifier is originally specified as OP-Amp, the term doesn't appear in the quoted specification.
No, I´m not sure. This just was my assumption.

My idea is that:
9. The gain must be between 1000 - 1500
..means the open loop gain.

and:
3. Frequency response DC - 10MHz + 2 db.
means the closed loop flatness.
But then the desired gain is missing.

For example: desired closed loop gain of +12dB. Will give a relatively flat gain of about 12dB from DC to (lets say 1MHz), then it could be +14dB at 5MHz (because of phase shift), before it begins to decrease because of GBW.
..But all this are assupmtions.

Klaus
 

About the temperature, I've read somewhere to glue the transistors back-to-back. I was thinking of gluing tiny heatsinks.
The purpose of gluing transistors together is to keep them at the same temperature (e.g. to minimize offset drift) whereas heatsinks are for cooling. Mounting them on the same heatsink will also help to reduce temperature differences but will also introduce a time lag in reaching thermal equilibrium.
 

As your Professor says, there are really three completely separate functions that require at least three separate stages.

The input stage will need to be a differential amplifier with a suitably high input impedance. There is no specification for dc offset voltage or dc drift with temperature which eases things a bit.

There is also no mention of input common mode range, which becomes important if this is to be used as a unity gain buffer with +/- 10v output swing.

The power output stage comes next, must be able to deliver +/- 10v into 50K, but then it also says 50 ohm maximum output impedance. So we need at least a +/- 200mA output drive capability right down to dc, preferably more.

We then need sufficient voltage gain to link the two together to get our X1,000 to x1,500 overall gain figure.

What is not clear is if the 10 Mhz bandwidth is the open loop requirement at full gain, or if it only has to do it closed loop as a buffer with unity overall voltage gain.

In other words 10Mhz GBW which I suspect is the real more realistic requirement.
That point needs some clarification.
 
Thank you all for the advice. I'll be starting to sketch some diagrams and update you all on my progress. Hopefully you guys can help me make it simpler and efficient, as in professional level kinda.

Thanks again!
 

So I've started a spice circuit of my project based on what I've read from textbooks. I've attached the picture of the circuit and the outputs at each stage. My question is to how to avoid clippings at max. Thanks! pspice.JPGstage 1.JPGstage 2.JPGstage 3.JPG
 
Hi,

My question is to how to avoid clippings
Reduce input voltage
....often this happens automatically when you connect the external circuit. Feedback.

An Opamp is not designed to operate without feedback.

Klaus
 

Thanks for quick reply. I tried that but it didn't yield good results, maybe I'm doing it wrong [feedback line with 1meg resistor]. Please advise.
 

Try increasing both the dc supply voltages, as these are not specified in the requirements, its a genuine fudge.
That should solve the clipping problem.

It needs to operate with say 10mV input to get 10v o/p (gain x1000).
And also as a unity gain voltage follower with 10v input giving 10v output.

The second case exercises the full common mode input voltage range, whereas the first case does not.
It needs to be able to do both without clipping.
 

Some additional comments:

- The initial doubts about the gain and bandwidth specifications haven't been clarified yet
- Also the other specifications are unclear as long as you can't say if they apply to open or closed loop amplifier operation
- The amplifier in post #12 needs frequency compensation to work stable with feedback
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top