Need suggestion on a discreet ldo design.

Thread Starter

opa627bm

Joined Nov 26, 2011
16
Hello everyone,
I got assigened to design a voltage source that is capable supplying +-100ma at -+10V. After I build the circuit, I notice that there is some fluctuation at the output side. I probed with scope and there is no oscillation at the gate or opamp. The Vref into the opamp is extremely stable 0.70160X volts. but my output fluctuates around 1mv. The mosfet is FR5305. I need some advise on getting this thing more stable....

Thank you!

Update 1. Yes.. I draw the arrow on mosfet backwards
Update 2. It cannot sink current... but I don't know how to do push pull since it will turn both transistors on in the middle region..
 

Attachments

Thread Starter

opa627bm

Joined Nov 26, 2011
16
Did the assignment say to do better than 1/10 of 1% stability from zero to full load? How much better?
Hello, I need to design a voltage source for my research lab. It need to be able to source and sink 100ma and output +- 10V. It need to be as quite as possible because we are using this to generate a B field to polarize the atom that is being probed by the laser. There is no such ldo on the market that is capable of source and sink current at same time...

Now I have a new idea...

Vref(adjustable)->Gainstage(so it output from near -12 all the way to the +12)->Bufferstage(referenced at middle of the rail so it outputs +- and deliver current)...
 

Thread Starter

opa627bm

Joined Nov 26, 2011
16
Did you notice that your op-amp being used as the error amplifier has absolutely no loop compensation?

HINT: loop compensation is the feedback network from the output back to the inverting input of an amplifier.

If you are trying to build an LDO regulator, you might find this article helpful:

http://www.ti.com/lit/ug/snoa826/snoa826.pdf
I build mine based on this article ....
http://www.maximintegrated.com/app-notes/index.mvp/id/3657

Could you tell me why the opamp is in inverting mode ?
 

THE_RB

Joined Feb 11, 2008
5,438
I'm missing something...

Why does it need to be an "LDO"?

The only time you need LDO is because you run from batteries and have very little headroom. If this is a lab power supply you don't need LDO it just needs to be a precision +/- 10v DC supply, which you can buy. Or google a schematic.
 

AnalogKid

Joined Aug 1, 2013
10,987
I did some stuff for some electro-chemists a while back. They, too, were blasting atoms with lasers, and needed a fast preamp with femtoamp input current. It's always about the noise with those boys.

My first thought also was a high-current opamp, but I don't think you'll get the noise performance you need.

Your original circuit and the one from Maxim is a source only. It will not sink anywhere near 100 mA of current, as that is being done only by the two feedback resistors. To do what you want you will need either a push-pull output or an audio class-A style output with a variable pull up and a constant current pull down (or vice versa). Both of these are medium-complexity circuits, a bit more than what you have now.

1. Is +/-12V your only option for the power source, or is something larger available (+/-15, +/-18, etc.)?

2. How clean is the source? Cheap linear, expensive linear, switcher...? A good regulator can clean up a bad source, but only up to a point.

3. What is your noise requirement for the 10V output. This usually is given in nV per root hertz over a specified bandwidth. Or maybe you need very low noise only at a certain frequency, or narrow frequency band?

4. What is the transient response requirement? That is, will the output see rapid changes in the load current, like 50 mA steps or pulses? One way to make a power supply quieter is to make it slower, but only if you know the minimum speed you need to support.

ak
 

Thread Starter

opa627bm

Joined Nov 26, 2011
16
I'm missing something...

Why does it need to be an "LDO"?

The only time you need LDO is because you run from batteries and have very little headroom. If this is a lab power supply you don't need LDO it just needs to be a precision +/- 10v DC supply, which you can buy. Or google a schematic.
Hello, we are going to power this from two 12V batteries to achieve lowest noise possible.
 

Thread Starter

opa627bm

Joined Nov 26, 2011
16
I did some stuff for some electro-chemists a while back. They, too, were blasting atoms with lasers, and needed a fast preamp with femtoamp input current. It's always about the noise with those boys.

My first thought also was a high-current opamp, but I don't think you'll get the noise performance you need.

Your original circuit and the one from Maxim is a source only. It will not sink anywhere near 100 mA of current, as that is being done only by the two feedback resistors. To do what you want you will need either a push-pull output or an audio class-A style output with a variable pull up and a constant current pull down (or vice versa). Both of these are medium-complexity circuits, a bit more than what you have now.

1. Is +/-12V your only option for the power source, or is something larger available (+/-15, +/-18, etc.)?

2. How clean is the source? Cheap linear, expensive linear, switcher...? A good regulator can clean up a bad source, but only up to a point.

3. What is your noise requirement for the 10V output. This usually is given in nV per root hertz over a specified bandwidth. Or maybe you need very low noise only at a certain frequency, or narrow frequency band?

4. What is the transient response requirement? That is, will the output see rapid changes in the load current, like 50 mA steps or pulses? One way to make a power supply quieter is to make it slower, but only if you know the minimum speed you need to support.

ak
1. We deiced to power them up with 2 12V batteries..
2. We have some good Agilent linear powersupply laying around.
3. As low as possible..
4. We dont have transient response requirement, We are going to adjusting the voltage by hand with 10 turn trimpot.
 

#12

Joined Nov 30, 2010
18,224
Let's see, are these the conditions of your request?

Completely rigid about which batteries to use and the end voltage you want out of them, don't know the definition of regulation, don't know the definition of transient response, don't know how to filter noise out, so, Please design a really, really, state of the art, noiseless circuit under severe limitations that only depend on the fact that I don't know the difference between noise and regulation.

Right?
 
Last edited:

AnalogKid

Joined Aug 1, 2013
10,987
Batteries aren't automatically better, as in lower noise, than offline supplies. There are some high-performance lab supplies that are switchers followed extra output filters. Before, I said that a high current op amp might not have the noise performance you want, but a true source/sink regulator isn't trivial. If you want to build rather than buy, a high current op amp followed by extra filters might be the best compromise. The problem is the +/-12V power. Most power op amps won't make 100 mA so close to the rail. Think about a couple of 15V 1A industrial linears, pi section filters, your final regulators, and another pi filter.

ak
 

#12

Joined Nov 30, 2010
18,224
Hey, KISS. Your link doesn't work for me.
Still, I looked it up by myself. Keithley+voltage
Good stuff. Looks expensive.
 
Top