Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

What Op amp to use

Status
Not open for further replies.

quintessential

New Member
hey guys,
so I'm doing crude PSRR testing on voltage regulators. My circuit is simple, a signal generator generates a ripple(say 200mv peak to peak) that i superimpose on the power supply of my regulators with the aid of an op Amp. with a spectrum analyzer i study the ripple at the output, my problem is i would like to test PSRR at frequencies up to 20Mhz and some of my regulators sap up to 1A of current so i'm wondering if there is an Op Amp out there that can source an Amp of current or greater and is unity gain stable up to 20Mhz?.......or if there is another way i can impose a ripple on a DC source and supply it to my regulators,thanks
 
quintessential said:
hey guys,
so I'm doing crude PSRR testing on voltage regulators. My circuit is simple, a signal generator generates a ripple(say 200mv peak to peak) that i superimpose on the power supply of my regulators with the aid of an op Amp. with a spectrum analyzer i study the ripple at the output, my problem is i would like to test PSRR at frequencies up to 20Mhz and some of my regulators sap up to 1A of current so i'm wondering if there is an Op Amp out there that can source an Amp of current or greater and is unity gain stable up to 20Mhz?.......or if there is another way i can impose a ripple on a DC source and supply it to my regulators,thanks

What you ask for out of an opamp is rather unusual and not at all common. This kind of problem is typically solved by designing your own power amplifier. With that said, someone has already done it.

Take a look at Apex opamps like this one:

**broken link removed**

You will not be happy with the price.

Here is the power amp selector guide page

**broken link removed**
 
Any other options

ahh great info!! The price is a little upsetting but this is the only method i know of to test PSRR, i just cant find any other way to generate a rippled voltage supply so i can measure the rippled response especially at such high frequencies
 
quintessential said:
ahh great info!! The price is a little upsetting but this is the only method i know of to test PSRR, i just cant find any other way to generate a rippled voltage supply so i can measure the rippled response especially at such high frequencies

Are you sure your rejection is load current dependent? You may be able to greatly reduce your load current and thus test with lighter loads (i.e. more practical amplifier circuits)
 
Why can't you just power your signal generator via an isolation transformer (If it's case is grounded via mains) and connect as per attached diagram? The capacitor is only needed if the generator can't handle DC on it's output. Most signal generators are designed to operate into a 50 ohm load. You'll also have to compensate for the DC voltage drop across the 3R3 resistor but I assume you are powering the whole test jig with an adjustable power supply anyway.
 

Attachments

  • ripple.GIF
    ripple.GIF
    3.1 KB · Views: 320
All the power opamps I know with that current output and GBW product aren't unity gain stable unfortunately. I'll poke around some and see what I can find.
 
thanks alot for the diagram, but if you would be so kind as to clarify some connections for me. I do have a variable power supply(Agilent 6632A) and from what i see of isolation transformers, they have a simple input here output there user interface.So in the diagram "Power in" is where my output from my agilent goes into the input of my transformer, the output of the transformer powers my signal generator and the output of my signal generator(after passing through a resistor and capacitor) becomes the positive node of my rippled power supply.is the 3r3 an internal component of the signal generator or would i have to take care of that parallel connection and also what purpose does it serve?
ripple-gif.13759
 
is the 3r3 an internal component of the signal generator or would i have to take care of that parallel connection and also what purpose does it serve?
It is a separate component (3.3 ohm 5Watt resistor). The 47 ohm is also a separate component (A 1/4 watt will do).
It forms a voltage divider with the 47 ohm resistor to match the 50 ohm impedance of the signal gen. You would generate 2.8Vac to get apx 200mVac ripple for your regulator under test. There will also be apx 3.3Vdc across it when the regulator draws 1 Amp.
but if you would be so kind as to clarify some connections for me
I've clarified it a bit for you:
 

Attachments

  • ripple2.GIF
    ripple2.GIF
    5.4 KB · Views: 184
kchriste said:
It is a separate component (3.3 ohm 5Watt resistor). The 47 ohm is also a separate component (A 1/4 watt will do).
It forms a voltage divider with the 47 ohm resistor to match the 50 ohm impedance of the signal gen. You would generate 2.8Vac to get apx 200mVac ripple for your regulator under test. There will also be apx 3.3Vdc across it when the regulator draws 1 Amp.

I've clarified it a bit for you:

thanks alot for that, the circuit is clearer to me now....why cant i just plug my signal generator into the mains?
 
why cant i just plug my signal generator into the mains?
Because usually the case of the power supply (and thus the - terminal) and the signal gen are both grounded via the AC plug. If this is the case, then sparks will fly when you connect the common (Grounded) output terminal of the signal gen to the positive output of the power supply. Scopes suffer from this too, so only connect the ground clip from the scope probe to the same points that connect to the - terminal of the power supply.
 
kchriste said:
Why can't you just power your signal generator via an isolation transformer (If it's case is grounded via mains) and connect as per attached diagram? The capacitor is only needed if the generator can't handle DC on it's output. Most signal generators are designed to operate into a 50 ohm load. You'll also have to compensate for the DC voltage drop across the 3R3 resistor but I assume you are powering the whole test jig with an adjustable power supply anyway.

the circuit does indeed work(produces rippled voltage supply up to 20MHz) without a capacitor at all,but my signal is still attenuated, the higher the frequency the higher the attenuation. with just resistors its hard to figure out why this is so,i mean i can understand attenuation with a cap but without a cap am i right in noting that attenuation is strange?
 
It is due to the input capacitance of the regulator under test and stray capacitance/inductance of the wires between the signal gen and the regulator. But you can easily compensate by adjusting the signal generator output as you vary the frequency.
 
kchriste said:
It is due to the input capacitance of the regulator under test and stray capacitance/inductance of the wires between the signal gen and the regulator. But you can easily compensate by adjusting the signal generator output as you vary the frequency.

oh no i havent even connected my DUT yet, right now i just have my scope probe tip connected to the (+) and the probe ground tip connected to the (-)
 
I may be getting over elaborate but what do you think about these additions to the circuit?
The goal is to one day have the PSRR test automated so ideally i was looking for the same amplitude of ripple on the output even when the frequency changes but this might be out of my reach....manual testing might be the only way
 

Attachments

  • pop-pop[2].jpg
    pop-pop[2].jpg
    60 KB · Views: 166
Last edited:
It will definitely change the frequency response of the circuit. Whether you can make it flat from DC to 10Mhz is another matter. How much attenuation are you seeing at the future DUT connection point? How much does the level change when you sweep from 100Hz to 10Mhz? Have you checked that the output at the signal generator is flat also?
 
kchriste said:
It will definitely change the frequency response of the circuit. Whether you can make it flat from DC to 10Mhz is another matter. How much attenuation are you seeing at the future DUT connection point? How much does the level change when you sweep from 100Hz to 10Mhz? Have you checked that the output at the signal generator is flat also?

from 3Khz to about 1Mhz the response is linear; from 1Mhz on the signal is attenuated quite significantly,lets say it gradually degrades to about 1/4 of its pre 1MHz strength, i've never used an RF choke before but i'm hoping it will compensate f i slide it in after the 3R3;despite the fact i cant tell where the high frequency attenuation is coming from.the signal gen response is flat and i pray to god the DC source has nothing to do with it.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top