Designer’s Guide Community

jorsh

New Member

Offline

Posts: 2

Why is not good this design for RF. �
Nov 22^nd, 2006, 3:27am

Hi guys,

I am new on the block!!.
Could somebody help me to identify why this design is not so good. We are having problems with Immunity EMC. If It is applied conducted disturbance through the PCB where the circuit is build the 5V output is experiencing an improper behavior. The power supply output is giving an DC offset when the disturbance is injected. 5.0 to 5.2V. An error of 0.2 is enough to consider the complete design as invalid.

I am not an expert at all but i can see that the schematic it is not usual. I can see the following topics:
1.- It is used the non inverting input of the OP-Amp as a feed back signal of the circuit.
2.- There is not used neither capacitor de-coupling nor electrolytic capacitor to filter the output signal.

The image attached shows the design of the 5V power supply.

Your help is highly appreciated.
Kind regards
Jorsh

Back to top

5V_ps.jpg

IP Logged

SATurn

Community Member

Offline

Posts: 53

Re: Why is not good this design for RF.
Reply #1 - Nov 24^th, 2006, 2:43pm

Hello,

This topology usually is used when it is not possible to use NPN transistor at the output stage. So, designer must use the non-inverting input of the OpAmp to stabilize the circuit. Of course, stabilizing this circuit is more difficult than a feedback uses inverting input of the OpAmp. The main reason for using an PNP at the output stage can be design of a low drop-out regulator (when the output regulated voltage must be close to the VCC).

Also, because of stability issues you must be very careful especially when you are going to put a filtering capacitor at the output.

SATurn

Back to top

IP Logged

loose-electron

Senior Fellow

Offline

Best Design Tool =
Capable Designers

Posts: 1638
San Diego California

Re: Why is not good this design for RF. �
Reply #2 - Nov 27^th, 2006, 12:22pm

The schematic shown is a "hybrid-discrete" version of an LDO (Low Drop Out) voltage regulator.

May I suggest - There are dozens of LDO's available with all the parts and pieces in one simple small package. Get one of those and plug it in in place of the what you have. SImpler, smaller, cheaper as well.

Jerry

Back to top

Jerry Twomey
www.effectiveelectrons.com
Read My Electronic Design Column Here
Contract IC-PCB-System Design - Analog, Mixed Signal, RF & Medical

IP Logged

jorsh

New Member

Offline

Posts: 2

Re: Why is not good this design for RF. �
Reply #3 - Nov 27^th, 2006, 1:53pm

Thank you very much for your answers,

Saturn,

When you say that I must be very careful using output filter capacitor you mean that it is �required special care or otherwise that it is not allowed?
In fact the circuit itself presents a kind of instability. The output voltage is oscillating 1.5Vpk-pk around Vcc at a frequency of 200kHz. �But If I add an output filter the �oscillation disappears.

Jerry,

Your suggestion is welcomed I will investigate further on this. But now I need a quick fix that solve the issue. Could you give me a link for the �hybrid-discrete� LDO. I have searched in google without success.

Regards
Jorsh

Back to top

IP Logged

loose-electron Senior Fellow Offline Best Design Tool = Capable Designers Posts: 1638 San Diego California	Re: Why is not good this design for RF. � Reply #4 - Nov 28^th, 2006, 9:21am The quick fix to a working system is to plug a pre-packaged LDO IC in place here. Doing your own version of something that exists from multiple sources does not make sense. Google "LDO voltage regulator" and you will find plenty of information. Order the part thru Digi Key and you are good to go. Jerry
Back to top	Jerry Twomey www.effectiveelectrons.com Read My Electronic Design Column Here Contract IC-PCB-System Design - Analog, Mixed Signal, RF & Medical IP Logged