Dear all memeber in this forum,
I do some stupid work.

I intend to have 3.3 V output by using the LM1117T 3.3V regulator (with 2 capacitor 20microfara). Unfortunately this regulator requires input of 4.75 to 5.25V.

I used a 7805 regulator (with 100micro Fara and 1micro fara capacitor) to obtain 5V, and then connect 5V output to the input of 3.3V regulator. Can you guess the result that I got?

I checked the voltage in two regulators, 7805 give 5.02V (good). But 3.3V regulator gives 0V output  :rolleyes: . Does anyone have experient in this issue? Can we use two regulator in Series connection? What is the common solution to supplied 5V input for LM1117T?

You have probably connected the LM1117T incorrectly.

What is the exact model number of the LM1117T? Is it an LM1117T-3.3 or LM1117T-ADJ? The National Semiconductor website shows an input of 4.75 to 5.25 vdc which is probably the ideal voltage. However, the maximum input voltage is 20 vdc as per the datasheet for the LM1117T-3.3.

Having 0 vdc output tells me that you either have a capacitor mounted backwards (polarity) or more likely the incorrect wiring or connections as Leon stated.  It may also be possible that you just have a "bad" LM1117T-3.3 regulator. Try another one with the correct connections and Capacitor placement.

Pthien,
The LM1117T-3.3 can take far more than 5 volts, as others have pointed out.  I myself use it on a board that feeds it 12 volts without any problems.

However, I can almost guarantee that you have a wiring error.  Why?  Because I have made the same mistake.  Everyone is familiar with the 7805, where the pins from left to right are:

Input  Ground  Output
Very logical.  The LM1117T-3.3, however, looks like this:

Ground Output Input
Completely different and not at all intuitive.  If you have wired the 3.3 volt regulator up the same way as the 7805, you have it wrong.