| Author |
Message |
veeraselvi
Joined: 21 Aug 2009
Posts: 6
Location: india
|
 24 Oct 2009 10:34 Amplifier switch on time |
|
|
|
|
Hi,
I am designing power supply sequencing for RF monolithic amplifier. I want to know the delay time of amplifier response. After switch on the power supply, how many nS that amplifier will take to give stabilized output gain?
What are all the parameter has taken into account to calculate ?
regards,
Veeraselvi.B
|
|
| Back to top |
|
 |
FvM
Joined: 22 Jan 2008
Posts: 5151
Helped: 766
Location: Bochum, Germany
|
24 Oct 2009 18:05 Amplifier switch on time |
|
|
|
|
As most RF amolifiers are AC coupled, you may want to check primarly the coupling time constants. A transient simulation
or empirical measurement is most likely necessary to determine the exact recovery time.
|
|
| Back to top |
|
|
biff44
Joined: 24 Dec 2004
Posts: 1833
Helped: 244
Location: New England, USA
|
27 Oct 2009 3:14 Re: Amplifier switch on time |
|
|
|
|
| Saw a wireless lan job recently that was all screwed up because the were trying to transmit before the DC bias levels had settled out in the transmit amplifier. Sorry, I do not remember how many ns it took to settle. It is probably a function of how much bias line capacitance you have to charge up. Unfortunately, with today's modern modulation schemes, you need a vector signal analyzer to measure error vector magnitude at turn-on to see the effect.
|
|
| Back to top |
|
|
Google
AdSense
|
27 Oct 2009 3:14 Ads |
|
|
|
|
|
|
| Back to top |
|
|
tlcfybj
Joined: 27 Oct 2009
Posts: 2
Location: China
|
28 Oct 2009 15:44 Re: Amplifier switch on time |
|
|
|
|
| amplifier response time (<5ns) is much shorter than bias stable time. you should make your bias stable time to be nano seconds level first. but it is challenging.
|
|
| Back to top |
|
|
vfone
Joined: 10 Oct 2001
Posts: 2326
Helped: 326
|
28 Oct 2009 16:11 Re: Amplifier switch on time |
|
|
|
|
This was a problem for the first CDMA IS95 handset power amplifiers. The system designers tried to shut down the PA during frames through the bias circuit, but the settle time of the bias circuit was too slow. So, all the CDMA PA’s (or WCDMA) have the bias always on.
For very fast settle time measurements you can use a crystal detector which has a response in a range of 200 picoseconds. Usually the measurement is done between 10% and 90% of the output power.
|
|
| Back to top |
|
|
