Voltage drops when PA turns on

[chihhao]

When I probed Vbat of PA by oscilloscope and I found the
voltage dropped when PA turned on.
If I removed decoupling capacitor from Vbat, it dropped even more.
Could someone tell me why ?
Thanks for your help.

 

[IanP]

What do you use as power source: a battery or a power supply? And if it is power supply, is it regulated or unregulated?

 

[biff44]

battery resistance

 

[habib83]

hello,
i think this problem is duo to oscillation of your power amplifier.
you can see frequency of your amp. with a spectroum analyzer.
yours.

 

[chihhao]

I used power supply to be the voltage source.
The question I mentioned is not a problem. It's a fact.
It does really drop a little bit voltage when PA turns on and it seems a normal
behavior.Please take a look at attached picture.
Blue line is Vramp signal & black line is Vbat voltage.
Vbat voltage dropped a little bit around 0.06V (from 3.4 dropped to 3.34)
when Vramp signal feeded in.
Tx performance is very good. NO abnormal behavior.
I just do not know why it dropped?
The current from power supply can not deliver to Vbat in time ?

PS: Power supply is Agilent E3642A

 

[vfone]

The DC resistance of the PCB trace between battery connector and PA is too high. Need a shorter and thicker trace. Anyway the design needs a low ESR capacitor ( >33uF), close to the PA.
Another reason of dropping (assuming the battery resistance is ok) could be imperfect contact between battery and battery-connector.
A maximum 100mV dropping during burst in GSM-900, full power, is ok. Beware that during bad VSWR the PA current is even higher.

 

[chihhao]

Hi vfone
One additional question is why the voltage did not always drop due to
resistance but it only dropped when PA turned on ?
Thanks.

 

[vfone]

Because between bursts the PA is OFF and is not taken any DC current. Between bursts you have receive, idle or sleep DC currents. They are very low currents (mA) compare to full power DC current that could be up to 1.7A (or 2A under bad VSWR). Higher DC current means higher DC drop.

 

[mckinson]

Hi ,Vfone . Waht issue maybe produce if the DC drop is too high !
BTW ,if use the battery , what DC drop is normal ?
Best regards !

 

[vfone]

First problem that can appear is to cannot get the required power. Also because the DC is lower than target supply voltage, to reach the output power the control loop increase the power, and put the PA in deep saturation, increasing the switching transients. At some point even the modulation spectrum and phase error could be affected (due to dramatically phase change that happened in deep saturation).

 

[mckinson]

Thank you Vfone .
I have a question ,how to judge a PA if work in deep saturation .Do you have any fast way to judge ?Do you think by testing the PA efficiency can judge ?