Power amplifier stability

[molloy]

amplifier stability

Hello,
im designing a broadcast power amplifier (so class AB 100W power output in 470-860MHz band). Randomly the amplifier faults without any reason only at a single TV channel (and so in just one frequency) and only in digital mode. I didnt find any oscillation and anyway this happens in some boards of the preseries. No strange behaviour in small signal tests. I make also some power tests without any results. Anyone can suggest some power tests to be sure the amplifier is not oscillating?
regards

 

[vfone]

power amplifier stability

Is time consuming but only a full load-pull test can help you.
You mentioned there are no strange behavior in small signal. Ideally would be to measure large-signal S-parameters:
http://www.nmdg.be/product_mt4463.html

 

[N1UL]

amplifier stability test

What do you mean by :

Randomly the amplifier faults

 

[BigBoss]

power amplifier stability test

If the amplifier stability is critical, by temperature or by avalanche effect, the amplifier may goes to instability...

 

[nandopg]

large signal stability power amplifier

I am assuming that by intermittance you meant that you have intermittent output power cutouts.

This type of intermittence might be caused by excessive current and/or voltage across the RF capacitors in the RF circuit in itslf and/or in the RF by-pass network. If it is the case the capacitor's ceramic will be slowly cooked what, over the time, will transform the ceramic into conductive material destroying the amplifier. During this process it is very commom to get some power cutouts.

PAs in 100 Watts range have considerably high current through the shunt capacitors in the output RF circuit. Did you check if the capacitors you are using withstand the current and voltage in every node of the output matching circuit? Please don't forget to check this values regarding the peak power (not RMS) to which the PA was designed for.

On the low frequency end, low frequency bypass capacitos grounds also high current maily due to sync pulses and other low frequencies components of the video signal. The same effect described above can appear here so that low frequency capacitors must be chosen to withstand current and voltage due to those low frequencies components.

In the long run, if the described above is the problem you are facing, these caps will be degrading to a point they will entirely destroy the amplifier's board, even more in high altitude sites.

NandoPG

 

[N1UL]

amplifier stability explained

To look at the voltage and even more at the current handling ability is the right way to go. I would also take a high output network analyzer , like up to 20 dBm and look at S11 under different level of drive and use a variable delay line to mismatch and to terminate the output for different VSWR over the frequency range, while under say 10 W output. Modern signals have both amplitude and phase changes, they require class A or AB1 operation, where the gain is higher then B or C. Many instabilities come from poorly regulated power supplies. That's where I would start.

 

[molloy]

power amplifier site:edaboard.com

Hello
thank you for your suggestions. Randomly i meant that a percentage of apparently similar (tested in both gain and phase) board faults.
The choice of capacitor was in fact difficult (i didnt find anything better than ATC180R) and anyway it seems they work well. I will check the behaviour of every capacitor in post matching.
The tests are performed keeping under control the flange temperature (there is a voltage controlled fan)
The power test: i connected a 10dB power attenuator with a phase shifter to amplifier output at 50dBm.
A thing that i forgot to tell is that anyway when the amplifiers faulted the power output was only 47dBm.
I will investigate every aspect you suggested to me (thinking of course even about a large signal para measurements)

thank you another time

 

[ChaoFan]

large-signal amplifier stability

Hello,
I think it's better to tell me which device you are using, and signal generater. I have saw this kind failures in customers.

 

[cjron]

in a power amplifier does the power output has a trade-off with gain bandwidth product?? If so what are the possible ways to eliminate that and to get a high GBP and o/p power??

 

[tony_lth]

Yes it has a trade-off.
And you can try balanced amp.

 

[cjron]

how to implement a balanced amplifier when working with a high power ampr (apex pa107dp)??

 

[tony_lth]

you can use power splitter + PA + power combiner to try.
I don't have experience on so low freq about the balanced amp. If I were you, I would do some simulation before design.

 

[cjron]

yeah will try that..thank u fr ur reply anyways

 

[DDavid]

u told only at digital
please give more info...