Hello Zhang,

I can answer part of your question. Hopefully someone familiar with amplifiers will tell you more.

Duty cycle (average fraction of time the device is on or is receiving a signal) multiplied by power of signal will give the average amount of power deposited into the probe or put out by the amplifier.

Some part of that power will have to be dissipated as heat. In the case of probe - almost all of that power goes into heat - if the probe is well tuned and matched.

Since each device has a limit in the amount of heat that can be dissipated, you can break the device by overheating it. So at any given level of power there is certain value of maximum safe duty cycle. If you lower power you can increase the duty cycle.

This post by Kirk Marat has a good explanation of how probes are damaged by high power irradiation.

Regards, Evgeny.

In addition to Evgeny's good and accurate answer:

Zhang,

It is interesting that you report the 1H amplifier to be damaged (twice, no less) rather than a probe getting damaged. That in itself is unusual (but certainly not impossible) because more typically it is the probe that gets damaged rather than the amplifier.

You also note "...long time or high power..." It does make a difference so please pick one or the other. ;)

It would also be helpful to tell the Wiki a bit more about the spectrometer where this amplifier is located: age, solid or liquid, 1H frequency, driver or final amplifier, if appropriate and so on.

In a properly designed RF power amplifier that is not pushing the boundaries of design, for example, extreme high power or frequency, perhaps the most significant failure mode is from excess heat (sometimes from clogged air passageways or weakening fan). Again, perhaps second in line is deteriorating electrolytic power supply capacitors (for aging units). Most vacuum tube amplifiers (used in high power instrumentation) are rather robust to so-called "bad loads" like NMR probes (tuned or untuned). Transistor amplifiers are generally a bit more twitchy but they are usually accompanied by fault detection circuitry which is intended to limit damage in bad situations.

If you can provide more background information then it might be able to help with the problem diagnosis and then with a possible solution.

Joseph

If the amplifier is connected to the probe, you may well check whther the output impedance of the amplifier and the input impedance of the Probe are matched well. If there are mismatches there could be reflecting power and this might add to the demand on the tolerance limits of the amplifier.

S.Aravamudhan