Testing a 50W 40KHz Ultrasonic Piezoelectric Transducer

[NZST205]

I am trying to test a 50W 40KHz Ultrasonic Piezoelectric Transducer and I don't know if I am doing it correctly.
The background. A Dual 50W 40KHz developed a pinhole leak in the stainless tub and water fell onto the PCS and shorted a few things out. I have repaired those and it now runs for a while then one of the transistor banks overheats, If I switch the outputs for the PCS to the Transducers over then the other bank overheats so I have concluded that one of the 50W 40KHz Ultrasonic Piezoelectric Transducer is faulty. I have received another from eBay and it is the same spec 50W 40KHz Ultrasonic Piezoelectric Transducer with static 3800 pF (except that the Fluke 289 says it is 4900 pF). I presume when they say static it means nothing connected to it ?
I thought I'd attach a 40 KHz 20v PP signal to each in turn and then see how they behaved, concluding that if the original good and the new questionably good one were similar versus the one causing the overheating were different then I'd go ahead and give it a try.
Unfortunately the two original ones both read 40 KHz and the voltage drops to 5 VRMS whilst the replacement one takes the voltage down to 1.5 VRMS. This has sort of blown my theory out of the water.
I don't want go and glue the new transducer to the tank and test it because then I can't return it if the 4900 pF is in fact faulty.
I thought I'd fire up the Old original and the new one (unglued) but I have read that if you fire them up without a load it will destroy them.
Can anyone throw some light on how to properly test Ultrasonic Piezoelectric Transducer please ?

[cpt.charlie]

As far as i know, the piezoelectric transducer must be glued and the tank must contain a suitable level of liquid in order to work, otherwise the transducer may fail.

Why must I keep solution at the tank's level indicator?
The solution level should always be maintained at the level indicator in the tank, with trays
or beakers installed. The ultrasonic cleaning system is a 'tuned' system. Improper solution
levels will change the characteristics of the environment, can affect the system frequency,
decrease effectiveness, and potentially damage the cleaner. Maintaining the proper solution
level provides optimum circulation of solution around parts, and protects heaters and
transducers from overheating or stress.

[NZST205]

Yes, thank you, it sort of makes testing a one way trip doesn't it. I do wonder why they specify that requirements as I've not found any definitive research as to why this is so.

[thm_w]

Yes, thank you, it sort of makes testing a one way trip doesn't it. I do wonder why they specify that requirements as I've not found any definitive research as to why this is so.

"Maintaining the proper solution level provides optimum circulation of solution around parts, and protects heaters and transducers from overheating or stress."

They are probably dumping a lot of heat into the solution. If there is no solution then you lose that heatsinking path, and I wouldn't be surprised if more power is reflected back as well.

I suspect this damage would not occur until some minutes of heating to high temperature damages crystal, but that is just speculation: http://uknowledge.uky.edu/cgi/viewcontent.cgi?article=1568&context=gradschool_theses

[SeanB]

If you want to test dunk it in a plastic tub with RO water and try it. That will both cool it and absorb the ultrasonic energy from it, while the high resistance of the water electrically will not damp it. Just make sure to dry it thoroughly after removing it so it does not corrode, and make sure it is clean by washing before and after with IPA to remove contaminants.

[cpt.charlie]

Just curiosity, i have seen those cheap ultrasonic generators on ebay like this one and seems quite simple (it won't offer the best performance, that's for sure but..) What kind of signal does a piezo transducer need? sine, square, triangular? what is the voltage of this signal Vpp?

Thanks in advance.

[NZST205]

Based on the measurements from the one I have it's a sine wave at about 800v peak to peak.