How about bonding on some small blocks to lower the resonant frequency? If the plate is vibrating in its fundamental mode, I'd've thought that a couple of symmetrically placed blocks bonded on (silver soldered? - I don't know what works on SS) could be used to "tune" it. These could then be ground down for final matching.

Just an idea.

Thanks John,

I have suggested that as an option but was shot down because of the bonding twice instead of once issues.

Instead of traditionan bonding of the blocks, what about using wax as a bonding agent for the blocks. After you are finished, simple head would dislodge the blocks.

Can you simply mask off one side of the plate during the plating process?

I am not an expert on these things and I don't know just what you guys are building but, nonetheless, a couple of questions/issues?

Why can't you buy slightly thicker plate and have a machine shop mill it down? That seems like it would be rudimentary.

Second, why can't you buy precisely tuned crystals? Seems to me that used to be a simple matter. If I remember correctly lot's of production gear depended on crystals all cut to very close frequency parameters.

Third, what are you all building? Is it one off or production? If a long production run I could see a problem milling a lot of metal.

Last! It is nice when you tell folks you are asking advice of what the advice is being applied to. That helps our education and perhaps the next inquirer.

j.

I suspect you are interested in tuning the thickness of the plate to match the resonant frequency of your PZT transducer. There are 3 important factors that affect the acoustic properties of your 0.250 thick plate. You have mentioned the thickness, but the elastic modulus and the density of the material are the other two important parameters. You might consider testing material lots at your vendor to locate plates that meet your requirement. You are not looking for a major change, and I suspect you could locate in-stock material that would match your acoustic requirement.

Welderman is correct in that we already have a purchased supply of material. It is my understanding that stainless in this thickness can only be purchased in larger quantities and we are stuck with what we got and thicker material is hard to find w/o having a whole lot manufactured.

As far as controlling the process of having the piezoelectric ceramics fabricated to match a particular frequency: We are actually using the 3rd and 5th overtone harmonic of the crystals. We have used 3 different vendors and with each vendor the best we can do is find an average center freq. that is based on a thickness.

To answer your question Jack, if I can with out saying to much (this is a very competitive business). We manufacture ultrasonic cleaning equipment and this latest project is still in the R & D phase. The 3rd and 5th overtone harmonic of these crystals is in the megahertz range and used in the semiconductor industry for cleaning substrates.

Thanks again for all of your comments and sorry for the tardy response, I have been out sick.

Anderbry,

Thanks for the fill in.

From what you say it seems you are using a manufactured piezoelectric device rather than a natural crystal. Therefore you can't do what could be done with natural crystal, if I remember correctly, e.g., cut the crystal to "tune" it.

The crystals, again as I understand it, are being attached to a steel plate which I imagine forms the base of a tub filled with liquid and wherein the material to be cleaned is placed.

You must realize that rather than use manufactured ceramic "crystals" you could switch to natural crystal which as I say can be tuned. But I will bet that natural crystals would cost a lot more having to be cut and tuned whereas I'll bet the manufactured devices are just molded and dumped out of the molds en mass.

The price of course is that slight impurities from one batch of ceramic material to the next are the cause of these devices variation in frequency.

As far as cleaning that variation would not matter a tinkers dam accept for the fact that the steel plate, which must resonate with the "crystal" is a fixed given, or is it?

You talk about altering the plate in terms of adding mass by plating on a couple thou. That would lower its resonant frequency. If you wanted to go the other way you could simply remove metal by drilling little dimples into the plate.

Why not take a welder and add, here and there, small amounts of metal thereby lowering the plates resonant frequency. You would have to rig to test each plate but that should be no big deal.

Am I right about the variability as to the piezo devices or are they all dead on frequency. Seems frequency wise you have a moving target. Why can the manufacturer not change his batch to manufacture devices right on frequency and uniform in frequency? Seems that would be a simple matter of careful quality control.

I'm mixing up a couple of parameter issues here but you should be able to see the direction of my thinking.

As to plating. Why can you not plate the whole plate and then attach the "crystals"?

j.

Anderbry,

Reading back I just noticed again you are using the third and fifth harmonics. I would think the same question I just raised applies. Changing the base resonant frequency would move the harmonics as well, presumably to the point they resonate those plates most efficiently.

j.

The idea is to plate one side of the plate and then bond the crystals on the side that was plated. This will allow us to use crystals that we already have in inventory. Removing material to make the freq. higher is not a problem. We have already exhausted the crystal in that frequency range.

You are correct Jack in you assumption about how the plate will be mounted to the bottom of a tank with the crystals bonded on the bottom side radiating up through the plate into the tank.

I am talking with a company right now that does electroless plating and think they might be able to do it, but it is expensive. The other problem with normal plating is the high and low areas that are unacceptable.

Anderbry,

"Removing material to make the freq. higher is not a problem. We have already exhausted the crystal in that frequency range."

As I understand it the above means you have used all the "crystals" (As I understand it not natural rock crystal but a synthetic ceramic that functions the same way) that were to high a frequency for the plate you have in stock. You achieved that by reducing the mass of the plate by removing material.

By the same token, therefore you have also exhausted all the "crystals" whose frequency match the mass of the plate.

Therefore the "crystals" you have left are too low a frequency for the remaining stock and you want to add mass, by plating such onto the remaining stock.

Are you aware that there are adhesives that will bond metal to metal and that may be thin enough and may bond tight enough not to interfere with the transmission of energy from the "crystal" to the plate.

The company that makes those adhesives is the same company that makes the LockTite products. They make adhesives that bond metal blocks so firmly that they can then be machined as though one piece.

You may already be using such an adhesive to bond the "crytals" to the plate.

Surely you are aware that shim stock, i.e., rolls of thin metals of various pedigree are available at .002 and .003 inches?

Have you tried anything like that, i.e., bonding thin foil such as shim stock to the plates with anerobic adhesives like those above mentioned? In that regard you said nothing in your post about my suggestion as to welding spots of a bit more mass to the plate; like the opposite of removing spots of metal with a drill, a technique sometimes used to balance spinning disks of one or another sort.

Which brings us to the last point of possible adjustment.

Have you enquired, as with natural crystal, if you can alter the frequency of those ceramic piezo devices you have left by grinding off a bit of mass, thus raising its frequency.

If I remember correctly that could be, and used to be, done with natural rock.

You now have me interested. I'm going to do some research here on line and see if my memories are correct.

j.

Your recollection is correct. The frequency of a PZT crystal can be increased by grinding off some material from one surface. The bad news is that the crystal must then be remetallized to provide electrical connection to the freshly ground side. The metallizing process is normally done at elevated temperature which depolarizes the PZT ceramic. The PZT can be repolarized, but you are looking at considerable expense and complication.

A good welding shop may be able to deposit additional metal that can be machined to provide the thickness you require. Also, metal can be added by various metal deposition techniques such as plasma spray, etc. I wouldn't rule out the use of a shim and adhesive. At the fundamental frequency or an overtone, the stress close to the joint surfaces is at a minimum, and the mechanical properties of the joint are less critical than they would be elsewhere in the structure.