How To Reduce Coefficient Thermal Expansion

It seems you are pre-supposing the solution? You can only change the CTE by changing the composition of the ceramic....

What circumstances cause it to crack? How is it mounted? What is it's physical size/shape? How are the connections made? What material are the connections?

Years ago I had a problem with a hybrid amplifer being very sensitive to the unit being flexed...it was mounted on 6 legs down each side, putting a zig zag bend in each leg completedly solved the problem. I likened it to standing with flexed legs when travelling standing up on a train...much smoother than standing rigid.

Dunno if this is any help

Del

AS Del said, before you can find a solution to your problem, you have to find the cause of the problem.

Is the CTE really the problem, or is mounting, shock loads, manufacturing defects etc the real cause.

Once you know the cause then you can find a solution. As Del also says (damn, why is he always right!) the only way to change the CTE of the material is to change the material.

As Del stated putting a flex in the leg, I use a loop or dog-leg vs sharp bends which would stress the wires and in a high vibration environment you may need to fasten the cap to the board and put flex in the leads. A few years ago I wired ocean-going racing boats, lots of flexing of fibreglass and shock loads on wiring and hardware.

The alternative is that you have an electronic problem where voltage transients overheat the cap to the point it explodes or the leakage thru the cap is so high it heats up and explodes. Look at the circuit with an oscilloscope for transients or voltages exceeding the breakdown voltage the cap was designed for which would cause it to heat up and explode.

I have seen DC control circuits where the coils of relays discharging were causing problems, the resolution was discharge diodes, if your circuit has large inductor/coils being switched without a discharge circuit,,,you have a problem you need to address with disharge diodes.

The type of capacitor for the circuit or the capacity or the breakdown rating is more likely to break the cap then vibration unless during installation it is physically stressed somehow?

Just a few more thoughts to move you in a direction that by now may be obvious?

The most common problems with ceramic capacitors that are surface-mounted in PCB are cracking of solder or separation of electrodes from the capacitor. These usually occur following temperature cycling that exceeds the capability of the mounting technique used, and are due to the TCE of the ceramic being much smaller than the TCE of the PCB. However, if the materials package, termination pads or process cycle are incorrectly defined, you can also see these problems during manufacture.

N.B. that there should be never be issues even for military or for automotive requirements if you can use capacitors whose maximum dimension is 2-mm or less. That means that, if your problem is because the design uses capacitors with larger dimensions it may be that the most economic approach would be to use a multiplicity of smaller components. If this is not practical, the situation can usually be improved by using the mounting techniques proposed in other postings, but these can sometimes introduce excessive series inductance, or make the equipment vibration sensitive. Only in this case will it make sense to use a substrate that provides a lower TCE; various substrates can provide better TCE match to capacitor ceramics, ranging from ceramic boards (so-called "hybrid packaging") to multilayer substrates that use low-expansion alloys (e.g. Kovar) as dimensional stabilisers.

If you will provide information that clears up the uncertainties we have regarding your problem, we may be able to provide clearer (and simpler) guidelines.

You have received a lot of interesting suggestions, but you have not gotten a direct answer to your question yet (interestingly, that is often the case on this board). The first zero CTE ceramic was Corningware. There may be others now. Zero CTO was achieved in Corningware by mixing very small particles of two ceramic materials- one with a positive CTO and one with a negative CTO. The application it was developed for was radar-transparent missle nose cones. (At least that's how I remember the story from Materials class decades ago.) It should be possible to purchase the material from Corning or someone else for use in your fabrication facility.

Bill Morrow

The question didn't actually state what part of the system would require the reduced TCE. Given the usual environments in which ceramic capacitors are used, it's most unlikely that it would be desirable to reduce the TCE of the capacitor. In any case, a capacitor is a layered structure of metal and ceramic**, so the TCE of the capacitor is not simply the TCE of the body material. Then, you need specific dielectric properties (usually high e_r) for the capacitor body, whereas the ideal for the nose-cone would be to have e_r as small (close to unity) as possible. You also need a material to which the conductor will adhere.
**As far as internal differential expansion is concerned, this layered structure is capable of withstanding temperatures up to 1300-^OC; terminations etc would be destroyed long before that temperature is reached.

Randall didn't answer the question as posed - but failing provision of additional detail I think we should assume that his answer would be the most appropriate solution to the actual problem.

Yes, you are probably correct. I did not mean to attack anyone with my comment:

"you have not gotten a direct answer to your question yet (interestingly, that is often the case on this board)".

I was just making an observation. Indeed, it is a good thing that there is such a wealth of information here that people can often see past the specific question and address the real issue/need.

:>)

Bill