Aluminum alone is not bad, low cost, machines easiliy, can be used with EDM, fairly high conductivity, good thermal resistance. There are various alloys that have different hardness. There are other more exotic materials, such as invar that are used for TCE issues. However invar is expensive and hard to work with, the machine operators much prefer aluminum. For high Q systems, this required high quality surface finish and metalugic plating of gold. Best to try and design around these issues to save cost. Nobody wants gold plated parts anymore.

Thank you for the information. The problem of diplexer and cavities is that the material needs to have small coeficient of change of dimensions with change of temperature. Another thing is that the materials needs to be plated by gold or silver (most common). The total costs is very important. If we use brass the plating will be inexpensive but the cost os brass is greater than aluminiun alloys. If we use aluminium alloy, the costs of aluminiun is small but the plating costs is greater than brass.
Then the problem is the minimum total  costs
Thanks
Marcos Olandoski

I have heard that the copper loaded metal matrix composites have made some inroads to the cavity and filter area.

The attributes:

Prescription TCE, low thermal resistance
Easily machined, pressed or molded to high precision
Low cost plating options.

I can not remember the company making these statements however.
 

Hi,

Use silver plated aLuminnum for both housing & resonators. The cheapest way to manufacture the diplexer is to cast both housing & resonators in one piece.

Follow this spec to silver plate your aluminum housing:

Electroless Nickel per AMS 2405B (200-250 micro-inch), then Copper per MIL-C-14550B (50-100 micro-inch) then silver plated per QQ-S-365D, type I (300-450 micro-inch).

Hope this will help.

Krytar

but be careful about using nickel underplate if you're worried about pim

Dear Bhawkins
I would like to know more about pim. Everybody says that you could not use nickel. Exactly in what application that is important?
We finished the development of diplexer for 400 MHz, 600 MHz and 900 MHz. Now we are working in diplexer for 1,2  and 2,4 GHz. We used aluminium and silver plate as specified by Krytar. The results were very good.
Thanks
Olandoski

I'm glad that you use my method.

Using silver plated aluminum to filter resonators, trnasmission line was downplayed by Bob Wenzel, a filter guru, but we had proven it worked well.

Some of our filters/ multiplexers (penta, quadra, tri, and diplexers) are functioning up there in at least 8 satellites.

Krytar

 

PIM:

http://www.wirelessdesignonline.com/content/news/article.asp?docid=a06f4ddc-8869-4544-bd04-717a41c44222

I would like to thanks Krytar and biff44 for all the help
Olandoski