Piezo quality

Message

gerryc · #1 Post by **gerryc** » Tue Apr 08, 2008 9:19 pm

OK. I'm about to round up some stuff to build a couple of Omni 10.5's with some melded arrays. One thing that seems a little disconcerting is that it sounds like the quality of piezos is somewhat questionable. I guess for about $1.50 a pop one can't expect MIL spec, but what is the fallout that I can expect (i.e. what percentage will wind up in the universal circular file)? 5%, 10%, more? Or maybe I just got the wrong impression. I can deal with having to check the polarity or sealing them, but what'll wind up as scrap?

Thanks

gdougherty · #2 Post by **gdougherty** » Wed Apr 09, 2008 12:49 am

Depends on how picky you want to be. Out of a batch of ~40 I might find 1-2 that are unuseable or flat out dead. To get reasonably well matched tone, I might set aside 1 out of 4-5 and for closer matched tone I think Leland mentioned something like 1 out of 3 gets pulled.

LelandCrooks · #3 Post by **LelandCrooks** » Wed Apr 09, 2008 6:57 am

If you're real particular 25-30% will get pulled. My criteria is within a couple of db average across the range. But that's because I'm making it for someone else. It should be as good as it can be if they're paying me or Les to build one. For personal use, I just test it to see if it works, it all averages out in an array.

gerryc · #4 Post by **gerryc** » Wed Apr 09, 2008 9:07 am

It sounds like it's not that bad as far as units that would be a total loss. As far as variation in output, I'm not going to worry too much about it. As you said Leland, it's more important if you're building for a customer (I'm not there yet, maybe..). I would think that larger arrays would be more forgiving because of the averaging effect. Thanks.

Tim A · #5 Post by **Tim A** » Wed Apr 09, 2008 9:53 am

I don't test for dB, just for working/not working. Of the hundreds I've installed, I've had two that didn't work. Both of them were due to a broken lead inside the housing, both repairable. And it was in the same array where I found two that sounded decidedly different, much darker, not as bright. I swapped them too.

gdougherty · #6 Post by **gdougherty** » Wed Apr 09, 2008 10:05 am

Tim Ard wrote:I don't test for dB, just for working/not working. Of the hundreds I've installed, I've had two that didn't work. Both of them were due to a broken lead inside the housing, both repairable. And it was in the same array where I found two that sounded decidedly different, much darker, not as bright. I swapped them too.

That's mostly what I check for beyond the dead tweeters. Once in a while you'll find one that sounds like it has no top end. Off the top of my head that's about 1 in 10. I also set the noise at a comfortable but still loud volume and listen for any harshness.

Paul Norman · #7 Post by **Paul Norman** » Wed Apr 09, 2008 10:59 am

Out of 100 1016's I had 9 that didn't cut it. I used pink noise and a DB meter. Set the output to 100db and measure each one the same. The ones that were not with in 5% of 100db where not used. This was pretty easy to do, but it mainly just checks that they do have a reasonable output.