Improving piezo sound

[Bigrob1234]

Hi all, after building 2 otop 12's I've been dialing them in over the last week, and I've noticed that the piezos were overpowering the low end so I made the recommended adjustment to the crossover by moving the position of the 4ohm resistor. Also for a bit of context i am using 6 goldwood 1016 piezos together with the Faital Pro 12FH520. I then did some very basic eq work which did help quite a but however the piezos still produce a harsh tone at certain frequencies and overall the sound being produced sounds "hollow " if that makes any sense. Any tips on improving the piezos sound and how to properly eq the boxes? Other than that they've blown me away by how efficient they are! Thanks !

[Grant Bunter]

I'm revisiting my melded arrays in my DR250's at present. It's 11 years of hard slog for them and it started with "HF seems down".

Here's what I've (re)discovered:
1. Goldwoods are the best.

2. The more elements there are in array, the smoother the sound.

3. You have to be very particular when soldering the elements.
I have replaced a heap from my spares box (more are on the way) that were non functional. The "new" elements tested perfectly, but after soldering, there is distortion in a couple. This tells me the elements are susceptible to producing distortion if you spend too long on the tabs with the soldering iron. Sometimes, that distortion displays itself as a very shrill sound.
From now on, I will use a couple of artery clips (as heat sinks) each side nearest the element on the tab when reattaching wires.
Test each element using the tube to the ear trick after disconnecting the Faital Pro to see if you have any distortion in any element.
I use a 3K test tone only.

Parallel/series wire the array for higher power handling. This also reduces sensitivity by a couple of dB.
Another option might be go melded instead.

I asked the same question in my build thread about moving the 4 ohm resistor, and Bill's answer was something like "if you have EQ, you probably don't need to move the 4 ohm resistor".

So, all that lot needs to be perfectly in order before you can think about EQ (not what you want to hear most likely).
Back when I built mine, I was in such a rush I never paid too much attention to the distortion, but my ears have been trained more now.
With EQ, unless you want to run sweeps all day, try pink noise from 2Kish up.
Move each EQ slider up and down to see if those offensive tones gets worse (if up) or better (if down), to identify the offensive frequencies.


Basically, you want to "notch filter" the offending frequencies, so you don't affect the frequencies either side too much.

I don't know your signal chain, or how or if you're processing, but notch filters have high Q so the effect is narrow but deep (more than 12dB if you can). Sometimes you can use a parametric EQ rather than a graphic EQ to sort that. There's a bunch of feedback destroyers out there that will notch filter up to say -60dB.

It's really worth getting to know your cabs well at this point of time. Hope this helps...

[Bigrob1234]

Thanks so much for the reply, it absolutely does help!
I'm afraid point 3 you made may be my problem. I do recall accidently moving one of the connectors on the back of the piezo. Unfortunately they are all glued together and would be a shame to rip em apart. And thanks for the pink noise advice, didn't occur to me that using pink noise would be much easier. Thanks again !!

[Grant Bunter]

You don't need to pull your array apart to replace an element. Unsolder the bad one, undo the 3 screws, put new one in. Resolder, done.

[Bill Fitzmaurice]

Key to not overheating piezo terminals is to use light gauge wire, no more than 18 gauge. Tin the wire with solder, then wrap it around the lug. It will take minimum heat to finish the job. Make sure you have at least a 40 watt iron, preferably more. Low wattage irons take too long to heat the joint. High heat for short duration works better.

[Bigrob1234]

Key to not overheating piezo terminals is to use light gauge wire, no more than 18 gauge. Tin the wire with solder, then wrap it around the lug. It will take minimum heat to finish the job. Make sure you have at least a 40 watt iron, preferably more. Low wattage irons take too long to heat the joint. High heat for short duration works better.

Once again I'm afraid I've used way to thick wire for the job. Luckily the piezos are very cheap so shouldn't be too hard to fix

[Bill Fitzmaurice]

From OTop 12 plans:
Tweeters draw almost no current, so they may be wired with as small as 24 gauge wire. Lighter gauge wire requires less heat for soldering, so you’re less likely to damage the tweeters as well.

[Bigrob1234]

From OTop 12 plans:
Tweeters draw almost no current, so they may be wired with as small as 24 gauge wire. Lighter gauge wire requires less heat for soldering, so you’re less likely to damage the tweeters as well.

I know I should've read over the plans a couple more times
I've managed to fix the shrill from the tweeters by swapping the crossover back to the original 4 ohm resistor placement and doing some more basic eq work and found that reducing I belive it was around 3.5 kHz it greatly Improved the shrill together with boosting some of the really high ends. It sounds almost perfect now, I don't think I'll replace the tweeters for now they seem to still do 99% of the job. Thanks again for the help

[Grant Bunter]

^ awesome...

[Seth]

Rob, are you tuning by ear or have you taken RTA measurements? Some of the guys swear by tuning by ear. Maybe I'm just not that good, I find the visual confirmation of measured results to be invaluable. Both, for tuning in person and for tuning over the internet.

Something else to double check when things just don't sound right is the polarity. If you haven't checked your array polarity, the directions for doing so are in the plans too.

[Grant Bunter]

Rob, are you tuning by ear or have you taken RTA measurements?

Nearly all Bill's designs have a significant hump rising from about 3K through to 5K if the cab is loaded with piezo's.
Doesn't surprise me in the least that Rob found dropping 3.5K made a noticeable difference.

That particular hump doesn't need measuring IMHO. Measuring tells you what the response is, but not how it sounds (until you've measured so many times you can correlate the result to how it sounds by ear).

A parametric EQ with the right amount of Q fixes that region very nicely...

[Seth]

Rob, are you tuning by ear or have you taken RTA measurements?

Nearly all Bill's designs have a significant hump rising from about 3K through to 5K if the cab is loaded with piezo's.
Doesn't surprise me in the least that Rob found dropping 3.5K made a noticeable difference.

That particular hump doesn't need measuring IMHO. Measuring tells you what the response is, but not how it sounds (until you've measured so many times you can correlate the result to how it sounds by ear).

A parametric EQ with the right amount of Q fixes that region very nicely...

99.9% agree


Still like to see an RTA if something's not right and I can't put my finger on it. Or, if it's tuning someone's system over the internet. IMO, measured response tells a lot about how a system sounds, even if I can't hear it in person. Not to say a perfect measurement couldn't still use a little sumpt'n sumpt'n to sweeten it up a little here and there, listening in person. That's the subjective part.

[Bigrob1234]

I've been using my driverack pa 2 and the p eq to make the adjustments. I don't have the rta mic and don't really want to spend the additional money to buy it, although I'll probably end up buying it sooner or later, when I get more serious about my system. I find tuning by ear is pretty effective by itself. And thanks also for suggesting to check polarity, didn't think of that!

Cheers for all the help as usual！

[Bill Fitzmaurice]

RTAs aren't optional, they're a necessity. At the very least you should have an app, like this: https://play.google.com/store/apps/deta ... yzer&pli=1

[Seth]

I have used Audio Tool by jjbunn, which has a function that uploads a calibration file that comes with the inexpensive (±$25) Dayton iMM-6 calibrated microphone. There may be other apps that have the same functionality. It can be used with phone or tablet. The microphone requires a headphone/microphone plug, or a headphone/microphone plug adapter for phones/tablets without one.


Also, you don't have to use the $125 dbx microphone for your PA2. You can use an inexpensive $35 Behringer ECM8000. Much more palatable.