New Build - T30s and then Otops to follow

[miked]

I have not tried EQing my OTops manually, but I can absolutely say that the before and after results of using the auto-EQ function on the Driverack PA+ are like night and day. The tops were "honky" and and "tinny" sounding in the mid and high range before. But after they are just...awesome sounding. Very "flat sounding" without SOUNDING flat, if that makes sense. I guess a better way to say it would be that all the frequencies coming out of them, instruments and voice are all even-timbered and of equal intensity. And jeez, are they freaking LOUD! Stupidly loud at 50 feet away and they don't sound strained at all. Sorta like King Kong putting his finger on your chest and you telling him how strong he is and you can't believe it and he's just rolling his eyes b/c he's not even trying hard.

EQ fixes OTops, period. Not they are broken to begin with, but all pro sound speakers NEED EQ. Of course, this assumes that both your tops are built the same, are air-tight, etc.

[escapemcp]

Ok, here's how I have fitted a tophat to my Otop without it showing through the reflector.

1) Buy a slim top hat (Adam Hall make them). http://www.bluearan.co.uk/index.php?id=ADHSM701
2) When making the baffle (i.e. the first step on the way to an Otop) route a few mm into the wood on the FRONT of the baffle - like these pics:


3) Extend the woofer spacer down to the base of the speaker... this will ensure a more solid baffle (as if it needs it!)... the real reason for this is because the screws from the top hat need something to bite into, and by extending the spacer down, the screws can bite into the spacer AND the baffle. Sorry, I didn't get a pic of this last night. This arrangement also means that the routing of the baffle done in step 2 now has another piece of wood (the extended spacer) behind it... so you could go even deeper, which means you might be able to hit the speaker's c.o.g. AND routing it that deep would eliminate the need for step 5 (only just realised this as I type!! )

4) Cut a hole in the base for the top hat to fit into.


5) Route a bit out of the horn reflector... this top hat is a TIGHT fit, so every mm you can pinch helps it fit.


6) Add in another bit of wood for the screws at the front to bite into


7) And if you have squeezed enough space in there, you should end up with something like this:


The top hat is hanging out a little on that last pic, but it does fit flush with the base of the speaker. I did want to countersink it, but that would have meant finding even more mm (or sixteenths of inches)... my maths said it would work, but I think would be taking too much from the wood and it would end up almost paper thin. This way, there is still plenty of strength in the wood.

Fitting the tweeter cable can be a pain in the ass though, but I managed to squeeze it all in. I know it looks a little rough, but it's all hidden, so I don't really care. Time pressures are getting to me at the moment, as I have a big free party on Saturday night At least the 2nd top (pictured) is now ready for duratex. Still got a list as long as my arm, and I have just decided to also build a flightcase for my new mixer Going to sleep well Sunday night!

[Radian]

EQ fixes OTops, period. Not they are broken to begin with, but all pro sound speakers NEED EQ.

Flattening out subwoofer response by proper application of active filters will provide an equally religious experience.

[escapemcp]

So without an RTA (which I will buy NEXT month, but it won't be in time for Saturday), is it best just to copy the settings in the "EQ Settings" bit of the forum. I have had to try to replicate the settings on my amp, but I only have parametric EQ and not 1/3 octave EQ, so it isn't perfect. Doing it this way, it still seems far too bright, so I guess I should just EQ by ear after that?

Is this the best approach for my current situation? Any other suggestions? I do have an RTA app on my phone (Galaxy Note 2) - it has calibrate on it, but how off is it likely to be? Will it be good enough to get the OTs dialled in close enough?

Thanks as always for any ideas

[byacey]

Do you have a HF shelving EQ in the parametric?

I would put an L-pad on the HF crossover until you get the desired attenuation in balancing the highs with the mids. The less EQ required, the better.

Too much EQ isn't a good thing as it creates unnecessary phase shift. I've been skeptical of the results after using auto EQ in the Driverack PA; sometimes one band will end up almost full boost, while the adjacent band is almost full cut.

[escapemcp]

Do you have a HF shelving EQ in the parametric?

Yes, have been using it to control the tweets, but still not happy.

I would put an L-pad on the HF crossover until you get the desired attenuation in balancing the highs with the mids. The less EQ required, the better.

An L-pad is for reducing the volume afaik. I currently have a 3.9Ω resistor on the + line just before the tweets (as per instructions)... should I just put in a bigger resistor here (/or another in series) to increase the attenuation? This is the way I have been thinking of going as it is the most elegant solution. Is an L-pad any different to this setup or is it just the fancy name for it? Should I use noise (pink/white?) in order to test the balance between the two? Haven't tried that yet and that could well get things close to where they should be. At the moment, I think that I am so far off, I am undercorrecting and then overcorrecting, then overrecorrecting, and then finally re-anti-over-correcting!

Too much EQ isn't a good thing as it creates unnecessary phase shift. I've been skeptical of the results after using auto EQ in the Driverack PA; sometimes one band will end up almost full boost, while the adjacent band is almost full cut.

Didn't know about the phase shift thing. Intresting to know this (I love all the theory stuff )

I am still wondering why my tweets are so loud. I followed the plans well enough, except for the extra tweeter pair in the array - by my back of a fag packet calculations they shouldn't make any where near as much difference as it has been making (looking like 4dB-ish)
For Saturday though, it's going to have to be the high-pass shelving filter to adjust those pesky tweets!

[byacey]

Try and get the box sounding good without any EQ to begin with. You'll find peaks in the response here and there, but they are just a natural response of the horn design. However, if there is a large difference between the mid horn and the HF tweeters, one of the two needs to be balanced; the most logical thing to do is attenuate whatever is to loud.

An L pad does exactly that. If you only change the series resistor value on the HF crossover, it will change the crossover point. Passive crossovers like to see a constant low source impedance from the amps, and the designed load impedance for the drivers. If you tinker around with any of this, the crossover point changes.

An L pad will allow you to tame the output of the tweeters while still maintaining a constant load impedance that the crossover is working into. Once you find a happy medium, you can remove the adjustable pad and build a non-adjustable L pad from fixed resistors.

Another method you can try is putting more of the tweeters in series with each other, rather than parallel. The only drawback is if one of the piezos goes open, you lose the whole string.

As a happy result of all of this, you can reduce the amount of EQ cut in the higher bands thus flattening out the corrective EQ curve.

[Bill Fitzmaurice]

An L pad will allow you to tame the output of the tweeters

You can't use an LPad with a piezo array.

[byacey]

No? Why not? As long as your parallel load resistor is in place to simulate a low impedance load, an L pad will work fine. It's simply a voltage divider.

[Bill Fitzmaurice]

No? Why not? As long as your parallel load resistor is in place to simulate a low impedance load, an L pad will work fine. It's simply a voltage divider.

If you change the load impedance you also will have to change the filter L/C component values, which will only deliver the desired filter response with the resistor values of the standard filter. That's probably beyond the ability of the OP.

[byacey]

That's the whole point of using an L-pad, it maintains a constant load impedance with respect to the crossover.

[Bill Fitzmaurice]

That's the whole point of using an L-pad, it maintains a constant load impedance with respect to the crossover.

LPads don't come in 30 ohm.

[byacey]

Nope, but it's easy enough to make one. If you want a 6dB drop, use 2) 15 ohm resistors in series in place of the 30 ohm resistor, and place the piezo load in parallel with one of the 15 ohm resistors.

If you want a 3dB drop, put a 22 ohm and a 8 ohm in series instead of the 30 ohm, parallel the load across the 22 ohm resistor.

There's nothing very complicated or difficult about this, and it doesn't change the crossover frequency because the crossover is in all cases still working into a 30 ohm load.

[byacey]

Since the resistor across the load in the plans is 30 ohm/10W, I might suggest putting 5) 6 ohm /2W resistors in series to make up the 30 ohm load. Connect the positive terminal of the tweeter array to the crossover common, and then connect the negative Piezo terminal to whatever tap at the junctions of the resistors you want for the desired attenuation. Technically this isn't an L-pad, but the piezos impedance is high enough that we can disregard the small change in resistance this creates. http://img.photobucket.com/albums/v354/ ... b82254.jpg

[Ryan A]

Note to self....keep byacey in mind next time I need to build a crossover.