AVS Forum banner

441 - 460 of 696 Posts

·
Registered
Joined
·
3,799 Posts
Discussion Starter #441
Isn't there a passive XO to blend the MF and HF pieces of that coax? That could probably be modified.

Resistors are cheap. I'd just buy the higher wattage ones and get: 2, 4, 6, and 8. Then arrange them in different Lpads like

Most attenuation

---8---!---DRIVER
!
2
!
Negative

To least attenuation

-----2----!----DRIVER
!
8
!
NEGATIVE


The first would offer the most attenuation and the second the least. I'd try something like a 4 and 4 and see what hat gets you. If you put in the Lpad and find you need to use boosts in your eq, raise it up to say a 2 and 6 or a 4 and 6. You really can't make a mistake with those resistors. Even if you use a 2 and 2 Lpad the impedance would still be safe for that amp, though I'd recommend a 2 and 4 as a minimum. If your worried about dropping the impedance to much just run it by me.

As for measurements. Isn't the MF and HF already worked out? So you just need to get the low blend right. I would think about the driveway if I were you. In your room getting useful results at 400hz is unlikely. You could use "at the seat" measurements. Normally I dismiss those as they're not representative for XOers in the midrange, but down that low it's all about the power response and getting it smooth at the seat is actually the point. Still important to know the anechoic response though. We could do it in July if you'd like.
At the moment, I think the timing/phase between the CD and woofers might be alright.

Here's a couple sweeps at 4ft, with the red being with the phase reversed:



Changing the delay in either direction by even .02ms doesn't have a positive effect, and it sounds ok to me.



Are these the resistors you're referring to?


http://www.parts-express.com/8-ohm-20w-resistor-wire-wound--017-8

http://www.parts-express.com/4-ohm-20w-resistor-wire-wound--017-4

I wasn't able to find a 2ohm or 6ohm equivalent, but I have no idea what I'm looking for :)
 

·
Registered
Joined
·
7,989 Posts
Oh forgot you have a passive XO on the mid and high. Is that the one MtG did? Only first order with nothing else?

I'd start by flipping the 8 and 4. That'll be less attenuation but won't screw with the other stuff as much. Maybe MTG could have a look and see.
 

·
Registered
Joined
·
2,137 Posts
You could try this:


Placing the resistors on only the HF reduces power consumption/waste in those resistors by a large amount by focusing only on the range that needs to be cut back to reduce noise. That small 0.33uF cap you see bridging the inductor of the MF helps to provide some extra filtering above 8k effectively giving a sharper low pass filter on the MF driver which may also help in reducing noise from the amp.


Here is before and after:


You can bump the 4 ohm up to a 10 ohm to give you even more padding of the HF level, but I would start with the 4 ohm first. Also you may want to check polarity on the HF driver, I can never remember if I had it wired wrong when I was measuring. Look for a notch around 6.5k wired one way but not the other, obviously correct way is without the notch.
 

·
Registered
Joined
·
3,799 Posts
Discussion Starter #447
You could try this:


Placing the resistors on only the HF reduces power consumption/waste in those resistors by a large amount by focusing only on the range that needs to be cut back to reduce noise. That small 0.33uF cap you see bridging the inductor of the MF helps to provide some extra filtering above 8k effectively giving a sharper low pass filter on the MF driver which may also help in reducing noise from the amp.


Here is before and after:


You can bump the 4 ohm up to a 10 ohm to give you even more padding of the HF level, but I would start with the 4 ohm first. Also you may want to check polarity on the HF driver, I can never remember if I had it wired wrong when I was measuring. Look for a notch around 6.5k wired one way but not the other, obviously correct way is without the notch.
Here's the current crossover with a 1uf cap on the HF and a 0.33mH coil on the mid. I did end up having to reverse the polarity on the HF to remove a dip.




So I need to add the resistors just to the HF, and I also need to add a 0.33uf cap to the mid?

Also, your drawing has a 1.5uf cap on the HF? I have a 1uf on there today, do I need to change that to a 1.5uf?
 

·
Registered
Joined
·
7,989 Posts
Not a bad idea to try what MTG suggested and just attenuate the HF unit. But I'm not totally sure it'll knock out the hiss enough. Not doubting it, just don't know. I've never done it like that. Personally, if you still get hiss I wouldn't be to worried about power getting soaked up on the MF unit. It's so effecient it'll hardly use any. Let us know how it goes.
 

·
Registered
Joined
·
3,799 Posts
Discussion Starter #450
Not a bad idea to try what MTG suggested and just attenuate the HF unit. But I'm not totally sure it'll knock out the hiss enough. Not doubting it, just don't know. I've never done it like that. Personally, if you still get hiss I wouldn't be to worried about power getting soaked up on the MF unit. It's so effecient it'll hardly use any. Let us know how it goes.
Are these the right parts that Matt mentioned?

http://www.parts-express.com/dayton-audio-pmpc-15-15uf-250v-precision-audio-capacitor--027-212

http://www.parts-express.com/dayton-audio-dmpc-033-033uf-250v-polypropylene-capacitor--027-404
 

·
Registered
Joined
·
3,799 Posts
Discussion Starter #452
@tuxedocivic @mtg90

Just to see how much of the hiss was coming from the mid, I unplugged the HF. The hiss was reduced a good bit, but still there, so I do need to pad the mid a bit too.

I ordered 4 of each of the 8ohm and 4ohm resistors, so I should have enough to pad both, right?
 

·
Registered
Joined
·
21,319 Posts
here is a calculator if it helps:
http://www.sengpielaudio.com/calculator-Lpad.htm


with a nominal 8 ohm load, and r1=4 and r2=8, spl is reduced by 6db.
the noise on the amp is actually pretty high at -66dbu, so with 118db sensitivity on the driver/horn that's about 50db in noise. :eek:
toeing in the horns is another way to reduce the upper end a little 'for free'.


also, some of the noise may not be coming from the amp, but from further upstream. with so much headroom to spare, attenuating the signal from the pre-amp may actually help reduce noise if that is part of the problem. (signal level isolation transformers/attenuators may help there depending on whether or not that is part of the problem).


also, perhaps a question for tux/mtg90, shouldn't more than 20w resistors be employed in this particular application?
 

·
Registered
Joined
·
436 Posts
It has to be resistive to do what it does - flat frequency response. The key is the statements in the manual that it has 12 settings and that if your amp can only tolerate 8 ohm or higher loads, than you can only put two in parallel.
My guess is that inside you will find a 12 position rotary switch and a bunch of high power resistors. The switch simply reconnects the resistors to get different values of attenuation while maintaining a 4 ohm or higher input impedance. It could simply be a linear string of resistors with the switch selecting the tap point on the string - a discrete approx to a L pad. That would have a constant input impedance, at least without a load. With the speaker attached, the input Z would change as a function of switch position so perhaps the switch also inserts more R into the output leg at the same time it moves a tap down an input string of resistors.
 

·
Registered
Joined
·
436 Posts
IMHO, I don't think enough attenuation is being suggested. The more the better. The amp can easily make up the difference. The hiss floor is pushed down by the amount of attenuation, so how many db quieter do you want it to be?
 

·
Registered
Joined
·
7,989 Posts
@tuxedocivic @mtg90

Just to see how much of the hiss was coming from the mid, I unplugged the HF. The hiss was reduced a good bit, but still there, so I do need to pad the mid a bit too.

I ordered 4 of each of the 8ohm and 4ohm resistors, so I should have enough to pad both, right?
I kinda figured that might happen. The "mids" are playing a lot of treble even with that passive low pass filter on there. I'd just try the L-pad right out in front, rather than on each driver.

also, perhaps a question for tux/mtg90, shouldn't more than 20w resistors be employed in this particular application?
I don't really think so. If he's attenuating around -20db at the amp, it's gonna take a lot of power to blow one of those. Resistors aren't the type of thing to fry as soon as they receive 21 watts either. Most speakers use a 10 watt resistor just fine. I have doubled or even tripled up a 20 watt resistor in PA applications, but even there it's hardly necessary. What does make this application a little concerning is the attenuation will be applied over such a wide bandwidth (450-20,000hz). Typically passive attenuation is used within the treble range only.

IMHO, I don't think enough attenuation is being suggested. The more the better. The amp can easily make up the difference. The hiss floor is pushed down by the amount of attenuation, so how many db quieter do you want it to be?
Depends how much it takes I guess. Originally he mentioned it was still a little audible. So might not take much. Would have been good if he could have gotten some 2 and 6ohm resistors to try. If the 4x8 L-pad isn't enough, he just flips to a 8x4 L-pad. That's why I suggested such a variety of resistors. IME it doesn't take much to knock out some hiss. I don't know how much he's getting though...
 

·
Registered
Joined
·
3,799 Posts
Discussion Starter #458
IMHO, I don't think enough attenuation is being suggested. The more the better. The amp can easily make up the difference. The hiss floor is pushed down by the amount of attenuation, so how many db quieter do you want it to be?
I kinda figured that might happen. The "mids" are playing a lot of treble even with that passive low pass filter on there. I'd just try the L-pad right out in front, rather than on each driver.



I don't really think so. If he's attenuating around -20db at the amp, it's gonna take a lot of power to blow one of those. Resistors aren't the type of thing to fry as soon as they receive 21 watts either. Most speakers use a 10 watt resistor just fine. I have doubled or even tripled up a 20 watt resistor in PA applications, but even there it's hardly necessary. What does make this application a little concerning is the attenuation will be applied over such a wide bandwidth (450-20,000hz). Typically passive attenuation is used within the treble range only.



Depends how much it takes I guess. Originally he mentioned it was still a little audible. So might not take much. Would have been good if he could have gotten some 2 and 6ohm resistors to try. If the 4x8 L-pad isn't enough, he just flips to a 8x4 L-pad. That's why I suggested such a variety of resistors. IME it doesn't take much to knock out some hiss. I don't know how much he's getting though...
Honestly, it's really not that much hiss. Yes I can hear it from the seats, but for anyone else I'd probably have to point it out for them to notice. If I had to I could live with it, but I think these resistors will do the trick.

If that's a lot of bandwidth to cover, I don't mind putting a 4/8 pair separately for the mid and also for the HF. I bought enough of them to do that.
 

·
Registered
Joined
·
3,799 Posts
Discussion Starter #460
Where is the hiss come from? The minidsp ?
I took the MiniDSP out of the loop and I'm now using a Lab Gruppen IDP 2400 with built in DSP for the BMS 4594's.

The noise floor on the LG is wayyyyyyy less than the mini, but there's still a tiny bit of hiss I want to remove.
 
441 - 460 of 696 Posts
Top