FAST Xover question...

[chrisb]

That's active IMO. The XO actively manipulates the frequency response. After the amps, it passively does the job.

Gotta love the English language and hobbyists' subculture jargon - for us greybeards, "active XO" means includes some type of powered (whether AC or battery ) gain stage or buffer between the source final line out and amplifiers - while a speaker system is described as active if there's no additional passive parts between the amps and speaker voice coil terminals. So you could be actively active, or passively active (or agressive, as I've been told) , but not actively passive?



Of course a fine expression would be an "Aktiv" system (trademark Linn Audio Products ) - dedicated electronic XO / frequency distribution network and a separate amplifier channel for each section of bandwidth - for example the Akubarik and Klimax (someone gets paid to come up with those names) include 6 onboard amps in each enclosure.

now that that's cleared up ...

 

[sideways]

Basically you would trace the circuit (you already have a pre-out to get you a starting point), cut the trace on the board, connect the power amp side to a set of RCAs.

A schema might help, board layout would make things easier.

dave

That's where I thought you were going LOL!!

 

[xrk971]

That's active IMO. The XO actively manipulates the frequency response. After the amps, it passively does the job.

Active means that an op amp is "actively" used as an integral part of the filter - like in a Sallen-Key filter. If you had a speaker with 30 kOhm impedance, you could put this after the power amp and it would do the same thing. The trick is that the power amp input is high impedance so very low currents are used thus components are smaller/cheaper.

Below are a couple of Sallen-Key active filter designs using TI's Filter Pro software for a 400 Hz cross-over.

 

[tuxedocivic]

I suppose if it uses a powered device to do the work, that's an active system. So maybe the minidsp is an active-active XO

 

[tuxedocivic]

40uF\10mH from my calcs...

I wouldn't want to buy that coil. That's huge. Calulators usually are crap. I'd go with about 3mH. If you have to, put 4ohms in series with the coil and you'll sorta have an impedance flattening first order.

 

[sideways]

I have the minidsp...have had the Behringer and a DEQX...I think they're all really good for what they are...

I'm trying a KISS method now...the Fostex's FAST is a prime version of that...

 

[chrisb]

Actually I'm only trying to get away from active on the Fostex side...or the high pass side...for the moment

I still would like to use the minidsp for the bottom end (SEAS CA22RNX's)...this way I can adjust levels easier to blend in better...

I was thinking it would be pretty simple...a first order cap...now possibly a second order circuit in the range of a 40uF\10mH from my calcs...or a line level which I have no idea how to calculate and if those values need to change if that amp changes...

May have posted this before in this thread, but if not:
TLS.org | Passive Line-Level Crossover

With a little help from Dave, even I was able to figure this out. In my particular case the component values worked out to .0068 (HP), .068 (LP,) and 10k. even with zoom-zoom parts, the cost of the hobby box and 6 RCA jacks would likely be more.

Below is not the best photo in the world, but representative of the simplicity of the project

Stupid reasons...my pre only has one input...the E is a nice amp that works well with the Fostex's...I could always buy a different preamp

Only one line in? -If you're not averse to modifying the preamp, and if there's room on the chassis for selector switch and input / output jacks, it's not that hard to add additional inputs and outputs for that matter. And as Dave describes elsewhere, it's usually not a huge problem to break into the circuit of an old school integrated to add the loop.

 

[sideways]

I dl'd Dave's spreadsheet...I got 70nF and 10k...thought I was cracked so that's why I posted

But you're getting the same thing...awesome!! I should ahve those values at work!!!

As for an extra input...I was thinking the same thing a few minutes ago cooking supper...there is room for another set of input...side mount selector switch...could easily work...yes...

 

[sideways]

K...so does this look right??

If I'm doing it right for a 250hz high pass...I can do an inline cap of 68nF with a 10k resistor in parallel with the Fostex's or with just the cap alone I can get away with an~ 6.8nF??



Input Lowpass Highpass
Crossover Frequency 60 250 hz
Desired Crossover Point
Amp Input Impedence 50 100 k Impedence of your amp

First Order
Select R1 10 10 k 5-10k suggested
C1 265.2582 70.0282 nf 9.09 R1 || Ramp
R2=R1' 11.1111 k C1_lp = C1_hp
Use Ramp as R1 in HP 6.3662 nf (no R1)

 

[planet10]

For the hi-pass i suggest just using the amp's input R with a single cap in front...

dave

 

[sideways]

K...so that would bring me to ~6.8nF using a 100k value of amp...trying to find specs on mine...it says:

Preamplifier section

Input impedance 47 kOhm
THD (1kHz 1V input, 1V output) 0,001%
Max. input > 14 V RMS
Channel separation (1V output at 1kHz) > 90dB
Equivalent input noise < 3 μV
Output impedance 100 Ohm

Amplifier section

Main voltage 120 V / 240 V. Clipping point of the amplifier is set to a level where total harmonic distortion (THD) is 0.2 %.
Output Impedance (20 Hz - 20 kHz) < 0,01 Ohm
Input sensitivity for rated output 0.4 V
Max. Peak current > 80 A
THD (measured at 1 kHz half power, 8 Ohm) < 0,001 %
THD (measured at 1 kHz -1 dB, 8 Ohm) < 0,001 %


So should I be using 47k?? if so then that gives me a single cap of 13.5nF...

And please explain to me why though when I use just a plain xover calc 6db slope I get a 80uF cap??

I can't wrap my head around the small value yet still allowing frequencies above 250hz to pass...

Thanks!!!

 

[planet10]

Your specs don't give the input impedance of the power amp. Do you have a schema?

And please explain to me why though when I use just a plain xover calc 6db slope I get a 80uF cap??

What load impedance are you using? Should be the input R of the amp.

dave

 

[sideways]

I don't have anything but I did find this:

Electrocompaniet

 

[picowallspeaker]

It reports an input impedance of 1 kΩ
(R1 is 1.8 kΩ)
Very strange ...

 

[planet10]

By that power amp input R would be R1=1.8k which is awfully low. 10k is usually considered low.

But you can't get at the power amp? You are using the integrated as a Power amp? Then you would use the pre-amp in.

dave

 

[sideways]

That was the original circuit my amp was based on...otherwise I Gots nothing LOL!!

 

[sideways]

Ok...lets back up a moment...

FAST - FF85WK's and some SEAS woofers...

ASL preamp to a minidsp at 250hz going to two power amps...

Would like to replace the ASL and one power amp with the Electrocompaniet...

BUT...it only has a preamp out with no main in...so it would be impossible to loop back using the minidsp...

Options...run the Fostex's full range but then I'm limited on levels and dynamics...I've already tried

Put on a cap or something to pass 250hz and up...just like what the mini was doing...

Problem is I can't figure out what size cap to use...

If I go with a regular 1st order 6db slope then I'm looking at something like 80uF in size...

A PLLXO puts the cap size way down but what is it?? And do I use a paralled 10k resistor??

 

[chrisb]

"based on" doesn't necessarily mean "exactly same as" - just ask ES, no, maybe don't


Dave - here's a dumb question - if the pre-amp's input was used, would the load impedance vary as the level control pot was adjusted?

 

[chrisb]

for the high pass section the optional resistor is to ground after the cap

As Dave has noted earlier, you can avoid using the R2 if you enter the amp's input impedance ( cel E4? on the spreadsheet) and use the value calculated in E19 ( labeled Ramp as R2 in HP) . It's probably not clear from the nasty photo I posted earlier, but that's what I did.

And yes, if 1K truly is the input impedance, I get a value for C1 of 636nF = closest readily available value likely .6mF - kinda high, but still much cheaper
than to achieve same filter function at speaker level

 

[sideways]

I think I may have missed something myself...PLLXO is not at speaker level but is a xover that comes before the amplification stage of said speaker???

I'm feeling pretty stupid at the moment