A B1 uses Idss as the operating quiescent current.
If you adopt BL grade 2sk170, then that Idss is between 6mA and 12mA.
A B1 runs @ between 6mA and 12mA.
10 stages would require between 60mA and 120mA for the quiescent current supply.
Add on the transient current requirements of each stage.
Since there are lots of capacitors in a B1 crossover there is lots of capacitance that needs to be charged and discharged. That requires current.
I would be surprised if the transient current requirements exceeded the quiescent current requirement, therefore I would guess that a 240mA supply could meet the demands of a 10stage B1 crossover.
BTW,
DCB1 does not equal B1 !
My error. I remember Papa saying a B1 (which I built and have) draws ~20mA for the stereo pair. That would agree with your 6-12/stage. What does a dcb1 draw? I (perhaps wrongly) assumed it would be the same.
Teabag's comment was that he was running a dcb1 @ 800mA which I assumed meant a stereo pair. The only explanation I see is Tony's:
Still incredible current. You "waste" ~780mA to get ~20? I thought 2:1 was good.I think that is the current in the shunt part of the shunt reg, not how much current the B1 part is drawing!
Tony.
I'm not sure what you are referring to? The dcb1 uses a bipolar supply, +/-10V instead of +18 and thereby avoids the need for dc blocking caps in the signal path. I have not seen current consumption for the dcb1, but given nearly identical components, how much different could it be?
If you are referring to the power supply which is essentially the SSLV, I realize THAT can be set to any current you want. My reading of that thread suggested that double the demand current was reasonable for the ccs. So my question remains: Why run 800mA for a circuit that only draws ~20mA? Or am I missing something?
Actually, the most recent 24dB XO has 8 B1 "stages". So @ 20mA/stage (stereo) that means a current draw of 160mA. If you want an overhead of at least 2x, you are up to 320mA which is certainly approaching the "hot rod" version of SSLV1.1 (500mA). I say go for it!... so, it would be nice to have an "hot rodded" DCB1 active XO
What I want is somebody to come up with a GB for a board! (XO only, use BiB for PS)
I plan on doing a GB, I am just not certain we have ironed out an effective schematic to employ for a board layout. I need to discuss probably some pointed questions to a couple of you offline for some fact finding. The big hurdle for me is if how much change to a board is needed to run 6,12,24,48db? I would think it would be a different layout of the circuit, but my knowledge of active X-over's is nill, and time to read about them short lately.
Hi Tea-Bag, basically you stack them to increase the order. But I don't believe 6db is an option. It would be 12db or 24db typically. you could cascade three to get 36db or four to get 48db but I've not seen anyone do that with analog active XO's.
So basically you can have the option for 12db or 24db with 24db basically just being 2 12db circuits in series.
Tony.
So basically you can have the option for 12db or 24db with 24db basically just being 2 12db circuits in series.
Tony.
I'll post the basic building blocks I've cobbled together tonight. Given that 3rd order filters are widely used, I believe we should have a 1st order (6dB/octave) and a 2nd order (12db/octave) section. Alternatively, if an efficient layout could be found, we may be able to get away with just a modular 2nd order section that can be run as 1st order by jumpers and only populating half the filter elements. Stacking such sections could get any slope or combination of slopes desired, 2-way or 3-way or even 4-way.
A DCB1 (or two) with optional global volume control should be ideal for powering and driving the x-over. BSC and relative level controls should be implementable before and/or after every section.
This is going to be tricky, but I'm convinced it can be done. From my experience, the pcb space needed and the resulting layout very much depend on the capacitors used.
A DCB1 (or two) with optional global volume control should be ideal for powering and driving the x-over. BSC and relative level controls should be implementable before and/or after every section.
This is going to be tricky, but I'm convinced it can be done. From my experience, the pcb space needed and the resulting layout very much depend on the capacitors used.
Hi Tony,
its non Sallen-Key, yes. Check out my post here: http://www.diyaudio.com/forums/pass-labs/156094-b1-active-crossover-7.html#post4255228
its non Sallen-Key, yes. Check out my post here: http://www.diyaudio.com/forums/pass-labs/156094-b1-active-crossover-7.html#post4255228
I'm not sure from your post if you plan to impliment an LR4 (24dB) or if that would require 2 boards. I, for one, would vote for 24dB option. It is easy to jumper that section if someone only wants 12.
Also, just for my edification, why is 18dB popular? It is not an LR topology and not phase coherent. I always thought 12 or 24 dB LR were the more ideal XO.
A 2pole 12dB/oct, 40dB/dec S&K filter can have the feedback omitted to give a single pole 6dB/oct/20dB/dec filter,
or have an extra RC tacked on the front to give a 3pole 18dB/oct/60dB/dec filter.
if the speaker driver had a 2pole roll off at exactly the frequency one wanted for the active and one added a 2pole S&K Butterworth to that, you end up with a 4pole with the correct slope and frequency for half of an LR4.
This could be applied to a Mid, or to a Treble, driver in a sealed enclosure.
If the drivers were in sealed enclosures that were far too big and the resulting roll-off frequency was far too low one could add a DC protecting capacitor giving a single pole roll off . This would then need a 3pole filter to create a LR4.
I was reminded that the cap protection does not give an accurate roll-off if the LF resonance occurs near the required LR4 frequency.
I would guess that the resonance would need to be at least two octaves lower, to be disregarded for LR4 accuracy.
or have an extra RC tacked on the front to give a 3pole 18dB/oct/60dB/dec filter.
if the speaker driver had a 2pole roll off at exactly the frequency one wanted for the active and one added a 2pole S&K Butterworth to that, you end up with a 4pole with the correct slope and frequency for half of an LR4.
This could be applied to a Mid, or to a Treble, driver in a sealed enclosure.
If the drivers were in sealed enclosures that were far too big and the resulting roll-off frequency was far too low one could add a DC protecting capacitor giving a single pole roll off . This would then need a 3pole filter to create a LR4.
I was reminded that the cap protection does not give an accurate roll-off if the LF resonance occurs near the required LR4 frequency.
I would guess that the resonance would need to be at least two octaves lower, to be disregarded for LR4 accuracy.
Last edited:
Here are B4 and B5 manual, you can find some explications of block diagram and differences between 1,2,3,4 order.
Anyway, if is it possible to implement 1,2,3,4 order + BSC, that would be great and highly flexible !
A speaker is also an hi-pass filter, so an electric third order hi-pass could be an acoustic 4 order when combined to speaker's acoustic response... I think. But I'm not sure
Anyway, if is it possible to implement 1,2,3,4 order + BSC, that would be great and highly flexible !
A speaker is also an hi-pass filter, so an electric third order hi-pass could be an acoustic 4 order when combined to speaker's acoustic response... I think. But I'm not sure
Yep. We will have to rally around an agreeable set of available, non exotic, easily obtainable capacitors. Wimas and Vishay MKP1837 etc come to mind, 1% tolerance stuff.
Cool thanks RodeoDave! It was Jaques original post that started me on my design (which diverged completely but still uses B1's) I'm particularly interested in the notch filter as I have a huge room resonance at around 40Hz
I've already made my boards but could possibly be an addition I could make as a separate board. Nice work! Tony.
I used POLYSTYRENE caps, you can buy them cheap and match them as needed, parallel, etc. In capacitor hierarchy, they are better than polypropylene in my opinion. I bought 100 each 1000 picofarad for $13.20 plus what ever values needed to get the slopes, parallel them.
Xicon POLYSTYRENE Film Capacitors | Mouser
The resistors may also need to be paralleled to achieve the value needed, double holes wouldn't hurt for the slope resistors - easy enough to do on top of the board though.
For 2700 Hz crossover:
C = C1 = C2 = C3 = C4 =.002
1st section: R5 = 0.9239/(2*Π*F*C) = 27.2k
1st section to ground: R6 = 1.0824/(2*Π*F*C) = 32k
2nd section: R11 = 0.3827/(2*Π*F*C) = 11.3k
2nd section to ground: R12 = 2.6130/(2*Π*F*C) = 77k
Butterworth low pass (Schematic 8)
R = R5 = R6 = R12 = R13 = 22k
1st section: C1 = 1.0824/(2*Π*F*R) = .0029 uF
1st section to ground: C2 = 0.9239/(2*Π*F*R) = .00247 uF
2nd section: C3 = 2.6130/(2*Π*F*R) = .007 uF
2nd section to ground: C4 = 0.3827/(2*Π*F*R) = .001 uF
Rush
open baffle with step equalization plus 20hz high pass
http://www.diyaudio.com/forums/full-range/204304-dx55-ob-micro-article.html#post2858923
http://www.firstwatt.com/pdf/prod_b5_man.pdf
http://www.firstwatt.com/pdf/prod_b4_man.pdf
http://www.diyaudio.com/forums/full-range/204304-dx55-ob-micro-article.html#post2858923
http://www.firstwatt.com/pdf/prod_b5_man.pdf
http://www.firstwatt.com/pdf/prod_b4_man.pdf
Last edited:
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.