Especially D1.???..
I realize that most of this is tweeter protection, but what the heck is D1 doing in series with the tweeter? Also if possible, could someone tell me the steps to bypass this mess of a protection circuit, I was planning on jumping D1, lifting a leg on D2 though D5 and jumping the relay. Does this sound good? Any info at all is appreciated. B&W wow, I am surprised at this mess.
I realize that most of this is tweeter protection, but what the heck is D1 doing in series with the tweeter? Also if possible, could someone tell me the steps to bypass this mess of a protection circuit, I was planning on jumping D1, lifting a leg on D2 though D5 and jumping the relay. Does this sound good? Any info at all is appreciated. B&W wow, I am surprised at this mess.
Attachments
I'd guess the tweeter impedance and crossover point is quite low here, hence the protection circuit.
You also have 2X LCR impedance correction at the woofer input and Zobels on the woofers.
Removing R8-11 would take the protection circuit out, but I'd sim the filter circuit first to see how it works out. At heart it's a 2.5 way filter, with a third order treble. L5 ,L6, C7, C8 is an allpass filter I have rarely seen used in audio. It adds time alignment and adjusts phase.
I have seen Zaph Audio do this sort of bass filter and treble filter.
I have got to wonder why you want to get rid of the protection. It looks well designed, and won't impact the sound much.
You also have 2X LCR impedance correction at the woofer input and Zobels on the woofers.
Removing R8-11 would take the protection circuit out, but I'd sim the filter circuit first to see how it works out. At heart it's a 2.5 way filter, with a third order treble. L5 ,L6, C7, C8 is an allpass filter I have rarely seen used in audio. It adds time alignment and adjusts phase.
I have seen Zaph Audio do this sort of bass filter and treble filter.
I have got to wonder why you want to get rid of the protection. It looks well designed, and won't impact the sound much.
Diodes D1 and D2 supply the relay protection circuit in proportion to the signal current via the series resistor dividing networks R8,R9 and R10,R11. D1 is not in series with the tweeter.It is a complicated circuit but not a mess.To remove the protection simply lift one end of both diodes from the circuit board.This is in answer to your question;not a recommendation to do so.
D1 is not in series with the tweeter, trace it out carefully.
D1 and D2 are a "wired OR" connection taping a fraction of the voltage to the drive units as a sense for the protection circuit. Just remove them and jump the really if you really want to remove it.
This looks like a crossover from the time when B&W just started using computer optimization and they were rather complex.
FWIW, I've just remembered where I've seen this sort of bass filter, followed by midrange trap/notch, followed by bass Zobel idea before...🙂
It's BBC polypropylene monitor days, as in the Rogers/BBC LS5/9, here documented and explained very well by Mark Hennessey:
Rogers Speakers - BBC LS5/9
The B&W Matrix 3 must have shockingly low and flat impedance being a 2.5 way...maybe it was designed for a valve amp?
It's BBC polypropylene monitor days, as in the Rogers/BBC LS5/9, here documented and explained very well by Mark Hennessey:
Rogers Speakers - BBC LS5/9
The B&W Matrix 3 must have shockingly low and flat impedance being a 2.5 way...maybe it was designed for a valve amp?
Last edited:
D1 is not is series with the tweeter. In fact, none of the diodes are. If you trace C5 backwards, you'll find the the tweeter is connected directly to the power amp output via the "closed" relay. In normal operation, the protection circuit does not affect the performance of the tweeter. This is based on the assumption that the relay contacts have not gone bad. When there's too much power, the protection circuit activates and the realy contacts open, thus disconnecting the amplifier from the tweeter.
If you really want to bypass the protection mechanism, simply remove diodes D1,D2 and D4. The relay will remain closed forever.
Briefly, this is the breakdown of the crossover. Starting with the MID/LF section, the first two LCR are Series Notch Filters. Looks like one is to tame the Resonance peak while the other is to notch a spike in the higher frequency.
It is then followed by a 3rd order Low Pass with some eq done, hence the parallel network in L2. The final series filter, R2, C3 is a zobel. That is for Impedance EQ.
What is unusual is the LF is taken from the MID/LF network instead of directly from the amplifier. The LF starts off with a 1st order, a 10mH (L3). That indicates quite a low crossover frequency Low Pass, perhaps 100Hz - 200Hz. R3 and C4 is again for Impedance EQ. But since it is taken after the MID/LF network, there will be another slope.
The tweeter crossover is 3rd order (C5,C6 & L4). I'm not certain about L5, L6,C7 and C8. My guess is L5 and C8 is the left side of a notch filter while C7, L6 is the other half. Perhaps this is deemed more effective than using a LCR to tame the tweeter resonance frequency. Lastly, R5,C9 is for Impedance EQ.
Personally, it's quite an impressive crossover. How it sounds like is another matter. Is there something you don't like about your B&W?
If you really want to bypass the protection mechanism, simply remove diodes D1,D2 and D4. The relay will remain closed forever.
Briefly, this is the breakdown of the crossover. Starting with the MID/LF section, the first two LCR are Series Notch Filters. Looks like one is to tame the Resonance peak while the other is to notch a spike in the higher frequency.
It is then followed by a 3rd order Low Pass with some eq done, hence the parallel network in L2. The final series filter, R2, C3 is a zobel. That is for Impedance EQ.
What is unusual is the LF is taken from the MID/LF network instead of directly from the amplifier. The LF starts off with a 1st order, a 10mH (L3). That indicates quite a low crossover frequency Low Pass, perhaps 100Hz - 200Hz. R3 and C4 is again for Impedance EQ. But since it is taken after the MID/LF network, there will be another slope.
The tweeter crossover is 3rd order (C5,C6 & L4). I'm not certain about L5, L6,C7 and C8. My guess is L5 and C8 is the left side of a notch filter while C7, L6 is the other half. Perhaps this is deemed more effective than using a LCR to tame the tweeter resonance frequency. Lastly, R5,C9 is for Impedance EQ.
Personally, it's quite an impressive crossover. How it sounds like is another matter. Is there something you don't like about your B&W?
Oh, I didn't know it's a 2.5 way. That accounts for no High Pass for the MID/LF unit. I thought it was quite brilliant taking the LF from the MID/LF network.
Wow! Thanks for all the great replies so far.
Here is the deal.
I am a mixing and mastering guy. I bought these speakers as an alternate set of references mostly for mixing, crosschecking in another room. The original crossover in one of the units was replaced by an unscrupulous Craigslist scammer who told me he had had the units recapped by the dealer and brought "up to spec" Haha! my own fault for not listening to these carefully before I took them, I do tend to trust people.
He had replaced the crossover in one unit with a "Pyle" generic crossover electrically taped in ! Almost funny if it were not so sad. Anyway I lucked out and found a correct set on eBay, which sound and measure differently per unit and when switched side to side to eliminate driver issues. The drivers are fine, tweeters were disassembled, ferrofluid was well in gap no congealed no sign of stress and in all drivers great shape actually. In short, no driver issues.We all know where this is going, 30 year old non polar 'lytic caps. At this point I will recap the electrolytics with same type as close as possible to maintain designed in esr.
So....since I will NEVER be using these at high volumes and always with their own dedicated amp, carefully eased down and up at each mastering session, I thought I would take this opportunity to get the help of you guys to help me clean these up, eliminate the parts of the crossover that could conceivably be gratuitous for my purpose in this carefully controlled environment. The protection circuit is a candidate for certain, there may be more, if so I would very much appreciate your opinions. These speakers did measure +/- 2 db from 70hz up and came from factory with individual plots. Not so bad for the time. Do I think you guys know this stuff a bit better now than they did at that time? well, yes.
Here is the deal.
I am a mixing and mastering guy. I bought these speakers as an alternate set of references mostly for mixing, crosschecking in another room. The original crossover in one of the units was replaced by an unscrupulous Craigslist scammer who told me he had had the units recapped by the dealer and brought "up to spec" Haha! my own fault for not listening to these carefully before I took them, I do tend to trust people.
He had replaced the crossover in one unit with a "Pyle" generic crossover electrically taped in ! Almost funny if it were not so sad. Anyway I lucked out and found a correct set on eBay, which sound and measure differently per unit and when switched side to side to eliminate driver issues. The drivers are fine, tweeters were disassembled, ferrofluid was well in gap no congealed no sign of stress and in all drivers great shape actually. In short, no driver issues.We all know where this is going, 30 year old non polar 'lytic caps. At this point I will recap the electrolytics with same type as close as possible to maintain designed in esr.
So....since I will NEVER be using these at high volumes and always with their own dedicated amp, carefully eased down and up at each mastering session, I thought I would take this opportunity to get the help of you guys to help me clean these up, eliminate the parts of the crossover that could conceivably be gratuitous for my purpose in this carefully controlled environment. The protection circuit is a candidate for certain, there may be more, if so I would very much appreciate your opinions. These speakers did measure +/- 2 db from 70hz up and came from factory with individual plots. Not so bad for the time. Do I think you guys know this stuff a bit better now than they did at that time? well, yes.
Last edited:
Thanks Michael. that was brilliant analysis. At this point I'm not sure about how I like them given the intended use. Obviously I do need flat FR as possible. When I get the recapping done I will know more, C4 and C5 sure look like they could wreak havoc if out of spec do they not?
I believe that B&W sold serialized sets in matched pairs, with the crossover components matched to each individual driver.
And... Wow! Seriously? A Pyle crossover was put in that fine speaker?!?!
Some peoples kids...
Mike
And... Wow! Seriously? A Pyle crossover was put in that fine speaker?!?!
Some peoples kids...
Mike
BTW, the protection circuit on this fairly decent old Kenwood power amp will, as it should, readily attest to the "shockingly low impedance" of these 2.5 units if I connect them in parallel for at any appreciable power level (-:
Very interesting project, Pete...🙂
I was wondering about the shape of the tweeters. Ferrofluid has its downside when it turns into gunge...
I found a couple of designs that will help you understand the filters:
Simpler 2.5 way bass filter:
Zaph|Audio
Time delay treble filter for flat baffle:
Zaph|Audio - ZD5 - Scan Speak 15W8530K00 and Vifa XT25
Impedance correction is another topic, but those speakers of yours have it. Hell of a crossover! Designer spared no expense! 😎
I was wondering about the shape of the tweeters. Ferrofluid has its downside when it turns into gunge...
I found a couple of designs that will help you understand the filters:
Simpler 2.5 way bass filter:
Zaph|Audio
Time delay treble filter for flat baffle:
Zaph|Audio - ZD5 - Scan Speak 15W8530K00 and Vifa XT25
Impedance correction is another topic, but those speakers of yours have it. Hell of a crossover! Designer spared no expense! 😎
Oh yeah, a Pyle generic on a B&W 2.5 and I let this man in my studio.... Doh!
Not to get off subject, but that Zaph is a great looking speaker on paper for sure! The Matrix 3's? the crossovers each weigh several pounds and they seem to be, as a lot of the B&W speakers from that era are, quite inefficient, a byproduct of all of that impedance work right? At any rate, at the levels I am using them and provided I can get them pretty flat again, that shouldn't matter. Probably going to get an old Hafler DH2** for them.
Ferrofluid.....Funny thing about ferrofluid, you really can't measure it out and it won't matter if you do.. The best way if it is "grunged" is to clean the units/gaps up, over fill them and then start wicking it out with paper. I've found that the tweeters will "tell" you when the amount is right because if over filled the coil will not fall into place. When it is just right, the coil will fall in and you can "feel" it center itself. Both tweeters need to "feel" the same when they drop in There is IMO, a tiny bit of lattitude in the amount you need, but not a lot. It is certainly "feelable" if you are a mechanical/physics sort of person. One thing for sure, no damn way they got an exact amount per tweeter at the factory, I have no doubt a LOT of mic measuring and matching per sets of speakers was done. B&W absolutely "gave a damn" on these units, one reason I wanted a pair for refs. Looking for another set of 801's from that era that need TLC as well.
There is a lot going on in this crossover, that much is certain. I am taking from all of your great input, that there really is nothing that can be eliminated without altering the specs of this speaker save for the tweeter protection, which it seems to me could conceivably have some audible artifacts, but perhaps I am mistaken. Also plan on jumping the relay in case there is any corrosion in the future.
One more point, there are apparently and visibly exactly four electrolytics in the signal path, which I plan on replacing with the closest match possible, the rest of the caps are huge non degrading paralleled (when needed) poly caps which add up exactly to the engineered specifications on schematic (things like 19.6uf right?) As I mentioned, there is at this time, certainly an audible difference between the crossovers. It seems to be centered in the high mid to hf but not the vhf area. There are but four electrolytics that are (as I think I have verified by you all), in the signal path.. They are as follows:
C10@63uf, C11@600uf, C3@20uf and C4 also@20uf. I am assuming C3 and C4 are the most suspicious regarding variations in FR?
Last edited:
Looking closer, and guided by what you guys have told me, I'm guessing C3 is very, very, very likely to be my FR deviation culprit given there are no electrolytics in the crossover anywhere near the tweeter circuit path, C5 through C9 are all gigantic brown parallel/ganged-up "chiclets" (-:
Thanks a lot, Hammond organ did something like this mechanically with the scanner for chorus in the 40's. goes to show that a circuit is a circuit, wires or pasta (-:
- Status
- Not open for further replies.