• TalkBass has been independent since 1998. Add your voice.
    Create a free account to reply to discussions, view embedded media, and browse with fewer display ads.
    Join free Log in
    Want zero display ads or expanded classifieds tools? Compare plans.

Crossover Voodoo

#22
Zooberwerx said:
Well, how come when I place an old toilet in my front yard as a planter, the civic league gets all bent out of shape?!? Porecelain is porcelain in my book.
Click to expand...
Toilets as planters! Psssh. The correct yard use of porcelain is for a Bathtub Mary.

2854694231_01594e12dc_o.jpg
 
#25
Weird.

I have run into people with backgrounds in RF and microwave who couldn't get their heads around the fact that impedance matching/maximum power transfer isn't the goal for most audio circuits. "What do you mean you want a low impedance driving a high one? That won't give you maximum power transfer!" No but in most cases audio circuits are based on maximum voltage transfer from supply to load. Maybe there's a bit of that going on because he clearly has some idea about electronics. Maybe just enough of an idea to be dangerous. :D

Are you etching those PCBs in house? How delightfully old school of you. :)
 
#26
Mark Reccord said:
Weird.

I have run into people with backgrounds in RF and microwave who couldn't get their heads around the fact that impedance matching/maximum power transfer isn't the goal for most audio circuits. "What do you mean you want a low impedance driving a high one? That won't give you maximum power transfer!" No but in most cases audio circuits are based on maximum voltage transfer from supply to load. Maybe there's a bit of that going on because he clearly has some idea about electronics. Maybe just enough of an idea to be dangerous. :D

Are you etching those PCBs in house? How delightfully old school of you. :)
Click to expand...

Of course, damping factor is load impedance divided by source impedance, but power transfer itself is about Ohm's law, nothing else.

As an aside, my understanding of the whole concept of the "transmission line" loudspeaker, in which the backwave is completely absorbed, is based on RF antenna technology, about which I know next to nothing.

Yeah, I lay out those circuit boards pixel by pixel. We have them silkscreened and etched locally, and finish them up here. I've been doing it for many years, and for me, it's kind of like a complex puzzle, in which overlapping parts can crisscross each other on both sides of the board, and all can fit.

The greater challenge is to design a board on which different circuits and combinations of components can all be assembled, based on requirements, on the same board. It gets a little hairy, but it's kind of fun, and hey- I do this for a living!

What I like about the current incarnation is that all point-to-point wiring has been eliminated. That was a bit of a challenge. I suppose it is sort of old school, but what's the alternative?

Andy
 
#27
The alternative is military-style binding posts but not many musical equipment companies use those any more. The second amp I have owned (and still do) used them - it was made by Carvin in 1976. I guess that is essentially point-to-point wiring. PCBs are a far better option I think. Especially if chips are involved. They usually aren't used in passive crossovers though, regardless of what Sergei is advocating. I get the impression he has no idea what passive crossovers are about.
 
#28
B-string said:
O M G..........someone sub nitrous oxide for his oxygen tank? Maybe he works for BOSE?
Quote " read about power supply design. Current limiting and impedance matching..." for a passive audio crossover? I laughed so hard I think I wet myself.

[Note for internet: Engage brain before putting mouth or hands in gear]
Click to expand...

That pretty much covers it me thinks.
 
#29
acmebass said:
Of course, damping factor is load impedance divided by source impedance, but power transfer itself is about Ohm's law, nothing else.
Click to expand...

As are most things in electronics at the end of the day. :)

As an aside, my understanding of the whole concept of the "transmission line" loudspeaker, in which the backwave is completely absorbed, is based on RF antenna technology, about which I know next to nothing.
Click to expand...

I think the math is directly applicable. The math involved in loudspeaker arrays is essentially equivalent to that of RF antenna arrays (directivity, etc.) with audio being actually more complex due to the short wavelengths involved. Oh god, integration in 3 dimensions.... :eek:

Yeah, I lay out those circuit boards pixel by pixel. We have them silkscreened and etched locally, and finish them up here. I've been doing it for many years, and for me, it's kind of like a complex puzzle, in which overlapping parts can crisscross each other on both sides of the board, and all can fit.

The greater challenge is to design a board on which different circuits and combinations of components can all be assembled, based on requirements, on the same board. It gets a little hairy, but it's kind of fun, and hey- I do this for a living!

What I like about the current incarnation is that all point-to-point wiring has been eliminated. That was a bit of a challenge. I suppose it is sort of old school, but what's the alternative?

Andy
Click to expand...

I'm currently a PCB designer in the telecom industry. Physically etching a PCB is old school to me. Everything I design is manufactured with a laser on a CNC machine (by necessity). The board I'm completing right now has 18 copper layers, over 6000 nets, devices with 1900+ pins and lots of 10-17GHz links, so I understand your puzzle analogy. :)

Agreed that eliminating point to point wiring is a good thing in terms of reliability, which is massively important for what you do.

Cheers,
Mark
 
#30
acmebass said:
Hey- I've got one for you speaker buffs.

I posted some pictures of one of my crossovers on my FB page, and got a few comments.

Check the comment I got from a guy:

"I would add some connectors and signal buffers to isolate the crossover and reduce current pulled from the driver."

To which I replied:

"Interesting observation, Sergei. I'd love to see a link to whatever it is you're talking about!

Isolate the crossover from what?

What is "current pulled from the driver?" How do you "pull" current?"

Obviously, I never heard anything like this in my life. So check out his response:

"Current pulled from the supply is the current carried throught the transmission line to the load. Load is the crossover. If you pull more current than the supply can provide you will "clamp" the output.l, thus change your signal. Buffer is the device that allows you use external supply (not the signal source) to drive your speaker - basically have a mini analog amp on you board and it will increase the performance."

Yeah, I know, it's the internet! "Increase the performance."

Wow! That sounds great, if it seems to echo the ad for Viagra that I heard.

Anybody have any idea what this dude's trying to say?

Thought it might be good for a laugh.

Andy
Click to expand...

Andy, never heard anything like that being used in a passive crossover network. I don't understand what he means either, just ignore it :)
Grtz,
Arjan
 
#36
Mark Reccord said:
As are most things in electronics at the end of the day. :)

I'm currently a PCB designer in the telecom industry. Physically etching a PCB is old school to me. Everything I design is manufactured with a laser on a CNC machine (by necessity). The board I'm completing right now has 18 copper layers, over 6000 nets, devices with 1900+ pins and lots of 10-17GHz links, so I understand your puzzle analogy. :)

Agreed that eliminating point to point wiring is a good thing in terms of reliability, which is massively important for what you do.
Click to expand...

I have a friend who's in that field here in Colorado. Those modern high-tech board are really amazing. Most of them, I believe, have negligible current requirements, and get away with incredibly small traces. That whole enterprise is obviously a lot different that what's happening here...

Andy
 
#37
acmebass said:
I have a friend who's in that field here in Colorado. Those modern high-tech board are really amazing. Most of them, I believe, have negligible current requirements, and get away with incredibly small traces. That whole enterprise is obviously a lot different that what's happening here...

Andy
Click to expand...

There are certainly a lot of signals that can use traces as small as 5 mils(!) but there are plenty of high current spots too. The core of the FPGA on the board I'm working on right now draws ~30A from the power supply. It's at 0.85V and all, but it still needs lots of copper so that there is very little voltage drop. I have 4 layers completely dedicated to DC distribution using planes as opposed to traces and the 0.85V straddles a pair of them. Lots and lots of vias. :D

My background is in audio and I've done lots of board design in that field as well. I miss working in audio. I guess it's something that I'm passionate about as opposed to developing products that facilitate ever faster porn downloads, which is what I do now... :D

Anyway, enough of that serious stuff, let's get back to the regularly scheduled silliness!
 
#38
Mark Reccord said:
There are certainly a lot of signals that can use traces as small as 5 mils(!) but there are plenty of high current spots too. The core of the FPGA on the board I'm working on right now draws ~30A from the power supply. It's at 0.85V and all, but it still needs lots of copper so that there is very little voltage drop. I have 4 layers completely dedicated to DC distribution using planes as opposed to traces and the 0.85V straddles a pair of them. Lots and lots of vias. :D

My background is in audio and I've done lots of board design in that field as well. I miss working in audio. I guess it's something that I'm passionate about as opposed to developing products that facilitate ever faster porn downloads, which is what I do now... :D

Anyway, enough of that serious stuff, let's get back to the regularly scheduled silliness!
Click to expand...

Yeah- what you said!

A
 

Latest posts

Top Bottom
Back