pinkmouse said:
I scoped it without the load (cable and power amp dissconnected)
I suspect the sockets; too many times I've got an IC in and out and the connection become lossy.
I will check with load and after I will solder the LM6172 to the board when I'll get the 6172.
Avi
gmikol said:
Hi,
I think I'll set up a dedicated webpage for this project......
I'll see what I can do....
The filese will be PDFs as I don't want to share my original files... Someone pointed out that eagle stores the serial number used for editing in the .brd and .sch files and I don't want my original and expensive serial floating on the www
\Jens
It may have come up here, but I haven't been able to pick it up - how big is this filter? PCB dimensions and approximate height.
The board is 80 x 100 mm. Build height depends on your choice of caps. Plan on 15mm or more for stacking purposes.
Thanks. I'll be using the Panasonics caps provided, so it'll fit nicely in the 44mm case I need for my volume control stepper.
I have a question on my application, though:
They're originally designed as ordinary rectangular boxes with a side-mounted woofer, but I just had to go and change it... The original bass filter is a 4.7 mH/74uF low-pass, but it has an extra 5.5Ohm -> 22mH -> 268uF in parallel. Dunno what it's for...
I'll be using Lars Clausen's NCD amps, derivatives of his LC Audio 700XE class D amps. I'll be running a simple two-way between the bass and mid/high, and let the M-Cap Supremes and all the other goodies do their job in the highs.
I have a question on my application, though:
An externally hosted image should be here but it was not working when we last tested it.
They're originally designed as ordinary rectangular boxes with a side-mounted woofer, but I just had to go and change it... The original bass filter is a 4.7 mH/74uF low-pass, but it has an extra 5.5Ohm -> 22mH -> 268uF in parallel. Dunno what it's for...
I'll be using Lars Clausen's NCD amps, derivatives of his LC Audio 700XE class D amps. I'll be running a simple two-way between the bass and mid/high, and let the M-Cap Supremes and all the other goodies do their job in the highs.
The "extra parts" sound like a notch filter. Probably to tame a woofer breakup mode. You'll need to measure the response to see what is needed in the active system.
It's a good idea to measure rather than try to reverse engineer the passive XO anyway since you are changing the box shape. Although visually interesting, I'm not sure about the tilt's effect on lobing - I'd measure the response at the intended listening distance and angle and design the XO accordingly.=
It's a good idea to measure rather than try to reverse engineer the passive XO anyway since you are changing the box shape. Although visually interesting, I'm not sure about the tilt's effect on lobing - I'd measure the response at the intended listening distance and angle and design the XO accordingly.=
Hi Bob,
Is it too late to order 4 active filter boards? Sorry, I only just discovered this thread.
If possible please let me now and I'll order through your site, otherwise I'll wait until AF5...
Thanks,
Andrew.
Is it too late to order 4 active filter boards? Sorry, I only just discovered this thread.
If possible please let me now and I'll order through your site, otherwise I'll wait until AF5...
Thanks,
Andrew.
Andrew, it is not too late to order. I didn't get the order placed Friday. I will buy some extras anyway. I still have a few power supplies left, too.
Oh God!
what have I missed out on?
Oh, it's OK. someone's running late.
Oh, hang on a mo, he's on about some-one else.
Well done Andrew (without the T.) you crossed the finishing line before they all went home.
what have I missed out on?
Oh, it's OK. someone's running late.
Oh, hang on a mo, he's on about some-one else.
Well done Andrew (without the T.) you crossed the finishing line before they all went home.
Thanks for calling it an "interesting" design, BobEllis, it's been called far worse 🙂 I went for the design pretty much because I wanted something completely different, and I figured pyramid shapes would be good for acoustics.
The speakers are only designed on UGS so far, so it's not a big problem to redo it, but I'd like to keep the general shape.
Would moving the bass to the front be a better option? Afterall, there is room for it, and the only reason I kept on the side is because that's where it was originally. Unless I've misunderstood completely, I could get by nicely with the shelving low pass filter on that one, and possibly a notch filter.
Any tips on the measuring you mentioned? I don't have any instruments for it, but could a decent sound card and a good microphone to the job?
The speakers are only designed on UGS so far, so it's not a big problem to redo it, but I'd like to keep the general shape.
Would moving the bass to the front be a better option? Afterall, there is room for it, and the only reason I kept on the side is because that's where it was originally. Unless I've misunderstood completely, I could get by nicely with the shelving low pass filter on that one, and possibly a notch filter.
Any tips on the measuring you mentioned? I don't have any instruments for it, but could a decent sound card and a good microphone to the job?
I really like your design visually. I don't think I have the patience to build something that complex and doubt I could sell it into her conservatively decorated house, but I like the look.
I use the Linkwitz circuit and modified WM-61A capsule - under $15 for everything at digikey. Speaker Workshop is a free program that will allow you to measure the response. For some it has a fairly steep learning curve, but Claudio Negro has published a very handy guide that walks you through measuring.
One of the perks of an active design is that within reason, you don't give a hoot about the drivers' impedances. So you can skip through the parts about measuring impedance, except for verifying woofer T-S parameters and box alignment.
Assuming that you plan to cross to the woofer below 100 Hz or so, side or front mounting makes little difference. If you want to cross it higher (and the woofer is clean higher) move it to the front. If you can bring the crossover point up to the baffle step, you can eliminate a shelving filter.
You'll need a low pass filter on your woofer, and maybe a shelving low pass for BSC. See Linkwitz's active crossover page for an explanation of filter types.
I use the Linkwitz circuit and modified WM-61A capsule - under $15 for everything at digikey. Speaker Workshop is a free program that will allow you to measure the response. For some it has a fairly steep learning curve, but Claudio Negro has published a very handy guide that walks you through measuring.
One of the perks of an active design is that within reason, you don't give a hoot about the drivers' impedances. So you can skip through the parts about measuring impedance, except for verifying woofer T-S parameters and box alignment.
Assuming that you plan to cross to the woofer below 100 Hz or so, side or front mounting makes little difference. If you want to cross it higher (and the woofer is clean higher) move it to the front. If you can bring the crossover point up to the baffle step, you can eliminate a shelving filter.
You'll need a low pass filter on your woofer, and maybe a shelving low pass for BSC. See Linkwitz's active crossover page for an explanation of filter types.
Thanks, Bob. I've placed the order.
AndrewT - I see you are almost as excited as I am about these boards??
AndrewT - I see you are almost as excited as I am about these boards??
Hi Arj,
yes,
I only just recently realised the topology reverted to EV S&K from the unity gain S&K.
They now mimic the original MOX boards and it makes component matching and Q adjustment so much easier.
The only part missing is easy conversion to MFB, but I'll save that till I have the PCBs in hand.
yes,
I only just recently realised the topology reverted to EV S&K from the unity gain S&K.
They now mimic the original MOX boards and it makes component matching and Q adjustment so much easier.
The only part missing is easy conversion to MFB, but I'll save that till I have the PCBs in hand.
Bushroot,
The email address you provided to this forum is no longer valid, so I will post my response here.
Each board can be a high & Low pass or a band pass. So to meet your 3 way + sub requirement you’d want 3 boards per channel for a total of 6 per stereo set. If you need a high pass on your sub for a peaking EQ or rumble protection you'd add one more board. (with three boards you'd have three Linkwitz transforms which you could use for inverting EQ by jumpering and omitting some parts if you want to avoid another board for just a single function.)
There are two connection options - that I call parallel and series. For parallel, you'd use the input buffer on one board and then feed all other boards from its buffered out. You'd use one for the extreme top and bottom, two for band pass. This would be basically the same thing Jens did with his AF1.
Some say series connection is better for preserving phase relationships between the drivers. I've always done it this way and haven't tried to see if the difference is audible. In this version, the first board would be the subwoofer XO. You'd feed its low pass to the sub and the HP to rest of the system. Then you'd have a woofer-mid connection where low pass goes to the woofer and high to the mid and tweeter. The band pass is formed by the first board's high pass and the second's low pass. Then the high pass goes to the mid-tweeter XO, where the low goes to the mid and the high to the tweeter.
This scheme allows you to work on phase correction mid to tweet and have it unaffected by your woofer-mid phase corrections. Also, this is the way Linkwitz sets his up, so I just followed his lead. If you need more than one all pass section per xo, in my experience the mid-tweet anomalies are the most noticeable, and the woofer-subwoofer phase difference is inaudible until it approaches 180 degrees.
I have only 5 power supplies left if you were considering the GB power supply. With 6 boards to power, you'll be on the edge of a single supply's capability if you use all EQ features. Two supplies would be a safer bet.
Bob
The email address you provided to this forum is no longer valid, so I will post my response here.
Each board can be a high & Low pass or a band pass. So to meet your 3 way + sub requirement you’d want 3 boards per channel for a total of 6 per stereo set. If you need a high pass on your sub for a peaking EQ or rumble protection you'd add one more board. (with three boards you'd have three Linkwitz transforms which you could use for inverting EQ by jumpering and omitting some parts if you want to avoid another board for just a single function.)
There are two connection options - that I call parallel and series. For parallel, you'd use the input buffer on one board and then feed all other boards from its buffered out. You'd use one for the extreme top and bottom, two for band pass. This would be basically the same thing Jens did with his AF1.
Some say series connection is better for preserving phase relationships between the drivers. I've always done it this way and haven't tried to see if the difference is audible. In this version, the first board would be the subwoofer XO. You'd feed its low pass to the sub and the HP to rest of the system. Then you'd have a woofer-mid connection where low pass goes to the woofer and high to the mid and tweeter. The band pass is formed by the first board's high pass and the second's low pass. Then the high pass goes to the mid-tweeter XO, where the low goes to the mid and the high to the tweeter.
This scheme allows you to work on phase correction mid to tweet and have it unaffected by your woofer-mid phase corrections. Also, this is the way Linkwitz sets his up, so I just followed his lead. If you need more than one all pass section per xo, in my experience the mid-tweet anomalies are the most noticeable, and the woofer-subwoofer phase difference is inaudible until it approaches 180 degrees.
I have only 5 power supplies left if you were considering the GB power supply. With 6 boards to power, you'll be on the edge of a single supply's capability if you use all EQ features. Two supplies would be a safer bet.
Bob
I have the luxury of having access to a CNC mill at the university, so whatever my mind can put into the computer, I can have made. And I'm gloriously free from female attachments, so that's not a problem. My flatmate's girlfriend might be less impressed, but that's not my problem 🙂
The woofer's cut at 250 Hz originally, so I think I'll stick with that. It uses a SEAS H1305/CA26RFX ported woofer and a pair of specially designed SEAS H1123 15cm mids. The passive filters are 12dB - would you recommend going up to 24dB with the active filters?
The woofer's cut at 250 Hz originally, so I think I'll stick with that. It uses a SEAS H1305/CA26RFX ported woofer and a pair of specially designed SEAS H1123 15cm mids. The passive filters are 12dB - would you recommend going up to 24dB with the active filters?
250 Hz crossover to a side mounted woofer? Are you sure they aren't a B*** design? 
In case you couldn't tell, I doubt you would be happy with them long term with a side mounted woofer. As for the filter order, let the measured response determine that. Zaph has a nice explanation of filter choice in one of his design details, I think ZD5.
The nutshell version is use a 2nd order ACOUSTIC slopes if the drivers will allow. Reasons for going higher order are nasty breakup modes that would not be adequately tamed or tweeter low end power handling.
Zaph is a little at odds with Linkwitz on this, though. Linkwitz points out that with less than a second order high pass electrical filter, a driver's excursion will continue to increase below cutoff even though its output will fall. Therefore, to minimize potential distortions, high pass filters should be at least 2nd order electrical.
An interesting historical note - in his early work (pre-dipole) Linkwitz used a bi-quad high pass in addition to the standard high pass filter. This compensated for the drop in tweeter output below resonance and made the acoustic slope continuous, rather than increasing below tweeter resonance. It appears that he has found that in the case of the current tweeters he uses the slope change is not audible and doesn't bother with the added complexity. It might be something worth investigating, since you'll have a spare bi-quad (Linkwitz transform) section available. Yes, I have been eyeing active speakers since Linkwitz' 1980 Speaker Builder articles.
Although she may limit my speaker designs, overall I prefer to be encumbered. 😉
In case you couldn't tell, I doubt you would be happy with them long term with a side mounted woofer. As for the filter order, let the measured response determine that. Zaph has a nice explanation of filter choice in one of his design details, I think ZD5.
The nutshell version is use a 2nd order ACOUSTIC slopes if the drivers will allow. Reasons for going higher order are nasty breakup modes that would not be adequately tamed or tweeter low end power handling.
Zaph is a little at odds with Linkwitz on this, though. Linkwitz points out that with less than a second order high pass electrical filter, a driver's excursion will continue to increase below cutoff even though its output will fall. Therefore, to minimize potential distortions, high pass filters should be at least 2nd order electrical.
An interesting historical note - in his early work (pre-dipole) Linkwitz used a bi-quad high pass in addition to the standard high pass filter. This compensated for the drop in tweeter output below resonance and made the acoustic slope continuous, rather than increasing below tweeter resonance. It appears that he has found that in the case of the current tweeters he uses the slope change is not audible and doesn't bother with the added complexity. It might be something worth investigating, since you'll have a spare bi-quad (Linkwitz transform) section available. Yes, I have been eyeing active speakers since Linkwitz' 1980 Speaker Builder articles.
Although she may limit my speaker designs, overall I prefer to be encumbered. 😉
Boards are ordered, expected ship date 12/06.
I have 32 extra AF4s coming.
4 PSU kits remaining. PSU parts ordered.
I have 32 extra AF4s coming.
4 PSU kits remaining. PSU parts ordered.
Nope, 250 Hz 12 dB is what they say, and it's a highly acclaimed speaker for it's price; the dynaBel S-33. Acoustically, it's said to roll of at about 18 dB.
How tall are those PSUs, without heatsink and with the standard caps? If they fit my 44mm case, I'd like two of them, so I can run dual mono all the way.
Ok, I might prefer having a girl in my life aswell, but I have to get the audio equipment right first, so she can't stop me 😀
How tall are those PSUs, without heatsink and with the standard caps? If they fit my 44mm case, I'd like two of them, so I can run dual mono all the way.
Ok, I might prefer having a girl in my life aswell, but I have to get the audio equipment right first, so she can't stop me 😀
- Status
- Not open for further replies.