Hey guys,
Just wondering if there is any interest out there for a group buy on some Linkwitz-Riley (LR) crossover PCB's. The circuit itself is very easy to design (straight from his site) but getting PCB's is another story.
We can get PCB's made from any one of the many fabricators (such as PCB Express) but there is a minimum quantiy required if you want to get good quality boards.
I've pretty much got the PCB layout finished, and I will post pictures of it for everyone to verify, etc.
There will be 2 boards essentially: one highpass and one lowpass. Both are 24db/oct (cascaded 12db/oct butterworth). Really, you could use any allignment that you want but I believe most would go with the LR allignment.
If I get enough interest I will start pricing the boads, but I believe they should be around $8 or $10 each. It really depends on how many people are in on the group buy. Prices may be cheaper if you guys know of a cheaper place (these prices were through PCB Express).
If there is interest in doing power supplies let's do some of those too. I already have a design based on the LM340/320 but we could always design a board around another part if prefered.
Anyway .. post a message if you think you'd be interested in the group buy(s). I'm just seeing how many of you would take part before we move forward with anything.
Thanks,
-Brian Cleven
Just wondering if there is any interest out there for a group buy on some Linkwitz-Riley (LR) crossover PCB's. The circuit itself is very easy to design (straight from his site) but getting PCB's is another story.
We can get PCB's made from any one of the many fabricators (such as PCB Express) but there is a minimum quantiy required if you want to get good quality boards.
I've pretty much got the PCB layout finished, and I will post pictures of it for everyone to verify, etc.
There will be 2 boards essentially: one highpass and one lowpass. Both are 24db/oct (cascaded 12db/oct butterworth). Really, you could use any allignment that you want but I believe most would go with the LR allignment.
If I get enough interest I will start pricing the boads, but I believe they should be around $8 or $10 each. It really depends on how many people are in on the group buy. Prices may be cheaper if you guys know of a cheaper place (these prices were through PCB Express).
If there is interest in doing power supplies let's do some of those too. I already have a design based on the LM340/320 but we could always design a board around another part if prefered.
Anyway .. post a message if you think you'd be interested in the group buy(s). I'm just seeing how many of you would take part before we move forward with anything.
Thanks,
-Brian Cleven
i've been working on a similar project for some time. i may be making a test unit tomorrow. Right now it is a 24dB/oct design with high, low on one board. standard dual opamp per 2-stage filter. Milled PCB, coated copper (some shiny metal...). Was looking at doing the following:
single amplifer biquad
gain setting resistor option
resisitor (CCS) option.
local adjustable regulator (lm317/337 pinouts)
locations for 100nf capacitors in closest location to power pins.
quasi-ground plane (milling cannot remove large areas of metal, leaving large unconnected metal surfaces. these can be used as a pseudo ground plane.)
customization -- i can mill the initials of whomever onto the top of the PCB.
i am currently looking into how cutting top traces as well as bottom (i can do 2 layer as easy as 1 layer) can reduce input and output capacitance, but at the further damage to the ground plane. also looking into the jfet CCS mod, but likely will not be in rev A, as it won't be tested with suitable JFETs.
if there is interest, i may redesign the board a bit to fit larger, audiophile capacitors and resistors. as well as an IC socket that was left out of my initial design.
a note to the op, I am not trying to take your buisness. i will not be able to make a large quantity of these PCBs, if i do make a run, i plan to charge:
$20 for the dual (high, low) board
$25 for a customized board.
+$10 for extra boards
+$15 for extra customized boards.
i will donate $5 + $2.50/extra board to the forums.
these boards take a fair amount of time to mill out. I will make a design tonight and hopefully a test board tomorrow.
Please note, that because of the time required per PCB, i may not be able to take all orders. I am a college student, not a buisness!
single amplifer biquad
gain setting resistor option
resisitor (CCS) option.
local adjustable regulator (lm317/337 pinouts)
locations for 100nf capacitors in closest location to power pins.
quasi-ground plane (milling cannot remove large areas of metal, leaving large unconnected metal surfaces. these can be used as a pseudo ground plane.)
customization -- i can mill the initials of whomever onto the top of the PCB.
i am currently looking into how cutting top traces as well as bottom (i can do 2 layer as easy as 1 layer) can reduce input and output capacitance, but at the further damage to the ground plane. also looking into the jfet CCS mod, but likely will not be in rev A, as it won't be tested with suitable JFETs.
if there is interest, i may redesign the board a bit to fit larger, audiophile capacitors and resistors. as well as an IC socket that was left out of my initial design.
a note to the op, I am not trying to take your buisness. i will not be able to make a large quantity of these PCBs, if i do make a run, i plan to charge:
$20 for the dual (high, low) board
$25 for a customized board.
+$10 for extra boards
+$15 for extra customized boards.
i will donate $5 + $2.50/extra board to the forums.
these boards take a fair amount of time to mill out. I will make a design tonight and hopefully a test board tomorrow.
Please note, that because of the time required per PCB, i may not be able to take all orders. I am a college student, not a buisness!
You can get a great LR 24db/octave stereo crossover complete with regulated power supply here. These are proven, and sound very good, and are extremely quiet. Bare boards as well as complete kits are available.
https://secure11.websitecomplete.com/sharrongrp/shop/showDept.asp?dept=528
https://secure11.websitecomplete.com/sharrongrp/shop/showDept.asp?dept=528
Hi hornlover,
Those PCBs that you recommend. RUBBISH. The output impedance of HF channel is 25k @ -6db. All the filters are based on unity gain blocks leaving no opton to change damping or gain. Inflexible and what about his suggestion to use 4VA transformer as the PSU. Does he cater for the Fidelity market or simple Disco music via a PA system?
Those PCBs that you recommend. RUBBISH. The output impedance of HF channel is 25k @ -6db. All the filters are based on unity gain blocks leaving no opton to change damping or gain. Inflexible and what about his suggestion to use 4VA transformer as the PSU. Does he cater for the Fidelity market or simple Disco music via a PA system?
I'm designing a 2way 24db L-R XO, using ESP's design and OPA2134s. Too much stuff right now, so I've been going slow. They will be similar to ESP's so that you can make the XO any frquency you want. I'm going to try to make each board have the 2 way XO on it. The RCA will be built into the PCB so that it can mount to the outside of the chasis if liked. The two boards will be on top of eachother. So from looking at it from the back the input will be on the right, then on the left will be the low pass, then another inch over will be high pass and the right channel will be above the left.
If anyone is interested in PCBs I'd be willing to share.
Thanks,
Josh
If anyone is interested in PCBs I'd be willing to share.
Thanks,
Josh
Hi,
I would like a -24db/oct(4pole) bandpass board that is capable of adjustment to butterworth via EVCsallen-key and MFB filter options.
Unity gain is too inflexible.
Reasoning is to allow a bandpass to be inserted into each power amp case and selection of butterworth or LR at time of stuffing boards. MFB to allow gain selection to avoid +4.002db that ECVs-k gives when damping is .541 (if memory is correct).
If we go this route I would be looking for about 18 boards.
I would like a -24db/oct(4pole) bandpass board that is capable of adjustment to butterworth via EVCsallen-key and MFB filter options.
Unity gain is too inflexible.
Reasoning is to allow a bandpass to be inserted into each power amp case and selection of butterworth or LR at time of stuffing boards. MFB to allow gain selection to avoid +4.002db that ECVs-k gives when damping is .541 (if memory is correct).
If we go this route I would be looking for about 18 boards.
I made my own boards in a modular form using unity gain. You can change the Q very easily just by changing the ratio of R1:R2 or C1:C2. I made space for 3 components for both Rs and Cs so that I can parallel them to get the desired values. If this person designs the boards in a similar way, as you are choosing a fixed Q at time of stuffing you should be fine and able to use his boards.
I would like to thank AndrewT for his insightful review of our crossover. However, I would like to take this opportunity to add a few comments. The crossover board we offer is a basic two way stereo crossover withan on-board regulated power supply. It wasn't designed to be everything to everybody, but a good, basic, high fidelity crossover at a bargain price. We are proud of the quality of the board and parts we offer, using 2% polypropylene capacitors, and MIL spec Vishay/Dale 1% metal film resistors, and believe you will not find equal quality for anywhere near the price anywhere else. We could have used cheaper parts, and sound quality would probably still be excellent, but we wanted to offer a kit where people didn't feel the need to throw out half of the parts, and substitute better ones. In fact, on the full kit, we actually loose money after taking labor into consideration. You may wonder why we do this, but we do it to encourage bi-amping, which we feel is absolutely necessary for a true hi fidelity system.
It is designed as unity gain, which is fine for most applications. The bandwidth of the hi pass section is -3db at about 900kHz, more than enough for the most demanding applications. The op-amps we supply are not the latest flavor, but are excellent performers, none the less. The 553x series of op-amps are designed with audio in mind, and offer very low noise, plenty of slew rate, and are stable. We provide sockets in the kit in case anybody feels the need to experiment with other pin compatible chips. We try new chips from time to time, but until we can notice a real improvement in sound quality, we will stay with the NE5532's.
The bare board is available, for those keen on mixing and matching their favorite parts.
While the kits we offer have only 3 crossover frequencies available, any frequency can be selected by substituting the proper resistor values. The documentation we provide gives all the information needed to calculate the necessary values.
As for the transformer, although small, it still provides about 3 times the current requirements needed by the board. We could have specified a larger one, but that would just add cost with no sonic benifits. You could, of course, use a larger one, if you feel the need.
We have received many positive comments from customers about the quality and ease of construction of the crossover, and the extremely low noise levels. We feel this is very important when using systems with very high efficiencies.
I would be glad to provide AndrewT with a sample of the board, if he wishes to base his review on actual use.
A fellow music lover
Ron Shaw
H.E.L.
It is designed as unity gain, which is fine for most applications. The bandwidth of the hi pass section is -3db at about 900kHz, more than enough for the most demanding applications. The op-amps we supply are not the latest flavor, but are excellent performers, none the less. The 553x series of op-amps are designed with audio in mind, and offer very low noise, plenty of slew rate, and are stable. We provide sockets in the kit in case anybody feels the need to experiment with other pin compatible chips. We try new chips from time to time, but until we can notice a real improvement in sound quality, we will stay with the NE5532's.
The bare board is available, for those keen on mixing and matching their favorite parts.
While the kits we offer have only 3 crossover frequencies available, any frequency can be selected by substituting the proper resistor values. The documentation we provide gives all the information needed to calculate the necessary values.
As for the transformer, although small, it still provides about 3 times the current requirements needed by the board. We could have specified a larger one, but that would just add cost with no sonic benifits. You could, of course, use a larger one, if you feel the need.
We have received many positive comments from customers about the quality and ease of construction of the crossover, and the extremely low noise levels. We feel this is very important when using systems with very high efficiencies.
I would be glad to provide AndrewT with a sample of the board, if he wishes to base his review on actual use.
A fellow music lover
Ron Shaw
H.E.L.
Hi Ron Shaw,
Thanks for that comprehensive reply particularly the ref to component quality.
My language was a bit strong but I stand by my strongly held view and the comment. Inflexible. Sure glad I went on to explain my reasoning.
Re post 15, Richie00boy the changes to cap ratios are rarely documented (can you refer me to a site that provides a calculation method) to achieve desired Q. Once these odd ratios are found you then have to // expensive tight tolerance caps to achieve very odd values. Then what happens if you want a new frequency? Oh just put all the resistors into sockets so they can be changed instead, I think not!
Thanks for that comprehensive reply particularly the ref to component quality.
My language was a bit strong but I stand by my strongly held view and the comment. Inflexible. Sure glad I went on to explain my reasoning.
Re post 15, Richie00boy the changes to cap ratios are rarely documented (can you refer me to a site that provides a calculation method) to achieve desired Q. Once these odd ratios are found you then have to // expensive tight tolerance caps to achieve very odd values. Then what happens if you want a new frequency? Oh just put all the resistors into sockets so they can be changed instead, I think not!
Having finally seen the schematic for Ron Shaw's board I have to say I can't really fault it with the exception of the 100k pot to attenuate HF, it needs a buffer after it. My only other gripe is that the capacitor symbols used are what I know as polarised electrolytics, which is obviously not what would be used given the small values, but still...
AndrewT, I don't have the formulas to hand, but I can offer you a very simple C++ program I wrote a long time ago to do the calculations for you. Check out the High Pass Filter Designer and Low Pass Filter Designer programs on my utilities page.
I do not have a problem with paralleling components and use standard 5% tolerance caps with no problems, and always arrive very close to the value I want. I will add that when I design an active crossover I tend to get the design pretty close first time (by spending time on the measurement and design) so the need to change frequency is minimal
AndrewT, I don't have the formulas to hand, but I can offer you a very simple C++ program I wrote a long time ago to do the calculations for you. Check out the High Pass Filter Designer and Low Pass Filter Designer programs on my utilities page.
I do not have a problem with paralleling components and use standard 5% tolerance caps with no problems, and always arrive very close to the value I want. I will add that when I design an active crossover I tend to get the design pretty close first time (by spending time on the measurement and design) so the need to change frequency is minimal
We decided not to add another buffer section to the output in order to minimize stages, since its a good possibility that the following stage will be buffered anyway. We have not had anyone have an issue with the input impedence of the stage following the hi-pass section, so so far, we will leave it like it is.
The cap sysmbols are IEC recognized, as are the symbols using straight lines for both sides. Polarized caps are shown with a (+) marking the positive plate. We dont use the symbol using both straight lines because that symbol is also an IEC accepted symbol for contacts, such as you may find in a circuit breaker. Sorry for any confusion this may have caused.
Ron Shaw
H.E.L.
The cap sysmbols are IEC recognized, as are the symbols using straight lines for both sides. Polarized caps are shown with a (+) marking the positive plate. We dont use the symbol using both straight lines because that symbol is also an IEC accepted symbol for contacts, such as you may find in a circuit breaker. Sorry for any confusion this may have caused.
Ron Shaw
H.E.L.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.