Yes of course. See the post above yours for details on the order process (send me PM with info, pay via Paypal).
Cheers, Maarten
Spice Model
Hello all,
Here is a LTSpice model for the following filter sections: summing circuit, 4th order Linkwitz Riley filter, 1 EQ section configured as notch filter, Linkwitz Transform section, Allpass section (phase) and an output buffer with volume control.
View attachment summing+LR4th+Notch+LTransform+allpass+volbuf.asc
-Simulation from 20-100Hz with 100 sample points/decade
-To change the LR4 filter frequency: adjust R4xx (in my current configuration adjustable from 30k to 130k (~113Hz-26Hz)
-I added values for quite an extreme notch as an example (something like center 50Hz with 3dB points on 48Hz and 52Hz and something like a 20dB cut). Use the sheet provided earlier to calculate values of your own. If you want to 'bypass' the notch, just enter a value of 0.1 (something close to zero) for R500
-The values for the current Linkwitz transform are for an Fsc of 33Hz with an F3 of 28Hz and target Fsc of 20Hz.
-Change PhasePOT from 1 to 100000 to change the phase behaviour. In this frequency range 220nF for C700 is quite ideal. You can test some other values.
-The volume control is just 2 modelled resistors. Change both in unison (keep total value the same) to test different volume settings. The current total is 92kOhm because the Alpha pot I tested has a maximum resistance of 92kOhm.
If you're familiar with LTSpice you know what to do. If not. A quick and easy instruction:
-Download & install LTSpice from the Linear website
-Load the schematic
-Press the 'run' button (small running figure)
-Click with your cursor on the output or any other net to measure the frequency and phase response
-Right clicking on the value of a component lets you change it. Beware: after changing component values you have to press 'run' again.
Hello all,
Here is a LTSpice model for the following filter sections: summing circuit, 4th order Linkwitz Riley filter, 1 EQ section configured as notch filter, Linkwitz Transform section, Allpass section (phase) and an output buffer with volume control.
View attachment summing+LR4th+Notch+LTransform+allpass+volbuf.asc
-Simulation from 20-100Hz with 100 sample points/decade
-To change the LR4 filter frequency: adjust R4xx (in my current configuration adjustable from 30k to 130k (~113Hz-26Hz)
-I added values for quite an extreme notch as an example (something like center 50Hz with 3dB points on 48Hz and 52Hz and something like a 20dB cut). Use the sheet provided earlier to calculate values of your own. If you want to 'bypass' the notch, just enter a value of 0.1 (something close to zero) for R500
-The values for the current Linkwitz transform are for an Fsc of 33Hz with an F3 of 28Hz and target Fsc of 20Hz.
-Change PhasePOT from 1 to 100000 to change the phase behaviour. In this frequency range 220nF for C700 is quite ideal. You can test some other values.
-The volume control is just 2 modelled resistors. Change both in unison (keep total value the same) to test different volume settings. The current total is 92kOhm because the Alpha pot I tested has a maximum resistance of 92kOhm.
If you're familiar with LTSpice you know what to do. If not. A quick and easy instruction:
-Download & install LTSpice from the Linear website
-Load the schematic
-Press the 'run' button (small running figure)
-Click with your cursor on the output or any other net to measure the frequency and phase response
-Right clicking on the value of a component lets you change it. Beware: after changing component values you have to press 'run' again.
Update on payments
Several people have payed their order now. In several cases Paypal charged fees to me so I come up short a few euro's. So I wanted to give an instruction, before I end up being short more than a few euro's:
-If your Paypal payment is done via your creditcard and/or you mark it as 'purchase payment', please add 3,4% + € 0,35 to the total amount.
-If you Paypal payment is done via your bank account or balance already present on you Paypal account AND you mark you payment as 'personal', you don't have to add anything to the total amount.
And the current status of the groupbuy:
Total number of persons that expressed interest on this thread or via PM: 27
Maximum total number of boards current interest: 57
Total number of persons that send me a PM with their address and final board number: 12
Total number of persons that payed their order: 7
I'll give another update after the weekend.
Cheers Maarten
Several people have payed their order now. In several cases Paypal charged fees to me so I come up short a few euro's. So I wanted to give an instruction, before I end up being short more than a few euro's:
-If your Paypal payment is done via your creditcard and/or you mark it as 'purchase payment', please add 3,4% + € 0,35 to the total amount.
-If you Paypal payment is done via your bank account or balance already present on you Paypal account AND you mark you payment as 'personal', you don't have to add anything to the total amount.
And the current status of the groupbuy:
Total number of persons that expressed interest on this thread or via PM: 27
Maximum total number of boards current interest: 57
Total number of persons that send me a PM with their address and final board number: 12
Total number of persons that payed their order: 7
I'll give another update after the weekend.
Cheers Maarten
I promised an update. So here we go:
Total expressed interest: 62 boards
Ordered and paid for: 32 boards
Next weekend (or maybe coming friday) I will post the updated schematic, BOM and board pictures. After consulting I decided on ground pours for both layers and adding some extra via's. Also after testing the sensing circuit I came to the conclusion this type of circuit needs a high input resistance (gate / base) on the switching transistor to charge the capacitor. So it needs a mosfet and not something like a BC547. These are the changes implemented in the update (and of course the beforementioned separate ground plane for the sensing circuit).
Hopefully many of you are still planning to place a final order, as the board price was based on 50 boards. So a minimum of 18 to go
.
Total expressed interest: 62 boards
Ordered and paid for: 32 boards
Next weekend (or maybe coming friday) I will post the updated schematic, BOM and board pictures. After consulting I decided on ground pours for both layers and adding some extra via's. Also after testing the sensing circuit I came to the conclusion this type of circuit needs a high input resistance (gate / base) on the switching transistor to charge the capacitor. So it needs a mosfet and not something like a BC547. These are the changes implemented in the update (and of course the beforementioned separate ground plane for the sensing circuit).
Hopefully many of you are still planning to place a final order, as the board price was based on 50 boards. So a minimum of 18 to go
.Hi all,
By now I have received orders and payment for 42 boards. I'll send the 11 people who expressed interest a PM to ask if they are still interested to participate. Hopefully orders get up to 50 pcs. to cover all costs. If they don't respond within a day or 2-3, I'll go ahead and place the order.
Here's the latest (and probably final) schematic and board:
Schematic:
Board top layer:
Board bottom layer:
And the updated BOM:
View attachment SubwooferFilter_BOM_Mouser.zip
The blue grey colored cells are components which you have to calculate or choose according to your personal preference. Please read the remarks carefully.
Here's the LTSpice model again, now including input buffers (input buffers are +6dB for better noise performance):
View attachment buffer+summing+LR4th+Notch+LTransform+allpass+volbuf.asc
Update log:
-Ground pours in both layers
-Many ground via's for low impedance coupling of both ground layers
-Separate ground for sensing circuit (ties to ground via jumper or resistor (R911)
-Larger mounting pads (so you can actually fit your screw in there without touching any components)
-Replaced the phase switch with a smaller form factor (is off board anyway and the previous mounting holes were a bit large)
-Replaced all <90 degrees trace connections with ones that are 90 degrees or larger
About grounding:
-The ground of the sensing circuit could possibly induce ground hum issues. First try to jumper R911. If this results in ground hum, you can try to tie the sensing side of R911 to a common power supply or off board star ground. Another option is to put a small valued resistor here (10R).
-Depending on your setup and the equipment you connect the board to, you still can encounter other ground hum problems. This could for example be due to ground connections via both the RCA LFE input and the speaker or L+R RCA inputs (2 ground connections to the same component or a pre-amp connected to the filter and to a power amp, which is in turn again connected to the filter). If you encounter such issues, please post it in this or a separate forum thread. There are always solutions, but these are of course situation specific..
About connecting it to your equipment:
The layout is made for the following 2 combinations:
-LR speaker inputs + LFE RCA input
-LR RCA inputs + LFE RCA input
So do not connect both LR RCA inputs and the LR speaker inputs at the same time!
Errors:
Although I checked and rechecked the schematic and board about a 100 times for errors and modeled the schematic in Spice, I am still just 1 person doing this as a hobby (and I am not an electronics engineer..). If anybody with some experience is willing to recheck my schematic, please do so. I am fairly certain it will work as designed but some extra error checking never hurts.
Oh almost forgot: HAPPY EASTER EVERYBODY!
By now I have received orders and payment for 42 boards. I'll send the 11 people who expressed interest a PM to ask if they are still interested to participate. Hopefully orders get up to 50 pcs. to cover all costs. If they don't respond within a day or 2-3, I'll go ahead and place the order.
Here's the latest (and probably final) schematic and board:
Schematic:
Board top layer:
Board bottom layer:
And the updated BOM:
View attachment SubwooferFilter_BOM_Mouser.zip
The blue grey colored cells are components which you have to calculate or choose according to your personal preference. Please read the remarks carefully.
Here's the LTSpice model again, now including input buffers (input buffers are +6dB for better noise performance):
View attachment buffer+summing+LR4th+Notch+LTransform+allpass+volbuf.asc
Update log:
-Ground pours in both layers
-Many ground via's for low impedance coupling of both ground layers
-Separate ground for sensing circuit (ties to ground via jumper or resistor (R911)
-Larger mounting pads (so you can actually fit your screw in there without touching any components)
-Replaced the phase switch with a smaller form factor (is off board anyway and the previous mounting holes were a bit large)
-Replaced all <90 degrees trace connections with ones that are 90 degrees or larger
About grounding:
-The ground of the sensing circuit could possibly induce ground hum issues. First try to jumper R911. If this results in ground hum, you can try to tie the sensing side of R911 to a common power supply or off board star ground. Another option is to put a small valued resistor here (10R).
-Depending on your setup and the equipment you connect the board to, you still can encounter other ground hum problems. This could for example be due to ground connections via both the RCA LFE input and the speaker or L+R RCA inputs (2 ground connections to the same component or a pre-amp connected to the filter and to a power amp, which is in turn again connected to the filter). If you encounter such issues, please post it in this or a separate forum thread. There are always solutions, but these are of course situation specific..
About connecting it to your equipment:
The layout is made for the following 2 combinations:
-LR speaker inputs + LFE RCA input
-LR RCA inputs + LFE RCA input
So do not connect both LR RCA inputs and the LR speaker inputs at the same time!
Errors:
Although I checked and rechecked the schematic and board about a 100 times for errors and modeled the schematic in Spice, I am still just 1 person doing this as a hobby (and I am not an electronics engineer..). If anybody with some experience is willing to recheck my schematic, please do so. I am fairly certain it will work as designed but some extra error checking never hurts.
Oh almost forgot: HAPPY EASTER EVERYBODY!
Last edited:
Are the potentiometers connected the wrong way input/output or is it just me?
I think it's just you
(except for the phase pot, which is debatable) .Some were connected the wrong way but in my latest post they are now connected correctly I believe, bear with me.:
In Eagle the board view is from the top (component side):
-The Linkwitz lowpass filter (left pot on the PCB): when the pot is in the left most position ('0'), resistance between the middle and the right leg (as seen in the PCB pictures, looking from the front side) is highest. In this case total resistance (including the 30k resistors) is 130kOhm. With a C of 33nF this results in X-over frequency of ~26Hz. When turning the pot to the right, resistance decreases to 30kOhm (only the resistance of the resistor, pot = 0 Ohm). With a resistance of 30kOhm, the X-over frequency is 113Hz. So turning the pot clockwise results in an increasing X-over frequency.
-The volume pot works as a standard voltage divider: the left leg is connected to ground, the middle leg is signal out (going to the + input of the output buffer, pin3 of IC9), the right leg is signal in (coming from rumble filter pin 7 of IC9). In the left most position, resistance between the left leg (ground) and the middle leg (out) is 0 Ohm, and the resistance between the middle leg (out) and the right leg is 100kOhm. This results in maximum attenuation. When turning the pot fully to the right, resistance between input (right leg) and output (middle leg) decreases to 0 and resistance between output and ground is at it's maximum. So turning the knob clockwise results in an increase in volume.
-Finally the phase: middle pin and right pin are connected. So turned to the left resistance is 0 Ohm. Let's take an arbitrary crossover point of 40Hz. When modeling the Phase in LTSpice this result in 80 degrees at the crossover point. To the right resistance is 100kOhm resulting in a phase of -80 degrees at the crossover point (range of 160 degrees). When switching the phase switch, pot in the left position results in -100 degrees at 40Hz turning it fully to the right: -260 degrees. So you have a total range of 320 degrees (full 180 is already sufficient). But for phase, it doesn't really matter where you start and how the pot turns. You could argue that with a centered pot, turning to the left decreases the phase and to the right increases it. Personal preference I think. Anybody suggestions / preference / 'correct way'?
Please check if my reasoning for the low pass and volume is correct.
Little update on the groupbuy process itself:
-Received payment for 56 boards, so costs are covered.
-Awaiting payment for another 8 boards (3 people).
-Awaiting response from 2 people (gave them till Thursday to respond).
Later this week (but before I send out the order), I'll do a final check on component placement (print out the silkscreen and fit all the actual components on the print).
Keep you updated.
-Received payment for 56 boards, so costs are covered.
-Awaiting payment for another 8 boards (3 people).
-Awaiting response from 2 people (gave them till Thursday to respond).
Later this week (but before I send out the order), I'll do a final check on component placement (print out the silkscreen and fit all the actual components on the print).
Keep you updated.
Update:
Received payment for 66 boards.
Awaiting response from 1 person.
The order will go out on friday/saturday morning.
I printed the silkscreen and started fitting some components. Made 2 changes to the design as a result of this:
-Moved C602 just a fraction further down (was just touching the mounting screw when using a capacitor with the largest footprint)
-Made as much room as the design would allow between the volume and phase pot
The rest of the components fit nicely.
Received payment for 66 boards.
Awaiting response from 1 person.
The order will go out on friday/saturday morning.
I printed the silkscreen and started fitting some components. Made 2 changes to the design as a result of this:
-Moved C602 just a fraction further down (was just touching the mounting screw when using a capacitor with the largest footprint)
-Made as much room as the design would allow between the volume and phase pot
The rest of the components fit nicely.
Order has been placed!
Hello all,
The order has been placed at PCBcart. Lead time is 12 working days. So including shipping it probably will take a little over three weeks before they are delivered to me. I will get them out to you asap.
Here's the final silk screen:
As you can see I added some more options:
-Pads for optional jumper connectors for input and output (some of you maybe prefer jumpers over screw terminals
-Options for 3 types of capacitor size for all capacitors with (possibly) large values (C600, C602 (Linkwitz transform) and C700 (Phase control). You can fit capacitors in these positions with 5, 10 and 15 mm lead spacing. You now can use EVOX PHE's of different sizes, WIMA MKP20's, WIMA MKP4's and many others types in these positions.
-The little plus (+) symbols adjacent to the jumpers and screw terminals indicate the connector side for the positive input for both the jumper as well as the screw terminal.
-Indicators for 12dB and 24dB/octave side for the S1 connector.
Hello all,
The order has been placed at PCBcart. Lead time is 12 working days. So including shipping it probably will take a little over three weeks before they are delivered to me. I will get them out to you asap.
Here's the final silk screen:
As you can see I added some more options:
-Pads for optional jumper connectors for input and output (some of you maybe prefer jumpers over screw terminals
-Options for 3 types of capacitor size for all capacitors with (possibly) large values (C600, C602 (Linkwitz transform) and C700 (Phase control). You can fit capacitors in these positions with 5, 10 and 15 mm lead spacing. You now can use EVOX PHE's of different sizes, WIMA MKP20's, WIMA MKP4's and many others types in these positions.
-The little plus (+) symbols adjacent to the jumpers and screw terminals indicate the connector side for the positive input for both the jumper as well as the screw terminal.
-Indicators for 12dB and 24dB/octave side for the S1 connector.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.