My experience with building completely on my own has been ups and downs with initially lack of complete understanding of various technical terms and the capability to have them measured. After many tweaks, trial and errors it was still difficult to know whether it sounded how it was suppose to sound. With the discovery of simulators and measuring software it again become complicated as it was needed to understand the holistic view of these and implement the same. In my current TL FS speakers, I would have done over 25 simulations and have implemented close to 11 of them in the cross-over to get them right (IMO) and have used reference speakers to compare its performance. I have yet to meausure them for their linearity.
The challenges i have faced when implementing the crossovers were,
1. Component tolerance:- I have measured even good polyester capacitors having tolerance of over 20% of their printed value. I have meausred 4.7
microfarad to have capacitance only 3.7 micorfarad. So you can imagine how deep will be the phase shift. Same for resistors having atleast 5% to 15% of tolerance - again cause level matching issues, impedance and phase issues. Inductors when wound will have to be measured as these can be source for further phase shift.
2. Getting the filter phase accurate for both the low pass and high pass filter. Since i mostly only use first order filters, the phase shift between them is seldom 90 deg. I am still struggling to get the high pass filter phase shift exactly to 90 deg.
3. Driver published and actual measured parameters: This challenge cannot be solved easily as if you work with the driver published parameter there are 90% chances that your cross-over will be incorrect. I have noticed atleast 25% to 30% difference in FR curve for the published and measured parameter. Free-air impedance and FR curve will be different once you mount the drivers in the box and will display different behavior all-together to the cross-over again further adding to the phase shift.
4.Getting the correct box volume and tuning, damping, stiffness for the box is another challenge which needs to be addressed which other wise can cause panel resonance, high acoustic impedance.
My current speaker took me over 6 months to get them to the current level when i began designing and yet it requires measurement and endorsement from other listeners to compliment its performance. I am sure most of the budget speakers and the so call expensive speakers will also not follow such stringent design cycle, iterations, tweaking, simulations and measurments due to cost reasons. This bring back the question whether to DIY or Not to DIY. IMO if you have the resources (read information) to do all these painful exercise and still want to enjoy doing all of this then do DIY. But never DIY if you think you can build a published design and get the same performance due to the factors mentioned above. The first basic investment i would say if you DIY is to buy a good quality LCR meter to measure the component values in absence of which it will be like firing in the dark. Most of the simulators and measurement software are freewares and can be downloaded. You still will require to invest in measurement microphones and good quality sound card. BTW, these investments could be quite high than buying a ready made speaker over the shelf.
Some one had said earlier - "The best speakers are the one which are never bought but are the ones that are made."
Will like other FM to share their DIY journey.
Cheers,
The challenges i have faced when implementing the crossovers were,
1. Component tolerance:- I have measured even good polyester capacitors having tolerance of over 20% of their printed value. I have meausred 4.7
microfarad to have capacitance only 3.7 micorfarad. So you can imagine how deep will be the phase shift. Same for resistors having atleast 5% to 15% of tolerance - again cause level matching issues, impedance and phase issues. Inductors when wound will have to be measured as these can be source for further phase shift.
2. Getting the filter phase accurate for both the low pass and high pass filter. Since i mostly only use first order filters, the phase shift between them is seldom 90 deg. I am still struggling to get the high pass filter phase shift exactly to 90 deg.
3. Driver published and actual measured parameters: This challenge cannot be solved easily as if you work with the driver published parameter there are 90% chances that your cross-over will be incorrect. I have noticed atleast 25% to 30% difference in FR curve for the published and measured parameter. Free-air impedance and FR curve will be different once you mount the drivers in the box and will display different behavior all-together to the cross-over again further adding to the phase shift.
4.Getting the correct box volume and tuning, damping, stiffness for the box is another challenge which needs to be addressed which other wise can cause panel resonance, high acoustic impedance.
My current speaker took me over 6 months to get them to the current level when i began designing and yet it requires measurement and endorsement from other listeners to compliment its performance. I am sure most of the budget speakers and the so call expensive speakers will also not follow such stringent design cycle, iterations, tweaking, simulations and measurments due to cost reasons. This bring back the question whether to DIY or Not to DIY. IMO if you have the resources (read information) to do all these painful exercise and still want to enjoy doing all of this then do DIY. But never DIY if you think you can build a published design and get the same performance due to the factors mentioned above. The first basic investment i would say if you DIY is to buy a good quality LCR meter to measure the component values in absence of which it will be like firing in the dark. Most of the simulators and measurement software are freewares and can be downloaded. You still will require to invest in measurement microphones and good quality sound card. BTW, these investments could be quite high than buying a ready made speaker over the shelf.
Some one had said earlier - "The best speakers are the one which are never bought but are the ones that are made."
Will like other FM to share their DIY journey.
Cheers,
Last edited:
