The mods I bought were $60. The most expensive part of that, by far, were the larger inductors for the woofers. The rest of the components are much cheaper. I think you could do the 2.1se crossover for less than $100, retail, with shipping.
But then it wouldn’t be completely flat in response. With the DSP, you could theoretically dial it in much tighter.
The zoudio amp has EVERYTHING built into it. 1 stop shopping. 50wx4
You can also change everything on the fly without re soldering anything, and make WAY finer adjustments much easier. I really find it difficult to understand the allure of passive crossover networks. Is it the satisfaction of creating something that really keeps them popular? Or is it the simplicity of only having to use 2 amp channels
Well, I have an amp or 2, and if I wanted to use a receiver, It’s not like the DSP amp would replace that - I’d need preouts, so Id need a more expensive receiver with an amp I couldn’t use. It’s just more convenient to have a device you can just plug into existing equipment.
However, I like the idea of using a DSP, now that the costs are coming down. Maybe one day I can figure out how to make that fit into my system. Maybe using an analog speaker-level signal converter to make a line-level input. That’ll be high quality.
AllenB, That's 72 possible components that you would have to pay for if going passive. Where are you buying components?
I have seen schematics of analog crossovers with time alignment. I’m not sure how necessary it is, but yeah, it would be complicated and costly.
Hi, here my learning process update, which by the way don't know the names of what i did. So what i learn is:
1. the yellow thing --> inductor + cap in series with a resitor in parallel will kill the pick to the left of the high pass filter.
2. the blue --> cap controls the high pas filter and controls the pick at this point
3. the pink --> cap in parallel at the end controls (lowers) the high frequencies (15k+) picks
here for comparison 2.1se and the one i just said. sorry for my lack of terminology
1. the yellow thing --> inductor + cap in series with a resitor in parallel will kill the pick to the left of the high pass filter.
2. the blue --> cap controls the high pas filter and controls the pick at this point
3. the pink --> cap in parallel at the end controls (lowers) the high frequencies (15k+) picks
here for comparison 2.1se and the one i just said. sorry for my lack of terminology
Last edited:
I've not seen the 'yellow' done before as you have. Normally, the rise towards Fs is treated by an LCR across the tweeter's terminals. Not that what you did does not work.
Pink- I would not leave that alone without a series resistor on it, as that will be shunting to a very low impedance in that range.
Wolf
Pink- I would not leave that alone without a series resistor on it, as that will be shunting to a very low impedance in that range.
Wolf
I have not seen that before. How come people don't use it more???
and no the (possibly? excessive) number of filter adjustment slots is just a finite value for filters and EQ that you can use.....for no extra cost to the user
I have spent MANY hours dialing in every inch of FR across the board within half a DB of a straight line, only to find it really doesn't make the perfect sound I thought it would. It seems that perfect sound I long for, I do not quite understand. Exactly what should the perfect measurement should look like ??
Last edited:
DSP has mysteriously had an impact on the kind of information that is generally seen as important.
Clearly you know that electronics can't fix a bad speaker.. that there's no point just taking a driver response, flattening it and applying filters.. that if someone gives us a pair of response measurements without giving us any more information about the speakers then it has little meaning. What I think you're looking for to change the underlying character of the speaker, is the acoustic design.
Clearly you know that electronics can't fix a bad speaker.. that there's no point just taking a driver response, flattening it and applying filters.. that if someone gives us a pair of response measurements without giving us any more information about the speakers then it has little meaning. What I think you're looking for to change the underlying character of the speaker, is the acoustic design.
I looked up the Zoudio, and I am not too excited about it. Seems low-power, needs a lot of DIY. Would be interesting to use in an auto setup, to get proper time alignment. Maybe good for a desktop system? Seems like active should have a lot of advantages, for sure.
As for it not sounding great when flattening the curve, maybe people prefer to listen to a non-flat freq. response.
As for it not sounding great when flattening the curve, maybe people prefer to listen to a non-flat freq. response.
Can you elaborate on the DSP??
What other information helps?
what else should I be looking at ?
Last edited:
My 2 cents, and this is my opinion as a newbie.
Perfection on a speaker is like looking thru a prism there are several opinions, so in theory the main slopes are:
1. Flat --> speaker will reproduce the sound that is given.
2. down slope or warm --> speaker will color the sound that is given on the lower frequencies.
3. hot or bright --> speaker will color the sound on to top end.
4. smiley or v or u shape --> speaker will have boost on top and bottom ( like BOSE originally)
Different people like different slopes. side note this is why BOSE is so popular to the regular listener.
The advantage of a flat speaker is that is easy to tune with the amp, dac, preamp, dsp, eq or what ever you pot on the signal path. this is why a flat response is typically the goal.
So if a designer is building a crossover or active crossover and already knows the slope he likes, his perfection will be the slope he likes so that he doesn't have to tweak the system to his liking.
In other hand a commercially designed speaker will carry the brand's slope signature.
DSP Digital Signal Processing --> What is DSP? Why do you need it? - YouTube
Perfection on a speaker is like looking thru a prism there are several opinions, so in theory the main slopes are:
1. Flat --> speaker will reproduce the sound that is given.
2. down slope or warm --> speaker will color the sound that is given on the lower frequencies.
3. hot or bright --> speaker will color the sound on to top end.
4. smiley or v or u shape --> speaker will have boost on top and bottom ( like BOSE originally)
Different people like different slopes. side note this is why BOSE is so popular to the regular listener.
The advantage of a flat speaker is that is easy to tune with the amp, dac, preamp, dsp, eq or what ever you pot on the signal path. this is why a flat response is typically the goal.
So if a designer is building a crossover or active crossover and already knows the slope he likes, his perfection will be the slope he likes so that he doesn't have to tweak the system to his liking.
In other hand a commercially designed speaker will carry the brand's slope signature.
DSP Digital Signal Processing --> What is DSP? Why do you need it? - YouTube
Great work.
The latest frequency response of that modified SE cross over is fantastic. Do we have any way to drop the price of all the components on it? The HIVI Swan 2.2A tweeter is REALLY good, and it would be great to get it under a bit of control. I have KEF LS50s, and I prefer the HIVI Swans for female vocals and most music.
The latest frequency response of that modified SE cross over is fantastic. Do we have any way to drop the price of all the components on it? The HIVI Swan 2.2A tweeter is REALLY good, and it would be great to get it under a bit of control. I have KEF LS50s, and I prefer the HIVI Swans for female vocals and most music.
Hi michaelcpu. I have was about to write up a follow up on my about the mods i have was simulating. i have come to a crossroad. i started working with the FRD files on this thread which are the measurements made by Alex with the drivers mounted on the cabinet. recently i went back and forward with the tweeter hence i found the lump and dip in the 2000 and 2700 hz threw me off hence i did not see this in the spec sheet's response graph of the manufacturer.
so i started learning to extract the graph (what a head ache). when applied this frd files to the crossover this was a hole different ball park.
this lead me to look at the b6.8b frd file from parts express and OMG rad change.... so extracted the graph from the spec sheet. Now i have 3 different FRD for the same driver all of the really different from one to the other. further more found a forth frd for the woofer from another driver reviewer with totally different result. the main difference on the woffer's files is at the (forgot the proper terminology) breakout/distortion point around the 1000-2700 and placed at different frequency in each different frd files.
at the end, i sent for my version 3.9uF .33mh on tweeter and just eliminate the woofers resistor until i buy a calibrated mic and generate my own frd files and build a crossover based on my drivers and box assembly.
Or if anybudy can direct me to the wright path.
all in all it's been a great learning experience to deal with all of this, loved it.
ps still haven't heard any of the versions hence just yesterday received the kit for the third time, this time it seems to have all the right parts and no damage on any parts. this is until i can test the drivers, but that is another story
so i started learning to extract the graph (what a head ache). when applied this frd files to the crossover this was a hole different ball park.
this lead me to look at the b6.8b frd file from parts express and OMG rad change.... so extracted the graph from the spec sheet. Now i have 3 different FRD for the same driver all of the really different from one to the other. further more found a forth frd for the woofer from another driver reviewer with totally different result. the main difference on the woffer's files is at the (forgot the proper terminology) breakout/distortion point around the 1000-2700 and placed at different frequency in each different frd files.
at the end, i sent for my version 3.9uF .33mh on tweeter and just eliminate the woofers resistor until i buy a calibrated mic and generate my own frd files and build a crossover based on my drivers and box assembly.
Or if anybudy can direct me to the wright path.
all in all it's been a great learning experience to deal with all of this, loved it.
ps still haven't heard any of the versions hence just yesterday received the kit for the third time, this time it seems to have all the right parts and no damage on any parts. this is until i can test the drivers, but that is another story
Allex, thank you for what you've provided here. A question: In XSim, which FRD files should I be using to (first step) replicate your simulations? I can load the individual FRD and ZMA files into the driver "Tune" panel in XSim and get close to your results but I'm thinking that I ought to be accounting for cabinet/port effects and don't know how to use the combined file.
Thanks again!
The combined file is used in “Frequency Response” window, where in the “Curves” menu you can see the option “Get File”. When you load that file you can see the FR plot of the two drivers combined (without any crossover), taken while the drivers were in the actual cabinet. Then, you have to tweak the “mod delay” from the “Tune” menu of the woofer until you get the two curves match as much (or as close) as possible. It should be a positive number, since this is actually the time delay of the woofer from the tweeter measured as a distance between them in inches. If my memory serves me right it was 0,85 in, but I might be wrong…
This is a standard procedure before you start designing your crossover and is a must if you want to get the phase tracking correctly. A good crossover design requires paying attention not only to the FR, but also to the impedance of the system and the phases.
Unfortunately, XSim is quite problematic software (at least in my experience) and I have troubles to get that number right, because often I end up with a different value when trying to repeat the procedure a second time. Good luck
Thanks. Eliminating the crossover by setting appropriate components to short or open, and displaying the curve from the combined file along with the modeled combined output, I get 1.78", which brings the two curves precisely together. I assume that's supposed to equal the distance between the acoustic centers. That seems like a lot. But I don't have the speakers in hand, but with a dome tweeter and depending upon the depth of the woofer cone, maybe it's not crazy.
Before doing this, I had to invert the tweeter in order to get reasonable system responses!
Before doing this, I had to invert the tweeter in order to get reasonable system responses!