capacitor replacement help

[ukspawn]

That’s got to be the sale of the century!.
A computer program that measures air within a box without a probe of some design. Can it tell if a bass port is present? Course not. If you have money to burn, my children are hungry, so am I

Martin

I'm not following you, Its not just a program its a physical device... i think you connect the box via USB to your PC and then the positive and negative terminals to the Inductor then it runs a known sweep/signal through it and gives a graph of the inductance and DC resistance over the frequency range.

My kid was hungry too i sent her to bed bacuse she knows no food after 8pm as an excuse to stay up later

 

[Martaine2005]

My kid was hungry too i sent her to bed bacuse she knows no food after 8pm as an excuse to stay up later

Your killing me mate, well done.


Martin

 

[ukspawn]

Anyway, goodnight mate, thanks for the chat.

have a good one

 

[ukspawn]

@Martaine2005
LCR meter arrived today. testing on a broken woofer coil, set to the 20mH range it shows 2.74...

is that just 2.74mH? how do i interpret it?

Putting a bit of metal in the middle changes the reading to 'more' so, all seems good and working just need to know what im looking at.

Thanks all

 

[kellys_eye]

Yes, 2.7mH.

Not sure what else you want to know! It's a coil!

 

[ukspawn]

thanks.

I just wasn't sure how to interpret the readout. EG is it 2.7 lots of 20mH cuz its set to the 20 range or.. i dont know.

And i think the resistance is important related to wire diamiter so i need one to replace with similar inductance AND resistance i believe.

Other than that i dont think there is anything else needed to be known in order to order one.

Cheers

 

[kellys_eye]

Back on topic (I think!) - if you want to test your changes to LC or R then you will need some pretty sensitive equipment to tell the microscopic differences that will result.

Audio spectrum analyser, distortion meter and all with specs of sub 1% accuracy (0.001% probably). There's no practical (ordinary-person-affordable) way to actually measure the differences - you'll be resorting to subjective results instead and that in itself could depend on all sorts of factors such as what you had to eat that day!

Steer clear of beans......

 

[ukspawn]

Back on topic (I think!) - if you want to test your changes to LC or R then you will need some pretty sensitive equipment to tell the microscopic differences that will result.

Audio spectrum analyser, distortion meter and all with specs of sub 1% accuracy (0.001% probably). There's no practical (ordinary-person-affordable) way to actually measure the differences - you'll be resorting to subjective results instead and that in itself could depend on all sorts of factors such as what you had to eat that day!

Steer clear of beans......

im hoping to just get a good enough value that it doesnt drastically change the frequency response [fr] as measured with my mic (i will DSP it back to my desired curve anyway) ±0.5dB

I want to test "what else" contributes to my perception of sound quality, all other things being as equal as reasonably possible.
its not ONLY fr that determines if something is good sounding i dont think. i want to test this theory

can capacitors sound "clearer" or better/worse - can i even percieve it in my room/setup resolution etc

Thanks

 

[Sidecar Bob]

You would need to put both speakers in exactly the same place in the room when comparing them because room acoustics can make as much difference as the components you are changing.
Ideally you would do a double blind test; Have someone else set up the speakers side by side behind a screen so that you can't tell which is A and which B and switch between them while listening. Make notes and when you think you are done have your assistant switch their positions and listen again WITH THE SAME PROGRAM MATERIAL AND AMPLIFIER SETTINGS as the first time to make sure it isn't the slight difference in room acoustics that you actually prefer.

Re capacitor choice: The capacitors used in crossovers must be bipolar because they need to pass more or less symmetrical AC waveforms If you can't get the value you need you can connect a pair of regular electrolytics in series back to back (note that the resulting capacitance can be calculated by the usual capacitors in series calculation).

BUT if the person who designed the crossover knew what they were doing the calculated values were only the starting point and the actual values used in production would be the result of repeated frequency response measurements with different component values until the desired frequency response (& preferably impedance response but unfortunately many designers never take that into account) is achieved, followed by the requisite double blind listening tests and design revisions as needed.
BTW: When I was designing speaker systems I almost never calculated the values of the crossover components. I knew the final values would be different anyway so it was easier to start with values I thought might be close and work from there.
Anyway, assuming that the components they use during the practical part of the design were the same as would be used in production any deficiencies in the type of capacitors &c used would be accounted for in the final design and replacing those components with "better" ones can only lead to aberrations in the frequency response (& impedance response) curves that can only be detrimental to the sound of the speakers.

Something to keep in mind when listening to speakers critically: I learned early on in my career that the listener should hear the music, not the speaker. I always strove for that goal because if they are aware of the speaker that means that it doesn't sound natural and the design is flawed.

 

[ukspawn]

thanks for the detailed reply. i have thought about a lot of what you say above.

Next i have taken pictures of the crossover and done a initial measure of the inductors while on the board (i will remove them and measure them off board before i order replacments)

Can anyone help me draw up this crossover please. i think iv got it mostly right in the picture below but if you could check it for me and help me figure out how to draw it properly.
I dont want to use the same board i will be making it in a larger and maybe 2 seperate boards so i can seperate the inductors better into LF and MF+HF boards.



https://drive.google.com/file/d/1WwLQ7wDIC_YeS4v1EbuxA-EkvJ9tBJ6K/view?usp=sharing

the file would not upload here so i put it on my Google drive in full size. the error here isnt specific and doesnt tell you why the image is unable to upload, its not particularly large and ive even tried to reduce the file size and image dimensions and it still wont allow it. anyway.

 

[Sidecar Bob]

The components on that board look perfectly fine and serviceable to me. And as I said, if the system was designed with the same components replacing any of them can have a very detrimental effect on the way the speaker sounds. Something to consider is that each of the capacitors and inductors has a base internal impedance in addition to the portion of its impedance that varies by frequency. If you substitute, for example, a coil of the same inductance that is would with thicker wire the base impedance will decrease and that can affect not only the level the component it is associated with plays at outside the crossover range but where in the spectrum the output of the speaker rolls off, which will add humps & bumps to the frequency response.
If you had the equipment to measure that you might be able to compensate for that but you would end up for all intents & purposes completely redesigning the crossover before it sounded close to right and you would probably not get it to sound as good as it did before you started tampering with it.

I would suggest that a project like this is beyond the understanding of someone who has not spent some time studying how crossovers work (if you understood them you would be able to knock out a schematic in a few minutes) and then done enough testing in a facility equipped to measure speakers to understand how the theory learned applies in the real world.

The changes you are talking about are akin to someone thinking
1) Cars go faster than bicycles.
2) Cars have wider tires than bicycles
3) Therefore If I modify my bicycle to allow 6" wide tires it will go faster.

 

[Martaine2005]

He doesn’t want to go faster.
He thinks the speakers will sound better if he spends lots of money on air core inductors and oxygen free wire!.


Martin

 

[ukspawn]

The components on that board look perfectly fine and serviceable to me. And as I said, if the system was designed with the same components replacing any of them can have a very detrimental effect on the way the speaker sounds. Something to consider is that each of the capacitors and inductors has a base internal impedance in addition to the portion of its impedance that varies by frequency. If you substitute, for example, a coil of the same inductance that is would with thicker wire the base impedance will decrease and that can affect not only the level the component it is associated with plays at outside the crossover range but where in the spectrum the output of the speaker rolls off, which will add humps & bumps to the frequency response.
If you had the equipment to measure that you might be able to compensate for that but you would end up for all intents & purposes completely redesigning the crossover before it sounded close to right and you would probably not get it to sound as good as it did before you started tampering with it.

I would suggest that a project like this is beyond the understanding of someone who has not spent some time studying how crossovers work (if you understood them you would be able to knock out a schematic in a few minutes) and then done enough testing in a facility equipped to measure speakers to understand how the theory learned applies in the real world.

The changes you are talking about are akin to someone thinking
1) Cars go faster than bicycles.
2) Cars have wider tires than bicycles
3) Therefore If I modify my bicycle to allow 6" wide tires it will go faster.

Thanks for the responsse.

Not to be dismissive but i have already decided I am doing this and I have already spent on it, some parts arrived some on the way... so its being done. Ill compare and measure the old and new crossovers, DSP them to the same frequency response to within about 0.5dB

Small changes to the frequency response are not a problem as i use DPS anyway to my desired house curve.

I am investigating for myself if crossover components can sound different even if the freqency response remains the same.
What else other than frequency response contributes to the perception of quality/clarity/tone and all those other audiofool terms we hear. i dont want to read about it, watch youtube or listen to anothers opinion on if its worth it or will be a failure.

If its a failure and ruins the sound then i will simply reinstall the original crossover and agree with you that it was pointless and a waste. Untill then i will proceed with the project and learn myself how to draw the crossover diagram.

I was just asking if anyoine could help me with that, thats all I want. I will get there eventually but if its a 2min job for someone else to look at that and save me the hours leaning it then thats all i was hoping for. dont ask dont get

So, again, can anyone help me with drawing this crossover diagram, does what iv indicated so far look correct?

Thanks all,

Mark

 

[ukspawn]

i tinker with things its what i like, i build a brand new top of the line PC and then i overclock it, make it unstable and crash untill i get it right then enjoy the 5% mopre performance that took me 100+hrs to achieve. everyone who doesnt overclock agrees its a waste of time, degrades components and risky for data loss...

Everyone that overclocks likes it..

I like it.. i like tinkering, i learn for myself the hard way always have.

If all i achieve is making the left and right tighter in tollorance moving from some 10% components to 3% ones then so be it. If i measure anything else with REW thats an improvment then BONUS!

 

[Sidecar Bob]

You can use digital equalization if you want but without the equipment to actually measure the frequency response it would be all guesswork.

I guess there are worse things you could do with your time but with many years of designing speakers behind me (and a lifetime since I left that field) it just seems like a waste of time to me, especially since without investing in good measuring equipment you will never really know what difference it actually makes.

 

[ukspawn]

You can use digital equalization if you want but without the equipment to actually measure the frequency response it would be all guesswork.

I guess there are worse things you could do with your time but with many years of designing speakers behind me (and a lifetime since I left that field) it just seems like a waste of time to me, especially since without investing in good measuring equipment you will never really know what difference it actually makes.

I use REW and calibrated mic so i can measure, my room is treated and speaker positions maked out etc.

 

[Sidecar Bob]

Good luck with that, then.

 

[ukspawn]

Good luck with that, then.

thank you, ill keep the thread updated and see where it goes.

now.

lets start with the crossover diagram. I believe learning first the components and how to draw them would come first and there are only 3 things so. How to draw a cap, inductor and resistor is the first port of call would you agree?

Then work out where the common ground is on that board?

 

[bertus]

Hello,

Perhaps this page of ESP will help you:
https://sound-au.com/articles/crossover-tables.htm
There are several cross-over schematics and compensation networks.

It will give you an idea how the filters are designed.

Bertus

 

[ukspawn]

thank you, ill start there

So a double Line like = sign is a capacitor.
A spring looking drawing is a inductor
A simple rectangle for resistor.

Nice! that was easier than i expected.

now to draw it out 'as it is' pysically just so i can get used to drawing, then i can start rearanging things in to 2 seperate boards (or whatever i learn is best).

im considering the posibility that changing lengths of connections between components could have an effect on time alignment of tweeter/mid/woofer

When designing a crossover is that a consideration or is it such an insignificantly small difference that its no difference? (like Phemtoseconds difference)

i have heard both arguments with speaker cables and i dont believe it matters unless using extreamly long cable runs. so im assuming it wont matter here either. but it just poped into my head that even 0.01ms delay can change the left/right balance in my setup from center to ~5degrees* off (*guess)