Hi to everyone...
When posting, it would be really helpful (to me anyway!) if you could direct your message to the person concerned, with a name first?
Would this be ok? Hope so. Might just make it easier! grant
Nearly 5000 page views, and no other contributors? Where is everyone?
When posting, it would be really helpful (to me anyway!) if you could direct your message to the person concerned, with a name first?
Would this be ok? Hope so. Might just make it easier! grant
Nearly 5000 page views, and no other contributors? Where is everyone?
Hi,
To make any progress in violinmaking takes lots and lots of practice, over and over again, and progress only comes in VERY small steps - the right tools and fine wood wont ever make me no "Strad"
My experience with speakers is about the same - hard work and lots of practice
I think, to become a good builder you need the urge allways to do it better
To make any progress in violinmaking takes lots and lots of practice, over and over again, and progress only comes in VERY small steps - the right tools and fine wood wont ever make me no "Strad"
My experience with speakers is about the same - hard work and lots of practice
I think, to become a good builder you need the urge allways to do it better
Hey, just made some crazy stuff
For a short while I have had some uncomfortness and a bit of hearing problems
Then I remembered it was after I had made some adjustments on my crossover - it sounded ok, but something was not completely right
I made another adjustment, and suddenly my ears relaxed completely
ADJUSTMENT: A ZOBEL on woofer consisting of 10uF/10ohm - resistor was reduced to 6.8 ohm, thats it
Crazy isnt it, well I am anyway
For a short while I have had some uncomfortness and a bit of hearing problems
Then I remembered it was after I had made some adjustments on my crossover - it sounded ok, but something was not completely right
I made another adjustment, and suddenly my ears relaxed completely
ADJUSTMENT: A ZOBEL on woofer consisting of 10uF/10ohm - resistor was reduced to 6.8 ohm, thats it
Crazy isnt it, well I am anyway
Hey Grant: I hear airfairs from Australia to the US are really cheap these days..........lol..........At any rate, processing the files this time around was a bit more efficient. I think it's simply a matter of the learning curve and......................well, being obsessed.........of course. I redid the TDFC and the CA15RLY, so now I am working with all files which I started in TRACE through to the Frequency Combiner..............Tinitus, I did indeed, experiment with the midbass in normal polarity, though only briefly. Each time I did, the summed responce dropped...........Some pretty wild stuff. I did not know that you could run a 12dB - 3 way with both midrange and tweeter reversed polarity. I guess it might be like running mid and tweet in normal polarity and having the woofer in reverse polarity.........................Go figure.........this DIY crossover stuff has really evolved, since the days of the textbook design...........Any observations, comments, suggestions, direction, ideas are welcome............I will keep on workin....................Grant, check your Email...............Respectfully..........Omni
Regarding my post:
"Hi to everyone...When posting....." , <-- this is clearly stupid, please ignore it! apologies, grant
"Hi to everyone...When posting....." , <-- this is clearly stupid, please ignore it! apologies, grant
Omni,
Thanks! I received your new CA and TDFC files, and kept the 'NEW' Peerless files. I modelled your new files with your latest xo values and DCR's. Except for the smaller 2 coils, I used Solen DCR values, for 14 gauge, ie, 0.11 ohms for both. This attachment is the sim for your xo with new data. A bit peaky and dippy.... grant
Thanks! I received your new CA and TDFC files, and kept the 'NEW' Peerless files. I modelled your new files with your latest xo values and DCR's. Except for the smaller 2 coils, I used Solen DCR values, for 14 gauge, ie, 0.11 ohms for both. This attachment is the sim for your xo with new data. A bit peaky and dippy.... grant
Attachments
Omni,.... I think your difficulties with polarity is related to your driver offset - polarity and phase is most important and the very key to the problems with a 3-way - and as you can see it has influence on your whole crossover
I know my way of doing it is not excactly textbook theory, but my reason for being here to try and support you is the fact that I have studied this particular curcuit and setup for more than 5 years now - believe me, you need dampening resistors on paralel components, and adjust components accordingly
And you need high crossoverpoint on midrange
Again I must encourage you to look at woofer lowend rolloff and use this curve as a reference - fact is that you are actually quite close
Another thing - acoustic center changes all the time, its kind of a uncontrollable player, which leads to compromises, and I dont think theres any certain way to deal with this problem
I hope to see some new and revised sims 🙂
I know my way of doing it is not excactly textbook theory, but my reason for being here to try and support you is the fact that I have studied this particular curcuit and setup for more than 5 years now - believe me, you need dampening resistors on paralel components, and adjust components accordingly
And you need high crossoverpoint on midrange
Again I must encourage you to look at woofer lowend rolloff and use this curve as a reference - fact is that you are actually quite close
Another thing - acoustic center changes all the time, its kind of a uncontrollable player, which leads to compromises, and I dont think theres any certain way to deal with this problem
I hope to see some new and revised sims 🙂
Omni,
So I thought I might do another rework, of a previous '2nd order' xo and came up with this new xo for your latest files. Strangely the tweeter now didn't need attenuation. Don't know why. The reponse
looks pretty good from 90Hz to about 8kHz within +- 1dB. The mid and tweeter 'appear' to xo at 2kHz, but I checked using the resonance formula. The mid LP and tweeter HP are both at 3kHz.
This seems to be the best I can get it for now. I might try a '1st order' for the mid sections tomorrow. best regards, grant
So I thought I might do another rework, of a previous '2nd order' xo and came up with this new xo for your latest files. Strangely the tweeter now didn't need attenuation. Don't know why. The reponse
looks pretty good from 90Hz to about 8kHz within +- 1dB. The mid and tweeter 'appear' to xo at 2kHz, but I checked using the resonance formula. The mid LP and tweeter HP are both at 3kHz.
This seems to be the best I can get it for now. I might try a '1st order' for the mid sections tomorrow. best regards, grant
Attachments
Omni,
the last circuit looks a little unclear, if you can't see all the values, I'll email it to you. grant
the last circuit looks a little unclear, if you can't see all the values, I'll email it to you. grant
I guess that if all drivers have offset to correct phase on acoustic center, measured from drivers dustcab/membrane to ears in listening position - then offset settings on sim program should be ZERO on all drivers, or have I misunderstood how it works ?
Omni,
I forgot to mention that I'm sure you won't notice the 1dB dip at 1500 and maybe even the 3dB peak at 14? kHz. ( I wouldn't, lol) grant
I forgot to mention that I'm sure you won't notice the 1dB dip at 1500 and maybe even the 3dB peak at 14? kHz. ( I wouldn't, lol) grant
Hi GN,
The 60R resistor looks superfluous.
Rzobel on the tweeter is far too low.
With no significant Rsource the tweeter zobel is superfluous.
🙂/sreten.
The 60R resistor looks superfluous.
Rzobel on the tweeter is far too low.
With no significant Rsource the tweeter zobel is superfluous.
🙂/sreten.
I have done listening test and concluded that tweeter sounds smoother, more extended and coherent with zobel, and that was with a Vifa XT25 which is 3 ohm DC and almost straight on its own - BUT a zobel with 2 ohm is too low, maybe a typo error ??
Are you sure you need attenuation on mid and nothing on tweeter - sounds strange to me, and I think it can only be finally determined by listening
You should solve the problem with the "dip" at 1.5khz....you might have to raise crossoverpoint....seems you have little overshoot because of too much overlap between woofer and tweeter
Tweeter will like a high crossoverpoint too as they generally tend to SOUND better
Just for the record.....but damm, every time I touch it the whole thing changes.....very frustrating
Component values at the present on my midrange-------series C 68uF with paralel inductor 4.0mH / 8.2 ohm ----------series Inductor 0.3 mH with paralel 8.2 uF / 2.7 ohm
Are you sure you need attenuation on mid and nothing on tweeter - sounds strange to me, and I think it can only be finally determined by listening
You should solve the problem with the "dip" at 1.5khz....you might have to raise crossoverpoint....seems you have little overshoot because of too much overlap between woofer and tweeter
Tweeter will like a high crossoverpoint too as they generally tend to SOUND better
Just for the record.....but damm, every time I touch it the whole thing changes.....very frustrating
Component values at the present on my midrange-------series C 68uF with paralel inductor 4.0mH / 8.2 ohm ----------series Inductor 0.3 mH with paralel 8.2 uF / 2.7 ohm
Experimented with crossover......and its finally back to "normal"......series coil 0.3mH, one winding off....... and standard zobel is 10uF/6.8ohm .... sounding MUCH better..... driver is ScSp 15W .... my years are healing again 🙂
I am definately appreciating the input here, and will do some things on sim program and will have results shortly.....Grant, on your first attempt with the new files, your simulation looked a bit like mine and in the midrange area, I need to round that down, however,on that nasty dip in your first graph, try reversing the polarity on the tweeter. With that same combination of components I got that same nasty dip, and just for the fun of it, I tried reversing the polarity, and wow, what a difference it made......Check out the graphs I posted with those crossover values and notice the difference................Sreten, what does superfluous mean?.......................Tinitus, I am glad you have taken the interest that you have, and you can be sure that I will have new and revised sims here shortly. For a little background info, I have been seeing your posts discussing damping resistors. A couple questions here:.........1: are the resistors in series with the components?...................2: which components should I add them to?.............3: Will they show differences on the response graph?..........4: how do I start to determine values for the resistors? I am definately open to checking out your resistor theme here as the sim program has a circuit adder to the program, which I can add resistors to the crossover sim and dial in various values. This is starting to heat up a bit........and that's cool.......Please give me a day or two, I'm not lettin this one go unfullfilled............Respectfully................Omni
Edit here............Tinitus, I reread your post about the damping resistors........All my questions are still valid except I saw you mentioned to put the resistors on Parallel components..........Will that be on ALL parallel components? I needed to clarify this........Respectfully, Omni
Hi Sreten,
Thanks for your comment. In that simulation I decided to sim all the Zobel component values, and choose the ones which gave the flattest response on screen. This was my only criterion at the time. I didn't thoroughly check Rod Elliot's treatise on this beforehand.. So yeah, I guess the resistors may look 'whacky', but they certainly did give the best result 'on screen' at the time assuming that the resistors were required. But I now realise, because of the values indicated, the
'effect' might be insignificant, and for example the 60R unnecesary/superfluous.
However, I note your comments, and I'll re-sim it pursuant to your suggestions. If there's any deviation in SW, can't see why, I'll let you know. Thanks grant
Thanks for your comment. In that simulation I decided to sim all the Zobel component values, and choose the ones which gave the flattest response on screen. This was my only criterion at the time. I didn't thoroughly check Rod Elliot's treatise on this beforehand.. So yeah, I guess the resistors may look 'whacky', but they certainly did give the best result 'on screen' at the time assuming that the resistors were required. But I now realise, because of the values indicated, the
'effect' might be insignificant, and for example the 60R unnecesary/superfluous.
However, I note your comments, and I'll re-sim it pursuant to your suggestions. If there's any deviation in SW, can't see why, I'll let you know. Thanks grant
- Status
- Not open for further replies.