Link-Ril. type or ??
Though a lot more involved are the DSP based X-over. You can measure speaker(s) in cabinet and make exact transition for those drivers --- flatten group delay, freq response, phase, time delay/align and the whole thing.... and also provide a x-over function.
-Richard
Last edited:
Oh. Only for the LXmini speaker. OK. Good approach for a specific spkr. But just handles amplitude EQ/xover frequencies(?). Not comprehensive enough for myself.
Doesn't help me or others. Fine. I might do the same with M2. Hate doing all that work by myself. Yuk. But the measured data is available to me already from H-K.
What will be the choice of complimentary jFETs?
THx-RNMarsh
Doesn't help me or others. Fine. I might do the same with M2. Hate doing all that work by myself. Yuk. But the measured data is available to me already from H-K.
What will be the choice of complimentary jFETs?
THx-RNMarsh
Last edited:
Possibilities of analog xover filters are very limited. Digital filters allow much more manipulation with phase, GD etc. More capable to help.
Yes previous to that he did, but if you recall before that he was not convinced that DSP would be as good. I understand that on the oplug forums a lot of people were sticking with the ASP.4 and a new board spin was done. I like it as it's an interesting study on a very complex active crossover.
Yes, there is a renewed interest in analogue, for instance Doug Self's new book. It would depend which is simplest, if it means additional conversions. Also I have a suspicion a down side to DSP may be it's power, at a cost, pre-ringing for example. There is no doubt it can do things analogue can't but it is questionable whether they are necessary or audible, in a good way.
Equation (2) of that paper is surely wrong? He has 1 + PI e(x) where he should have PI (1+e(x)) - which can be approximated by 1 + SUM e(x).john curl said:
Equation (10) shows that this is a small effect, maximised around the open loop LF rolloff and then reducing with increasing frequency. As it is a small effect it will result in much the same frequency components as 'AM' distortion, but with slightly modified phase.
Thanks for reminding us about this, but it really is a non-story as the effect is so small.
Your test equipment will lump them together, which won't matter because the FM is small and almost entirely at the same frequencies as AM. Hence there is no need to separately measure the FM.
Oh, forgot to say, this effect will be present however little feedback you use so it is a good job it is such a small effect. Like all manifestations of re-entrant distortion if you are worried about it then you have to use enough feedback to suppress it; otherwise use no feedback at all, remembering that local feedback (including degeneration) obeys the same mathematical laws as global feedback.
Did I say they did, is the heat getting to you? You seem a bit grouchy today 😛
It's obvious from first principles of course....😉
So much effort and expense is used to prevent and remove e/m coupling issues, ground loops, interference in systems.
And yet, magnetic coupling of tens of amps at frequencies into the tens of kHz is ignored when all these agressors are stuffed into 100 square inches of fully enclosed metal with the victims, and everybody has to reference a single potential located somewhere in that tornado of fields.
I am reminded of a WWF cage match, or a demolition derby.
At least with class A you know where all the currents run and can at least close the current loops a bit, but with B, two return paths make it more difficult. And for two channel amps? Sheesh.
For me, the biggest problem is one of metrics. If you can't measure it, how can you possibly fix it?
We try to make two channel systems where the final user is sensitive to 2 to 5 uSec interchannel delays, but we can't measure that when two complex partially correlated signals are exercising the system?
I keep harassing the controls guys here....do NOT try to fix a mechanical problem using software!! Fix the hardware!! Our problem is even more complex as the systems are extremely non linear, position dependent, and are highly coupled nonlinearly with position dependence on the coupling. And yet, they have been trying to solve this using simplistic linear thinking.
Feedback is not the cure to a layout that couples internally due to lack of control of currents.
Jn
Last edited:
I am sure the analog xover will sound very nice in Bi or Tri-amp mode. I did that ages ago. BUT … those large peaks that didn't get EQ'ed out are a problem IMO. Close but no cigar. Nice easy DIY project.
THx-RNMarsh
THx-RNMarsh
yes. It was an issue with the Damir designed amp build also. The EE/Techs at mfr could not get it right and distortion was higher than it should be. I had measured it a lot lower. Only doing a couple trips to the factory and spending time with them --- re. how to do the wiring properly to prevent coupling. Now the distortion at the bottom of test equip.
THx-RNMarsh
Last edited:
What happens when a shield is fitted around high current wire pairs...IOW high power/big copper microphone cable used as power distribution inside an amp, for example ?.
Dan.
- Status
- Not open for further replies.