Site Links
Howdy, Stranger!
It looks like you're new here. If you want to get involved, click one of these buttons!
Quick Links
Categories
In this Discussion
Please review the site Rules, Terms of Service, and Privacy Policy at your convenience. Rules, TOS, Privacy
Get familiar with the reaction system: Introducing the Reaction System
In-Line Filters and measuring tweeters?
HI ALL. My first post here. I signed up to learn.
I was just wondering if using in-line filters as described in the video I linked would be a good way to keep a delicate tweeter under test from seeing the first few octaves. Would there be any advantages to rolling off the low bass this way rather than placing a cap between the amp and tweet?
Comments
Probably better off just using a cap between the amp and the tweeter for testing. If there is any turn on "spike" from the amp, the cap would have a chance of catching it.
I foresee it being a bit fiddly with regard to figuring out your actual input impedance. Then, since the values are rather small, any variation in the actual value from one to another would make for quite the swing in the rolloff frequency. Leading to trying to match up caps. Though once you've got it dialed in I imagine it could work well for limiting woofers.
Also I'm wondering how does removing the low end before the amp as opposed to using a cap after the amp affects the measured phase?
I imagine both would have a similar effect.
Thanks DrewsBrews
Until I watched the video I'd never heard of in-line filters. My first thought was using it to protect delicate or tiny tweets. At least now I know there are more options than just cap before tweeter or no cap.
The cap at the tweeter shifts the phase or adds delay in the electrical signal to the tweeter relative to any other drivers measured without the cap or a T0 marker in a two channel measurement. The filter pre amplification would not add a delay to the signal for any other drivers or T0 marker.
Wrong! The only thing that changes when adding a cap in a two channel measurement, is decreased SNR on the low end because the signal is rolled off, literally.
Here is impulse and resulting frequency response and phase for some tweeter. Green and blue lines overlaid for two measurements, one measurement was completed with an 18uF cap in series. Can you tell the difference in time that the impulse occurred? Phase error in the order of 1us lol.
For the topic on hand here, main difference with line filter vs standard filter will be achieving a proper filter rolloff, where a simple capacitor provides a rather poor rolloff characteristic for normal tweeters with impedance peaks on the low end. Whether the filter is pre or post-amp, it is a passive filter, so any change in phase resulting from the filter will be the same regardless of which implementation you use.
As a "protection" mechanism, filter between amp and speaker is the best solution, provides protection all the time, from amp turn on/off thump, to someone doing silly things like hot swapping RCA cables making that lovely "buzz". The cap protects what is upsteam of it, so best protection would be to locate it as close to the device being protected as possible.
Beyond the protection mechanism, I would say to provide the filter in digital domain rather than fuss around with line level filters. REW provides good features for this, allowing both frequency range limits for the measurement, and the ability to measure with any filter built with the EQ function. The EQ provided filter is a poor safety function though, as it's very easy to just hit the measure button and measure without it accidentally.
That is my understanding, but I was an economics major (way) back in the day...
@dcibel what you are saying is the cap at the tweeter adds no delay (phase shift) from T0? Holy smokes you might want to test that theory in your favorite sim program.
Not saying, showing with measurements. Properly done dual channel measurement will compensate for the cap with no extra steps required. Response with or without cap is the same, so is timing and phase.
A filter that rolls off the response will change phase. Phase is a function of frequency response and time. When you compensate for the cap by restoring the original response, phase is shifted the opposite direction by the same amount. Net result is no change, see above.
Only a dummy would use a tweeter response we with a cap in series without compensating for it, so simple to do these days even with single channel USB mic systems.
Could you be a little more terse and insulting, please? I just don't feel it coming through yet . . .
Maybe I’ll just stop posting here, y’all can be ignorant without me.
How do we compensate for the phase shift of the series cap in the final output? [They will exist using DSP, pre-amp as well as post-amp passive].
I am utterly lost at this point in the discussion and defer back to the OP's original inquiry.
Please help us understand . . .
I noticed the first word in the first response from Drew that he starts with the word probably. I like that. It seems appropriate for matters such as this.
Okay, So I've learned it's not a good idea to use an inline filter to reduce bass for measuring because of the possibility of damaging the tweeter. So that leaves us with a cap in series just before the tweeter and the phase shift that naturally comes along with this cap. BUT this phase shift can be compensated for even when using USB mics. So how do you compensate using a single channel system like Holms for example? I've read that some people will just ignore the cap phase shift and take their tweeter with cap measurement straight into Xsim. Will performing a HBT with tails in xsim do anything to help?
As soon as the Omnimic is back in stock I plan to start measuring and designing my own xovers. So, these are the kinds of things i need to know first.
Sorry I am just a new guy. I do not use SIM's. I use my Omnimic and my limited knowledge.
I see TroyH says it'll be 2nd quarter this year before the Omnimic 40k will be released.
InDIYana Event Website
Wolfy's Photobucket
My dog can’t wait.
The main benefits of the new system entail HD to 20k, and the measured time of flight system.
InDIYana Event Website
Wolfy's Photobucket
The HD to 20k will be great but the FR to 40k will be a disaster.
I've used ARTA in single channel mode, but was never able to match those up to measurements of both midbass and tweeter in parallel to find the Z offset. So my sims were WAY off. I then tried ARTA in dual mode and couldn't get it to work at all...I couldn't even get it to produce a sweep. I guess I'm one of those dummies dcibel talked about. But that was a few years back and I've learned a lot since them. I'm sure I'll have little problems with OmniMic. I wish I had one now. Maybe someone getting the new V3 will sell me their old V2 at a good price.
?
It's that time of flight measurement in the V3 that makes it attractive to me. Maybe it'll be worth the wait.
Being able to measure distortion with the V3 doesn't hurt either.
How do you get the video to show on the page? I've tried in a few posts but only see the link.
Just so there is no confusion, OmniMic V2 measures distortion quite well. It is just limited on the high end.
Whenever I see Danny Richy's face I know there's gonna be some serious BS blather...
Yes, but even a broken clock…. this one provides some usable info