crossover network impedance question

Status
Not open for further replies.
Ok, I aquired this panasonic 5.1 amp, it sais that speakers that are 6 to 8 ohm only should be used.

The speakers i wanna use with it are kindof special, they have a 4 ohm woofer, a 6 ohm mid and a 8 ohm tweeter to match their sensibility.

So that means that at 0 to 200 hz, the speaker impedance will be 4 ohm, from 200hz to 2000hz 6 ohm and from 2000hz and +, 8 ohm.

Evrything is second order too.

Could that cause problems to the amp? Its a class H, and i'd guess they use premade ic's like stk's, and those have like tons of protections..

Any ideas on this one?
 
Ilianh said:
I aquired this panasonic 5.1 amp, it sais that speakers that are 6 to 8 ohm only should be used.

This is one way of saying that the amp has limited capability of providing the current needed to drive lower impedances -- the power supply has been skimped on to get the price down.

It could well have some significant problems driving your 4 ohm woofers.

dave
 
oh well.. that greatly sucks...
ok... so what i'd do is... hum...
add a 2 ohm resistor to the woofer, the mid and the tweeter?

I dont see how a too high impedance (the amp is 6ohms to 8ohms and if i add a 2 ohm to the tweeter is gets 10ohms) could do something to the amp unless its a tube amp, but its not...

but that still sucks... bah...

think it'll work?
 
Have you actually tried the amp out with your speakers?
not yet.. im still trying to figure out what to do


Are the speakers your own DIY?

yhea.. my own diy


So here comes a question.
On a crossover network, lets say a 3 way, when you have a frequecy that in the LP section, its the woofer's impedance, when its in the BP section, its the mids impedance, and when its in the HP section, its the tweeters impedance.

But audio isnt only one frequency at a time, its alot of frequensies at the same time, and if you have mids, bass and highs, your total speaker impedance (if evrything is 8 ohm) goes down to like 2.67 ohm.

Ive seen comercial speakers having a "3 way" crossover thats simply 2 caps, one for the mid, one for the tweeter and the woofer is left to his natural rolloff, they are all 8 ohm drivers. and the speaker sais its 8 ohm on a sticker in the back. same thing here, its basicly never 8 ohm, its almost always under that...

The speakers I made have an external crossover so I can change stuff on it and make tests easily.
For now its a second order LP, a first order BP and a second order HP. I wanted to try simple a cap on the mid, to see whats his natural rolloff and how does it perform.

So, all this may sound stupid but its one of those things that bugs me off...
 
Okay, just for the heck of it, can you give us details on your woofer, the Thiele-Small parameters, which sized box, ported or sealed, and tuned to what frequency using which sized port?

I want to run your speaker in Subwoofer Simulator, a freeware program, (written by F4ier, a member here), which gives impedance curves when it has that information.

Also, can you give us details of your crossover from woofer to mid?

Some 4 ohm speakers end up having fairly high impedances. If you are near 6 ohms anyway, who knows, maybe you can pass.
 
Ilianh said:
So here comes a question.
On a crossover network, lets say a 3 way, when you have a frequecy that in the LP section, its the woofer's impedance, when its in the BP section, its the mids impedance, and when its in the HP section, its the tweeters impedance.

But audio isnt only one frequency at a time, its alot of frequensies at the same time, and if you have mids, bass and highs, your total speaker impedance (if evrything is 8 ohm) goes down to like 2.67 ohm.

You only need worry about the lower impedance in the woofer section. Because of various factors, but mostly because of the XO, the other speakers have very high impedances out of their band and don't significanly impact the impedance of the in-band speaker.

get that info to kelticwizard and he will curn out some pictures with MUCH more meaning than the simple 1 dimensional nominal impedance number.

dave
 
ok, well its in a 1.771 cu ft box tuned at 31 hz.

thats the woofer part, its separated form the mid and tweeter part.

the crossover is a second order LP at 200 hz fx.
and the mid has a simple first order BP from 200 hz to 2000 hz

heres the woofer tiele-small params
 

Attachments

  • woofer 8inch specs.jpg
    woofer 8inch specs.jpg
    37.6 KB · Views: 474
Simulation with an Re of about 4. The 3.8 Ohms reading will change this only marginally.

Forget about the low impedance near 200 Hz. Your crossover components will riaise that. Your inductor is probably about 4.7 mH or so.

The impedance vaalley at the tuning frequency is probably just a bit higher than shown. It is a rule of thumb that the impedance at the tuning frequency of a bass reflex is between 1 and 1.5 ohms higher than Re. However, most of the time it measures a bit higher.

I would say that you could probably hook these up as they are. To be safe, I would consider an 1 ohm resistor in series. It will cut your midpoint sensitivity by 2 dB, but will put you safely over the 6 ohm line. You could even consider parallelling two 1 ohm resistors to give a 1/2 Ohm value. That would cut midrange sensitivity by less than 1 dB.

Here is the chart.

The light blue line is your response on 2.83 Volts. I watt for an 8 ohm speaker, 2 watts for a 4 ohms speaker.

The purple line is the impedance.

The yellow line is the impedance phase angle.
 

Attachments

  • ilianh woofer impedance.gif
    ilianh woofer impedance.gif
    20.5 KB · Views: 449
kelticwizard said:
I would say that you could probably hook these up as they are. To be safe, I would consider an 1 ohm resistor in series.

Keep in mind that the 6 ohm limit is for using the amp with all 5 channels driven. If you are only using it to drive a pair of speakers the current that would be available for the other channels is now available for the two channels used so you have a little more leeway.

Also, they will have given themselves some margin for error. One of the reasons being that a speaker will often be speced at a higher impedance than it really is so that the sensitivity spec can be exaggerated.

I'd just try them. Listen for strain as you turn the volume up. Then stop and back them off. That is the limit of your amp. Then you can try it with the resistor. If it doesn't play louder with it in then you don't have a problem. If it does then the amp is very sensitive to low impedance loads.

dave
 
By the way, your parameters did not contain a spec for Le. I modelled the speaker with an Le of 1mH, but that was a guess.

I noticed your Thiele-Small parameters had such numbers as Fc, Qmc, etc., that indicates that this speaker was used in a closed box. Do you have a response graph of the speaker either unenclosed or in a closed box?
 
The multimeter I have has a resolution when measuring inductance of 0.01 H

I could test its Le, but at this resolution..

Whoa. From your first post, I thought that these speakers were already built, crossed over, housed, finished, and up-and-running!

I said this in the 4th reply


The speakers I made have an external crossover so I can change stuff on it and make tests easily.



Do you have a response graph of the speaker either unenclosed or in a closed box?

Not yet, I look forward to using speakershop for this but my sound card is half duplex (well, theoriticaly, it should be full duplex...but.. speakershop sais its not)
well I'm getting a new sound card soon because I burned the preamp stage of this one...
 
Ilianh:

A) Oops, sorry, missed your 4th post where you said that. That explains a lot.

I mean, I was just wondering, if you had not built the speakers yet, why you would be trying to use woofers that you were thought might not be suitable. I thought, "If this guy is worried, why doesn't he just use 6 or 8 ohm rated woofers to use in his speakers?" 😀 😀

As for the response graph, I noticed that your parameters contained such numbers as Fs and Fc, Qms and Qmc, Qes and Qec, etc. Those numbers that end in c mean that this speaker was enclosed in a sealed enclosure, and those measurements were taken and listed along with the parameters in free air. I was just wondering if your speaker came with a response graph in a sealed enclosure, since those "c" numbers indicate that measurements were taken in a sealed enclosure.

By the way, what is the brand and model of your woofer?
 
Full duplex sound card? I don't know what your finances are, but you can get a decent full duplex sound card for $10 delivered at www.pricewatch.com. Just click on Multimedia-Sound Cards and take a look. that is how I got my sound card.

I ordered a generic sound card with the Crystal 4280 chipset and I used a freeware program to arrive at a frequency response given in this thread:
http://www.diyaudio.com/forums/show...&perpage=15&highlight=sound card&pagenumber=2

The link to the test freeware is given on that page as well.

At first glance, the response doesn't look good, but then you look closely and see that the sound card is -3 dB at 35 Hz and only -6 dB at 20 Hz. Not bad performance, but more importantly, predictable and therefore can be used to test equipment under 20 Hz.

The sound card came without softwaare, but I downloaded the driver at www.drivers.com without a hitch.

PS: If you do download the freeware for the 4280, for some reason I found it easier to put it into a folder in Program Files and then install the sound card through the Add Hardware part of the Control Panel. When the time came to choose, I chose Have Disk and specified the C drive and the specific folder the 4280 was in. For some reason, it was only interested in the "inf" file.
 
Kelticwizard:

Well theres the CMI 8738 4.1 channel and its 14$ cdn and the Diamond DT-588 4 channels too for same price, at a store 1 street away.
The CMI8738 is full duplex, I think i'll get this one.

I was just wondering if your speaker came with a response graph in a sealed enclosure, since those "c" numbers indicate that measurements were taken in a sealed enclosure.

Well its a AR 8"speaker and its writen d01-0007 on the back, I had to find the thiel-small parameters myself because AR didnt respond to my emails. I used the sealed enclosure method to fnd the specs.
This driver is used in the ar-318ps..
http://www.supremevideo.com/audio/acoustic_research/ar318ps.htm
 
Status
Not open for further replies.