Author Topic: Possible to view speaker crossover points on scope  (Read 4737 times)

0 Members and 1 Guest are viewing this topic.

Offline adauphinTopic starter

  • Supporter
  • ****
  • Posts: 128
  • Country: us
Possible to view speaker crossover points on scope
« on: January 25, 2019, 03:02:47 am »
If there are tutorials on this, please direct me as I would love to learn about it.

What I'm trying to accomplish, and I understand the results change with different loads on the crossover, is to see the actual crossover points from a frequency sweep or direct frequencies as they are coming out of a crossover.

My home speakers have had the caps and  Inductors replaced, I still need to verify the inductance, but I would like to run a signal through the crossover, and see how it behaves.

I've used a calibrated mic and listened to the room, but would like to verify the crossover is functioning properly and also note any changes based on inductor placement and types of caps or resistors.

I'm using a Rigol 1054Z, I dont have a function generator yet, I can feed sine waves through the preamp into the crossover.

Also would like to see if and how my preamp and amp may or may not influence the frequencies from the perspective of the scope and not my ears.

 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Possible to view speaker crossover points on scope
« Reply #1 on: January 25, 2019, 07:13:30 am »
Hi adauphin

The equipment you will need is a scope (which you have) and a sine wave frequency generator that can be continuously adjusted over the frequency range 5Hz to 32kHz (20Hz to 20kHz is the maximum frequency range of human hearing). Why have a frequency generator with a range wider than the range of human hearing: for two reasons. Firstly, it is not wise to have any equipment running at the extremes of its performance and, secondly, the additional frequency range will be useful for other testing you may wish to do in the future.

The frequency generator would also need to have a continuously adjustable output voltage amplitude of 0V to 1V RMS (2.83V peak to peak)  at least. The output voltage from the frequency generator needs to stay constant over the entire range of 5Hz to 32kHz. Make sure you do not get a frequency generator where the output voltage/frequency bounces when you change frequency or output voltage, as this will make the frequency response testing exercise tiresome.

To check the frequency characteristics of your pre amplifier:
Connect the scope to the preamp output.
Set the the signal generator to 0V output at a frequency of around 440Hz and connect the signal generator to the preamp input.
Adjust the signal generator output until you get a sine wave of around 100mV RMS (283mV peak to peak) approximately displayed on the scope. Adjust the scope trigger to lock the sine wave on the scopes screen.

Then adjust the frequency of the oscillator from 20Hz to 20kHz and observe the amplitude of the sine wave on the scope.

To check the frequency response of your power amplifier(PA) do the same thing, but, to avoid damage to your amplifier and speakers (the speakers should be connected to the PA output for the PA frequency response tests), you must observe the following precautions without fail:
  • Always ensure that the output voltage from the oscillator is set to 0V, and the frequency is set to around 440Hz, before connecting the oscillator to the PA
  • Never adjust the signal generator output voltage so that the output voltage from the PA in more than 1VRMS, give or take a touch, for amplifiers and speakers with an RMS power rating of 10W or more. For lower power amplifiers/speakers scale down accordingly
  • As a precaution, keep testing at frequencies above 5kHz to a minimum
To check the frequency response of your Xovers, you will need to get inside the speaker cabinets and do the same test as for the PA, but with the scope connected across each pair of speaker chassis terminals in turn. For the tests the speakers should be fully assembled as they would be for normal listening (this is most important for the bass speaker). The speakers should also be in a normal listening mode: that is not facing a wall or with anything on the speakers to suppress the sound output. If you plan to do a lot of Xover testing it may be advantageous to permanently wire-in a set of terminals connected to the speaker chassis terminals.

This test assumes that the frequency response of your PA is flat, which it should be, except possibly for frequencies above 20kHz and below 20Hz.

By the way, when doing audio frequency response tests it is conventional to use 440Hz (middle A in music) as the reference frequency.

Also, to mimic the human ear's perception of loudness, a logarithmic scale is used with deciBells (dBs) being the working unit. So, if you draw graphs of the various frequency responses, convert the peak to peak reading that you see on the scope to RMS (multiply by 0.35) and convert the RMS value to dBs. Also plot the frequency on a logarithmic scale. An idealized frequency/dB graph for a three-way Xover is attached below as an example.

Finally, remember the warning about damaging your PA and speakers. Why all the drama: PAs and speakers are not necessarily designed for reproducing continuous sine waves (normal music has a surprisingly low RMS value) and, with too higher voltage input, you stand the risk of overheating the speaker chassis voice coils and wrecking the speaker chassis. This is especially the case with tweeters. If you take a typical 100W RMS speaker, the woofer may have an RMS rating of 150W, the squawker 50W, but the tweeter only 5W. Also, by their very nature, tweeters are delicate transducers with minuscule clearances. So the message is beware of high frequency testing.

If there is any doubt about the stability or otherwise of a PA, always connect a 15R or higher, 25W or higher, resistor between the PA output and the speakers for initial, look-see tests.

The above procedure is the low-cost knife-and-fork approach, but if you are doing frequency response testing on a regular basis, you may want to investigate test equipment that automates the frequency sweep and produces a nice frequency response graph for you.

Manual frequency testing  may sound like a palava but, with a bit of practice, it is quite easy to do.
« Last Edit: January 25, 2019, 01:58:54 pm by spec »
 

Offline adauphinTopic starter

  • Supporter
  • ****
  • Posts: 128
  • Country: us
Re: Possible to view speaker crossover points on scope
« Reply #2 on: January 25, 2019, 11:57:36 am »
Thank you spec for the very informative response...much appreciated.

I will surely heed the warnings about the output voltages.

As far as signal generators, not sure if these will suffice, but I'm debating either a GW Instek SFG1003 or possibly going for a Siglent SDG 2042X. The price of the 1003 is far more attractive, but wondering if the 2042 would be more of a permanent piece for quite a few years to come....not sure if I need everything the 2042 offers, but still an option.



 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Possible to view speaker crossover points on scope
« Reply #3 on: January 25, 2019, 01:53:40 pm »
Thank you spec for the very informative response...much appreciated.
My pleasure.  :)

I will surely heed the warnings about the output voltages.
Very wise- I have blown up some really nice speakers though the years but not by frequency testing, mainly by testing prototype PAs which oscillated.

As far as signal generators, not sure if these will suffice, but I'm debating either a GW Instek SFG1003 or possibly going for a Siglent SDG 2042X. The price of the 1003 is far more attractive, but wondering if the 2042 would be more of a permanent piece for quite a few years to come....not sure if I need everything the 2042 offers, but still an option.
The requirements for a  sine wave oscillator for frequency response testing are really basic and could probably be met by one of the $7 ebay units.

But, if you are into high end audio, you may be looking for a very low distortion sine wave oscillator for testing for distortion, for example, in the future. Afraid I am not well versed in signal generators, but there are quite a few other members who know the score.
« Last Edit: January 25, 2019, 02:05:16 pm by spec »
 

Offline ArthurDent

  • Super Contributor
  • ***
  • Posts: 1193
  • Country: us
Re: Possible to view speaker crossover points on scope
« Reply #4 on: January 25, 2019, 03:50:46 pm »
A simpler way to do a quick check of a crossover is to use a dual trace scope with one channel connected to the input to the crossover and the second channel connected to the output of the crossover with a resistive load instead of the speaker. If you manually adjust the 1st channel to always have the same level trace you don’t need a leveled oscillator or a flat amp. Start the oscillator at a low frequency and increase the frequency until you see the 2nd channel waveform start to dip while keeping the 1st channel level constant by adjusting gain. The ‘knee’ will be more or less pronounced depending on whether you have a first or second order crossover network. The crossover frequency might be considered to be where the output waveform is about 70% of the amplitude of the input waveform.

If you have a crossover for the woofer then the test procedure is similar but you sweep the frequency down, not up.
 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Possible to view speaker crossover points on scope
« Reply #5 on: January 26, 2019, 08:09:51 am »
A simpler way to do a quick check of a crossover is to use a dual trace scope with one channel connected to the input to the crossover and the second channel connected to the output of the crossover with a resistive load instead of the speaker. If you manually adjust the 1st channel to always have the same level trace you don’t need a leveled oscillator or a flat amp. Start the oscillator at a low frequency and increase the frequency until you see the 2nd channel waveform start to dip while keeping the 1st channel level constant by adjusting gain. The ‘knee’ will be more or less pronounced depending on whether you have a first or second order crossover network. The crossover frequency might be considered to be where the output waveform is about 70% of the amplitude of the input waveform.

If you have a crossover for the woofer then the test procedure is similar but you sweep the frequency down, not up.
You didn't mention that you need a signal generator with an output impedance of zero Ohms and capable of driving around 1 Ohm reactive load.

Also, the loudspeaker and cabinet characteristics are very much part of the Xover function, especially in the bass area. A resistor does not represent the load of a speaker.

If the PA is not quite flat you can use the same approach that you say, simply adjust the input signal to compensate, so you do not need a PA with a flat frequency response, although it would make the tests much simpler and quicker if you did not have to adjust the input for every frequency.

You also missed out the practical aspects of taking the back/front of the speaker cabinet off, and unsoldering the Xover from the speaker chassis. You would also need to buy two or three power resistors, depending on the number of speakers involved, and solder them to the Xover as the sections in a Xover do not operate in isolation, they interact with each other.

If you want a really quick, simple, and zero-cost approach, just make an FLAC, or other suitable audio file format, with the required frequency sweep and amplitude to give a flat response. Then play that back and listen to each speaker or measure the speaker terminals with a scope using 2 or 3 channels. :)

But I get the impression that the OP wishes to examine the crossovers in detail, especially the areas where the Xover takes place.
« Last Edit: January 26, 2019, 12:10:06 pm by spec »
 

Offline ArthurDent

  • Super Contributor
  • ***
  • Posts: 1193
  • Country: us
Re: Possible to view speaker crossover points on scope
« Reply #6 on: January 26, 2019, 05:06:39 pm »
spec - "You didn't mention that you need a signal generator with an output impedance of zero Ohms and capable of driving around 1 Ohm reactive load."

You misunderstood. I never said to drive the crossover directly from the generator. You use the generator and amp. Also the crossover frequency is determined by the R, C, and L that you choose. Whether you are testing audio or R.F., a dummy load is a valid way to test a frequency dependent circuit.


The OP states that after making repairs they "would like to verify the crossover is functioning properly." I have used my method on 12db/octave crossovers I've designed and it works. There are more complicated nearly lab methods of testing and design but I suggested a method that is simple and works for what the OP asked. 
 

Offline Zero999

  • Super Contributor
  • ***
  • Posts: 20357
  • Country: gb
  • 0999
Re: Possible to view speaker crossover points on scope
« Reply #7 on: January 26, 2019, 06:26:55 pm »
The impedance of the speakers will not be constant. It will be frequency dependant and will also contribute to the frequency response.


In my opinion, an oscilloscope isn't the correct tool to use. An attenuator connected to the sound card line in could arguably do a better job. Another possibility is to just the calibrated mic to monitor the sound from each driver, but it might not be possible to do that to each driver in isolation, depending on how the crossover network is designed.
« Last Edit: January 27, 2019, 09:34:46 am by Zero999 »
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: Possible to view speaker crossover points on scope
« Reply #8 on: January 27, 2019, 01:21:58 am »
Also with a dual channel oscilloscope measuring the input and output of the crossover, the phase response can be measured which usually give a better indication of the frequency breakpoints.

If your oscilloscope can do an FFT, then the frequency (and phase) response can be measured in one shot using impulse testing but this either requires an impulse source or better, an oscilloscope which can calculate the FFT of the derivative.
 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Possible to view speaker crossover points on scope
« Reply #9 on: January 27, 2019, 01:26:54 pm »
+AD

spec - "You didn't mention that you need a signal generator with an output impedance of zero Ohms and capable of driving around 1 Ohm reactive load."
You don't with my method.

You misunderstood. I never said to drive the crossover directly from the generator. You use the generator and amp.
You gave the impression that you were driving the Xover from the signal generator and made no mention of using an amp. In fact you implied the opposite.

Also the crossover frequency is determined by the R, C, and L that you choose. Whether you are testing audio or R.F., a dummy load is a valid way to test a frequency dependent circuit.
True where the load is fixed and clearly defined, but not true when dealing with speakers.

The OP states that after making repairs they "would like to verify the crossover is functioning properly." I have used my method on 12db/octave crossovers I've designed and it works.
Sure, but does it give accurate results rather than a quick look see.

There are more complicated nearly lab methods of testing and design but I suggested a method that is simple and works for what the OP asked.
We will have to agree to disagree on this one. :)
 

Offline spec

  • Frequent Contributor
  • **
  • Posts: 833
  • Country: england
  • MALE
Re: Possible to view speaker crossover points on scope
« Reply #10 on: January 27, 2019, 01:35:38 pm »
The impedance of the speakers will not be constant. It will be frequency dependant and will also contribute to the frequency response.
Yes, that is the point.

In my opinion, an oscilloscope isn't the correct tool to use.
True but it is the instrument that the OP has and what he asked about. A better instrument would be an RMS voltmeter, with a flat frequency response from 10Hz to 20kHz

Another possibility is to just the calibrated mic to monitor the sound from each driver
That would work, but there is the danger of a whole load of variables clouding the issue, especially in the bass region- I'm not sure it is easier either.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf