Hi, I'm an audio engineer by profession. I particularly interested in learning about the electronic side of the trade. I have been recently working on learning about filter circuits and cross over design. I'm competent with basic calculation like impedance etc. I'm just going through some circuit designs trying to figure out what does what
I have attached a screen shot of the cross over circuit which I'm going through (I sketched it myself as I couldn't find a schematic for the circuit). Couple of question
1. What is the use of the combination of R1 and C2?
2. What is the use of the combination of C3 and R2?
I would like to know a way of measuring the cut off frequency - I could get access to an oscilloscope and a spectrum analyzer at my old college. Would I need to connect a dummy load?
Is there a way measuring this at home? Recently I download a neat software called spectrum lab.
Very grateful for your responses in advance. Comments/advice on going further will really be appreciated.
Thanks
p.s I left out the values of some component as I couldn't figure it out (from the screenshot).

 

hi Yami,
As you are planning to study filters, I would suggest you look at the LTSpice simulator.
It is a free download from the web, there a number of members who use LTS, so there is plenty of help on offer.

You could use it to analyse and study your posted circuit.
E

EDIT:
This is an example of your circuit, I have assumed the inductor values, post them if you know them.
Note the cut off off frequencies of the R/C filters.

 

hi Yami,
As you are planning to study filters, I would suggest you look at the LTSpice simulator.
It is a free download from the web, there a number of members who use LTS, so there is plenty of help on offer.

You could use it to analyse and study your posted circuit.
E

EDIT:
This is an example of your circuit, I have assumed the inductor values, post them if you know them.
Note the cut off off frequencies of the R/C filters.

Hi Eric, Thanks so much, I have downloaded LTSpice. I don't know the inductor value. I reckon to find it out I would have to de-solder them. Could you please direct me on how to measure the values of the cutoff frequency from the circuit (I mean using the actual circuit). Thanks a million again

 

hi,
I have attached the LTS .asc file so that you can download and run it.
On the LTS plot image, the Y axis is in dB's and the X axis in frequency.
Look for the points on the plot lines where the dB's fall by 3dB and 6dB
E

 

hi,
Look at this LTS method of stepping Cap values, note the effect.

 

how to measure the values of the cutoff frequency from the circuit (I mean using the actual circuit).

The best is to leave the speakers attached and measure the voltage across the speakers with a constant voltage AC signal over the frequency range of interest.
That will then include the speaker impedance, which can vary significantly with frequency.
Note that you need to use an oscilloscope or AC voltmeter that can go to the highest frequency for the measurement.

 

hi,
I have attached the LTS .asc file so that you can download and run it.
On the LTS plot image, the Y axis is in dB's and the X axis in frequency.
Look for the points on the plot lines where the dB's fall by 3dB and 6dB
E

Thanks downloaded and messing around

 

The best is to leave the speakers attached and measure the voltage across the speakers with a constant voltage AC signal over the frequency range of interest.
That will then include the speaker impedance, which can vary significantly with frequency.
Note that you need to use an oscilloscope or AC voltmeter that can go to the highest frequency for the measurement.

Totally newbie question...would I be able to use a sine wave from a function generator with the voltage at around 5 - 10Vpp(as an input) and measure the output using an oscilloscope?
If so would it be possible to use my audio interface output and input to see the cut- off frequency?
Here is the specs of my audio interface : -

Analogue Output Performance
Nominal Output Level 0dBFS = 10dBu, balanced
Frequency Response 20Hz - 20kHz +/- 0.2dB
THD+N -100dB (measured with -1dBFS input 20Hz-20kHz filter, un-weighted)
Output Impedance < 10Ω
Power Output into 150Ω 15mW
Power Output into 50Ω
30mW

 

hi Yami,
The signal generator and scope is often used in the way you describe in order to get the 3dB or 6dB points on the response range.
Use one scope channel for the Sig Gen output and the other channel across the speaker of interest.
Manually sweep the Sig Gen output frequency over the required range.
E

BTW:
Try not to overdrive the circuit, ie: keep the output sinusoidal.

Bedtime reading:
http://novo.press/understanding-speaker-specifications-and-frequency-response/

 

hi Yami,
The signal generator and scope is often used in the way you describe in order to get the 3dB or 6dB points on the response range.
Use one scope channel for the Sig Gen output and the other channel across the speaker of interest.
Manually sweep the Sig Gen output frequency over the required range.
E

BTW:
Try not to overdrive the circuit, ie: keep the output sinusoidal.

Bedtime reading:
http://novo.press/understanding-speaker-specifications-and-frequency-response/

Ah I see, I remember someone saying to me that it wasn't possible and I would have to use an amplified signal as the output from the function generator or my audio interface would not be enough to get a meaningful reading.

 

hi Y,
I am not sure what they mean by 'amplified signal', my SG is able to output a 10Vppk sine wave over the full audio range.
You have to decide if you want to measure the frequency response of the cross over network or the audio amplifier that drives the cross over circuit.
They could have very different response curves.

E

 

hi Y,
I am not sure what they mean by 'amplified signal', my SG is able to output a 10Vppk sine wave over the full audio range.
You have to decide if you want to measure the frequency response of the cross over network or the audio amplifier that drives the cross over circuit.
They could have very different response curves.

E

True makes sense alot now.

 

I am not sure what they mean by 'amplified signal', my SG is able to output a 10Vppk sine wave over the full audio range.

Yes, but it is probably not able to drive an 8 Ohm speaker to that level.

Bob

 

Totally newbie question...would I be able to use a sine wave from a function generator with the voltage at around 5 - 10Vpp(as an input) and measure the output using an oscilloscope?

Yes, but it likely wouldn't be able to provide that amplitude with a speaker load, as Bob noted.
You can try and see the maximum signal it can generate with that and if that's enough to get a measurable output at the desired frequencies.
The generator output will likely change with frequency, so you want to measure the generator output at each frequency to accurately determine the voltage drop through the filters.

But if you have an audio power amplifier you can use, that would likely simplify the measurement.

 

Yes, but it is probably not able to drive an 8 Ohm speaker to that level.

Bob,
As I understand the TS requirement, he wants to measure the cross over network frequency response, no mention of power ratings.
If the TS signal generator is not able to give a 10Vppk sinusoidal out as mine will, it is not a problem.
A much lower SG output level can be used to measure the cross over network frequency response.

As far I can see he has not stated any technical details of the speakers, other than one is a tweeter
No mention speaker impedance, speaker or amplifier power rating etc ....

If the TS finds that he needs more power drive, he can easily do as @crutschow as suggested
But if you have an audio power amplifier you can use, that would likely simplify the measurement.

E

 

Since the passive crossover has a second-order lowpass filter and a second-order highpass filter the phases of the signal to the tweeter and the signal to the woofer have opposite phase as shown in the graph from LTspice. If the tweeter and woofer are connected with the same phase then the sound will have a cancellation at the crossover frequency that sounds bad. If the tweeter and woofer are connected with opposing phase then both play the crossover frequency in phase causing a boosted level at the crossover frequency that sounds bad.

 

Bob,
As I understand the TS requirement, he wants to measure the cross over network frequency response, no mention of power ratings.
If the TS signal generator is not able to give a 10Vppk sinusoidal out as mine will, it is not a problem.
A much lower SG output level can be used to measure the cross over network frequency response.

As far I can see he has not stated any technical details of the speakers, other than one is a tweeter
No mention speaker impedance, speaker or amplifier power rating etc ....

If the TS finds that he needs more power drive, he can easily do as @crutschow as suggested
But if you have an audio power amplifier you can use, that would likely simplify the measurement.

E

Yes, I just wanted to check the cut off freq of the cross-over network, Some points I would like to clear though
1. If I connect the speaker (surely with 10Vppk would not drive the speakers) what effect would it have on the measured output from the oscilloscope. (not from a mic).
2.Can I measure across the output with the speakers connected?

Lets say that I wanted to amplify the signal, some question about the amplifier matching. Currently I don't have the amp which came with the speaker. The transducers (i.e the woofer or the tweeter) doesn't have information on it. But doing some online browsing I came across some technical spec for the speakers.

• Dimensions: 4.75 by 8.38 by 5 inches (W x H x D)
• Frequency response: 57 Hz to 30 kHz
• Recommended amplification: 10 to 150 watts <-------so can I use any amplifier which outputs this much of power??
• Efficiency: 89 dB
• Driver complement: 1 x 4.5" bass/midrange driver; 1 x 4.5" planar low-frequency radiator; 1 x 1" aluminum dome tweeter
• Magnetic shielding: Yes
• Finish: Gloss black

On a side note : Eventually the direction I want to go is try some cross over design and measure the phase response etc with the speakers connected. The signal chain would be
Cross-over ---> Low amp -----> Low driver
|
|
High amp ---> Tweeter
Any ideas/comments/advice on doing this would be greatly appreciated.
All this information is very valuable thanks all for your info and time really appreciate it.
Thanks
Yamin

 

1. If I connect the speaker (surely with 10Vppk would not drive the speakers) what effect would it have on the measured output from the oscilloscope. (not from a mic).
2.Can I measure across the output with the speakers connected?

hi Yami,
Q1,
The only reason I would connect a SG directly across the speaker is to check if the speaker works.
When you say 'measured output' do you mean maximum power or frequency range.

Q2,
Do you mean the output of the cross over network with the speakers connected.
In order to get meaningful measurement results the network needs to be loaded by the speaker impedance

It would help if your questions were more specific, so that we can give specific answers.

A point to be aware of is the claimed power rating of amplifiers and speakers.
Some claims are rated at Peak to Peak, Peak, RMS or Average.

As you may know woofers and mid range speaker impedances are typically 15R 8R 4R

Eric

 

hi Yami,
Q1,
The only reason I would connect a SG directly across the speaker is to check if the speaker works.
When you say 'measured output' do you mean maximum power or frequency range.

Q2,
Do you mean the output of the cross over network with the speakers connected.
In order to get meaningful measurement results the network needs to be loaded by the speaker impedance

It would help if your questions were more specific, so that we can give specific answers.

A point to be aware of is the claimed power rating of amplifiers and speakers.
Some claims are rated at Peak to Peak, Peak, RMS or Average.

As you may know woofers and mid range speaker impedances are typically 15R 8R 4R

Eric

Q1 I meant the frequency range.
Q2 So to check the frequency range I would have to connected the speakers?
Thanks

 

hi Y,
Yes, to measure the frequency response of the cross over network it should be connected to the two speakers.
Measure the input frequency amplitude and that across the speakers.
It does not have to be driven at full power.

Comparing the amplitude of input sinusoidal signal to the network to the signal amplitude across the speakers.
Plot the Freq versus Amplitude response, check for the Power drop at 3dB down [ 0.707 down]

Post your plot and we can advise.

E

EDIT:
Ideally the speakers should be mounted in their enclosures/cabinets.