Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

EQ / spectrum analizer revisited

Status
Not open for further replies.

MrDEB

Well-Known Member
After seeing a thread about the LM3914 bar/graph display and planing on etching a couple of boards I remembered that I still have a bunch of parts for the EQ / spectrum analizer unit I and a boy scout were designing .
Well to make a long story short, The scout is going to away to ISU come September.
I have decided to etch the boards for this project. But before I do I copied and pasted an abreviated schematic together (right channel, one frequency)
most if not all of this schematic has been reviewed by Audioguru etc. but for my piece of mind I hope I can get a quick review before I finalize the PCboard layouts.
THANKS!!
most of the lines are straight. Tried importing into PAINT but way to much erasing etc. unless there is an easier way??
 

Attachments

  • Capture3-4-2009-2.36.43 PM.jpg
    Capture3-4-2009-2.36.43 PM.jpg
    192.4 KB · Views: 253
I can't see your schematic with the crooked lines, odd font and text colliding with lines.
 

Attachments

  • Mr DEB.PNG
    Mr DEB.PNG
    14 KB · Views: 224
I will recheck and make changes

I tried your idea of MS PAINT but the editing using eraser etc.
 
I use MS Paint for schematics. Straight lines are made with the SHIFT key down.
The text is any font that I want. I select an ENGLISH font.
I place the text away from the lines so it can be seen.

Most schematics are made with copy and paste. Simple.
 
problems and more problems

the screen shot utility (hoversnap) says I havn't enought room to save screen shot (have LOTS of room 1 terabyte of room)
took and tried putting file in openoffice draw program to convert to pdf
not real clear pic.
printed then scanned then saved to jpeg(figure out what happened tomorrow)
here (hopfully) a good pic of schematic
will get back tomorrow morn--time for bed
 

Attachments

  • spectrum analizer.jpg
    spectrum analizer.jpg
    242.8 KB · Views: 207
looks good but its rotated.

to give MS PAINT another try or maybe the pdf version?
 
got the screen shot utility working

here is a screen shot with the corners in red dots, the boarder is red as well
I straighten up most if not all the lines.
this is only one of 5 filters (total of ten filters, 5 each channel)
went with ten of the 3916's for the displays. Multiplexing is a pain.
thanks for all your expert advice on this project.
 

Attachments

  • Capture3-4-2009-8.30.19 PM3-5-2009-5.25.34 AM.jpg
    Capture3-4-2009-8.30.19 PM3-5-2009-5.25.34 AM.jpg
    196.1 KB · Views: 195
You have most opamps as an inverting type that uses two extra resistors and has a very low input impedance which needs huge coupling capacitors for good bass response.
An example is the input with the inverting opamp U1 pins 1-3 with a 1k ohm input impedance. Then the 220nf value for C45 cuts 727Hz by -3dB and also cuts frequencies below 3635Hz.
If U1 pins 1-3 is a non-inverting opamp then its input impedance can be 220k ohms and then the 220nF input capacitor will cut 3.3Hz by -3dB and also cut frequencies below 16Hz.

Why is opamp U1 pins 8-10 an inverting type?
Why does it have a load resistor R5 as low as only 10k ohms?
Then coupling capacitor C1 cuts 73Hz by -3dB.

What is the function of U3 pins 5-7 and U4 pins 8-10? Why two opamps there?
Maybe one of them is the mixer for all the frequencies but the schematic shows no mixer.

The filter looks odd. I think its bandwidth is much too wide since it is a type that is used to boost or cut mid-frequency sounds in a bass, mid and treble tone controls setup.

The CD4066 has a very high supply voltage of 18V. They are spec'd at a max of only 15V.
 
been so long but I see what your talking about

I can't find alot of my bookmarks for the different aspects of this monster but found the mixer inputs and the biophonic filter circuits

the inputs U1 pins 1&2 are wired for a mic (my oups)
**broken link removed**

the filters are based on this-about half way down it shows what my circuit is
**broken link removed**

the op amps U3 5 & 7 were inserted why??. I recall something about isolating the filter outputs so the 3916's would work right?
U4 pins 8 & 10 I have no clue. something about gain?? I think.
will alter the mixer inputs.
 
As I suspected your filter has an extremely wide bandwidth and cannot be used.
The Headwize article shows that a 1kHz filter affects from 50Hz to 12kHz.
If you have 10 filters across the audio band then their bandwidth must be very narrow.

A boost of 20dB is too much for an equalizer. It is 100 times the power. Use a max of 10db for 10 times the power.
 
the 20 db boost?

I am assuming your refering to R9 on U4 pins 8 & 9?
only having 5 bands-64hz, 595hz, 1804hz, 8757hz, 18045hz
the thread on May26th 08 has a pic in red (url I got it from?) shows 10 filters like this on I am planning on building hopefully one day.
suggestions??
 
pic of planed filter

took a screen shot of planned filter
not sure of where I got this from
 

Attachments

  • 3-5-2009-2.20.19 PM.jpg
    3-5-2009-2.20.19 PM.jpg
    167.1 KB · Views: 192
The Headwize article shows that the bandwidth of the filter is way too wide.
The Q of a filter determines its bandwidth. That filter does not have an adjustable Q like the filters in equalizers and in audio spectrum analysers have.

Here are articles about equalizers that use Multiple-feedback Bandpass filters:
**broken link removed**
**broken link removed**
The Oriental equalizer circuit also used this filter.
 
schematic I am pretty sure I used

as a model for this EQ monster

I posted a screen shot of TINA simulation with 2 filters
the VF2 and VF3 are outputs to the peak detector/display circuits
hope this helps with why U4 - pins 8-10
U3 pins 5-7 is a buffer that was recommended by ? but could have been a leftover from when I was trying to multiplex?
could still be a buffer to seperate the filter outputs to display?
 

Attachments

  • 3-5-2009-8.16.17 PM.jpg
    3-5-2009-8.16.17 PM.jpg
    311.8 KB · Views: 193
  • 3-5-2009-8.16.33 PM.jpg
    3-5-2009-8.16.33 PM.jpg
    318.4 KB · Views: 179
larger screen shot of filter

posted in May of 2008
this is basic filter design.
eliminated R4 and C3. neither did anything to wave form.
pic of revised input as suggested (my oups, used mic inputs)
will run simulation with out U4 pins 8-9
 

Attachments

  • 3-5-2009-8.31.05 PM.jpg
    3-5-2009-8.31.05 PM.jpg
    159.5 KB · Views: 162
  • 3-5-2009-8.34.05 PM.jpg
    3-5-2009-8.34.05 PM.jpg
    179.8 KB · Views: 169
Here are my comments:
 

Attachments

  • equalizer and spectrum analyser1.PNG
    equalizer and spectrum analyser1.PNG
    32.1 KB · Views: 185
I tried omitting R15

the 47K resistor at the input R15 (R1 in TINA)
the wave form peak drops slightly?
Vf5 is showing what happens when I take the last op amp out. unless I need to actually remove?
don't know if these wave forms help or hinder??
 

Attachments

  • 3-6-2009-5.42.57 AM.jpg
    3-6-2009-5.42.57 AM.jpg
    189.4 KB · Views: 184
  • 3-6-2009-5.43.11 AM.jpg
    3-6-2009-5.43.11 AM.jpg
    180.2 KB · Views: 161
Woh Audioguru you are unbeliveable

I removed U4 pins 8-9 and Vf5 now has some peaks!!
If I remove U3 pins 5-7 won't the display circuit mix the filter outputs all together?
U3 5-7 acts as a buffer I recall?
 

Attachments

  • 3-6-2009-5.54.08 AM.jpg
    3-6-2009-5.54.08 AM.jpg
    310.7 KB · Views: 173
The 47k resistor R15 on the schematic forms an attenuator when it feeds the 10k volume control. You don't want an attenuator.

The frequency response curves in TINA have an extremely large linear voltage range instead of having a reference and a range that is measured in decibels so nobody can see exactly what is happening. Also, we do not know what are the settings of the EQ controls.

Your schematic does not have the mixer circuit you added to the TINA simulation.

The first curves have no low frequencies and the last curves have plenty of low frequencies. I don't know what you did to change them since they should be the same.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top