revised schematic 4 EQlizer display

[MrDEB]

after several suggestions I think I have a workable design??
only portions of all the filters etc are showen but you can get the idea of whats going on.
display circuit is similar to:**broken link removed**
any advice would be helpful

 

[Nigel Goodwin]

An obvious possible problem is lack of brightness on the LED's, as you don't have any drivers from the LM3915.

 

[MrDEB]

any suggestions?

I kinda figured that the transistors/ Vcc would provide enought?
wouldn't the 3916 via the transistors from the 4069 be plenty or perhaps I should lower the value on the resistor array?
seeing how the the 4069's are sequencely turned on via the 4017 so only ten leds max per 4017 count are on.
Still need to calculate max current for entire eqlizer set up w/ displays.
going to try and be safe by using the current if ALL 200 leds are lit at once (stereo- 1 led array per channel is probally overkill but it would make one heck of a display)
20ma x 200 = 4 8,000 ma + all the op amps. Kinda alot of current?
should be safe if just going w/ 4,000ma power supply?
12v duel 4 amp supply?
not real sure??

 

[audioguru]

I see only 16 LEDs. Where are 200?
The max output of an LM39xx is about 30mA. If 10 LEDs are in parallel then some will be bright and some will be dim. The average current in each LED is only 3mA and the low duty cycle will make them look much dimmer.

If all outputs of an LM39xx have a current of 30mA then the LM39xx will melt if the LED supply voltage is higher than 6.3VDC.

 

[MrDEB]

Didn't draw all the leds

any suggestions on how to improve the brightness?
thought about a 2803 to increase the current?
just might go w/ smd LEDs.
guy on Ebay has some in colors we want-red, green, yellow.
Whats your opnion on using the 1n4148 diodes to isolate the outputs from the filters to the 4066 switches?

 

[audioguru]

I think each bandpass filter should have its own peak rectifier and LM3915.
Then each frequency column will show its own peak level and the 100 LEDs will be nice and bright.

 

[MrDEB]

would be easier but cost?

I would use 10 3916's but this is a boy scouts project that he is paying for.
need to check Ebay etc.

 

[MrDEB]

proposed circuit w/ 20 transistors

the circuit using 20 transistors etc costs around $5 per channel where the 3916's are $2.25+ per chip x 10
cost is a real factor as well as a challenge
need help on power supply=how much needed?

 

[audioguru]

The sample and hold capacitor in the peak detector needs to have a resistor to discharge it. Since you want the hold time to be long enough to be seen then the multiplex scanning rate must be much too slow, because you have only one peak detector and only one LM3916.

 

[MrDEB]

add more peak dectors or more peak dectors AND 3916's?

hopfully just one peak dector per filter but?
may just have to go w/ one display if we need one peak dector and one 3916 per filter.
this could get expensive
any comments on power supply, current needed etc?
remember I am basically a rookie, just know enought to get into trouble

 

[Nigel Goodwin]

hopfully just one peak dector per filter but?
may just have to go w/ one display if we need one peak dector and one 3916 per filter.
this could get expensive
any comments on power supply, current needed etc?
remember I am basically a rookie, just know enought to get into trouble

You can multiplex the display (as the Velleman one does), but you still need peak detectors for every channel.

 

[MrDEB]

just add peak dectors to each filter?

that sounds doable
question on a circuit I have been looking at
https://www.electro-tech-online.com/attachments/audio-spectrum-analyzer-en-pdf.15065/
this circuit has no peak detectors? bascally it looks like the Velman circuit?
just curious
If I add the peak detectors per filter do I need the resistor you were refering to?

 

[Nigel Goodwin]

that sounds doable
question on a circuit I have been looking at
https://www.electro-tech-online.com/attachments/audio-spectrum-analyzer-en-pdf.15065/
this circuit has no peak detectors?

What are D1 to D10 then?.

 

[MrDEB]

????Peak Detectors????

I am going by the peak detector that is showen on the 3915 data sheet fig 6 under helpful hints.
using the 1n4148 diodes (D1-D10 )are they peak detectors?

 

[MrDEB]

wish I could just chat with you on this circuit.
You and Audioguru seem to be the only knowlegable persons willing to help.
Thanks

 

[MrDEB]

oups the data sheet fig is 4

 

[audioguru]

The Velleman circuit uses an opamp in an active half-wave peak detector circuit per frequency channel to reduce the diode's loss of sensitivity at low levels. An active detector is accurate at low levels.

The Oriental circuit has just a diode charging a capacitor per frequecy channel (D1-D10) but the driving opamp has its bias voltage raised one diode voltage drop to cancel the voltage drop across the peak detector diodes. This circuit is not accurate at low levels but with an LM3914 with its narrow dynamic range (only 20dB) it won't be noticed.

 

[MrDEB]

going with lower cost cermic caps

talked with young scout who is going to build this eqlizer.
Mentioned that the sound using ceramic caps vers poly caps would not be noticiable unless you have test equiptment hooked up to unit.
going to redo the pc board running the power traces first then adding jumpers to the low signal traces.
now just need to contact the guy on Ebay w/ the smd leds and compute current draw

 

[audioguru]

Before you were making an audio spectrum analyser that shows the levels at 10 frequencies. It doesn't have audio going through it. You don't hear its output.

Poly caps are available inexpensively at 5% tolerance which is the accuracy that is needed. Ceramic capacitors are not available in many values that are needed and have very wide tolerances.

 

[MrDEB]

now Im confussed???

the outputs from the LM837 feed into the 3916. This is not audio??
I get from your answer that I should stick with the poly caps?
decoupling caps on power pins?I should include but his is a duel power supply. decoupling between the V+ and V-? or both to ground as the norm?
Another question arose = the 10K duel pots are on another board using 2- 10 pin ribbon cables. the ground is on a 4 pin cable. should I use any decoupling caps?
This audio stuff is confusing. I have dealt mostly with simple digital circuits.