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.

Problem with chained LM3915

Status
Not open for further replies.
fezder said:
i still have that oscilloscope, remember that one which we shot with freezer-spray? I totally forgot that, suppose i could start redoing something with it, only i'm not sure about ''necroing'' thread, but then again, it was confusing thread

Maybe start all over, but use links to the pics posted in the old thread with a summary That repair is particularly unusual. I have a few repairs to do myself.

October/November is a busy month for me because of yard work. I typically get 40-50 bags of leaves from essentially one tree. Most drop by the end of November. Some hang on till February. I got two bags today. Then there is another clean-up in the spring. The tree is easily 80 feet high and 54 years old. It's home to a few squirrels. I haven't measured the circumference lately Hand to hand won't wrap around 1/2 the tree.

I have some minor roof repairs to complete, Glue the corners of some asphalt shingles on the roof ridge vent. I did one side of the vent already.
 
Last edited:
Hmm, i now noticed something that's not intended behaviour: bar has less leds lit in peak-mode, shouldn't it detect most-highest peaks? instantaneus-mode has more
 
Hmm, opamp should have gain of 1, as in your peak circuit, but in mine i measured with generator:
input:416mv
output:820mv
so somehow i have gain messed up or something?

i don't have 2n3904 so i used bc547
 
oops, i measured opamp gain totally wrong, it is correct
 
Here's shots, both sections go via opamp amplifier
so, yellow is generator, blue is opamp amplifier input, also other 3915's input
peak: NewFile3.jpg
instant:NewFile4.jpg
 
hehe, feeling quite dumb here now, i placed 330r resistor in place of 330k in paraller with c8, well now it works! sorry for false alarm, quite EE hobbyist i am -_-
 
Only now there's another problem, lower set, -30db is lit even without signal applies, checked with scope and there is 1.2v dc bias so no wonder they are lit. Placed capacitor, dc is gone, but now it behaves even more strangely: at first, leds light like they should, but when second set starts to light, lower ones start to fade?

so, instant-mode works, but when i put peak-circuit, no matter before or after amplification, things go sour, only first few leds light. I don't have virtual ground anywhere
and both amplification and peak-detection is built in same lm358
what is strange also, when i switch that 330k to 330r, it works better in peak-mode, more leds are lit?
 
Last edited:
The 330k resistor being 330 ohms explains the very high ripple and quick release time shown on the 'scope.
The fairly noisy LM358 might be causing the -30dB LED to be lighted all the time.
Yes, please post your schematic.
 
sorry guys, forgot that schematic earlier in a hurry
here's what works fully as intended, even with speaker hooked:
Lm3915_-60dB_instant.jpg
Then these variations won't work, first peak detetor, from there to amplifier, some leds lit even when signal is grounded (dc in amplifier output)
lm3915_peak-amp.jpg
then backwards, first amplifier, then peak, works but only first two leds light in first, lower set of leds, nowhere near instantaneus-mode (only other lm3915 under test this way, i have only one lm358 on board now).
lm3915_amp_peak.jpg
 
You made a few mistakes with the peak detector that I corrected for you.
Your reference voltages are only 2.1V which is a low sound level in a speaker (the 20th LED will light when an 8 ohm speaker power is only 0.28W). Then the 1st LED lights when there is no input because it will show the noise and the input offset voltage of the peak detector opamp. With the 20th LED lighting at only 2.1V then the 1st LED lights with only 3.0mV. If you change the voltage reference to 10V (6.3W into 8 ohms) then the 1st LED will light when the input signal or the input offset voltage of the opamp is 14.1mV. The max input offset voltage of an LM358 is 7mV.
 

Attachments

  • peak detector.png
    peak detector.png
    28.5 KB · Views: 182
Thanks for pointing stuff, again, out guru! that ''this must be grounded....i have it grounded, but missed it when drawing schematic. I'll try those what you corrected. I never listen more than 1w at home even when alone, so i adjusted reference voltage for that. But, i'll try those.....
 
Your second schematic correctly shows the peak detector feeding the 30dB amplifier and the SIG input of the high levels LM3915. But first fix the peak detector.
Your 3rd schematic is completely wrong and the opamp will be damaged by the sound signals feeding its input way below 0V.
The 30dB amplifier is supposed to amplify the output of the peak detector (not amplify the signal) so that the first LM3915 shows low levels.
 
Your 3rd schematic is completely wrong and the opamp will be damaged by the sound signals feeding its input way below 0V.
ouch, good point...well, luckily i have spare ones, no wonder that didn't work then
 
Hi guru, sorry to keep you waiting.....had trouble with car (cooler fan started at same time when engine starts, fault was coolant temperature sensor which controls fan via ECM)
But, back to this one, i have now all those resistors value of 4.7k (r7, r9, r10), and here's shot
blue=input from generator
yellow=output from peak detector
ac-coupling set at scope, with dc coupling, yellow shows dc-offset
and, only peak-detector now in use here, changed opamp just in case
NewFile5.jpg
 
and i made mistake in drawing that schematic earlier, i mean other that that grounding: r9, r10 were 100k (seems copy-paste got me when i coped resistors....so it was complete replica from what you posted: r7, 9, 10 were 100k)
 
But, just to clear things out, i now built that your original version, and, no idea why, but output is quite noisy to say the least....
peak-detector-png.95063

NewFile6.jpg
 
It would be better to Build this Peak Detector.
A TL071 or TL081 can be used, But it Does require a Plus / Minus power supply

PEAK-2.PNG
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top