Go to Mouser or Digikey to download spec sheets. In general green, yellow and red leds fwd voltages are around 1.7 volts. Blue and white are around twice that. If a chip drives them you need to have a current limiting R. The LM3914,5 has current regulation built in if memory serves.
LM339 and a simple 4 segmented VU meter
- Thread starter Potato Cakes
- Start date
Help Support GroupDIY Audio Forum:
No need for a limiting R as the current is set by the constant current source of Q1, R13, R14 and R15. Again the supply voltage and LED specs are needed in order to set the resistor values. LED specs include not just the voltage but also the current.
luckily you have internet access so your should be able to find data on components like that pretty easily.
JR
Well JR, I don't consider it my job to go hunting for specs or completely design the solution. Besides, there are as many LEDs as there are solutions so guessing at the specs and the supply voltage is not very helpful. This is Electronics 101 and I hope to point the poster in the right direction and help them learn from the experience. If the OP was to post back asking questions about my schematic and provide some insight as to the items I had listed as missing then I could help them out.
sorry it sounded like you were expecting Ian to do that....
Complaining about free advice is rarely productive.
JR
Potato Cakes
Well-known member
Thanks for everyone's input and assistance.
I wired up the 4 LED circuit I last posted and after overcoming several dyslexic wiring episodes (and blowing up the LEDs) I can confirm it does work but with a bipolar power supply as pointed out by Ian. The project that would use this meter runs off of +/-24V and I thought it would be simpler to have to meter section run off of 24V/0V and be within the voltage operating range of the op amp and comparators. The LEDs being used are these:
https://www.mouser.com/ProductDetail/Lite-On/LTST-C230KRKT?qs=62Z1DqX9N4RCGhBcfl7T/g==
The forward current is 20mA and forward voltage is 2V.
TJ, I thank you for working up the schematic but to fit the meter circuit within the board constraints for this project I will wind up using the last schematic that I posted which I have verified works. I'll just use some small, low current regulators to get the voltage down to either +/-15V or +/-12V.
Thanks!
Paul
I wired up the 4 LED circuit I last posted and after overcoming several dyslexic wiring episodes (and blowing up the LEDs) I can confirm it does work but with a bipolar power supply as pointed out by Ian. The project that would use this meter runs off of +/-24V and I thought it would be simpler to have to meter section run off of 24V/0V and be within the voltage operating range of the op amp and comparators. The LEDs being used are these:
https://www.mouser.com/ProductDetail/Lite-On/LTST-C230KRKT?qs=62Z1DqX9N4RCGhBcfl7T/g==
The forward current is 20mA and forward voltage is 2V.
TJ, I thank you for working up the schematic but to fit the meter circuit within the board constraints for this project I will wind up using the last schematic that I posted which I have verified works. I'll just use some small, low current regulators to get the voltage down to either +/-15V or +/-12V.
Thanks!
Paul
You can adapt the circuit in the first post to work with a single rail PSU. But you’ll have to recalculate resistors ladder for LEDs to show you correct values. Ian is spot on about biasing the op-amps at 1/2 supply. So your 1/2 V is your point of reference for the ladder.
mobyd
Well-known member
That's true from an ultimate sharp pencil analysis but ignores the NRE (non-recurring engineering) cost of programming a MCU. I am an old analog dog who went over to the dark (digital) side only this century, but one of my projects was a master section meter for a well respected (APB) analog console. My PIC based meter design was cheaper than the previous analog meter circuitry and added extra features (like simultaneous Peak and VU).
JR
PS; I used to have people get angry with me for saying I could do that cheaper with a MCU, but now decades later is stirs up less pushback...
mobyd
Well-known member
Re the NRE costs, once you've done it, you've done it and after a few hundred iterations, those costs disappear into the long grass. Have you seen the price of LM391x these daysThat's true from an ultimate sharp pencil analysis but ignores the NRE (non-recurring engineering) cost of programming a MCU. I am an old analog dog who went over to the dark (digital) side only this century, but one of my projects was a master section meter for a well respected (APB) analog console. My PIC based meter design was cheaper than the previous analog meter circuitry and added extra features (like simultaneous Peak and VU).
JR
PS; I used to have people get angry with me for saying I could do that cheaper with a MCU, but now decades later is stirs up less pushback...
(not that I've ever liked them that much).
I never used a 391x in a commercial design but over the decades I have designed scores of LED meters using everything from op amps, comparitors, and even PIC mcus.Re the NRE costs, once you've done it, you've done it and after a few hundred iterations, those costs disappear into the long grass. Have you seen the price of LM391x these days
I won two US patents (US04166245 Roberts, US05119426 Roberts) related to simultaneous peak/VU LED displays. For chuckles I prototyped a LM3915 (the 3dB step version) with multiplexed dot/bar display. In other words I tricked the LM3915 to simultaneously display Peak and VU..This involved separate peak and average input signal capture and electronic switching between the two inputs.
Then I moved on because the LM391x series was just not cost effective.
JR
The LM3915 in a DIP 18 pin pkg has retired. Obsolete. But you can find them dirt cheap on Amazon. The smt package is available.
Last edited:
ruffrecords
Well-known member
They are still in production in China , branded HRF. I just bought some on eBay. Not tested them yet.
Cheers
Ian
