Author Topic: Making an analog bar graph voltmeter 1-10V using 10 LEDs  (Read 2292 times)

0 Members and 1 Guest are viewing this topic.

Offline BeaminTopic starter

  • Super Contributor
  • ***
  • Posts: 1567
  • Country: us
  • If you think my Boobs are big you should see my ba
Making an analog bar graph voltmeter 1-10V using 10 LEDs
« on: July 03, 2019, 01:58:36 pm »
Is it possible to make a bar graph style volt meter with LEDs by attaching a transistor to each LED with the Base resistor having high and higher values to get the transistors to turn on the LEDs as the voltage rises? So the first LED would have the smallest resistor allowing the transistor to conduct at 1 volt then higher value for 2 v etc etc. Would this waste power when not in use or only use power proportional to the LEDs being lit?
Max characters: 300; characters remaining: 191
Images in your signature must be no greater than 500x25 pixels
 

Offline Audioguru

  • Super Contributor
  • ***
  • Posts: 1507
  • Country: ca
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #1 on: July 03, 2019, 02:26:00 pm »
A single resistor biasing the base of a transistor will turn on some transistors but not other transistors, even if they have the same part number because transistors have a wide range of current gain.
Also, temperature affects transistors, heat turns them on and cold turns them off.

A transistor turns on gradually, not suddenly unless it is in a circuit that switches it on and off quickly at a certain voltage. The circuit is called a comparator.

An LM3914 IC has a voltage divider feeding 10 comparators and it is made to do what you want, making an LED voltmeter graph. Sorry, they are obsolete and are not made anymore.
 

Offline sokoloff

  • Super Contributor
  • ***
  • Posts: 1799
  • Country: us
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #2 on: July 03, 2019, 02:45:57 pm »
TI still shows them as an active and in-stock part:
http://www.ti.com/product/LM3914
 

Offline Siwastaja

  • Super Contributor
  • ***
  • Posts: 9336
  • Country: fi
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #3 on: July 03, 2019, 02:52:53 pm »
LM3914, an absolute classic, works fine if you don't mind their design bug which causes one of the LEDs to be on at very low current all the time, annoyingly visible with modern high-efficiency LEDs.
 

Offline Audioguru

  • Super Contributor
  • ***
  • Posts: 1507
  • Country: ca
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #4 on: July 03, 2019, 03:11:19 pm »
The LM3914 is available today in a tiny surface-mount package that can be soldered in an oven by a robot.
The datasheet says that a 10k resistor can turn off the LED that is dimly on all the time.
 

Offline sokoloff

  • Super Contributor
  • ***
  • Posts: 1799
  • Country: us
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #5 on: July 03, 2019, 04:03:42 pm »
The LM3914 is available today in a tiny surface-mount package that can be soldered in an oven by a robot.
If you need just a couple, they're available in a breadboard-friendly DIP package or you can use a PLCC-20 socket to DIP adapter
If you need 100s, they're available in a PnP-friendly PLCC package.

Seems like you're covered either way...and it's surely less work than rolling your own from scratch, right?
 

Offline BeaminTopic starter

  • Super Contributor
  • ***
  • Posts: 1567
  • Country: us
  • If you think my Boobs are big you should see my ba
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #6 on: July 08, 2019, 04:34:23 pm »
The LM3914 is available today in a tiny surface-mount package that can be soldered in an oven by a robot.
If you need just a couple, they're available in a breadboard-friendly DIP package or you can use a PLCC-20 socket to DIP adapter
If you need 100s, they're available in a PnP-friendly PLCC package.

Seems like you're covered either way...and it's surely less work than rolling your own from scratch, right?


That seems like a fun and unnecessary addition to add to projects. But doing it the hard way is sometimes my goal as it requires more learning. I don't mind if the leds get brighter or are you saying all the transistors would have to be matched otherwise they won't light in order? Or is this only an issue if you want them to turn on successively with sharp cut offs?
Max characters: 300; characters remaining: 191
Images in your signature must be no greater than 500x25 pixels
 

Offline GerryR

  • Frequent Contributor
  • **
  • Posts: 256
  • Country: us
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #7 on: July 08, 2019, 05:21:31 pm »
LM3914, an absolute classic, works fine if you don't mind their design bug which causes one of the LEDs to be on at very low current all the time, annoyingly visible with modern high-efficiency LEDs.

Not a design bug; see the attached data sheet.  It was done purposely to avoid ambiguous displays.  I recently used one of these to make a stepped, variable gain, amplifier.  Worked out pretty well.
Still learning; good judgment comes from experience, which comes from bad judgment!!
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #8 on: July 08, 2019, 06:16:25 pm »
That seems like a fun and unnecessary addition to add to projects. But doing it the hard way is sometimes my goal as it requires more learning. I don't mind if the leds get brighter or are you saying all the transistors would have to be matched otherwise they won't light in order? Or is this only an issue if you want them to turn on successively with sharp cut offs?

I made something very much like you described in my very early teens as a "VU meter", and it does kind of work if your expectations are not high.  The LEDs do not turn on and off sharply and as the input voltage is increased and more LEDs turn on, the already lit LEDs get brighter and brighter.  Since you are using the Vbe drop of a transistor as a reference, the circuit will behave differently at high and low temperatures.  The change in brightness can be somewhat mitigated by a common resistor on the anode side of the LEDs which will drop more voltage as more LEDs get lit.  However if you want something to reliably indicate a voltage rather than a flashy LED toy then forget it.
 

Offline mariush

  • Super Contributor
  • ***
  • Posts: 5170
  • Country: ro
  • .
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #9 on: July 08, 2019, 07:23:54 pm »
Considering that Microchip sells a PIC programmer for around 15$ and you can buy 16pin PIC microcontrollers for under 1$ ... seems like a no-brainer to use a microcontroller to replace old chips or lots of transistors/comparators.
They have build in oscillators, voltage references, so it would be enough to just use a voltage divider to get the voltage under 4.096v or 5v and you can use the output pins of the micro to power the leds. Some micros can do 25-50mA per pin, and something like 150mA in total.. so you can still use resistors for each led or you could pwm each led to reduce brightness or maintain uniform brightness on all leds depending on how many leds are on or off at one time.
 

Offline StuartA

  • Regular Contributor
  • *
  • Posts: 103
  • Country: gb
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #10 on: July 09, 2019, 04:13:23 am »
The Velleman company make a kit for an audio VU meter using LED's and dicrete componenets, see http://www.velleman.co.uk/contents/en-uk/p41.html The manual for it shows the circuit diagram (which is probably copyright, so not shown here). That might form the basis for what you want. I have built that kit and it works well.

S
 

Offline mikerj

  • Super Contributor
  • ***
  • Posts: 3382
  • Country: gb
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #11 on: July 09, 2019, 03:13:53 pm »
Nowadays the cheapest way would be to use a cheap MCU (with an IO-based dual slope ADC, no hardware ADC/DAC/COMP/PWM needed) to implement LM3914.

As much as I feel a little sad (or maybe just nostalgic) that small micros have replaced so many functions these days I have to agree, even micros with built in 8-10bit ADC are very cheap.  This also gives the options of features such as peak hold, custom scaling etc.
 

Online ledtester

  • Super Contributor
  • ***
  • Posts: 3282
  • Country: us
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #12 on: July 10, 2019, 06:23:29 am »
In the January 1980 issue of Popular Electronics Forrest Mims describes how he used a simple chain of LEDs to get a bar graph display of a voltage.

See pp 77-80 (starting on page 73 in the PDF) in:

https://www.americanradiohistory.com/Archive-Poptronics/80s/1980/Poptronics-1980-01.pdf

Update: The article mentions a "Gregory Kovacs, a student at Eric Hamber Secondary School in Vancouver, British Columbia" who built a oscilloscope based on the ideas in the article and it even had  a storage capability. A quick google search reveals the person is most likely this Professor Emeritus at Stanford:

https://engineering.stanford.edu/people/gregory-kovacs
« Last Edit: July 10, 2019, 06:36:08 am by ledtester »
 

Offline shanezampire

  • Contributor
  • Posts: 19
  • Country: us
Re: Making an analog bar graph voltmeter 1-10V using 10 LEDs
« Reply #13 on: July 10, 2019, 06:35:52 am »
Is that the same thing as something like - https://www.vellemanusa.com/products/view/?country=us&lang=enu&id=350546

The leds light as more power is given























 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf