LM3914 LED bar driver

[NHSA]

Hi,

After getting my amplifier / fullwaverectifier circuit more or less running

https://www.eevblog.com/forum/beginners/problem-with-full-wave-rectifier-using-tlp082/

I now added a LM3914 and a 10 LED bar in order to signalize the vibration.

Now I have searched everyhwere for a LM3914 calculator in order to calculate the R1 and R2 resistor. Found this one  https://www.electro-tech-online.com/tools/LM3914V2.php but what I practical see to the theory of the calculator does not fit ...

Taking the formula from the TI datasheet " Ref Out V = 1.25*(1+(R2/R1)) is a bit better but only when supplying a voltage below 12 V. As I want to drive my amplifier/full wave rectifier circuit as well as the LM3914 LED bar driver with 18 Volt evrything look different. The formula from TI does not take into consideration the supply voltage while the calculator does.

Anybody knows of a better calculator or is this a Trial & Error exercise   

Thanks in advance

regards Rainer

[StillTrying]

The V+ supply doesn't come into it.

You have to decide on the LEDs current first which gives you the value of R1.

Then you can work out R2, which raises the whole 1.25V Ref up to the voltage you want for RHI.
What RLO and RHI voltages do you want.

Page 9 on the data sheet. https://www.ti.com/product/LM3914

[David Hess]

Like StillTrying writes, the LM3914 datasheet has everything required and the calculations are simple and reliable.

[NHSA]

Hi,

Thanks both

Rainer

[NHSA]

Hi,

Still trying to get a result when driving the LM3914 with 18V.

The circuit depicted in this message https://www.eevblog.com/forum/beginners/problem-with-full-wave-rectifier-using-tlp082/

is powered with +- 9V and GND for the operational amplifiers TL082.

Now the problem is with the LM3914 which allegedly works up to 18 volt but when I connect it to the 18V power supply I do not get any signal. All Leds are lit up.

Powering it with half the voltage e.g. +9V and the middle ground it works trouble free.

I am using a 3K3 for the LED current (pins 7 to and a variable potentiometer 10K for adjusting the voltage (pin 8 to GND) which works OK for adjusting the sensitivity.

Still trying to find out why it does not work with 18V. 

[Benta]

From the references to other threads, you're NOT connecting the LM3914 to +18 V, but to +/- 9 V. That's not the same. As you've seen for yourself, it works when connected to 0 and +9 V, which is exactly as it should be.

[StillTrying]

What Benta said.
If you're connecting the LM3914's GNDs to -9V, the SIG input would now appear to be at +9V to the LM3914, so all LEDs on.

If in bar mode 18V would also give a dissipation problem when nearly all the LEDs are on.
Is there a good reason why it can't be powered from the 0V and +9V.

[NHSA]

From the references to other threads, you're NOT connecting the LM3914 to +18 V, but to +/- 9 V. That's not the same. As you've seen for yourself, it works when connected to 0 and +9 V, which is exactly as it should be.

Hi Benta,

Did not think about that. Another lesson learned.

Thanks   

Rainer

[NHSA]

What Benta said.
If you're connecting the LM3914's GNDs to -9V, the SIG input would now appear to be at +9V to the LM3914, so all LEDs on.

If in bar mode 18V would also give a dissipation problem when nearly all the LEDs are on.
Is there a good reason why it can't be powered from the 0V and +9V.

Hi,

Well my reasoning is that one battery will drain sooner then the other ¿ or ?

My thinking was to go down from 18V (2X 9V square battery) to a 12 V power brick and after reading Benta's answer there is then the same problem.

Interesting Stuff.   

Thanks Rainer

[StillTrying]

Well my reasoning is that one battery will drain sooner then the other ¿ or ?

Yes, but running the LM3914 off +/- 9V wouldn't save any of the current from the +9V at all, at least with the -9V now being little used not lighting the LEDs you could swap the +9V and -9V batteries half way though their life.
And the LM3914 should run cooler having to drop only 7V on the LED outputs rather than 16V.

I've got a few rechargeable full voltage PP3s 8.3V to 9.8V, the lower voltage down to 7.2V versions probably have higher mA hours and would do for your TL op amps.

[Old Printer]

I recently had a project where the LM3914 looked like a good candidate, I was making a VU meter for a small audio amplifier. I was disappointed to find that the DIP packaged version had been discontinued and it was SMD only from the large vendors, DigiKey Mouser etc. That put me in the realm of suppliers on ebay and the like where i am always wary of counterfeit chips. Were you able to find a reliable source for good DIP packaged chips, or did you use the SMD versions?

[David Hess]

I was disappointed to find that the DIP packaged version had been discontinued and it was SMD only from the large vendors, DigiKey Mouser etc.

Even worse TI now only produces it in the 20 pin PLCC package.  Was there no SO (small outline) package they could have used?

[Zero999]

I don't think I would bother with LM3914. Nowadays, I'd use an MCU, or if I really had to go analogue: 3 quad op-amps/comparators, a cheap reference, such a TL431 and a butt-load of resistors.

[NHSA]

I recently had a project where the LM3914 looked like a good candidate, I was making a VU meter for a small audio amplifier. I was disappointed to find that the DIP packaged version had been discontinued and it was SMD only from the large vendors, DigiKey Mouser etc. That put me in the realm of suppliers on ebay and the like where i am always wary of counterfeit chips. Were you able to find a reliable source for good DIP packaged chips, or did you use the SMD versions?

Hi,

I have some here so I do not plan to buy them new right now.

regards Rainer

[NHSA]

I don't think I would bother with LM3914. Nowadays, I'd use an MCU, or if I really had to go analogue: 3 quad op-amps/comparators, a cheap reference, such a TL431 and a butt-load of resistors.

Hi,

I still have some LM3914 laying around so it was logical that I wanted to use them.

Using op amp instead of LM3914 means 1 op amp for every Led and so I would need 10 additional op amps for my 10 LED bar. Lots of resistors and lots of space. Anyhow interesting idea for a future project.

May I ask which MCU ¿?

Thanks Rainer

[Zero999]

I don't think I would bother with LM3914. Nowadays, I'd use an MCU, or if I really had to go analogue: 3 quad op-amps/comparators, a cheap reference, such a TL431 and a butt-load of resistors.

Hi,

I still have some LM3914 laying around so it was logical that I wanted to use them.

Using op amp instead of LM3914 means 1 op amp for every Led and so I would need 10 additional op amps for my 10 LED bar. Lots of resistors and lots of space. Anyhow interesting idea for a future project.

The TL084 & LM324 have 4 op-amps, so 3 ICs would give 12 op-amps, 10 of which can be used as comparators, with the remaining two as op-amps for the rectifier. They're much cheaper than the LM3914.

May I ask which MCU ¿?

Thanks Rainer

Any MCU with an ADC input and at least 10 output pins for the LEDs. Here's a link to a bargraph meter with 20 LEDs, using the PIC16F686.
http://www.electronics-lab.com/project/bargraph-voltmeter-0-5v-dc-range-pic16f686/

[NHSA]

I don't think I would bother with LM3914. Nowadays, I'd use an MCU, or if I really had to go analogue: 3 quad op-amps/comparators, a cheap reference, such a TL431 and a butt-load of resistors.

Hi,

I still have some LM3914 laying around so it was logical that I wanted to use them.

Using op amp instead of LM3914 means 1 op amp for every Led and so I would need 10 additional op amps for my 10 LED bar. Lots of resistors and lots of space. Anyhow interesting idea for a future project.

The TL084 & LM324 have 4 op-amps, so 3 ICs would give 12 op-amps, 10 of which can be used as comparators, with the remaining two as op-amps for the rectifier. They're much cheaper than the LM3914.

May I ask which MCU ¿?

Thanks Rainer

Any MCU with an ADC input and at least 10 output pins for the LEDs. Here's a link to a bargraph meter with 20 LEDs, using the PIC16F686.
http://www.electronics-lab.com/project/bargraph-voltmeter-0-5v-dc-range-pic16f686/

Hi Zero999,

Thanks. I will think about the opamp solution which is somthing new, but now ¿ where is the fun by buying ready made stuff ?   

As I mentioned earlier it is also a question of the size of the ready made enchilada.

Thanks again and regards

Rainer

[Zero999]

I don't think I would bother with LM3914. Nowadays, I'd use an MCU, or if I really had to go analogue: 3 quad op-amps/comparators, a cheap reference, such a TL431 and a butt-load of resistors.

Hi,

I still have some LM3914 laying around so it was logical that I wanted to use them.

Using op amp instead of LM3914 means 1 op amp for every Led and so I would need 10 additional op amps for my 10 LED bar. Lots of resistors and lots of space. Anyhow interesting idea for a future project.

The TL084 & LM324 have 4 op-amps, so 3 ICs would give 12 op-amps, 10 of which can be used as comparators, with the remaining two as op-amps for the rectifier. They're much cheaper than the LM3914.

May I ask which MCU ¿?

Thanks Rainer

Any MCU with an ADC input and at least 10 output pins for the LEDs. Here's a link to a bargraph meter with 20 LEDs, using the PIC16F686.
http://www.electronics-lab.com/project/bargraph-voltmeter-0-5v-dc-range-pic16f686/

Hi Zero999,

Thanks. I will think about the opamp solution which is somthing new, but now ¿ where is the fun by buying ready made stuff ?   

As I mentioned earlier it is also a question of the size of the ready made enchilada.

Thanks again and regards

Rainer

Who said anything about ready made stuff? Check out the link I posted again. It's someone's bargraph display project. It looks like it's bought off the shelf, because they got the PCB made professionally. I consider using an MCU more of a challange in this case. Where's the fun in taking an IC, copying the schematic off the data sheet and soldering it to a board?