I made myself a VU meter and amplifier with a 7.2V supply and discrete components. I plan to double the voltage as the discrete components can handle it.

Now for better lighting effects, I want to take the output of the VU meter and connect it to a CD4017 IC for like a special led effect on every music beat.

I checked the data sheet and It says the IC works up to 15-18V so I'm safe there, but what about LED output? Am I safe to connect the leds and 3.3K resistors in series and directly to the output pins? And what would the internal resistances be to the output pins from within the chip?

And is it normal if the IC warms up somewhat due to the high-ish voltage going through it?

I'm used to working with IC's on a 5V supply not on a 14.2V supply.

 

You don't say what the forward voltage of the LEDs are, so I will assume ~2 V. If your able to actually get 14.4 V out of the IC, then the resistor will limit the current to ~3.8 mA. That's pushing the chip pretty hard. The data sheet says that the minimum source/sink current is only 3.4 mA at Vcc = 15 V and that's allowing for a 1.5 V deviation in the output voltage. It will most likely work, because the typical limits are twice this.

Each output will likely be dissipating about 1.5 V * 3.3 mA ~= 5 mW, which is well below the 100 mW absolute limit. The total limit per package is 500 mW and even if you have all ten outputs on, you will be an order of magnitude below that.

Chips warming up is normal, but I wouldn't expect to see any significant heating of this part since the overwhelming majority of the power will be dissipated in your resistors and half to three-fourths of the rest in the LEDs.

I'm assuming you are going to be counting super fast, so the dynamic power dissipation shouldn't be too much, either. At a Vcc=15 V, you don't get to a dynamic power dissipation of 1 mW until you are above 20 kHz.

I'm not too happy about pushing the current drive specs as hard as you are, but I think for your application it should work okay. You might, however, see if you can find LEDs that are happy with 1 mA or 2 mA and use 6.8 kΩ to 8.2 kΩ resistors.

 

Provided that you keep below the 100mW power dissipation, you can use the Rds(on) of the 4017 as the current limit. If you put LEDs in series on the 4017 output with no limiting resistor, Rds(on) will limit the current. If the LED voltage >10V then the IC stays within its power limit.

 

I checked the data sheet and It says the IC works up to 15-18V so I'm safe there, but what about LED output? Am I safe to connect the leds and 3.3K resistors in series and directly to the output pins?

At 15V, the outputs will source 3.4-6.8mA (a little less at 14.4V), so 3.3k current limit resistors will be operating close to the limits of the counter's specs.

what would the internal resistances be to the output pins from within the chip?

From the typical specs, I calculate about 220 ohms.

is it normal if the IC warms up somewhat due to the high-ish voltage going through it?

It's the current that matters. Since only one output is being loaded at a time, power dissipation at 15V is minimal

I'm used to working with IC's on a 5V supply not on a 14.2V supply.

If you're working with TTL, higher than 5V isn't an option. Some of the CMOS 74 series will operate voltages higher than 5V, but not all.

 

The output current from an ordinary CD4017 is very low and its maximum allowed heating for each output is also very low.
Then the LED brightness is dim.

I made LED chasers powered from a 3V battery and a 6V battery and the LEDs are very bright because the 4017 is a CD74HC4017 that can easily produce output currents of 20mA into LEDs for high brightness when the supply is 5V or 6V. Do not exceed 6V.

The CD74HC family is high speed and high current Cmos logic.

 

It would be safest to use the 4017 outputs to drive an open-collector buffer (ULN2002 family for example) or a bunch of transistors (2N2222). Then the 4017 could run from 5V or so, and the LEDs could be strung in series and supplied from as many volts as necessary.

 

It would be safest to use the 4017 outputs to drive an open-collector buffer (ULN2002 family for example) or a bunch of transistors (2N2222). Then the 4017 could run from 5V or so, and the LEDs could be strung in series and supplied from as many volts as necessary.

Yes. But not "supplied from as many volts as necessary".

Also, Mosfet arrays can be used and won't require much current (microamps) from the CD4017.