LEDs in parralell with only one resistor

What is actually stopping me from doing this?

The 74HC595 is just a shift register.
I’d like to visualize what is being shifted out and preferably using a single LED.
I think the brightness will change depending on the number of leds on… but assuming that I’ve got enough ohms to hold all 8 I’m not going to blow them.


Hi Pix
OK if you have only one LED at a time. But if not they will just have to share the available current.

You will have to size the resistor to suit one LED and take what you get if more than one is on at once.

The application note suggests 560Ω per LED. Assuming a drop of 2V across each LED that is only 5.5mA. so I would check the IC source current capability. If you allow 10mA / LED you would need 300Ω

Cheers Bob


Thanks Bob.
Seems like this method is fine for testing but not for application production.
Not dangerous, just not ideal.

As usual @Robert93820 I really appreciate your experience.

PIx :heavy_heart_exclamation:


I tried this and have some thoughts for future users.

  1. It works and saves alot of time.
  2. As predicted by Bob above, more LEDS HIGH causes brightness of the set to drop.
  3. The single resistor value you use makes a real difference. You want it as high as you dare so the logarithmic effects of the distance square law kicks in and the differences in brightness is minimized.
  4. If you’re going to do this I recommend using the same color LED for all. Different colors already need different ohms and the added unevenness starts to make the product look “unfinished”.

This would be perfect if your goal is to maintain a consistent amount of lumens from the board (just changing where the light is coming from).

For the record, I even though I tried it I did end up choosing to give each of my LEDs it’s own resistor. The result was much better and allowed me to use many colours.

Pix :heavy_heart_exclamation:


Hi Pixmusix,

I once had a setup using an 8 segment LED using only one resistor. It worked fine.
One can actually buy 8 segmented displays that use one resistor.


The only issue was that depending on the amount LEDs lit the display would dim or brighten slightly.

I think some very old 80’s LED lit clock alarms used this method as a production cost saver. When you scrolled the time setting one would clearly see this phenomenon.

1 Like

Hi Eddie
That link clearly states that 8 current limiting resistors are required and the use of one common resistor is the INCORRECT way to use this display.

Not exactly a phenomenon. It is easily and clearly explainable.
Cheers Bob


Oh cool! :grinning:

1 Like

Hi Pixmusix,


Hi Eddie,

My favourite part about this is just how many alarm clocks you were able to source. It’s nice to have something already on hand to demonstrate what something


Hi Eddie
A few dead segments in that lot. Only the green one complete. The flickering is to do with the refresh rate and the movie frame rate being different or no mathematical relationship.
Cheers Bob


Very interesting indeed.

When I see it with my eye it all looks naturally constant.

I get back to you…

Hi Eddie

That is normal. Not how the camera sees it however. The human eye has persistence that a camera has not got. That is how you view a movie or TV picture without flicker.
Cheers Bob

Hi Eddie
Just actually ran those videos. There are no dead segments I don’t think. The segments that seem to be missing are just not on when that frame ws taken.
Cheers Bob

Hi Jack,

Absolutely. I’m not a hoarder, just saying.

Hi Robert,

The more I look into this the more questions appear.

I have been using a very reasonably priced smartphone to record these clips.

I have a few more here. The first few are slowed down.

Then I thought I could compare the LEDs to something else that uses LEDs. I have a battery pack with 4 blue LEDs and a USB red LED charging indicator.

The clock appears to behave unusually, however the LEDS from the battery pack and USB red LED charging indicator seem unaffected.

Hi Eddie
I don’t know how this post got to the pros and cons of digital clocks.
In an earlier reply you state

Which is what has been said all along as the situation when you use 1 resistor for multiple LEDs

Nothing unusual about it. The camera is catching the display at an instant that all of the required segment are not lit as the display is refreshed. The human eye does not see that flicker.

Because they have a steady DC voltage applied and are not refreshed periodically as are the clock displays.

Cheers Bob


I wonder what those clocks would look like when shot in high speed (slomo) on a phone capable of it, like a halfway recent Samsung.

Phenomenon != unexplained phenomenon.

All the things are phenomena…

Phenomenal, I know

1 Like

Sometimes I find these forum posts amusing in the twists and turns they take.
The original question by @Pixmusix has been answered.

The introduction of Digital Clock displays by @Eddie112471 is a different question.

These use a single chip that does everything and each segment is displayed individually at a rate the human cannot detect due to persistence of vision making it look like all segments are active at the same time. One load resistor for all segments and no change in brightness. Simple and effective.

The datasheets don’t go into the internal hardware design of the chips. but I remember studying how it could be done when I was learning electronics too many years ago now.
The chips are highly dedicated to being a digital clock and are not much use for anything else.

If I was @Pixmusix I’d look at 7 segment display drivers, you program it to display what you want and the chip does all the rest. I have successfully used these kind of drivers with just LEDs. The MAX6954 would be a good starting point. This can drive up to 128 LEDs.


EDIT: The device I used was a MAX7219 driving an 8x8 LED matrix controlled by an ATMega328P…


Hi James
You are right.
In my last post I did query how the discussion got around to digital clocks.
I think the thread got out of hand a bit.

Pixmusix does not want to drive a 7 segment display. All he wants to do is monitor the outputs of a shift register. Which as you say his question was answered some time ago.
Cheers Bob


Hi Kimmo,

Yes, indeed it would be very interesting to see.

I did run a slow motion capture on my very reasonably priced xiaomi m9 phone.

However, being a hobbyist and not really smart at that.
I can’t say for certain if I’m actually doing it correctly.

I did open up the old Goldair clock. Just because I’m curious as cats.

Yes James46717, I think you are right regarding the operation of the clock IC chip.

It’s all very dried up and the cables and connectors are a bit crispy. I main IC is a Texas Instruments (sorry I can’t seem to upload the photo). The LED display seems to go straight into the IC chip. Not many passive components about.

I looked up this datasheet…

Unfortunately it doesn’t have any circuit examples more a block diagram showing the method of operation.

It talks about using 2 parts of the AC cycle to alternative light the LED Display (page 10).
Could that be what in the above video…hmmmm…?

The TI Chip seems a common part taking on many variations in a plethora of LED clocks of the day. I think Sanyo made a variant.

Thank god this is only a pastime and not my day job!