555 timer output

I set up a 555 timer using a .01uf capacitor, a 1ko resistor and a potentiometer to help tune the 555 to a frequency of 38kHz.

That all works well - however my output seems to fluctuate. What will happen is I can generally tune it to around 38kHz - like 37.5kHz. The signal will start off there, but slowly grow, I've watched it go up to about 38.5kHz.

What would be the reason for this? I am thinking it is the potentiometer, or is my setup just not grounded properly.

What is the most reliable way to get a steady output from the 555 timer or is it always going to fluctuate?

Would using a .001uf capacitor help?

I'm assuming you're using that nominal 38kHz for an IR carrier?

I found that my 38kHz TSOP receiver responds with no problem between about 35 and 40kHz, so unless your fluctuation is causing a problem with the receiver, I wouldn't worry about it.

Never seen this happen in my experience.

How are you measuring the frequency?

Is this a CMOS 555?

What duty cycle do you have?

Jim I was actually using your PDF as some guidance :slight_smile: Think I am going to steal your idea of 2 555 timers to modulate the signal so I know it is the beacon. But yeah, the slight variation shouldn't make a difference from what you have told me and my own few experiments, however I would like to get a steady frequency.

Larry - I am using a multimeter set to frequency. Don't have an oscilloscope

NE555 -not sure if that is CMOS or not with a duty cycle of 63.2%

What is the output current/load. Does the IC heat up.
Is the supply reasonably constant, and more than 4.5volt.
Film or ceramic timing capacitor.
With or without control voltage capacitor.
Did you use any supply decoupling.
The last IR transmitter I made with a Cmos 555 only had few hundred herz drift.
AFAIK, the duty cycle should be 50% for maximum range.
A Cmos 555 only needs one cap and one resistor for that, apart from the supply decoupling.

Try it without the LED connected. Does it still drift?
I would use a smaller value capacitor and higher resistor anyway.


Carbon composition resistor? Have you used the right circuit?

If you can avoid carbon resistors completely, they drift a lot with temperature,
metal film / metal oxide are much better.

Definitely go for 10k resistor values and smaller cap, then you can't get
any significant self-heating. Use plenty of decoupling with a 555 by the way.


Can you please post a copy of your circuit, in CAD or a picture of a hand drawn circuit in jpg, png?

So we can see your circuit configuration.
The 1K resistor has me worried as you are charging and discharging.
If you are after 50% duty cycle there is a special 555 circuit to do it.

Tom... :slight_smile:

From what you are all saying I will get drift unless I do some coupling and decoupling. Not necessarily sure how to go about doing that. When I get home tonight I think I will try to use 2 pots to get a 50% duty cycle.

I think it's a good start - but would definitely like to create a more reliable circuit with a steady frequency.

Here is my circuit:

Place a 100 nF ceramic capacitor as close as possible from pin 8 to pin 1.

You can try to move pin 4, reset, to pin 8, Vcc.

What does it matter if it is closer to pin 1? That's just ground - using my breadboard I just place the cathode directly in the ground column

The supply really needs decoupling with an electrolytic capacitor.
9volt batteries can have a fairly high internal resistance.
The LED spikes could upset things.
Use ~100uF.

C2 can be ceramic, but C1 (timing) needs to be film.

Short leads means less inductance, one criteria needed in proper decoupling.
Bread boards do not have any real resemblance to a proper ground plane.

100uF electrolytic caps are notorious for inductance.
Read this:

Not at that frequency.
This is not a "problem" IC that needs perfect decoupling.
You need some buffer capacitance for the 35mA LED spikes.

You need both!


This has a circuit to get 50% duty cycle.

Tom..... :slight_smile:

Fig. 4.4.6 is a good way to achieve the 50% Duty Cycle.

I don’t know why the OP wants this continuous mode circuit.

A while ago I designed/build several curling rock timer gates with continuous transmitters.
They had to work for a looong time on batteries on the ice.
I choose two AA batteries, because it’s silly to waste 9volts on a single 1.2volt LED.
One charge lasted for more than a week, with long days of use.
Note that this has to be a cmos IC. Normal 555s don’t work on 2.5volts.
The 50% duty cycle oscillator is from the Maxim 7555 datasheet.
The LED is NOT driven from the output, but from the discharge pin.
The output did not have the ability to sink the LED current at that low voltage, but the discharge pin did.
Transmitter range with the narrow beam LED is 7meters.
A normal wide angle LED might be a bit less.
Schematic attached.