555 monostable timer, Pulse duration changes with input voltage

Thread Starter

ekuo1

Joined Jun 17, 2021
3
Hi all,

I'm trying to use a 555 timer to do a single pulse to control a MOSFET for 4 seconds using the following schematic:
1632879991923.png
Using a SE555 timer, not the NE555 shown above.

The input voltage for this PCB can range between 6V and 17V, which is why I chose that specific SE555 timer, as it said max recommended voltage is 18V instead of 16V like most other timers I saw.
I thought it would have the same 4sec output for all voltages in the given range, however at 6V, the output pulse is about 4seconds long which is good, but at 17V the pulse duration drops to about 3 seconds. The timer datasheet itself only mentioned a very small drift in timing, not the 25% drop I saw, which was very unexpected.

Looking at other forum posts, inaccuracies were mentioned as a result of lack of decoupling capacitors or using electrolytic capacitors, however I added the decoupling caps and only used ceramic caps, and am still seeing the same issues.
Is there anything I might be missing?
 

Thread Starter

ekuo1

Joined Jun 17, 2021
3
It works perfectly in the simulation at all voltage values, but after I designed and populated a PCB for it, the issue started occurring.
The trigger is just a basic resistor and capacitor, with a waveform like this:
1632884661629.png
1632884934601.png
 

Thread Starter

ekuo1

Joined Jun 17, 2021
3
I switched the timing cap from 10u to 1u and it seems to have reduced the inaccuracy significantly. The tolerance on the 10u should have been 10%, so not sure why it caused such large variations depending on the voltage, but is good news anyway.
 

Ian0

Joined Aug 7, 2020
3,754
Is your 10uF capacitor an electrolytic? If so, leakage current is most likely your problem. It increases with voltage.
Try a 50V-rated capacitor.
You could try a 10uF multilayer ceramic. Leakage won't be a problem, but the initial accuracy might be as large values of ceramic capacitor have tolerances of -20%+80%.

By the way, I would strongly advise using a voltage regulator for the 555, as the IRF7401 is only rated for ±12V maximum gate voltage.
 

LowQCab

Joined Nov 6, 2012
1,308
A 5-Volt, LDO-Voltage-Regulator will solve all of your problems.
The SE555 is rated for operation down to ~4.5-Volts,
and the FET has a Logic-Level-Gate.
.
.
.
 

UweX

Joined Sep 2, 2020
12
The C10 capacitor may be the problem. Please look up the datasheet of the capacitor: the ceramic capacitors have a high voltage dependency. The 10% tolerance may be valid only for room temperature and a specific test voltage, not over the full range of voltage and temperature. As a rule of thumb: a mechanically bigger capacitor has a lower voltage dependency. Usually a higher voltage rating will lead to a bigger package.

This effect matches with your results: at higher voltages the capacitance becomes smaller.
 

Yffig

Joined Aug 25, 2019
17
Hi !
As far as I understand, the Vcc voltage of your circuit seems to vary from 6 to 17V, doesn't it ?
The trigger and thresold voltages of the SE555 are directly related to the Vcc voltage by mean of resistors dividers (see Simplified Schematic fig4 of the TI datasheet). Therefore the triggering levels will depend on the Vcc voltage.

LowQCab answer (LDO 5V tripod regulator) should be a good way to solve this issue as long as the power consumption of your SE555 is low enough ( you may have to increase the decoupling caps to provide enough transient current when switching).
 
Last edited:

UweX

Joined Sep 2, 2020
12
For 6 V supply you need a low drop regulator, so that the supply stays at 5 V even for 6 V input voltage.
Also the bipolar NE555/SE555 has some drop in the output stage. To overcome that an external pull-up of 1 k to the 5 V supply may be a good idea, then the FET is driven really between 0 and 5 V.

But exchanging a capacitor should be simpler solution compared to the additional effort of the voltage regulator.
 

Ian0

Joined Aug 7, 2020
3,754
But exchanging a capacitor should be simpler solution compared to the additional effort of the voltage regulator.
Unfortunately not, because the MOSFET is limited to 12V gate voltage.
An unregulated supply of >13.5V is going to damage the MOSFET
 

UweX

Joined Sep 2, 2020
12
Ian0, you got me. So my new idea:
Considering the slow switching in this case a resistor between pin3 and the gate and a zener diode from gate to source will do.
So it is about the same number of components.
From a technical point of view I consider both solutions are valid, so make your decision from a cost point of view.
 

crutschow

Joined Mar 14, 2008
28,161
Therefore the triggering levels will depend on the Vcc voltage.
True.
But it's the ratio of the charging voltage to the threshold voltage that determines the 555 trigger time, and that stays constant with a change in voltage.
It always triggers at the same point on the RC charge curve, independent of voltage.
 
Top