LM555 oscillator doesn't work as expected.

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
The hope was to get extended charge time out of an LM555.

LM555 normal operation.

On startup the capacitor is discharged “trigger level” is met. Discharge is released and the capacitor starts to charge. Eventually the capacitor charges up to the Threshold level (2/3 of VCC, or about 1 time constant). When Threshold is met Discharge goes low and the capacitor starts to discharge. When the charge gets down to about 1/3 of VCC the Trigger point is hit and Discharge is released again, starting the cycle over again.


Circuit A is the standard oscillator design. R (10K) and C (100 µF) has a time constant of about 1 second. In this standard we should, and do, get a time cycle of about 3.5 seconds. About 1 second charge time and about 2 seconds discharge time.


Circuit B. The hope was to get about two RC time cycles on the charge by feeding back only a portion of the capacitor charge back to the Threshold pin. To my surprise I get a cycle of about 10 Hz.

Yes, I know I am losing the discharge RC time. I was still expecting a cycle of about 2 Hz.


What went wrong?
 

Attachments

Alec_t

Joined Sep 17, 2013
11,505
According to LTspice the period should be ~1.75s. Perhaps one of your resistor values isn't what you think it is? Or the cap is leaky?
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
According to LTspice the period should be ~1.75s. Perhaps one of your resistor values isn't what you think it is? Or the cap is leaky?
Same cap and chip used in both circuits. I went back and forth between the two circuits twice. I thought the two added resistors whose original values were 24K and 75K might be loading the cap down so I went to 240K and 750K, then 2.4 M and 7.5 M. No change.
The 1.75 sec would fit my expectations. Could somebody else build this and see if they get a different result?
My concern at this point is questioning my understanding of the 555. There are other designs I could go to in order to get the desired results in timing. I am wondering where I went wrong.
.
 

AnalogKid

Joined Aug 1, 2013
8,534
In circuit A the charge impedance is 34K. In circuit B the charge impedance is 9.99K. So the cap is charging up over 300% faster, while the voltage seen by the threshold input is decreased by only 25%. Without getting into all of the exponential math, 75% of 300% is 225%. Even without the discharge impedance being close to 0, B is much faster than A.

Also, your description of circuit A is incorrect. It's 2 seconds charge time and 1 second discharge time.

ak
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
According to LTspice the period should be ~1.75s. Perhaps one of your resistor values isn't what you think it is? Or the cap is leaky?
Thoughts while walking the dog. I don't have a data sheet for the 555 that details the limits of the voltage comparator for the Threshold pin. Could I be pushing the operating limit of the voltage comparator?

In the past I tried extending the charge time by adding a 1K resistor from the Control pin to +V, moving the Threshold point on the internal voltage divider. If memory serves me right I would have moved it almost a volt higher than the design shown here would have. It didn't work at all. "Threshold" was never reached and I just figured I was beyond the limits of the comparator. Could this design just be on the edge of operation?
 

Dodgydave

Joined Jun 22, 2012
9,307
What time constant do you need, charge up and discharge times can be made individual using pin 3 as the charge/discharge with diodes.
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
In circuit A the charge impedance is 34K. In circuit B the charge impedance is 9.99K. So the cap is charging up over 300% faster, while the voltage seen by the threshold input is decreased by only 25%. Without getting into all of the exponential math, 75% of 300% is 225%. Even without the discharge impedance being close to 0, B is much faster than A.

Also, your description of circuit A is incorrect. It's 2 seconds charge time and 1 second discharge time.

ak
Thanks for the wisdom. And yes I misspoke on the timing of circuit A.
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
What time constant do you need, charge up and discharge times can be made individual using pin 3 as the charge/discharge with diodes.
No specific need for the circuit. Just trying to fudge a little extra time out of the charge cycle if I should ever need it. Tinkering in retirement.
 

Alec_t

Joined Sep 17, 2013
11,505
The CMOS version of the 555 would allow higher resistor values to increase the time constant, but there are limits. With very high values the charge current is comparable to the leakage current of electrolytic caps, so charge time is not stable.
 

dannyf

Joined Sep 13, 2015
2,197
What went wrong?
It should work - I simulated it (keep R15 in, for obvious reasons) and it worked just as expected - Circuit B got extended periods. Two points:

1) you will need to make sure that the added divider has sufficient resistance vs. R14/R15 so that the threshold pin will reach the threshold indeed;
2) but those resistance cannot be too large vs. the threshold's input resistance so that it can never trigger the threshold pin.
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
It should work - I simulated it (keep R15 in, for obvious reasons) and it worked just as expected - Circuit B got extended periods. Two points:

1) you will need to make sure that the added divider has sufficient resistance vs. R14/R15 so that the threshold pin will reach the threshold indeed;
2) but those resistance cannot be too large vs. the threshold's input resistance so that it can never trigger the threshold pin.
Interesting that the simulation says it should work. In my limited understanding I thought it should work. Not the first time reality did not agree with my theory. :)

Re: Reaching the Threshold
The problem seems to be that it reaches the threshold way too soon. Instead of getting a cycle in seconds I get an oscillation of about 10 Hz.
 

Thread Starter

hp1729

Joined Nov 23, 2015
2,304
Nope.

If you understand how the first circuit works, you understand why the 2nd circuit has faster cycle time, putting aside the threshold issue.
I understand how the first circuit works. I miss why the second one is faster.
In my thinking the Threshold value of 2/3 VCC shouldn't be hit until the charge on the capacitor reaches a higher value.
 
Top