Stupid question: are there any cheap crystal-driven timers out there that can outmatch the 555 ?

 

The CMOS version of the 555, e.g. the LMC555, is worth looking at too.

http://www.ti.com/lit/ds/symlink/lmc555.pdf

As an astable it has a higher maximum frequency and its outputs switch faster (10 ns rise and fall vs. 100 ns for the regular 555). It draws less operating current (150 μA vs. 10 mA) and has much lower shoot-through current during output transitions. Bias currents on the Threshold and Trigger inputs are less than a ten-thousandth of what they are on the 555 (10 pA vs. 0.5 μA), meaning much larger timing resistor values can be used, allowing longer pulse widths and lower frequencies. They don't have quite the output oomph of the regular 555 (50 mA sink current vs. 200 mA), but good enough for a lot of applications.

 

The 555 was the right part at the right time, and the time was the thing. In the early 70's, 1% resistors were not pennies, they were dollars in today's money. At a time when IC's were relatively new and viewed as being inherently inaccurate for timing applications, the 555 had three *relatively* precise resistors inside that made it's timing accuracy almost completely dependent on the external components; it contributed almost zero error. Compared to the 74121 and 74123, the only well-known IC monostables, it was a bargain even without that output stage. The 8-pin package forced some compromises, but for its performance it was cheap at way more than twice the price.

ak

 

Ok, another timer related question for you guys. As one new to all of this stuff, is there some kind of "rule of thumb" to use when choosing R and C values for them? Say for micro and milli seconds. There are 'calculators' online but they ask for both values and give the time output. So a 'rule of thumb' would give me a starting place.

 

is there some kind of "rule of thumb" to use when choosing R and C values for them?

The rule for an astable 555 is T = 1.1 RC. The starting place is, "which 555?" You choose the highest resistance you believe the input requirements can stand, then calculate the capacitor. If the answer comes out absurd, like 10 megs and 8 picofarads, you guess again.

Say, 1000pf will be large enough to swamp out any stray pfs on the circuit board and the resistor comes out as 80k.

 

Ok, another timer related question for you guys. As one new to all of this stuff, is there some kind of "rule of thumb" to use when choosing R and C values for them? Say for micro and milli seconds. There are 'calculators' online but they ask for both values and give the time output. So a 'rule of thumb' would give me a starting place.

Since resistors are accurate at all values, you are best off picking a stable capacitor (polymer film) which are usually under 1uF. Ceramic capacitors have significant error when there is a constant DC offset applied to them - as is the case for an astable 555 (always at least 1/3 Vcc).

http://m.powerelectronics.com/site-...ackaging_interconnects/packaging/705PET21.pdf

 

Ok, another timer related question for you guys. As one new to all of this stuff, is there some kind of "rule of thumb" to use when choosing R and C values for them? Say for micro and milli seconds. There are 'calculators' online but they ask for both values and give the time output. So a 'rule of thumb' would give me a starting place.

I remember that there's a valid range of values for R and another one for C. But I can't seem to find that info anywhere

 

Ok, another timer related question for you guys. As one new to all of this stuff, is there some kind of "rule of thumb" to use when choosing R and C values for them? Say for micro and milli seconds. There are 'calculators' online but they ask for both values and give the time output. So a 'rule of thumb' would give me a starting place.

For a 555 the basic limits come from two sources -- the input bias current requirements for the timing pins and the leakage current of the capacitor (particularly if you are using electrolytics to get large time constants). For a garden variety 555 you should allow for about 1uA of current into the threshold pin and trigger pins, so you want your current in the timing resistor to be well above that. If you are using an electrolytic cap and don't have the leakage current specs on it, a rule of thumb is about 5 uA plus an additional 1 uA/10 uF, so a 100 uF cap might have about 15 uA of leakage current. If the voltages you are using are in the 10 V range, then the upper limit on the resistance value you can use with this size cap is going to be in the (very roughly) 100 kΩ range.

I mention electrolytics because people start really running into problems when they are trying to get long delays by using large caps, which often drives them to choose a large electrolytic and also a large resistor value and what happens is the timer stalls because all of the current through the resistor is feeding the leakage current.

 

Page 10, fig. 11

 

Page 10, fig. 11

That's more or less what I was referring to. The graph shows the valid range of caps according to the value of R. But isn't there a limit to the absolute value of R as well? Like 10MΩ >= R >= 1KΩ, for instance.

 

That's more or less what I was referring to. The graph shows the valid range of caps according to the value of R. But isn't there a limit to the absolute value of R as well? Like 10MΩ >= R >= 1KΩ, for instance.

I think that maybe WBahn has answered my question already...

 

That's more or less what I was referring to. The graph shows the valid range of caps according to the value of R. But isn't there a limit to the absolute value of R as well? Like 10MΩ >= R >= 1KΩ, for instance.

I think WBahn said a lot about that in post #28.
IIRC, low limit is about the 150 ma max current the discharge transistor can carry. If the time required to dump your cap @ 150 ma is going to mess up the timing, don't do that. High limit is about input impedance of the sense pins and the leakage of the capacitor.

 

Not wanting to hijack a thread, http://forum.allaboutcircuits.com/threads/time-delayed-switch-trigger-by-opening.113144/ it was perfect for a question I have. Why doesn't anyone use any of the other timers, other than a 555? It seems to be the only one people use.

The cd4538, cd4098, and many others seem to do things that the 555 can't. The main one being able to use either a rising or falling edge as a trigger. They can be used as an astable or monostable and have many of the same traits as the 555.

Another of my 'million dumb questions'.

Thanks for asking this question, it put me onto the 4538 and it was the perfect chip for a project I was working on. The rising edge triggering with non-retriggerable and retriggering as well as the Q and not Q outputs all made the project very easy. Mainly, no capacitive coupling to 555 to allow a very short mono-stable. Having trouble getting a short enough pulse to trigger the 555 while still having a reliable pulse to ground to get the 555 to trigger at all. Very easy with the 4538 and I could also use the NOT Q output to avoid a transistor to invert the signal of a 555.

 

I use a miniature timer MIC1557.
www.micrel.com/_PDF/mic1555.pdf
Example:

 

The MIC1555 et al are *great* parts. We don't use 555's often, but when we do we us the Micrel parts. Talk about K.I.S.S.

And back to someone's question about where to start when picking R-C combinations, consider the noise environment of the application. A part with a 4-decade range of acceptable values has 2 or 3 decades in the middle where overall performance is best. In a high noise environment, start at the lower impedance end of the range to minimize external influences.

ak

 

Thanks for asking this question, it put me onto the 4538 and it was the perfect chip for a project I was working on. The rising edge triggering with non-retriggerable and retriggering as well as the Q and not Q outputs all made the project very easy. Mainly, no capacitive coupling to 555 to allow a very short mono-stable. Having trouble getting a short enough pulse to trigger the 555 while still having a reliable pulse to ground to get the 555 to trigger at all. Very easy with the 4538 and I could also use the NOT Q output to avoid a transistor to invert the signal of a 555.

Glad to be of help. As a newbie to this stuff I was looking at all the different ways of doing things. Some of my methods defy what most people do but I've done that most of my life in every thing I try. This is the first time ever for me, looking forward to winter, to have time to start again on the EDM project.