Investigating Erratic 555 Timer Behaviour with Digital ICs

In summary, the 555 timer with 5V Vcc as the clock behaves erratically. I've mostly been using it to trigger a 74LS90 asynchronous 4-bit counter. Sometime it works right (triggers on every positive edge), sometimes it triggers on both edges, and sometimes it triggers twice on each positive edge (counting by twos each time). Sometimes it works fine for a few numbers, then counts on each edge for a couple before going back to normal. And it's always on the same numbers in the count too, so something's weird. Also, sometimes two units of the same chip (in two versions of more or less the same circuit, one soldered on a PCB the other on a
  • #1
Jiggy-Ninja
309
1
Since I don't have a function generator, for some experiments at home with digital ICs I've used a 555 timer with 5V Vcc as the clock, and I noticed something problematic. It behaves rather erratically.

I've mostly been using it to trigger a 74LS90 asynchronous 4-bit counter. Sometime it works right (triggers on every positive edge), sometimes it triggers on both edges, and sometimes it triggers twice on each positive edge (counting by twos each time). Sometimes it works fine for a few numbers, then counts on each edge for a couple before going back to normal. And it's always on the same numbers in the count too, so something's weird.

Also, sometimes two units of the same chip (in two versions of more or less the same circuit, one soldered on a PCB the other on a breadboard) hooked up to the same timer will act differently. One will count by twos and the other will work fine.

Something that seems to make the problem go away is to load the timer with a couple LEDs, one Active LOW, the other Active HIGH. I've no idea why that works, since I have no idea what's going on to cause the problem in the first place.

At school there's boards like this http://www.elexp.com/tst_0dig.htm" [Broken] with a clock conditioning circuit using 7414 Schmitt Trigger Inverters. Would that make sure the problem goes away for good?

Unrelated small question: Is there any chip still being made that does the same thing as an XR-2206? It seems like such a cool little chip, but it looks like it's not being produced any more, and I can't find any ICs that make sine waves.
 
Last edited by a moderator:
Engineering news on Phys.org
  • #3
http://ca.mouser.com/Search/Refine.aspx?Keyword=max038" [Broken]

I've been doing some googling myself, and haven't come up with anything. The only ones I've turned up have been discontinued, and the ones left for sale are pretty expensive.

It's not super critical, I'm just looking for a VCO that does sine waves.
 
Last edited by a moderator:
  • #4
Jiggy-Ninja said:
http://ca.mouser.com/Search/Refine.aspx?Keyword=max038" [Broken]

I've been doing some googling myself, and haven't come up with anything. The only ones I've turned up have been discontinued, and the ones left for sale are pretty expensive.

It's not super critical, I'm just looking for a VCO that does sine waves.

Maybe a signal generator kit would already have the IC with it and would fit the bill?
 
Last edited by a moderator:
  • #5
berkeman said:
Maybe a signal generator kit would already have the IC with it and would fit the bill?
There is a function generator kit at SparkFun with an XR-2206, but it's $35. Not worth it just for the chip.

I could make do with a square wave VCO (of which there are plenty), it just seems like sine wave generators are really rare.

Anyway, I'm more concerned with the 555 timer clock problem right now.
 
  • #6
The following project may be of interest.

Let me have an email address (by PM) that can receive larger files if you want a higher resolution scan.
 

Attachments

  • Fgenerator1.jpg
    Fgenerator1.jpg
    45.4 KB · Views: 395
  • #7
Jiggy-Ninja said:
Anyway, I'm more concerned with the 555 timer clock problem right now.

I think you may have the right idea of using a Schmidt trigger buffer after the 555. I'm not that familiar with the 555, but if it has slow output slew rates, that could cause oscillations around the input trigger point of the logic that you are driving. Especially if there is much noise or ripple on the power supply, and if the grounding is not real low-impedance for the 555 and following logic.

You can also get multiple triggering problems if you have fast logic that is driving unterminated traces that have some length to them, but that doesn't sound like the issue that you're having.
 
  • #8
berkeman said:
I think you may have the right idea of using a Schmidt trigger buffer after the 555. I'm not that familiar with the 555, but if it has slow output slew rates, that could cause oscillations around the input trigger point of the logic that you are driving. Especially if there is much noise or ripple on the power supply, and if the grounding is not real low-impedance for the 555 and following logic.

You can also get multiple triggering problems if you have fast logic that is driving unterminated traces that have some length to them, but that doesn't sound like the issue that you're having.
Slow slew (say that 5 times fast) is something I thought about, but the why would the problem go away after adding the LEDs?

The timer was only set to a few Hertz so that I could see the counter go with my own eyes. Definitely not fast enough for any of the high frequency weirdness you're talking about.
 
  • #9
Jiggy-Ninja said:
Slow slew (say that 5 times fast) is something I thought about, but the why would the problem go away after adding the LEDs?

The timer was only set to a few Hertz so that I could see the counter go with my own eyes. Definitely not fast enough for any of the high frequency weirdness you're talking about.

Not sure about the LEDs. They may have lowered the impedance of the line, and helped to lower the effects of noise.

Does adding a 'scope probe to the line change the behavior? How fast of a 'scope do you have access to?.

Also, are there any unused CMOS inputs anywhere in the circuit? Like if the IC has a dual FF in it, are you tying off the other one's inputs?
 
  • #10
berkeman said:
Not sure about the LEDs. They may have lowered the impedance of the line, and helped to lower the effects of noise.

Does adding a 'scope probe to the line change the behavior? How fast of a 'scope do you have access to?.

Also, are there any unused CMOS inputs anywhere in the circuit? Like if the IC has a dual FF in it, are you tying off the other one's inputs?
This is the circuit I'm referring to: http://www.elexp.com/kit_0edp.htm" [Broken]

I spent a couple hours one day following the traces, seeing what's hooked up to what, and figuring out how it works. A simple circuit, but it was fun. Then I simplified the gate configuration and built it on my breadboard to verify that I understood it properly.

The ICs are 1 74LS90 (asynchronous counter) and 3 74LS00 (quad NAND gates). TTL, not CMOS, and all the gates are used. When I reconfigured it, it needed less gates so there were several unused, but that wouldn't effect other gates. Besides, the breadboard one is the one that usually worked, with the PCB one screwing up.

To try and synchronize their operation and see them work together, I powered them both from the same power supply (5V from an old computer supply I have), and I connected both their CLK As to the output of the 555. 2 of the same kind of chip, running off the exact same power supply, hooked up to the exact same clock signal, sometimes behaved completely differently.

I don't have a scope at home, but there are a few benchtop ones at school.
 
Last edited by a moderator:
  • #11
Jiggy-Ninja said:
This is the circuit I'm referring to: http://www.elexp.com/kit_0edp.htm" [Broken]

I spent a couple hours one day following the traces, seeing what's hooked up to what, and figuring out how it works. A simple circuit, but it was fun. Then I simplified the gate configuration and built it on my breadboard to verify that I understood it properly.

The ICs are 1 74LS90 (asynchronous counter) and 3 74LS00 (quad NAND gates). TTL, not CMOS, and all the gates are used. When I reconfigured it, it needed less gates so there were several unused, but that wouldn't effect other gates. Besides, the breadboard one is the one that usually worked, with the PCB one screwing up.

To try and synchronize their operation and see them work together, I powered them both from the same power supply (5V from an old computer supply I have), and I connected both their CLK As to the output of the 555. 2 of the same kind of chip, running off the exact same power supply, hooked up to the exact same clock signal, sometimes behaved completely differently.

I don't have a scope at home, but there are a few benchtop ones at school.

It looks like the PCB one has a 9V battery clip and simple Zener diode regulator. Does it only misbehave when connected to the PC power supply?
 
Last edited by a moderator:
  • #12
The performance of 555 circuits are very dependent upon the type of 555 you are using.

The original (still available) type draws a very large current spike from the supply when it switches. This can disturb both the supply and other parts of the circuit.

Later, CMOS and HCMOS versions have significantly reduced this spike at the expense of reduced output current drive capability.
 
  • #13
berkeman said:
It looks like the PCB one has a 9V battery clip and simple Zener diode regulator. Does it only misbehave when connected to the PC power supply?
Not sure. Don't think I tried to run it off the battery with the 555. I just hook the 5V from the power supply directly across the Zener and run it off that. It's a 5.7V zener, so it's not high enough to make it conduct. I broke off the flimsy batter connector a couple times already, and got tired of soldering it back on.

It might be the transition time screwing it up. The data sheet for a 74LS00 lists the times for the outputs: tr = 13 ns, tf= 3 ns. The typical rise and fall time of the 555 is 100 ns, a full order of magnitude greater. Still wondering what the LEDs do about that though.

@Studiot: It's an LM555 from National Semiconductor, and the datasheet specifically says, in two separate places: "..or dive TTL circuits" and "Output and supply TTL compatible"

Er...maybe not. I just pulled the chip out and it's labelled NE555N. Should have known not to trust the datasheet they gave us. Stupid school.

EDIT#2: That datasheet says TTL compatible too.
 
Last edited:
  • #14
Jiggy-Ninja said:
This is the circuit I'm referring to: http://www.elexp.com/kit_0edp.htm" [Broken]

The ICs are 1 74LS90 (asynchronous counter) and 3 74LS00 (quad NAND gates). TTL, not CMOS, and all the gates are used. When I reconfigured it, it needed less gates so there were several unused, but that wouldn't effect other gates. Besides, the breadboard one is the one that usually worked, with the PCB one screwing up.
.
I think the circuit you are building is similar to the one in

http://www.electroniq.net/other-projects/electronic-dice-schematic-circuit.html

In my experience, bypassing the Vcc to ground with a 0.01 uF cap at every chip is good protection against coupling (double pulsing, etc.).

Bob S
 
Last edited by a moderator:
  • #15
Bob S said:
I think the circuit you are building is similar to the one in

http://www.electroniq.net/other-projects/electronic-dice-schematic-circuit.html

In my experience, bypassing the Vcc to ground with a 0.01 uF cap at every chip is good protection against coupling (double pulsing, etc.).

Bob S
Similar, yes, except I had modified it to use a 555 for the clock rather than the triple-inverter setup on the bottom, so that I could actually see the counting and demonstrate it.

Well, I didn't get a bunch of 0.1uF caps for nothing. Next time I build the circuit to show a friend of mine, I'll see if the problem still happens, and if the cap helps.

Should I bypass the timer, the counter, or both?
 
  • #16
Jiggy-Ninja said:
Similar, yes, except I had modified it to use a 555 for the clock rather than the triple-inverter setup on the bottom, so that I could actually see the counting and demonstrate it.

Well, I didn't get a bunch of 0.1uF caps for nothing. Next time I build the circuit to show a friend of mine, I'll see if the problem still happens, and if the cap helps.

Should I bypass the timer, the counter, or both?

Bypass everything. I put a small bypass cap (0.1uF SMT or 0.01uF TH) at each power supply pin, connected tightly between the supply and ground. And add a few larger caps sprinkled around the layout, to act as medium-range charge storage.
 
  • #17
If you can, get a CMOS 555 and use that. CMOS 555s give a square wave output which goes very close to the supply rails whereas the normal (old) one can give a volt out when it is supplosed to be low.

CMOS 555s include the ICM7555 and the TLC555. Even Radio Shack sell them.:)
http://www.radioshack.com/product/index.jsp?productId=2062595

These also take a lot less current and have less effect on the power supply rails when they switch.

Another advantage of the CMOS 555s is that you can use much larger resistors in the timing circuit and either smaller capacitors or get lower frequency output.
Decoupling capacitors are always good insurance and I always put 0.1 µF monolithic ceramics in. These are small, cheap and easy to use, but they can be pulled apart if you are rough with the leads.
 
  • #18
vk6kro said:
but they can be pulled apart if you are rough with the leads.

:bugeye:
 
  • #19
Yes, after a few times you learn to bend the wires without putting pressure on the capacitor body.

Great capacitors, though. Ours are always blue for some reason.
 
  • #20
LOL. I couldn't resist poking some fun...
 

1. What is a 555 timer?

A 555 timer is an integrated circuit (IC) that is commonly used in electronic circuits to generate stable and accurate time delays or oscillations. It consists of a voltage divider network, a comparator, a flip-flop, and a discharge transistor.

2. What causes erratic behavior in a 555 timer?

Erratic behavior in a 555 timer can be caused by a variety of factors, including incorrect connections, faulty components, noise interference, and incorrect voltage levels.

3. How can digital ICs affect the behavior of a 555 timer?

Digital ICs can affect the behavior of a 555 timer by introducing noise or interference into the circuit, which can disrupt the timing and cause erratic behavior. They can also affect the voltage levels, which can impact the operation of the 555 timer.

4. What are some methods for investigating erratic 555 timer behavior?

There are several methods for investigating erratic 555 timer behavior, including using an oscilloscope to visualize the signal, checking for correct component values and connections, and isolating the circuit from external noise sources.

5. How can I troubleshoot and fix erratic behavior in a 555 timer circuit?

If you are experiencing erratic behavior in a 555 timer circuit, some steps you can take to troubleshoot and fix the issue include checking for loose connections, replacing faulty components, adding decoupling capacitors to reduce noise, and adjusting the voltage levels to meet the specifications of the 555 timer.

Similar threads

  • Electrical Engineering
Replies
9
Views
2K
  • Electrical Engineering
2
Replies
49
Views
2K
  • Electrical Engineering
Replies
29
Views
9K
Replies
5
Views
2K
  • Electrical Engineering
Replies
6
Views
9K
  • Electrical Engineering
Replies
2
Views
2K
Replies
4
Views
20K
  • Electrical Engineering
Replies
1
Views
7K
  • Engineering and Comp Sci Homework Help
Replies
1
Views
1K
  • Electrical Engineering
Replies
8
Views
5K
Back
Top