My first scope + my first 555 circuit == Questions

[kfitch42]

I recently got a TEK 2336YA 100Mhz scope. Now I need some signals to feed into it .
So first I used the 'trigger view' button to see the 60Hz mains. This scope has an LCD readout you can use to measure things like periods. To the best of my ability I am getting about 17.4ms period. Pretty close to 5% too long. Guess I will want to calibrate the horizontal on my scope at some point.

Next I breadboarded an astable NE555 circuit with two 220K (5%) resistors and a 68 pico cap. This should be about 32khz with 66% duty cycle. Measuring it on my scope I get about 25Khz. Not sure about the tolerance of the cap... so I don't know if that is a reasonable difference.

The real fun started when I decided to try to find the rise time for this 555 circuit (just to see if I could). zooming in to 1microsecond per division (at least I am pretty sure that is what I had it set too) I get the attached picture, which seems weird.

Is this normal for a 555 circuit on a breadboard?

Is my scope funky?

Using the cal. on the scope front I see a bit of overshoot/undershoot, but NOTHING like with I see with the 555.

[vk6zgo]

Do you have a device anywhere in the house which has a NTSC composite video output,such as a Digital Set Top Box,DVD player,etc.

The Field rate in NTSC is very closely controlled at 60Hz,so there is a second check for your period.
The line rate is approx 15.750 KHz *,which gives you a line duration of 63.49 microseconds.
The line sync pulse width is around 4.7 uS.--Please forgive me for not knowing this precisely,as I live in a PAL country!

If you terminate your Oscilloscope in 75 Ohms,& connect it to the composite NTSC video source with a 75 Ohm coax
cable ( for a short length,you can get away with a 50 Ohm cable),you should be able to get a fair check of your 'scopes performance using known signals.

These signals will not test the high frequency limits of your scope,but will let you see if there is anything basically wrong with it.

*I say approx,because they fiddled it slightly when NTSC colour was devised.
It was exactly that figure for the original B&W System.

[ju1ce]

Do you have a decoupling capacitor in your circuit? Try to add a 100 nF capacitor between the supply rails close to the NE555 and see if it gets any better. Breadboards connections can have high inductances, so the fast edge can make the supply rail funky without the cap.

Moreover, they have large stray capacitance (2-25 pF per connection point according to Wikipedia), so the frequency won't be accurate. Not so sure about the error of your mains measurement though.

[kfitch42]

Do you have a decoupling capacitor in your circuit? Try to add a 100 nF capacitor between the supply rails close to the NE555 and see if it gets any better. Breadboards connections can have high inductances, so the fast edge can make the supply rail funky without the cap.

Good call on the decoupling. I only had decoupling (10nf) on the control pin. Since that is what the schematic I found had. I added a 220nf electrolytic cap (my random parts bin is a bit pathetic right now) near Vcc and things got better. The attached pics are at 0.2 us per division.

I will definitely have to try the ntsc composite video thing a try soon.

[IanB]

To give you a point of comparison, here's what I got on a breadboard with an NE556 from TI. I had R_A=1k, R_B=10k and C=560 pF. The frequency should have been about 120 kHz but apparently I was getting 100 kHz. Maybe just the effect of component tolerances?

The main thing though, is a horrible square wave. The rising edge is ugly although the falling edge is not too bad.

I'd guess what you are seeing is real and your scope is OK.

[IanB]

So a question for the more experienced people around here, do you think this ugly square wave is a construction problem or an inherent NE555 limitation?

I have a 1 uF tantalum across the Vcc and GND pins and removing this makes the ringing very slightly worse, but not enough to say the capacitor is a definite improvement.

I am surprised at how bad the ringing is. At kHz frequencies it doesn't seem that it should be too hard to get a clean square wave.

[KTP]

Both of you:  Put the other scope probe on pin 2 of the 555 (the trigger pin which is the charge on the timing capacitor) and show us the output relative to this pin.  (Trigger the scope on the output of the 555 shown on channel 1 of the scope and put channel 2 of the scope on the trigger pin 2 of the 555 chip).

Another nice thing to show:  One scope probe on the power supply voltage, the other probe on the output of the 555.

[IanB]

I will do that later today.

Yesterday when I put the probe on the trigger pin I observed a clean curvy sawtooth consistent with the charging and discharging capacitor, although I didn't post that here.

[Simon]

they rarely show decoupling caps on datasheets. but you will always need one particularly if your expecting fast rising outputs

[IanB]

Both of you:  Put the other scope probe on pin 2 of the 555 (the trigger pin which is the charge on the timing capacitor) and show us the output relative to this pin.  (Trigger the scope on the output of the 555 shown on channel 1 of the scope and put channel 2 of the scope on the trigger pin 2 of the 555 chip).

Another nice thing to show:  One scope probe on the power supply voltage, the other probe on the output of the 555.

Pictures below.

• The output overshoot
• The power supply voltage next to the chip
• The threshold/trigger voltage on the capacitor

I've tried various bypass capacitors on the power supply but I can't reduce the voltage bounce.

[G7PSK]

I think that this is due entirely to the breadboard construction.Or the frequency chosen. I used some vero board to build a 555 oscillator for circa 1000 to 1200 hz and did not have any problem, dead square wave form. I have not used any decoupler on the power supply.

[IanB]

Can you try it at 100 kHz instead of 1 kHz? Also zoom right in on the rising edge?

[kfitch42]

So, I finally got some time to play with my scope again, and, well ... it turns out the decoupling cap helped ... but only because it was acting as a 'regulator' for my unregulated power supply (in hind sight I should of mentioned I was directly using two AAAs as my power supply). I am very serious about fulfilling my role as a beginner properly, and I intend to make as many beginner mistakes as I can

I hooked up my Vcc to the scope, and there was ~200mV of sag right at the edges. It had this weird little stepped pattern that matched up with the steps I was seeing on the output.

I pulled out the 'decoupling' cap, and the 'sag' turned into a giant mess including over a volt of sag and overshoot!!

I am now going to order a proper linear regulator.

[G7PSK]

I pushed my 555 up to 0.2687 MHZ. I only replaced the timing cap. with the smallest I had which was 22pf. It had more ringing than Quasimodo. I will have another go sometime and see just how high a 555 will go, but it would look like you need a squaring circuit on the output at high frequency.

[G7PSK]

I managed to push the 555 to 0.3 MHZ then played around a bit further and then found that I had 1MHZ but nothing like a square wave I don't think that it is possible to get a square wave at high frequency with a 555.

[G7PSK]

Even higher and it gets really strange. It becomes voltage sensitive with the frequency going down with increased or decreased voltage optimum around 7.6 volt. Also have higher frequency coming out of the power rails than the output pin.

[HardBoot]

555 is terrible for high frequency, you should make a simple Inverter+ICI oscillator, fairly stable and fast, can give a nice square with an output inverter.
If you don't have a crystal, could salvage one.

Speaking of inverter... you probably want one on your 555's output, the thing has slow rise/fall.

[G7PSK]

Yes I agree the 555 is not good for high frequencies, all I was doing was trying to see how high it would go. The 555 is only rated to 100khz .
I wonder if any one can get it over 3MHZ.