Problem using DM74LS123N to produce uS pulses

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
..and i have found out one more (very important, i believe!) thing: when C is smaller, shape of the pulses generated changes - than smaller capacitance than larger slope of the pulse edges..so that eventually it becomes triangular instead of rectangular, for example, when C=10pF and with even smaller C it stops generate pulses at all.
...a problem is certainly capacitance related but what it is and how this could be avoided? :rolleyes:
 

JoeJester

Joined Apr 26, 2005
4,390
Peter,

First on the scope operation ... if you see it at millisecond pulse widths, the component is working as advertised. The problem is when you use the scope on a slow sweep [20 seconds] your trying to see a pulse that is 0.0000005 times smaller than that sweep rate. You probably won't even sample it. That's why I suggested you trigger your scope internally viewing only the output pulse.

Onward to the second inquiry ... when you viewed the millisecond outputs, you only got one pulse per negative transistion of the 0.2 Hz signal. Your stated requirement was for a pulse group [multiple pulses] on negative transistion. This could be a miscommunication, but that's how I read your posting about the signals. If I misread, then your circuit is ok.
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Onward to the second inquiry ... when you viewed the millisecond outputs, you only got one pulse per negative transistion of the 0.2 Hz signal. Your stated requirement was for a pulse group [multiple pulses] on negative transistion. This could be a miscommunication, but that's how I read your posting about the signals. If I misread, then your circuit is ok.
Agh...circuit seems allright (according to the book) when generates millisec pulses, but in practice it fails to produce microsec pulses...any ideas, please! :)
 

Ron H

Joined Apr 14, 2005
7,063
Well, now I have found out that chip starts working well and produces pulses of about hundred millisecond duration each with repetition time 5 sec, triggered by input pulse train. But it does not produce very short pulses, of duration few microseconds (with the same repetition time) - what I really need. When I change timing capacitor from 220microF to 0.0066microF it stops to produce pulses. :( Maybe any ideas? Any advice is very much welcome!
You may not have the scope set up to sample fast enough to catch short pulses. Try triggering the chip at faster rates, or modifying your scope sweep rate. The chip triggering capability does not diminish at slow rates.
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
You may not have the scope set up to sample fast enough to catch short pulses. Try triggering the chip at faster rates, or modifying your scope sweep rate. The chip triggering capability does not diminish at slow rates.
Well, it's easier for me to switch scope into microSec (or even nanoSec) range. However there are still no microSec pulses there (voltage does not change from low to high at all). Perhaps it is because some stray capacitance is present in the circuit which becomes larger than value of small timing capacitor. I guess, it might be a reason why my circuit does not produce microSec pulses. But how it could be fixed?
 

thingmaker3

Joined May 16, 2005
5,083
Two suggestions:

1) varythe resistor instead of the cap.

2) try the other side of the chip, or another chip.

And a question:

What is the bandwidth of your PicoScope? If you are using the 2202 or the ADC-212/3 then you can't measure microsecond pulses.
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Two suggestions:

1) varythe resistor instead of the cap.
Yes, I had tried this but the result is similar to that when I change cap.

2) try the other side of the chip, or another chip.

And a question:

What is the bandwidth of your PicoScope? If you are using the 2202 or the ADC-212/3 then you can't measure microsecond pulses.
It's Pico ADC-200, I think it's suitable for microsecond pulse observation.

Anyway, though I got stuck like elephant (poor me!) I still have hope and ...a question :) A very general question. For example, someone is making a project where a chip is supposed to generate rectangular pulses whose width is determined by two passive components - R and C. This
chip generates pulses of good shape for some range of relatively large C but when C gets smaller pulse edges get slope and its shape becomes really first trapezoid, then triangular as C decreases due to this distortion..and then disappers at all when C is small enough.
I have attached pic of such pulses (taken from my PC screen connected to PC scope Pico ADC-200). What could bring this kind of distortions, in general? (sorry for foreign English :D )
 

Attachments

JoeJester

Joined Apr 26, 2005
4,390
Peter,

Looking at your drawing tells me that as the pulse width gets shorter the sampling is not truely representing the output pulse.

I have a few questions concerning your equipment setup.

1. Are you triggering the oscilloscope with the 0.2 Hz signal?

2. Are you triggering automatically with the oscilloscope or are you using internal?

3. If you are triggering internally, which slope and what is your trigger level?
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
Peter,

Looking at your drawing tells me that as the pulse width gets shorter the sampling is not truely representing the output pulse.
I took them at different time scales. Otherwise it could be impossible to see their shape distortions in detail.
I have a few questions concerning your equipment setup.

1. Are you triggering the oscilloscope with the 0.2 Hz signal?

2. Are you triggering automatically with the oscilloscope or are you using internal?

3. If you are triggering internally, which slope and what is your trigger level?
1. I do not trigger scope with 0.2Hz.
2. I just use A channel of the scope and oscilloscope uses internal, I think.
3. (If I got your question correctly): level of incoming signal of 0.2Hz is about 4V and my DM74LS123 triggers on falling edge of incoming signal.
A bit more about microsec signals on my scope:
For example, oscillogram of 4MHz signal taken on 10uS scale of the scope looks as on the pic attached.
 

Attachments

JoeJester

Joined Apr 26, 2005
4,390
Peter,

PicoScope 200/10 [10 MHz BW] has a sampling rate of 20 MS/S.

I agree you can see the 4 MHz signal when setting your time base to 10 uS.

If you had the oscilloscope triggering on the negative transistion and looked at just the Q output pulse, you would miss it because your not triggering on the start of the Q output pulse.

Your original illustration shows your triggering on the negative slope, and your 4 MHz illustration shows a triggering level of about negative 0.7 volts.
 

thingmaker3

Joined May 16, 2005
5,083
I don't know if it will help or not, but I found this in the manual for the Pico AT-200/20:

The ADC-2xx can either start collecting data immediately, or it can be programmed to wait for a trigger event to occur. The trigger event can occur when the channel A or B input crosses a threshold voltage, or on a change of state of the external (digital) trigger input. The trigger event can be either a rising or a falling edge.

The ADC-2xx can be programmed to place the trigger event at the beginning of the buffer, like an analogue scope, or at the end of the buffer (pre-trigger), or any point in between.

The external trigger input is the same as the signal generator output, so these two functions cannot be used at the same time.
 

Thread Starter

Peter Pan

Joined Mar 24, 2005
122
I have got it! Pulses are there!:) To see them this time I have used external triggering of my scope and selected option: Trigger: Auto, 'Rising', and got a nice pulse of 900nS width which appears on screen every 5 seconds- its width is in good agreement with theory (tp=0.37CR). I have attached pic of the pulse.
Many thanks to all who helped me with this! Your help was really helpful :) Nice weekend!
 

Attachments

Top