Using the 74AC14 to generate a pulse with fast rise time

Thread Starter

Dritech

Joined Sep 21, 2011
888
Hi all,

I was watching the video below where the 74AC14 is used to generate a fast rise time pulse for a TDR application.

#88: Cheap and simple TDR using an oscilloscope and 74AC14 Schmitt Trigger Inverter

I have three questions regarding this circuit:

1) The first inverter is used as an oscillator to generate pulses at a per-defined frequency, but I didn't quite understand why the other inverters are used. Are they somehow helping in getting a faster rise time at the output? If so, how?

2) Since for a TDR, only one rising edge is required, why use the first inverter as an oscillator? Why not make it to generate just one rising edge (ex: when a switch is pressed)?

3) Will the circuit below achieve the same fast rising edge?

1650146619752.png
 

MrSalts

Joined Apr 2, 2020
1,762
it will create a fast falling edge but you'd need a pull down resistor to make sure the input is connected to ground (through a resistor) to make sure the input is low before the button is pressed.

FOR FALLING EDGE

76E19AAD-984F-4B5E-AF6A-888244533079.jpeg
if you want the INVERTER to make a rising edge, put the resistor between the input and positive supply - put the switch between input and ground

FOR RISING EDGE INVERTER CIRCUIT
D0E941E0-9C8A-43D8-8296-C3AD5C00CE4F.jpeg
 

eetech00

Joined Jun 8, 2013
3,146
Hi all,

I was watching the video below where the 74AC14 is used to generate a fast rise time pulse for a TDR application.

#88: Cheap and simple TDR using an oscilloscope and 74AC14 Schmitt Trigger Inverter

I have three questions regarding this circuit:

1) The first inverter is used as an oscillator to generate pulses at a per-defined frequency, but I didn't quite understand why the other inverters are used. Are they somehow helping in getting a faster rise time at the output? If so, how?
The paralleled inverter output current capacity is additive, so the combined outputs will provide more drive current. It also inverts the oscillator output. However, it could change the signal quality if there is much variance in rise time or delay between the parallel inverters.
2) Since for a TDR, only one rising edge is required, why use the first inverter as an oscillator? Why not make it to generate just one rising edge (ex: when a switch is pressed)?
Capturing one rising transition is not going to be much of a measurement.
Better to assess/test the consistency of the rising edge over many transitions.
 

crutschow

Joined Mar 14, 2008
29,755
One inverter doesn't have enough drive current for a 50Ω load, that's why the inverters are paralleled.

Why do you want to keep pushing a button rather then let the oscillator give you a continuous train of pulses?
 

crutschow

Joined Mar 14, 2008
29,755
it could change the signal quality if there is much variance in rise time or delay between the parallel inverters.
I would expect the outputs from a single chip to be very similar in delay and risetime, not enough to significantly affect the output waveform.
 

upand_at_them

Joined May 15, 2010
896
One inverter doesn't have enough drive current for a 50Ω load, that's why the inverters are paralleled.
And to prove this to himself, OP should do the calculations. V = I x R for V = 5V, R = 50 ohms...Can the chip drive that?...See the datasheet output current spec. Do V = I x R for Alan's resistor value, V = 5V, R = 220 ohms(I think). Can the chip drive that?

OP should also be aware of switch bounce in his proposed circuit. Build it and scope it, if you have one.
 

DickCappels

Joined Aug 21, 2008
8,704
The single edge would work if you have a scope that can capture a single edge and store it, but then you would have to set up the scope to capture that single event.

A lot of trouble so you can replace a dirt cheap resistor and capacitor with a push button.
 

SgtWookie

Joined Jul 17, 2007
22,226
As DickCappels mentions, having a free-running astable multivibrator saves you the trouble of setting up a storage 'scope, and most hobbyist-range 'scopes don't have that feature.

The circuit as shown in the video schematic works quite well, and is designed to be within tolerance of the design of the IC.

As shown, the circuit has a 44 Ohm output impedance. If you were going to use the circuit for RF coax, you may wish to increase the 220 Ohm resistors to 249, which would net a nominal 49.8 Ohm output impedance, a close match for common TX/RX coax impedance values.

Don't forget to include the 100nF (0.1uF) capacitor across the IC supply pins, and keep the capacitor leads very short. Failing to do so will result in a *LOT* of ringing on your O-scope's display, none of it useful.
 

Thread Starter

Dritech

Joined Sep 21, 2011
888
Hi all. Thanks for the very informative replies. I just tried the circuit where a step from low to high is being generated. In theory, the scope should read the voltage of the step (5V) x2 when the end is not terminated (5V of the incident wave + 5V of the reflected wave). The cable is around 30m (98ft) long. Somehow I am only seeing 5V. What can be causing this please? From my understanding, the the scope should see 10V after some time, irrelevant of the rise time or type of pulse (single step or free-running astable multivibrator)

Any suggestions please?
 

Attachments

MisterBill2

Joined Jan 23, 2018
11,874
If the 50 ohm series resistor shown in the single inverter circuits is used the rise time and fall time will be greater because of the RC time constant being greater than if no series resistor is used.
 

Thread Starter

Dritech

Joined Sep 21, 2011
888
For the time being, I am using the circuit below as a fast rise time step generator. When the switch is closed, I am only seeing 5V on the scope. If the 30m long coax cable is not terminated, voltage should be 10V, correct?
 

Attachments

crutschow

Joined Mar 14, 2008
29,755
If the 30m long coax cable is not terminated, voltage should be 10V, correct?
No.
For a matched impedance between the generator and the cable., the incident cable voltage to 1/2 of the generator pulse voltage.

Below is the ideal simulation of what you should see:
(R2 is for simulation purposes only).

What is the horizontal time scale of your oscilloscope?
Show us a picture of the oscilloscope results and where you are measuring.

1652542492726.png
 
Last edited:

Thread Starter

Dritech

Joined Sep 21, 2011
888
@crutschow , thank you for the simulation. Correct, the voltage will be 1/2 that of the pulse since the output of the generator is 50ohms, characteristic impedance of coax is 50ohms (voltage divider). The time base was set high and I was not seeing the step. First photo below shows a delta of 322ns, which gives a cable length of 31.8m. This is very close to the actual length of the cable as it is slightly longer than 30m. There is a little noise and what looks like an over-shoot of the reflected wave. I need to understand what is causing this.
 

crutschow

Joined Mar 14, 2008
29,755
There is a little noise and what looks like an over-shoot of the reflected wave. I need to understand what is causing this.
It looks like possible discontinuity in the connector impedance.
Perhaps the connecter is 75Ω rather than 50Ω.
(It's difficult to tell the difference between a 50Ω and a 75Ω BNC connector.)
Simulation below with a 0.5ns (length) 75Ω connector:

1652564704430.png
 

DickCappels

Joined Aug 21, 2008
8,704
Signals through transmission lines are slower than light in a vacuum. In the specifications this is called velocity factor. RG-58/U has a velocity factor of 0.66 which comes out to 1.54 ns/foot.

From memory, I have seen velocity factors from almost as low as 0.5 to a high of something in the range of 0.8 to 0.9.
 

Thread Starter

Dritech

Joined Sep 21, 2011
888
Hi again. I have another question regarding the circuit shown/used in the video (link below). All five outputs were connected to a 220 ohms resistor to form an output impedance of around 50 ohms. My question is, does the output of the schmitt trigger has its own impedance? If so, then the output resistances have to be set accordingly. How can one find the output impedance of a components? i could not find any reference in the datasheet.

#88: Cheap and simple TDR using an oscilloscope and 74AC14 Schmitt Trigger Inverter
 
Top