question about analog delay

[capacitor1]

Hi
i'm new in this forum and i have a question

if i have a pulse (V=3V,t=10msec), and i want to store this pulse and use it after 0.5 sec .

any one have an idea about this subject, rc circuit ? any thing ??

thank you

 

[dainiss]

so suddenly I'd say it would be necessary to store the pulse in the condenser. then applied to a gate of a transistor with an RC circuit time constant set to 0.5 seconds. after that time the transistor circuit would triggered the stored pulse

 

[jiripolivka]

To save one pulse like this you can use a sample-hold circuit, with a capacitor as a memory. For a more complex signal, you need a more complex solution.

 

[capacitor1]

Dianiss ,Jiripolivka Please take into account that the only source I have for the circuit is the pulse itself, so after the pulse passes I am left with no other voltage supply to bias a transisitor or such?,

Thanks in advance

 

[dainiss]

if the pulse is the only source, storing the same signal at the output you would never have all the 3V, because off losses. you would need an amplifier.

Jiripolivka is right,
try to look at this

https://www.analog.com/static/imported-files/tutorials/MT-090.pdf

 

[capacitor1]

dainiss, the sample and hold circuit will do the job but the amplifiers in that circuit need a seperate power supply and that is not available to me? what should i do?

 

[dainiss]

i think you need something like this. i can only tell you that the time constant of a capacitor tc = C*R, like n this circuit.


or maybe you need something like this:

**broken link removed**

 

[jiripolivka]

i think you need something like this. i can only tell you that the time constant of a capacitor tc = C*R, like n this circuit.
View attachment 77698

or maybe you need something like this:

**broken link removed**

You cannot make a perpetuum mobile and use your input pulse signal to power your circuits. If you "cannot use a power supply", use a battery. I do not see any other solution.

 

[jeffrey samuel]

the only thing is that when the pulse reaches a capacitor it starts to charge till it reaches the maximum voltage when the input pulse is turned off the capacitor starts to dissipate immediately that is the actual processes if you want a delayed output try using some delay circuit which consumes fairly very less input voltage and that can also be applied as a feedback to them

 

[dainiss]

true.
i cannot see other solution too.

 

[jeffrey samuel]

but in general what our friend states is really and actually true
cos no signal can be replaced as such without using an external power supply in dc form from either a source or a battery

 

[capacitor1]

All of our solutions needs a permanant power supply in the circuit which I don't have and I also cannot use any kind of battery.
Thanks All

 

[jeffrey samuel]

why dont you initially try it out with a charged capacitor at t=0 to trigger the delay system and by the time the pulse reaches the capacitor the capacitor would have completely discharged thereby stores the signal when the time delay is reached the pulse will leave the capacitor and there by you get your output after a delay and power is needed externally

dont have an idea whether it will work but this is worth trying

 

[keith1200rs]

You probably need to explain more about your system. While you say you don't have any power supply, what are you going to do with the resultant pulse?

Is the pulse width and amplitude always the same and fixed or will it vary? What feature of the signal is important to you?

A very long transmission line would do the job but would be ridiculously long and lossy. A lumped approximation to a transmission line is possible but could need rather a lot of stages. Also, transmission lines need to be terminated which results in a 50% signal loss.

Keith.

 

[capacitor1]

Yes the amplitude and pulse width are fixed. The voltage is an important part of the signal. Can a lumped transmission have a delay of 0.5 seconds?

 

[keith1200rs]

If the pulse width and amplitude are fixed then why do you need to delay it and what are you doing with the resulting signal? If the pulse characteristics are always the same then there is no need for an analogue delay - it could be digital in that you could just reproduce the signal 0.5 seconds later. Assuming the signal is not of infinite impedance you could possibly self power a CMOS circuit from the signal itself.

You can delay a signal any amount with a lumped transmission line approach but the number of stages required depends on the incoming pulse width and the delay required. I have use them occasionally in the past and you can probably still buy ready made ones, although they will certainly be for shorter delays than 0.5s

www.rhombus-ind.com/dlcat/app1_pas.pdf

Keith.

 

[capacitor1]

If the pulse width and amplitude are fixed then why do you need to delay it and what are you doing with the resulting signal? If the pulse characteristics are always the same then there is no need for an analogue delay - it could be digital in that you could just reproduce the signal 0.5 seconds later. Assuming the signal is not of infinite impedance you could possibly self power a CMOS circuit from the signal itself.

You can delay a signal any amount with a lumped transmission line approach but the number of stages required depends on the incoming pulse width and the delay required. I have use them occasionally in the past and you can probably still buy ready made ones, although they will certainly be for shorter delays than 0.5s

www.rhombus-ind.com/dlcat/app1_pas.pdf

Keith.

Keith.

Yes the signal has no infinite impedance and it is always the same. How can I reproduce this signal by a cmos circuit?
thx

 

[keith1200rs]

Use a Schottky diode from the pulse to charge a capacitor which is then used as the power supply to some CMOS circuity. The CMOS circuit needs to be a very low power monostable based circuit which produces a 10ms pulse after being powered (or detecting the incoming pulse). The exact circuit would depend on the source impedance of the signal and repetition rate of the pulses.

I have done a similar thing with a signal with 100k ohm source impedance. I used it to charge a capacitor and supply a low power consumption dual opamp (LMC6442) which takes 0.95uA per amplifier. There are probably better ones now - the LPV521 takes only 0.35uA for example.

There are low power comparators as well which would be suitable for making a monostable - you just need to make sure the resistors you use are considerably larger than the signal source impedance.

Keith.

 

[capacitor1]

Use a Schottky diode from the pulse to charge a capacitor which is then used as the power supply to some CMOS circuity. The CMOS circuit needs to be a very low power monostable based circuit which produces a 10ms pulse after being powered (or detecting the incoming pulse). The exact circuit would depend on the source impedance of the signal and repetition rate of the pulses.

I have done a similar thing with a signal with 100k ohm source impedance. I used it to charge a capacitor and supply a low power consumption dual opamp (LMC6442) which takes 0.95uA per amplifier. There are probably better ones now - the LPV521 takes only 0.35uA for example.

There are low power comparators as well which would be suitable for making a monostable - you just need to make sure the resistors you use are considerably larger than the signal source impedance.

Keith.

Keith,
A schematic diagram will be a good start for me if it is available?
Thanks