Laser Pulse vs RF-Trigger Pulse in Oscilloscope

What kind of 'scope are you using? If it's an old analogue one, is it set to ALT or CHOP?

Thank you so much for your kind reply let me clear the question actually the pulse width of the square shape pulse is 1 micro second. the laser pulse width is 100 nanosecond and the time when the laser pulse starts is 3.7 micro second from the start of the square pulse. The trigger pulse switches the RF signal which is sent to the qswitch by Rf -driver on and off. In the qswitch we have a quartz crystal when the RF passes through the crystal it deflect the light and we don't have pulse. When the RF is off we have laser pulse. So the pulse should be bulit in the interval when we don't have RF it means in the 1 micro second interval. But it has a delay. Thank you again.

Ok, at these speeds it is very easy to have nearly anything from signal path length to sensor propagation delay be the cause of your unexpected delay. This will take several iterations of experimenting with the equipment to ascertain what is happening. Be very careful with cable lengths and cable impedances being properly terminated. It would be preferred for you to use the original matched set of oscilloscope probes that came with the instrument instead of just coaxial cables for the connection to the oscilloscope.

You do not give enough information to even begin to answer your question.

In the scope triggering while using dual trace scope - again you have alternate mode and chopped mode. First set it to chopped mode. Here again you may have problems with chopping at such high speeds. Anyway - if you have very good scope- it may not be a problem.

First set the scope as suggested and then look at other delay issues.

Shyam is on target with response time of the photodiode. And a GA from me.

A clue is where the OP is seeing the laser pulse as something resembling 1/2 of a sine wave. The leading edge as well as the trailing edge should be much steeper; i.e. much of a square wave itself.

The OP is correct in saying that when the RF pulse in ON, the laser is OFF (in holdoff mode); and vice versa.

RedFred is right on; you have to deskew the two scope channels to compensate for the propagation delays of the two signals. Most scopes have deskew fixtures to provide accurate deskew and calibration.

Low frequency ripple can come only from poor quality power source or power injection network. Source of multiple pulses is in this zone.

you can make a small digital delay using a variable monostable 74Hc123 and use it for trigger on its falling edge. You can get very large delay.

Keep the repeat rate high.

Perhaps you have channel chopping oscilloscope which can miss pulse during chopping so try for different scan rates for horizontal trace.

Perhaps your signal is getting blanked out in chopping. You can try alternative trace mode rather than chopping and use high repeat rate and auto mode and not normal trigger mode. do not use a+b trigger mode.

Shyam in the previous has good suggestion; but still the OP still is left with a poor representation of the laser output. As drawn, it is not as it should be, and we're back to the photodiode response characteristics.

Photo diode tto be selected for the very fast rise time and its output can be terminated to <1k Ohms to avoid stray capacitance effect. Also use photo-diode in charge injection mode or reverse bias mode and not in photo cell mode with no biasing.

In negative bias, photodiode can respond much faster. Only problem could be its breakdown voltage and leakage current. Using a capacitor output and again terminating with low resistance can give good shape.

Any idea, which photodiode is in use? and laser pulse duration and light flux / intensity and its wavelength?

Oscillations may be in laser pulse or in charge amplifier. use clean Q switch. This one I use is 250ns 362V into 100pF.

I guess that delay is because of crystal in qswitch. When the Rf apply to fused silica crystal in qswitch it takes a time for rf to propagate through the crystal.the rf speed in crystal is around 109ns/mm . so for our qswitch that have a 60 mm crystal it takes 30×109 ns for rf pulse to reach in the middle of the crystal which is around 3 micro second . and it is reasonable. Is my guess right. Thank you again.

perhaps your q-switch driver has some problem. I force drice in push pull. 3 microseconds sounds too much. what is resonant frequency for this device? in kHz?

the qswitch is run in 27MHZ and the trigger pulse is 10khz

in that case the power to the Qswitch is getting a lots of ripple. This must be same to that of the diode pulses. Your power supply to Qswitch is very weak.

How do you arrive at that conclusion?

Power to the Qswitch is often through DC-DC converters of audio or ultrasound frequency range. For small Q pulse, buffer charge storage capacitor delivers power but for long Q pulse, charge runs out and capacity of the HV supply becomes much below to handle the power and its ripple can easily appear in the Q pulse.

I am using different power rated HV Supplies but none of them can reach the Q pulse peak current level of 100A on rise time. However, charge storage capacitor can do this job with 100nF to 1uF capacitor. However, the remains nearly drains most of the charge, and then if switch remains ON for long then power becomes oscillatory.

One can also see oscillations if power from Q switch is delivered through a long wire forming large inductance and LC oscillations then can show up in the tail portion of the Q pulse. These come from undamped rise time with large energy getting stored in inductance.

As I record Q pulse shape along with power supply changes, I can easily figure out the source of the problem. Often this part of the design is a small trade secrete and only few people master out the pulse shape generating scheme properly.

Reflections seems to me to be a much more plausible reason to me than a power supply problem for the ripples you seem to think may are the problem. Now the reflections could be from anything from measurement probing, cabling, signal path reflections, ground plane bounce, de-coupling capacitor signal paths, etc. All of these plausible sources of reflections can only be deduced by a skilled on site analysis of the experiment.

Q switch operates in very high frequency and the switching ON time can cause a lot of RF reflections, resonance and feedback but all this will be in several MHz range. If observed reflections is in the detected signal then obviously it is a good thinking.

Bad quality Q-switch can generate such pulses that will appear like amplitude modulated signal and can cause several pulses with frequency in MHz range.

I don't think there is enough information yet to come to your conclusion. I note your post #23, and I don't think we can go much further (without baseless speculation) until at least some of it has been answered.

You are right.There is not enough information.

I have listed two possibilities of having the device to produce oscillations. While Q-switch itself can produce high frequency oscillations due to fast rise time, inductive coupling and high side and low side MOSFET switching time classing, another source is Q-switch HV power supply having low power or poorly designed power injection network can produce low frequency oscillations due to power ripples or RC network ripples.

Oscillations also come from feedback of the Q-switch power pulse affecting the Q-switch MOSGFET driver input stage and its low voltage power sources and by making the control signal itself oscillatory due to oscillations in the power. I isolate all power sources.

DC-DC converters having large leakage capacitance used in MOSFET/ IGBT driver to RF power source circuitry cause serious problems.

This is not easy to pint to problem unless real information is made available and one monitors all signals at desired points. I have listed many points that do produce their own typical oscillations that I know and these come from design problems and are unavoidable unless design itself is modified.

Due to bad power supply in the Q-switch modulator? Even if a switcher, the ripple/noise would be apparent with high load (q-switch on), not off.

Normally people use 50kHz to 100kHz switching power supplies but some of these have control function working in lower frequency range of few kHz, and that becomes visible in the Q switch pulse if Q-pulse is very large and Load remains active.

Inductive ripple are of very high frequency in few MHz range (see answer 7, 10MHz small ripple after rise time and fall time) and hence, it is all due to low power delivery capacity in the HV power supply that is used in Q-switch. DC Power supplies can have all shorts of low frequency ripples, depends on how they are designed.

Another source of ripple can be the RC network used between HV DC Supply and Q-switch. Its time constant for f=1/RC and the laser pulses that appear may be same. If Q pulse is < 1us and its rise time and fall time sharp then this may not happen. I suspect that Q-switch remains ON for 100s of microseconds or even milli seconds.

Give following details please

1) The duration for the RF output pulse nd its shape (not that of trigger pulse to RF generator)

2) Load drive capacity of the RF pulse

3) Modulation if any used in the RF pulse

4) Ripple in the RF Power Supply during RF Pulse switching

5) Frequency of multiple pulses seen in the photo diode output

6) Input stage and Output stage design of the Q-switch or RF Generator