Frequency Counter / Pulser Hybrid?

[Plasmateur]

I have a particular problem where I need two frequencies to pulse in such a way that they look exactly the same in the time domain, every pulse, so as the average doesn't go to zero at a particular point.

I am able to do this so long as the each frequency, f1 and f2, are a multiple of the 10MHz refference clock.

f1 = 2.600 GHz - channel 1
f2 = 2.610 GHz - channel 2

I can pulse this repeatedly with a pulse delay generator triggering both the signal generators and an oscilloscope, and in the time domain the wave train looks roughly the same every time. When each pulse is averaged I get a clean sinusoidal wavetrain.

But if I have

f1 = 2.600 GHz - channel 1
f2 = 2.608 GHz - channel 2

Channel 1's will have a clean sinusoidal wavetrain, however channel 2 will average out to zero.

The only way I can think of to get around this is to use a a frequency counter which counter the 10MHz clock, and sends a pulse signal to the signal generators and the oscilloscope once it has counted a set amount of counts.

Does such a device exist or is there another way to cheaply get around this problem?

[jadew]

Hi Plasmateur,

I can't figure out your setup from what you're describing, or why in the first case you're getting a stable signal, because you shouldn't, not unless some other conditions are met.

Now.. I might have gotten this wrong, but it sounds to me like you only want to pulse those signals when they are in phase. If that's the case, then there are more ways of going at it, but it depends on what the requirements are. Are you looking for a piece of lab equipment that can do this, for a component or for an entire solution? What are the constraints, budget, etc?

I guess what I'm trying to say is that more context would help

[Plasmateur]

Hi Plasmateur,

I can't figure out your setup from what you're describing, or why in the first case you're getting a stable signal, because you shouldn't, not unless some other conditions are met.

Now.. I might have gotten this wrong, but it sounds to me like you only want to pulse those signals when they are in phase. If that's the case, then there are more ways of going at it, but it depends on what the requirements are. Are you looking for a piece of lab equipment that can do this, for a component or for an entire solution? What are the constraints, budget, etc?

I guess what I'm trying to say is that more context would help

Hey. Thanks for replying! I'll try to provide more context.

I uploaded an example of what I am seeing on my oscilloscope. The upper left picture shows the signals I would see after a single pulse. There is a 90 MHZ signal and a 91MHz signal.

Upon a second pulse you can see that the phase of the 91MHz signal is at a different location.

If I continue to pulse, then the 91MHz averages towards zero.

This is the case with my lab equipment. I need to be able to pulse in such a way that the 91MHz signal is at the same phase every time.

Now if I did this with a 90MHz signal and a 100MHz signal, there wouldn't be a problem because it is a multiple of the reference clock - 10MHz. The dual signal generator that is producing the 90MHz and 91MHz signals is referenced to the 10MHz clock of a pulse delay generator.

So the only way I can think of to make it so I can pulse these signals multiple times without the 91MHz signal averaging out to zero, would be able to know the clock cycle of the 10MHz. If I can count the clock cycles and send a pulse to trigger the dual signal generator at the correct time, then I should be able to make sure both signals are at the same phase, at each pulse.

If there is better terminology to describe what I'm getting at, I apologize for not knowing it. I hope this better conveys the problem I am facing. 

[capt bullshot]

I think I've got it:
You're looking for a frequency divider to create the greatest common divider (of the two frequencies) from the reference clock. It's output would be used to trigger the instruments to ensure both signals are captured at the same phase relation.

For your second (90 and 91MHz) example, this would be a simple divider by 10 (since 1 MHz is a common divider of 91 and 90MHz).
One could create such a thing by using a simple integrated circuit (DIY on a prototype board), e.g. the 74HC90 (a fixed divider by 10). Speaking of test instruments, I'd use another signal generator locked to the reference frequency, set to your desired common divider frequency (1 MHz). Since its output frequency is locked to the same reference as the other two signals are, everything should be nicely aligned.