Hi friends,
I want to design a high frequency function generator.
The purpose is to check my logic analyser up to 400Mhz signal.
The output level should peak to peak 5v. At the same time I also like to be a variable frequency.
Thanks
Nadeem Hashmi

 

So.... what's your question?

What experience do you have designing and implementing circuits at those frequencies?

What kind of output signals are you wanting? Normally, "function generator" implies continuous waveforms, like sine waves and such. But since you are using it to check a logic analyzer, that implies digital signals. But then you are talking about peak-to-peak signals and digital signals are typically unipolar.

So it's very unclear just what you are trying to do and what it is that you are wanting to discuss.

 

Be more specific about the signal voltage levels and frequency range you want.

 

Many high frequency sources will not do 5Vp-p. Many are 1V. (ECL) Do you really need 5V?

 

So.... what's your question?

What experience do you have designing and implementing circuits at those frequencies?

What kind of output signals are you wanting? Normally, "function generator" implies continuous waveforms, like sine waves and such. But since you are using it to check a logic analyzer, that implies digital signals. But then you are talking about peak-to-peak signals and digital signals are typically unipolar.

So it's very unclear just what you are trying to do and what it is that you are wanting to discuss.

Thanks for early response,
Some about me, I am a hobbyist but I left this for many years due to my job. I have a good knowledge of basic concepts of electronics but not like you......... . Now I have some spare time and started working on Micropython as well as Arduino.
I am experimenting on logic analyser with raspberry pi pico-2 by Dr. Gusman. The good thing is that he worked a lot on this logic analyser and it is open source for hobbyist like me.
I have prepared that logic analyser which is working fine.

Now I am facing some problems.
1. Raspberry pi pico accepts 3.3v logic whereas I also use Arduino which is 5v.
2. How to verify the peak frequency i.e. 400Mbps. I have "Hantek DSO2D15 Digital Oscilloscope 150Mhz" which has built Function generator with max frequency of 25Mhz which is too low to check this logic analyser.

Now the solution of problem No. 1 is to use logic level shifter, I have resolved this with 420Mhz logic level shifter chip by TI which is working fine.

But the second problem still stands i.e. frequency to test max level of logic analyser.

So I requested to help me to resolve the problem. And why i wrote the amplitude level of the frequency should 5v, you could understand.

At the last, may be I am expecting or writing some thing wrong. But this is the world of knowledge and I have less knowledge but want to gain more from friends like your with the best knowledge.
Hope you all encourage me, and help me.

 

Many high frequency sources will not do 5Vp-p. Many are 1V. (ECL) Do you really need 5V?

Yes sure I need 5v, rest of the detail mentioned in my response to " WBahn "

 

Be more specific about the signal voltage levels and frequency range you want.

thanks for response, please read my "quote" to " WBahn "

 

Thanks for early response,
Some about me, I am a hobbyist but I left this for many years due to my job. I have a good knowledge of basic concepts of electronics but not like you......... . Now I have some spare time and started working on Micropython as well as Arduino.
I am experimenting on logic analyser with raspberry pi pico-2 by Dr. Gusman. The good thing is that he worked a lot on this logic analyser and it is open source for hobbyist like me.
I have prepared that logic analyser which is working fine.

Now I am facing some problems.
1. Raspberry pi pico accepts 3.3v logic whereas I also use Arduino which is 5v.
2. How to verify the peak frequency i.e. 400Mbps. I have "Hantek DSO2D15 Digital Oscilloscope 150Mhz" which has built Function generator with max frequency of 25Mhz which is too low to check this logic analyser.

Now the solution of problem No. 1 is to use logic level shifter, I have resolved this with 420Mhz logic level shifter chip by TI which is working fine.

But the second problem still stands i.e. frequency to test max level of logic analyser.

So I requested to help me to resolve the problem. And why i wrote the amplitude level of the frequency should 5v, you could understand.

At the last, may be I am expecting or writing some thing wrong. But this is the world of knowledge and I have less knowledge but want to gain more from friends like your with the best knowledge.
Hope you all encourage me, and help me.

400Mbps for a logic analyzer or function generator (even more) is not a trivial circuit. Even the PCB design will be critical at frequencies that high.

 

You could get some high frequency oscillators. One at 400mhz, then another at 300mhz..... Probably simpler than making something variable.

 

Hi friends,
I want to design a high frequency function generator.
The purpose is to check my logic analyser up to 400Mhz signal.
The output level should peak to peak 5v. At the same time I also like to be a variable frequency.
Thanks
Nadeem Hashmi

Good luck, you'll need it !

Square and pulse waveforms at 400 MHz are difficult......however a LA usually doesn't replicate waveforms but instead uses a comparator to ensure levels are met so you might even be able to use a 400 MHz sinewave.....not sure but that's where I would start that journey.
You still have to verify frequency so counter circuitry is also required and it preferably should be calibrated.

Or you could buy something off the shelf with known capability that you can also use for other projects.
A hacked SDG6022X will give you 500 MHz sinewave and 150 MHz pulses to 5V.

 

A signal generator for 400mhz will run $2000 to $6000usd. You could build something using a FPGA development board. Some FPGAs have good PLLs. If you were close, you could barrow my generator. I am retiring and might never use it again.

Ask us before you buy a signal generator. It is easy to get tricked into the wrong instrument. It is very confusing. There is a max for sinewave, ARB, pulse, clock. (each different numbers)
I have a 60mhz signal generator that brings out some of the internal test points. I think the clock might come out. That is 0-250mhz. I need to look.
--------edited----------------
There are several limitations to generators. One: is the ability to make a signal. Two: the limitations on the output stage. The output that makes a clean sine wave and the one that make a good sharp pulse are not the same amplifier. My 60mhz generator, on the back, has some "3.3V TTL" outputs that are fixed voltage. There is no amplifier. Not 50 ohms. Not built to drive coax. I know it makes much faster digital signals.

 

Now the solution of problem No. 1 is to use logic level shifter, I have resolved this with 420Mhz logic level shifter chip by TI which is working fine.

How do you know the voltage level translators are working fine if you don't have yet a clock generator (and a scope) that goes beyond 400 MHz?
The TI voltage level translators used for the LA are specified for min. 125 Mbps in down translation mode. They may go somewhat higher but 420MHz is a stretch.

But the second problem still stands i.e. frequency to test max level of logic analyser.

The normal sampling speed of the LA goes up to 200 Msps, only in blast mode it's supposed to do 400 Msps. What I don't understand is the use of sampling a 400 MHz signal once per period. What do you intend to test exactly?

Another issue is length matching of the traces on the board as it looks like the length of the traces on the Pico are not included.

The translators have Schmitt-trigger inputs, they are probably perfectly happy with a 13...17dBm @ 50 Ohm sinewave input signal if you provide an appropriate DC offset.

 

How do you know the voltage level translators are working fine if you don't have yet a clock generator (and a scope) that goes beyond 400 MHz?
The TI voltage level translators used for the LA are specified for min. 125 Mbps in down translation mode. They may go somewhat higher but 420MHz is a stretch.




The normal sampling speed of the LA goes up to 200 Msps, only in blast mode it's supposed to do 400 Msps. What I don't understand is the use of sampling a 400 MHz signal once per period. What do you intend to test exactly?

Another issue is length matching of the traces on the board as it looks like the length of the traces on the Pico are not included.

The translators have Schmitt-trigger inputs, they are probably perfectly happy with a 13...17dBm @ 50 Ohm sinewave input signal if you provide an appropriate DC offset.

Thank for response,
I agree my words "working fine" are really creating confusion. I should write "working fine with lower frequencies etc."
I am using "SN74LXC8T245" its max frequency is 420 as per data sheet, for the time being I am not discussing how 420Mbps, because I have no any source of higher frequency to verify good and bad results of my project. So at the first stage I want higher frequency source then dive into the project to face and resolve the problems. I 100% agree that it is not an easy task, but my first barrier is higher frequency source, than I could start my journey. I agree that if I have a required frequency and perfect parts to travel the frequency, the designing of pcb may be a big task for it, I may have to design multiple prototypes to achieve the target and so on.
So at first I am focusing on frequency, when I achieve 400 plus frequency target then war will begin.
I think you may under stand me now,
Best regards for your words, waiting for more from you and my other friends.

 

A signal generator for 400mhz will run $2000 to $6000usd. You could build something using a FPGA development board. Some FPGAs have good PLLs. If you were close, you could barrow my generator. I am retiring and might never use it again.

Ask us before you buy a signal generator. It is easy to get tricked into the wrong instrument. It is very confusing. There is a max for sinewave, ARB, pulse, clock. (each different numbers)
I have a 60mhz signal generator that brings out some of the internal test points. I think the clock might come out. That is 0-250mhz. I need to look.
--------edited----------------
There are several limitations to generators. One: is the ability to make a signal. Two: the limitations on the output stage. The output that makes a clean sine wave and the one that make a good sharp pulse are not the same amplifier. My 60mhz generator, on the back, has some "3.3V TTL" outputs that are fixed voltage. There is no amplifier. Not 50 ohms. Not built to drive coax. I know it makes much faster digital signals.

Thanks dear,
As you quoted that proper generator may cost thousands of $, so I am trying make my own generator which may produce required pulses to only check the efficiency of my project, which I already has briefed that I want to check a logic analyser made with Pi Pico-2. It is very clear that my target is not a generator, my target is to check the ability of my project, for this reason I never want to invest in function generator.
But at the same time designing a such high frequency pulses circuit will also boost my knowledge and will be a good challenging experience.

 

Good luck, you'll need it !

Square and pulse waveforms at 400 MHz are difficult......however a LA usually doesn't replicate waveforms but instead uses a comparator to ensure levels are met so you might even be able to use a 400 MHz sinewave.....not sure but that's where I would start that journey.
You still have to verify frequency so counter circuitry is also required and it preferably should be calibrated.

Or you could buy something off the shelf with known capability that you can also use for other projects.
A hacked SDG6022X will give you 500 MHz sinewave and 150 MHz pulses to 5V.

thanks dear,
I have checked SDG6022X, a good device, But my primery target is not a generator, so I never want to invest in it. I need higher frequency to check the good and bad of my project as I briefed in discussion.

 

400Mbps for a logic analyzer or function generator (even more) is not a trivial circuit. Even the PCB design will be critical at frequencies that high.

Thanks a lot,
You have reached at the target, I appreciate you, the logic analyser shown in the video is clamming the frequency of 400Mhz (not 400Mbps, as you know Mbps is twice of the frequency) and using the logic level shifter TXU0104, as per data sheet the max frequency of TXU0104 is 200Mbps. So how it is possible to achieve the target of 400Mhz.
So in the initial stage I have replaced the logic level shifter chip with actual 420Mbps capability. So the enterance of the frequency may have a clear path. The second phase of the project the check the project at real atmosphere. If it achieve the target then it may be a good tool for my lab.
At the same time I agree with you that it is not an easy task to design identical pcb for the project. But it would be a next stage of war.
best regards
Nadeem Hashmi

 

So how it is possible to achieve the target of 400Mhz.

It isn't. The creator of the video either confuses the frequency of an input signal with the sample rate or the "400 MHz" are just plain clickbait.
The data rate specs for the SN74LXC8T245 voltage translator are - best case scenario - min. 100 Mbps and 210 Mbps typical. Mbps is not MHz. I don't see the LA being able to deal with a 200 MHz input clock let alone 400 MHz.

 

It isn't. The creator of the video either confuses the frequency of an input signal with the sample rate or the "400 MHz" are just plain clickbait.
The data rate specs for the SN74LXC8T245 voltage translator are - best case scenario - min. 100 Mbps and 210 Mbps typical. Mbps is not MHz. I don't see the LA being able to deal with a 200 MHz input clock let alone 400 MHz.

I agree with your words " just plain clickbait " for this reason I am trying get real results real atmosphere at the same time I salute Dr. Gusman for his hard work, but now it is our turn to get best of the project.

 

 

420 Mbps are only valid for translation from 3.3 V to 5 V output. You need 3.3 V out in which case the max. data rate drops to typ. 210 Mbps.