Sounds like this took quite a bit of work
Bud, so I'd like to say thanks for your investigation and posting the results in advance.
Looking forward to following the thread as it develops.
Yes it took a while and unfortunately takes long to get the info to you guys. Sorry.
Just updated Part 2, you can start reading.
Yes it took a while and unfortunately takes long to get the info to you guys. Sorry.
Just updated Part 2, you can start reading.
Nice write-up - interesting!
Hi, great work.
With the test pads and the zero-ohm resistors on the bus would it not make sense to remove the resistors and fit a small micro to the bus to program the correct settings? Take the 3v3 from the NCP1117 and have it once set the freq. it could even be tied to the lock pin and monitor that as well?
Thanks again
Thank you so much for the time and effect that went into this. I can't wait for part 3. Very keen to get the scope working as it should.
Skip the analysis and just post pictures of the fix so people can improve their equipment right away!
Better to include the analysis so that people can judge the competence and validity of the statements. That will distinguish the statements from the vast majority of unsupported assertions seen on the net.
Anyone wanting to skip the analysis (and jump to conclusions) is then free to do so.
Skip the analysis and just post pictures of the fix so people can improve their equipment right away!
I don't understand this. Clearly the analysis is something he wants to share with us, it seems he thought it was interesting and thought we might as well. I'm not sure if I'm the only one who thinks this, but this "skip the analysis" sounds very rude to me, like he did all this work and you're not willing in return to put up with the way he wants to present it...
Keep it up Bud, I'm enjoying the analysis as well.
EEV forum is more of a time sink than a video game:)
Just stumbled across this thread and will try your fix after others try or fail.
Now, your wire inductors, are they at right angles? Hard to tell from picture angle. Or maybe they are far enough apart to prevent coupling?
Anyway, good stuff, read it all. I'll also have to do the 50 to 100mhz mod while I'm at it.
What about calibration, will it need a cal lab adjustment after?
With the test pads and the zero-ohm resistors on the bus would it not make sense to remove the resistors and fit a small micro to the bus to program the correct settings? Take the 3v3 from the NCP1117 and have it once set the freq. it could even be tied to the lock pin and monitor that as well?
That is right, this can be done. In fact i considered as the final solution and i also done a lot of testing with an external PIC connected to the PLL. That way i was able to try different PLL settings. As i mentioned in Part 2, a configuration with a higher PFD frequency (i tried up to about 7MHz) produced a better clock with less phase nose (>10dB improvement), but in the end i decided to leave the factory programming because of two reasons:
- further improvement in the PLL noise would not have a material effect overall because it would be swallowed by the ADC noise which is the second biggest offender responsible for overall system noise. The first and biggest offender is the front end amplifier, i will talk about it later on in the article.
- not many users would be willing to replicate that solution. What i proposed in the article should be good enough and fairly easy to do and without cosmetically disturbing the PCB.
EEV forum is more of a time sink than a video game:)
Could not say it any better
Now, your wire inductors, are they at right angles? Hard to tell from picture angle. Or maybe they are far enough apart to prevent coupling?
I ended up with them not being exactly at right angle because i tweaked them while watching the SA screen for best result (S/N ratio). Making them by hand i would not know the exact inductance value. If they were SMT parts i would mount at right angle in order to minimize change in the part's nominal inductance. The ADI datasheet formula says should be 3.18 nH so we want about 3nH assuming the other 0.2nH or something will be contributed by the ground vias.
Now kill the "bat"
You bet, there should be a silver bullet for it
Thank you so much for the time and effect that went into this. I can't wait for part 3. Very keen to get the scope working as it should.
Was that your software to process waveform data files and display FFT ? I did not try a offline FFT after the mod. May be someone in this thread could post Before and After pictures of an offline FFT.
Great Project!
But sad, that someone on a forum has to do the job of RIGOL...
Thank you so much for the time and effect that went into this. I can't wait for part 3. Very keen to get the scope working as it should.
Was that your software to process waveform data files and display FFT ? I did not try a offline FFT after the mod. May be someone in this thread could post Before and After pictures of an offline FFT.
Hi bud,
I think there are 2 or 3 different projects to download the scope memory and do an FFT, one of them was mine . You can find it by looking through my post history. You need the labview RTE (it's free) to run my exe so it's a bit 'bloated' but that's the price for quick development in labview.
I am very interested in this project because cleaning up the mess that occurs in the FFT with the stock pll is a goal of mine. I would be very interested in seeing the improvements you've made to that effect.
When you release the rest of the documentation and when I get around to doing the mods on my 1054z I will definitely post a comparison FFT with the full 24mpoints.
Skip the analysis and just post pictures of the fix so people can improve their equipment right away!
This is easy - just bookmark this thread and come back in a month or so and go straight to the last part.
EDIT: the fix part of it ends in Part 4, the rest is going to be exploratory information about the scope capabilities and some reverse engineering of the front end, which a curious reader may find useful.
@AlessandroAU
Ok so i start thinking that a higher PFD solution with an external PIC could make sense for offline FFT to get to the best of its capability...
@ All:
So you guys see, if you did not read the complete document with analysis, you would not get what we may be talking about in this thread, example being this post.
So you guys see, if you did not read the complete document with analysis, you would not get what we may be talking about in this thread, example being this post.
Nice work
Bud, and patiently awaiting for more.
BTW, I particularly like those high-end inductors you installed in the VCO section.
Amazing at how simple a solution can be if you know what you're doing.
Caused me only a 3 hr sleep this morning...
Caused me only a 3 hr sleep this morning...
You should not loose time slepping!!! That's what Yagol Yngineers did!!!!
Do not sleep again, and release Part3
Just Kidding, and thank you for not sleeping and present us with this great work.....
Fantastic work! Also I envy your test setup, 2 spectrum analysers, wow. Can't stand the suspense though, those bat wings look awfully like low level amplitude modulation of the carrier/clock.
Bud with
MarkL's assistance bought the poorly implemented Yaigol PLL clock to our attention nearly 18 months ago (for those that were taking notice), there is much about it in the previously linked thread and since then has documented the ordeal in this thread for the benefit of all.
That he has taken so long to present his findings in detail is confusing to me and I wait with great interest for any gems not yet revealed.
Carry on Bud, the floor is yours.
BTW you could probably get the 2 pdfs in one post by posting one and then the other by way of a "Modify".
Not sure, but it'd be worth a try rather than have to download the 1.5 mb which will a PITA for some.