Products > Test Equipment
Affordable <200MHz PDN analysis / impedance measurement hardware?
inevitableavoidance:
--- Quote --- :-DD :-DD I am near retirement and still feel like a first year student. :-DD
--- End quote ---
It's nice to never stop learning. :)
--- Quote ---I placed a 4.7uF 0603 cer. across C10, C39, C41, C43. I placed a 10uF 1206 cer across C61 & C62 (doubt these made any difference, I was shotgunning). I placed a TVS on each port to ground, at the connector. The parts I had on-hand are about a half puff. While they did not effect the performance, I would look for a better part. Digikey has some with much lower capacitance.
--- End quote ---
Looking at the schematic yours has more decoupling already! Two 100nF's per mixer compared to one over here. Looks like we had similar strategies, I added 3x2.2uF 0603 to each of the mixers (scratching away some solder mask from the ground plane and sliding them right up to the IC pin - aiming to minimize inductance) and 1x2.2uF to the x5351. Mixers on mine are the SA602A. So the difference in the graph in my previous post could very well be similar to the difference you've seen over there.
joeqsmith:
Thanks for checking those mixers. I had seen they had also used a different part and I wanted to make sure we didn't have yet another variable in the mix.
With you basically replicating my finding, it's all pointing to that original synthesizer. I'll install the latest firmware on friends original NanoVNA and run a baseline with it before I start pulling parts.
***
Was that last data you collected using 1kHz IFBW? I try to remember to include that when I post and have mine setup to default to 1k for now. Wish the firmware allowed me to set it over the remote link.
joeqsmith:
If you wanted to try adding a bit more bulk, I had also added 4.7uF on the front side of the ferrite. Doubt it made much of a difference.
Showing the jist of the bypass mod. Sorry about the sloppy soldering. I had pulled the mixer three times and was getting a bit impatient trying to hunt that problem down.
Guessing you removed the mask where L3 silkscreen is and went right to the pin. I was a bit concern I may damage the mask and short out the supply so I elevated them a bit when I soldered them.
Kean:
Joe, did I understand that post from Hugen correctly that there is a special firmware needed to get the best performance from the MS clock generator?
If so, have you found that firmware posted somewhere? I don't know if the firmware can autodetect which generator is installed and compensate...
joeqsmith:
--- Quote from: Kean on November 20, 2024, 01:21:38 am ---Joe, did I understand that post from Hugen correctly that there is a special firmware needed to get the best performance from the MS clock generator?
--- End quote ---
Correct, there are two versions, one for the Si the other for the MS.
--- Quote from: Kean on November 20, 2024, 01:21:38 am ---If so, have you found that firmware posted somewhere?
--- End quote ---
Yes, they are both in the release area. I am currently testing with NanoVNA-H4-MS_20240220.dfu
--- Quote from: Kean on November 20, 2024, 01:21:38 am ---I don't know if the firmware can autodetect which generator is installed and compensate...
--- End quote ---
As I understand it, they are the same firmware with different defaults for the two synthesizer but you can manually change the type using the firmware's UI. I assume they offer two flavors in an effort to help those people who can't/won't read.
See my prior post on how to select it:
--- Quote --- You can also select Config, Expert Settings, More. At the top is should show the synthesizer selected (MS5351 or Si5351).
--- End quote ---
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version