Products > Test Equipment
Affordable <200MHz PDN analysis / impedance measurement hardware?
joeqsmith:
Swapping the MS for the Si synthesizer had little to no effect. This was with the exact same firmware installed in mine with the same settings. :palm: I rechecked all my mods on the new H4 and can't explain why this one performs so poorly. I had added all the same caps at this point. There is one other potential difference...
--- Quote from: joeqsmith on November 20, 2024, 02:02:55 pm ---
--- Quote from: inevitableavoidance on November 20, 2024, 09:25:39 am ---
--- Quote from: joeqsmith on November 19, 2024, 06:10:17 pm ---Guessing you removed the mask where L3 silkscreen is and went right to the pin.
--- End quote ---
Right there indeed. .... with the inductance that's still between the IC's power pins.
--- End quote ---
I did not pay attention to the layout when I pulled the mixers.
--- End quote ---
Pulling the mixers from the new H4, the ground plane runs under them going directly to the pads of the two bypass caps. Placing the caps to the plane where you did would add to the return path and suspect adds a fair amount of inductance. But again, I doubt it matters much at the frequencies I would use the H4 at. More just an FYI.
Remember I had swapped all of the mixers from my original NanoVNA into my H4. Last thing to try, swap the mixers on the new H4 with the ones from their NanoVNA.
Oddly enough their new mixers were again marked a little different. Sorry for the fuzzy photo but showing the new H4 with same setup after all the rework. It is still a couple dB off, but now very close to the performance of mine.
****
--- Quote from: ftg on November 20, 2024, 08:52:40 am ---All SA602/SA612/NE602/NE612 have had the same die inside since the Philips fab making 602's burned down.
--- End quote ---
Fact or internet fiction? Vertical marked Thailand mixers shown in the previous photo seem to be the hot ticket if you want to work close to DC.
https://www.eevblog.com/forum/testgear/affordable-lt200mhz-pdn-analysis-hardware/msg5720009/#msg5720009
joeqsmith:
Sanded down one of the mixers. Sadly, a bit too much damage and not enough resolution to see much.
***
Added photo before sanding. This part was dead and with it marked NE vs SA, I am not sure if the die markings would be the same for what appears to be the better parts.
tszaboo:
--- Quote from: joeqsmith on November 14, 2024, 02:03:46 pm ---
--- Quote from: tszaboo on November 13, 2024, 01:04:20 pm --- nano/liteVNA
--- End quote ---
Over the years, I have attempted to explain that below 300MHz, the original NanoVNA can often out perform the V2Plus/4 and LiteVNA/64. See my previous post:
https://www.eevblog.com/forum/testgear/affordable-lt200mhz-pdn-analysis-hardware/msg5670833/#msg5670833
I have seen people post about their disappointment after they purchased one of these higher frequency VNAs and attempt to use them for narrow band measurements. Even if you crank down the IFBW, take a day per sweep rather than seconds, it will still produce artifacts that we can't address. I've always said the fix, get the original NanoVNA.
OP was using an H4. The video I linked used my original NanoVNA. These are both much better suited for PDN measurements. You bring up the Lite though. Using the latest 3.2 hardware and released firmware, 100Hz IFBW, 801 points, normalized. Swept from 1.6kHz to 50kHz with ports terminated. Note the poor dynamic range compared with my H4. All of my LiteVNA64s are also very non-linear once you get below about 25kHz. I never use them this low so it's not been a problem for me.
Hopefully this clears up some of the misconceptions about these low cost VNAs.
--- End quote ---
I mention the LiteVNA because that's what I have. I only ever used a Agilent E5000 series VNA that went down to 100KHz, Megiq VNA that goes down to 400MHz (not a typo) and now this LiteVNA. The low frequency measurement didn't really interest me up until now. I'm starting to understand the issues with the shunt trough method, and the applicability to other low cost VNAs. Indeed if they go down to 1.6KHz they sound useful. I watched your video about measuring PDNs again, where you use the transformer to overcome the ground loop problem. I was thinking that active amplifiers can be used to make the second port of the VNA into a semi-differential input. TI suggest this in their application note:
www.ti.com/lit/an/sluaai3/sluaai3.pdf
I think it should be relatively straightforward to make:
A driver for the first port to isolate the load, DC block, and set the desired excitation currents for the DUT.
A semi-differential to single ended converter for the second port to turn the measurement into a 4 wire measurements.
And do this to cover the KHz to ~100MHz range. IMHO amplifiers should be easier to tune for a flat response than common mode transformers. Plus they can amplify, limit, and do other useful functions.
nctnico:
When using a VNA or oscilloscope, just calibrate the response using a calibration fixture. Like a low inductance 0.1 Ohm resistor. Then you won't need a flat response. Just enough signal to get above the VNA's (or oscilloscope's) noise floor.
I'm contemplating turning two (Kelvin style) tweezers into PDN probes. One to inject current and one to measure the voltage across the supply. The measured impedances are too low and the components are too small to use a single tweezer reliably. I want these to be non-polarised so I can clip them at any point on a crowded (live) PCB without creating a short. I have a project where I want to measure like 15 supply rails with power applied. Soldering wires to the board is going to be too tedious & time consuming. I already got some parts (my differential probe is among them) but the tweezers need more time until delivery.
tszaboo:
--- Quote from: nctnico on November 28, 2024, 01:24:51 pm ---When using a VNA or oscilloscope, just calibrate the response using a calibration fixture. Like a low inductance 0.1 Ohm resistor. Then you won't need a flat response. Just enough signal to get above the VNA's (or oscilloscope's) noise floor.
I'm contemplating turning two (Kelvin style) tweezers into PDN probes. One to inject current and one to measure the voltage across the supply. The measured impedances are too low and the components are too small to use a single tweezer reliably. I want these to be non-polarised so I can clip them at any point on a crowded (live) PCB without creating a short. I have a project where I want to measure like 15 supply rails with power applied. Soldering wires to the board is going to be too tedious & time consuming. I already got some parts (my differential probe is among them) but the tweezers need more time until delivery.
--- End quote ---
True, though I was thinking of the CMRR and not the actual gain. The CMRR of these transformers on the low end is impacted by the physical size limitations. They are also fairly expensive, or time consuming to build and characterize. Compared to that an amplifier is easy (or it requires different skills).
Making RF probes is difficult. I tried making Z0 probes in the past with some sucess, ones that go on 2.54mm headers. I'm guessing having two ,with 0 Ohm instead of the resistance would solve that issue. Or making it into a 2x SMA through connection on the probe. If you have space for UFL connectors, placing them in the design is a great way to probe power rails. Just place the vias right into the footprint, when you solder these manually it doesn't matter.
Navigation
[0] Message Index
[#] Next page
[*] Previous page
Go to full version