diff --git a/Documentation/DeveloperInfo/VersionsAndModifications.md b/Documentation/DeveloperInfo/VersionsAndModifications.md new file mode 100644 index 0000000..f646a77 --- /dev/null +++ b/Documentation/DeveloperInfo/VersionsAndModifications.md @@ -0,0 +1,46 @@ +# Versions + +So far, two hardware versions of this project exist. There are only small differences between them and the firmware is the same for both versions. + +The easiest way to differentiate between them is to check for a DC jack input next to the USB-C connector. The DC jack is only present on the second prototype. +## First Prototype +PCB top | PCB bottom +:--------------------------------------:|:------------------------------------------: +![](../Pictures/firstPrototypeTop.jpg) | ![](../Pictures/firstPrototypeBottom.JPG) + +The gerber files for this version can be found [here](https://github.com/jankae/VNA2/releases/download/v0.1.0-alpha.1/PCB_Gerber.zip). +### Mistakes and Limitations +* MOSI/CLK for FPGA configuration reversed. This has to be fixed, otherwise the firmware will not start properly. +* Traces on the bottom side routed through aluminium shield. They have to be isolated to prevent shorts. +* The source signal amplifier does not work well at low frequencies +### Modifications +Two modifications are mandatory to get a working device: +* Swap resistors to correct for MOSI/CLK mixup +* Cover exposed tracks on bottom side with kapton tape + +MOSI/CLK and amplifier modification | Kapton tape for exposed traces +:--------------------------------------:|:------------------------------------------: +![](../Pictures/firstPrototypeMod1.JPG) | ![](../Pictures/firstPrototypeMod2.JPG) + +To improve performance, these optional modifications are possible: +* Add an inductor to the signal amplifier. This will increase the output level at low frequencies (<10MHz or so) and improve dynamic range. +* Drill four holes for additional screws. This is a bit risky because the holes have to be drilled at the correct positions (otherwise they might cut internal traces). However, it can improve isolation significantly, leading to approximately the same RF performance as the second prototype. +![](../Pictures/firstPrototypeAdditionalHoles.JPG) + +## Second Prototype +PCB top | PCB bottom +:--------------------------------------:|:------------------------------------------: +![](../Pictures/secondPrototypeTop.JPG) | ![](../Pictures/secondPrototypeBottom.JPG) + +The gerber files for this version can be found [here](https://github.com/jankae/VNA2/releases/download/v0.1.0-alpha.2/PCB_Gerber.zip). +### Changes from previous version +* Additional DC jack to supply the device if the USB host does not provide enough current +* Additional screws for better isolation +* RC lowpass filters in digital control lines +* Attenuators in the LO feedline to the first stage mixers of port 1 and the reference to decouple them a bit more +* minor footprint and routing adjustments +### Mistakes and Limitations +* Pin mixup on the DC jack. Unfortunately, an easy work-around is not available and the DC jack can not be used at all. +* The source signal amplifier does not work well at low frequencies (same as on first prototype because the problem was discovered too late) +### Modifications +* Same amplifier modification as for the first prototype is possible to improve signal output level at low frequencies diff --git a/Documentation/FAQ.md b/Documentation/FAQ.md index ba9d297..e1c3322 100644 --- a/Documentation/FAQ.md +++ b/Documentation/FAQ.md @@ -1,13 +1,21 @@ # FAQ ### Can I buy one? -No, not at the moment. So far this is just a private hobby project. +No, not at the moment. However, thanks to Hugen, some initial engineering samples have been distributed and if testing goes well, this project might be for sale at some point. ### Can I build one? Absolutely. All required files are available and if you are not afraid to solder a lot of small components you can certainly build your own. The total BOM cost (without the aluminium shielding) is about 200€ when only buying parts for one PCB. There are some [basic instructions](DeveloperInfo/BuildAndFlash.md) to help get started on the software side. +### Why the dual stage conversion? +There are some advantages to having the second stage mixers (mostly to compensate for limited fractional dividers in the MAX2871) but for most operations, a single stage conversion would probably work almost as well. However, compared to other parts the cost for the second stage conversion is not that high and without it certain functionality would not be available (e.g. spectrum analyzer mode with low RBW). There is a bit longer explanation [here](https://github.com/jankae/VNA2/issues/2#issuecomment-700218226). + +### Why are there no aliasing filters in front of the ADC? +The actually are RC lowpass filters in front of the ADC but they are not very sharp and even for signals way above the nyquist frequency they only provide a small amount of attenuation. With my combination of final IF and ADC samplerate, proper alias filters are difficult to implement. Thankfully, they are not absolutely required in a VNA (same reason why image reject filters reject filters after the mixers are not required: the device already knows which frequency to look for, because it generates the stimulus signal itself). + +However, in spectrum analyzer mode, all these missing filters are a problem and lead to a lot of images that have to be removed by signal identification. + ### Why is the final IF so high? I am using 250kHz as the final IF frequency. Compared to some other VNA projects this is rather high and requires a fast (more expensive) ADC. However, it also makes the sweep rate faster: each measurement point takes at least one IF period (more if noise should be reduced to practical levels).