I have been using the educational version that does not require (or ask for) a license. Maybe you downloaded the commercial version. Gowin recently updated the software to a non-beta version. The file I downloaded for Windows have name Gowin_V1.9.9.03_Education_x64_win.zip. It contains a conventional Windows installer.

My background is probably similar to yours. I don't know VHDL. I started with Verilog because of its C-like syntax. If you know C but also understand how clocks work with flip-flips, getting started with Verilog is quite easy, and its module concept works well. I think it is also less verbose than VHDL. But I find things not to like about Verilog. One is that it is not strongly typed. This can lead to errors. Another is that the "wire" vs "reg" definition is pretty weak from a language design point of view. But I think I will stick with Verilog out of inertia. Verilog on a cheap FPGA board can be fun for software folks (like us) who have worked close to the hardware. Cheers.

Verilog puts you on the pipeline to SystemVerilog and SystemC so definitely go verilog. I first learned VHDL and let me tell you, it is awful and has a strict syntax and design in the worst ways. The type system is awful.

@@mattymerr701 When I worked in the defence industry it VHDL was the standard to use. However as an embedded C programmer, I'll take the verilog recommendations, thanks both!

Rust gains traction as it fails compile time, which is better - I see vhdl as the more safe variant. I did however start with verilog, but now i just use whatever is used already :-)

Unless you're doing DSP and math heavy twos compliment /signed and unsigned arithmetic, there's no advantage to VHDL versus Verilog these days. It is more verbose and while 'safe' it looks similar to ADA syntax wise. If you're already comfortable with C , then Verilog will make a bit more sense.

I am happy that you find it useful. Thanks!

I found the FPGA pins connected to stuff on the board like the buttons by reading the Tang Nano 9K schematic available from the sipeed wiki (just google Tang Nano 9K). It's weird that they don't document this more conveniently. Since you asked, I just pushed a pinout diagram that I annotated with some of these pins. It's now on the github mentioned in the video description above. I have been using this diagram as my cheat sheet. Cheers.

I am glad the video was helpful. I have since acquired a Tang Nano 20K. Getting started with it is very similar.

The litex project proves what you say about doing more on even the Tang Nano 9K. It's easy to get it to spit out a much more complete and capable SoC than mine-- including the PSRAM. But if you want to add to it, you have to understand and buy into Litex's tools and methods. For me it's too much abstraction while I am still getting experience with FPGAs. Thanks for your comments.

LiteX automates the process of creating an in-FPGA SoC design, letting you create full systems. The Nano 9K is in its supported boards list, but that doesn’t mean you can build a fully spec’d system on one. There is nothing wrong with wanting to do your own thing, and the suggestions were there as things you could do to enhance it. You might want to think about how you could support uploading arbitrary code to it at runtime for example.

Yes, it's good that you point this out. There also appears to be a Tang Nano 20k. The sipeed wiki lists a number of Tang FPGA boards.

I don't have a Tang Nano 20K, but I can say a few things. The 20K uses a Gowin FPGA that has more resources and can thus support larger designs. There are many board-level differences also. The two boards have differences in connectors and external pin counts. Pin constraint files for one will not be compatible with the other. I think that either the 9K or the 20K are good for learning about FPGAs. On Amazon, the 9K was cheaper. Hence my choice.

I should also have said. I think this PicoRV32 project would work on the 20K if you change the device type and the pins in the constraint file. To figure out the pins, look at the pinout diagram and the schematic on the Sipeed wiki (Google finds it if you search for Tang Nano 20K).

Thank you for this kind comment. Good luck with the 25K.

Nano 25K? Did you mean the Primer 25K? As far as I know the Nano range stops at 20K

I am not an i2c expert, but I suspect you are correct. The i2c spec says that the transmitter releases SDA after each byte so the receiver can ACK by pulling SDA low. The spike could be the point at which the transmitter releases SDA. All is well as long as the receiver can pull SDA low in time to meet setup time. Thanks for the comment. At the time, I didn't think about it for very long.

I am glad you found it useful.

Good suggestion. I didn't know how to add chapters. One minute of research solved that-- just type time stamps and titles as text into the description. Thanks for the good idea.

Cool. I was attracted to the Tang Nano 9K because it was cheap. I don't actually need an FPGA board! I just wanted to learn about FPGA programming. I think I will try to build a simple SoC on it. There is quite a bit of material out there on that topic.

I think either is OK, but I'd go for a vertical mount regardless (not a horizontal dipole-- but vertical dipoles should be OK, J-pole for example).

@@electronics.tinker Thanks, I will buy the DPD Production Airband Antenna.

Thanks for the comment. I have no ganged pots so I just changed the resistors in pairs.

Thank you for this comment. I made these videos because I did not see much content about the ADALM2000, but I think it is a useful device and significantly cheaper than the Analog Discovery. I am actually using my ADALM2000 as I type this to view an i2c signal. I hope Analog Devices continues to support and update the ADALM2000.

@@electronics.tinker same here. I am hopeful that someone will, considering that it is open source.

Glad you like them. I decided to make videos about the Tang Nano 9K because it is so cheap and there is not much content about them. But I am new to FPGA programming, so I am just trying to figure out how to use various features.

My guess is that the GW1NR-LV9 used on the Tang Nano 9K is larger in silicon area than a small Arduino (ATmega328). But most of GW1NR-LV9 is not used to blink the LED. Maybe the very smallest sort of FPGA could give an Arduino a run for its money in blinking an LED. But much smaller microcontrollers exist also. They likely win the prize. The flexibility of FPGAs can cost 10x in area versus a dedicated design.

Hi, I see your comment about "no luck posting" but do not see any sign of a previous comment. It does not look like it was somehow held for review by me. I have had a similar experience where I try to comment on some other persons channel and never see the comment. My comment that failed to appear contained a web link. Could KZbin silently reject those? I don't know. I was not able to figure it out. Sorry you are having trouble. I appreciate it when people take the time to comment.

I think it can measure inductance of a coil as long as it is within a reasonable range. I don't know much about stepper motors though. I am not sure how much inductance measurements apply.

Gowin has some i2c soft ip that might simplify the test, but it looks like the slave was written for a different FPGA than the one on the Tang Nano 9K. I plan to look into it. It may take some time. I think I must figure out bidirectional signals in a simpler context first.

I have not tried apicula. Maybe I will someday. I have found the Gowin educational tools edition to be imperfect but easy enough to get started with.

I see you have experience with these things! There are indeed some things that don't work with the Gowin EDA (at least on Linux), and things that require incantations!

Very interesting. What did you change to impedance match? Also, what analog multiplexer did you use? Most people use an fst3253 which has lower "on" resistance than the part I used. This affects the input impedance and size of capacitors. I paid little attention to this. I just experimented with the size of the caps. NA5Y uses a bandpass filter and, thus, more sophisticated matching. I wonder how much the arbitrary nature of the antenna wire slung over a tree affects this. NA5Y, being a HAM, might be using proper 50 Ohm antennas.

I used an MCP6002 without giving it much thought. I wanted the convenience of powering everything from 3.3V. I was just playing around with it and was surprised that it works as well as it does. Dan Tayloe suggests using a precision or instrumentation op amp but for a toy receiver like this, my bet is that other basic op amps can work. I think NA5Y used an lm324 in a tayloe detector. It's convenient to use a dual op amp. Dr. Volt used an amp that is especially good for driving sound cards. Just watch your power supply level, pay attention to biasing the antenna signal (like to 1/2 of the op amp's supply voltage). But also don't overdrive your sound card's input. If you have a scope, watch out for clipping. And use a long piece of wire for the antenna, ideally running outside. Cheers and good luck.

I found it to be a fun thing to do. I hope you do also. Check out Dr. Volt's video and NA5Y's channel also. There is a lot of material out there. Getting quisk working takes a little guess work. Let me know if you folks out there want more information on that. I have found quisk to work better than other SDR software I have tried with sound card based receivers.

Well, in the interest of full disclosure, a good SDR receiver contains much more: low-noise amplifier with gain control, multiple HW band-pass filters and HW to select among them, analog to digital conversion is included (instead of using sound card), and it works at a far higher sampling rate, and USB interface. I think the SDR Play RSP1A is an example of a quite good but reasonably cheap SDR receiver. And the rtl-sdr V3 dongle is not as good, but you can do a lot with it for a very low price. Cheers.

Which SDR radios do you mean? I agree that the quisk software requires guess work. Do you have any specific questions? The thing with SDR is that quite basic hardware (like an rtl-sdr v3) enables many things from listening to AM radio to receiving data from satellites. There is quite a bit of youtube content, but it takes experimenting for sure.

I don't have an atmega 328 so I am not sure. The task is to generate two 50% duty cycle square waves with one shifted in phase by 90 degrees from the other. You want the frequency to be 10-20 KHz different from the radio station you want to receive. The problem is that microcontroller timer hardware has limits to how it multiples and divides clocks to generate the final output frequency. I bet the atmega 328 can do it for *some* frequencies but maybe not for the frequency near the station you want to receive. This was the problem with the ADALM2000, actually. The Raspberry Pi Pico is unusually flexible in how it can multiply and divide clocks, but even it has problems. The way it divides can cause jitter. Many people who build simple radios like this (including Dr Volt and NA5Y) use an Si5351 chip to generate the quadrature clocks. This is a good idea. Si5351 modules you can use on a breadboard are cheap and available from Amazon. You would control the Si5351 from the Arduino using i2c-- but you really have to study the Si5351 to generate quadrature signals the best way. But many on youtube have done it so there are videos to study. I might do one more if I think I am adding something. I'll mention another option. You can generate a clock with 4x the frequency you need and then use a pair of flip-flops to convert that into the two quadrature signals that feed the mixer. There are also youtube videos on this. For example: kzbin.info/www/bejne/hX2YY5ljoJicf6csi=1-T5_9JH60DV_vtn. I see a good video on mixers on this channel also.

Thank you very much, I liked the flip flop version @@electronics.tinker

Yes, though Dr Volt did a great job I think. To really improve the receiver, I suspect one needs to add an RF bandpass filter and maybe a low-noise RF amplifier. But that would not be simple. Check out this channel I recently found. www.youtube.com/@M0NTVHomebrewing/videos

@@electronics.tinker Indeed. It's the difference between the most minimalist design that can be reasonably functional, and taking a particular design as far as it can go. The best possible RF amp, mixer, LO, filtering, and audio amp and processing can take a DC receiver pretty far.