Open Research Institute Weekly Reports, 4 Apr 2023

Previous Meeting Next Meeting

⏯

youtube image

►

From YouTube: 4 April 2023 ORI FPGA Meetup

Description

MATLAB class outline final review and Remote Labs updates.

A

A

Welcome everybody to the stand-up meeting for open Research Institute for fpga efforts and remote Labs. What we do in this meeting is we talk about what we've done over the past week, what we have scheduled or what we would like to do over the next week if we need any resources, any anything, material or otherwise, and if we have any specific roadblocks that where we need to call in some people to swarm over it, so that's kind of the outline of this meeting.

A

So what I'm going to do is I'm going to turn it over to James first off so that I know he's he's got some school pressures here. James is our lead for remote, lab South and it's been the site of some very exciting things over the past couple of weeks because of tornadoes and also is going to get a whole bunch of stuff, soon lots of lab equipment. So thank you so much James for being here and give us your report and let us know what you need.

B

Happy to be here Michelle, as Michelle mentioned, I'm James, Keel Julie at 7 kilo, Delta Echo I am a sort of lab tech here with Ori I work here at remote, Labs South uh situated in central Arkansas. We have been dealing with a few storms lately, but nothing that has hampered any of our projects. uh All the people, all of our people down here, have been safe. So far from all the storms, which has been really good, uh I've been getting extra work done on things.

B

I've been I've been slightly delayed in my own efforts, because I've been taking care of schoolwork with a new term coming on, but we've been continuing our efforts overall and the property upon which the permanent installation from olap South has been coming along very nicely.

B

uh Yeah I think that's the the main report from Irma labsa.

A

All right, thank you. So much and uh yeah lots of ongoing work there, it's very fun. So, if you're into fpgas or uh radio work or if you're, into inferometry radio astronomy, that's uh that's one of our main sites so catch up with us. We have plenty of materials and lots of uh things going on in remote, lab South, okay, okay, so hello, Mike um I see Mike's iPhone. uh If you have a report for us, let us know: is there anything that you need or go ahead and introduce yourself.

C

Yeah I'm sorry this is my KT 7D I'm just eavesdropping and trying to see what you guys are up to these days. Oh.

A

Hi there, oh hello, okay, now I recognize you, okay. Yes, uh today we're gonna discuss the proposal from Matlab uh for fpga training.

A

We went over it uh last week, but today is our deadline for getting back to MathWorks the company that sells sells you the wonderful product of Matlab and simulink, and they are um we're working with them for training, because we have a startup license at extremely generous and wonderful rates uh from them and Matlab has been extremely supportive of our work in open source, Digital radio and we're going to go over that next.

A

So thank you so much for being here and uh Hey. Do you? If there's any updates that you have about any of your work, then uh please feel free to to share them. If you were able to.

C

Okay, yeah well, the uh with the cats at work, the satellite, the University of Arizona's uh working on. uh We had our Rockets stolen by the space force, so uh we hope to launch later this summer as a result of that, but uh that that's my main update and and uh uh back to you.

A

Okay, so you're dealing with rocket rustlers, wow, okay, so rocket wrestlers have taken the cat, sat they've, got the cat out of the bag and have run away with it, and it looks like that cat sat will be launched later this summer. You know sometimes.

C

Yeah yeah, basically the operational responses space organization. uh Their goal is to launch a rocket within one day of of a call up and uh so they're going to try to demonstrate that with the next Firefly launch.

A

C

Wow, okay, so like they.

A

Want like a one day turnaround and they they needed the the slot.

C

A

C

And so that would be interesting to see if that happens, but uh in the meantime they they grabbed our rocket.

A

Oh wow, okay! Well, you know at least you're part of something very interesting, um that's good and and uh well uh I I'm. Just we're we're all uh very supportive of this particular Mission and we're very interested in it happening, and um you know never a dull moment right.

C

That's about right, I guess.

A

Oh well, okay,.

C

Trouble knowing whether I'm on mute or not so I I'll shut up.

A

Oh well, you know never you should never shut up and you sound good, but yeah I understand the uh the challenges with mute. It's also an original moment with um with it or AV okay. So, let's talk about this class that we are uh we have access to. So the customized training is part of the deal that we get with uh Matlab they're wonderful they've been totally supportive.

A

um We have a startup license with Matlab and simulink. We have all of the toolboxes and the big Advantage there for people that like to do fpga work is that all of the toolboxes includes a something like HDL, coder and HDL stands for Hardware, descriptive language. This is the broad classification of the software that you write in order to tell fpgas field programmable gate arrays. What to do. Hdl is also used to make chips.

A

So, if you're interested in open source or amateur radio chips being available to do the things that we want to do in digital Communications being more available or available at all, then this is a conduit to that this. These are the tools or one set of tools that can let you do that. That's not the only way that at the top of the mountain, but it's definitely one of them and today is our deadline for our verbal commitment back to MathWorks on this particular class and the class outline that they've.

A

Given us, we went over it last week and we're going to go over it today. This is the sort of the final review so I'm glad to have some people here that are very familiar with digital Communications. We already have some feedback. That's received over email. We've tried to get this outline out to as many people as we can for feedback we've gotten some some useful feedback, and so I'll go ahead and I'll try to share the screen and and just go straight to the outline.

A

So let's see you should be seeing uh you're invited or I'm at lab and fpga training. Does everybody see that? Okay, let me know when it happens.

D

A

Nope, oh good, okay, all right! So so, if you're seeing the screen, then I'll go ahead and okay course proposal all right. Here we go so this is the proposal there's three days worth of class material. Here the course is drawn from four different classes that Matlab offers and they build up these classes in order, so that folks, that can go from sort of a basic I know some math I might be an engineer. I might be studying as an engineer and they take you all the way up to software-defined radio with fpga.

A

So what we've done is we've said: okay, let's assume that you you've got some some practical knowledge that you're already a practitioner of of radio, Digital, radio and you're you're grappling with fpgas, and you want to contribute and you're interested in contributing open source designs. What is the class that you will need in order to use the Matlab ecosystem?

A

And yes, Matlab is a proprietary company? Yes, simulink and Matlab an HCL coder and all of that are closed, Source tools, they're all they're, also, industry, standard tools and the feedback that we've gotten is that this is the way to cut through from math from your algorithm to intellectual property, the really attractive thing about HDL, coder and GPU, GPU coder and all of the coders. The translators in in the Matlab toolbox is that the output you own, so anything that we make with these tools is publishable as open source work. That's great.

A

This is a big accelerator and you know when you're after an open source design, whatever it takes to get that open source design published where the design is open, source I think is legit. So that's why we're we're putting some effort here into learning these tools? We get lots of feedback that this is what people are using. We need to know how to use them.

A

They cut the time down greatly and great okay, so we're going to offer a class not just about Advanced, Digital, Signal, processing and multi-rate, filter, design, cortex and all of that, but also how to use HDL coder to turn your Matlab code into HDL.

A

My experiences so far, I've done done some smaller projects with HDL coder. The output is very human, readable, it's very high quality and it, and if you set up your Matlab code with with all of the comments, um then those comments and the explanations and the structure appears in the in the HDL.

A

This is great because I know in the past that, with both hls uh and and other translator, programs that you can put in um C code or or Matlab code or or some sort of math and what you get out might be really only usable by a machine, and we don't really want that. We want HDL output, that's human readable, that's formatted correctly and that we can publish pretty much, as is as an open source module. So that's kind of what we're after so day, one is talking about.

A

Simulink simulink is on top of Matlab, so so Matlab works and is a you. Programmatically interact with the Matrix laboratory or Matlab to achieve maths, wonderful things but simulink has blocks, and this is drag and drop, connect them up like a flow graph or a state diagram, and they have a whole state flow system. So this is all on top of Matlab we're talking about using simulink and Matlab to produce HDL code.

A

So if you can express it in Matlab and simulink, we need to learn how to produce the HDL code that goes to an fpga or or an Asic day. One is talking about this exact thing. What we do is we prepare a simulink model for HDL code generation along with a test bench because we're all about verification and validation. At every step.

A

We don't want to just swap out some code as an open source organization that can't be tested or verified so matlab's all over this, and that's part of this course from the get-go literally from the first lesson.

A

So this talks starts out with a simple model where you generate you take start out with simulink, generate HDL code and a test bench, and you verify it with the simulator. So this is good stuff, but it requires somebody walking into this class to already kind of know that this is important, so not a beginner class, but something that I think really should appeal to current amateur radio experimenters people that want to do spacecraft, design with a fpgas that sort of people it goes on to fixed Point, Precision control.

A

So what you need to be able to do is to be intelligent about your fixed Point scaling, inheritance designer workflow, the tools. What is fixed, Point, impose upon you and how to Grapple with it? That's that's the next part. That's like two hours.

A

The next four hours is how to optimize your Genera generated HCL code with pipelines specifically and I'd, really like feedback like. Is this the right direction to go? I think it is, but we need to know also. You notice that also resources for area optimization and using the workflow advisor So within the HDL coders, the HDL workflow advisor. That's how you get most things done with this particular toolbox so day.

A

One second half you're grappling directly with the tool that we're talking about um knowing and verifying that you're optimized HDL code is bit true, psycho accurate, that's something that really popped out to me.

A

This is something that we all deal with as fpga and Asic people and then finally, mapping simulant blocks to Dedicated Hardware resources on the fpga, and that's it for day, two or day, one day, two, we start off with the signal flow graph, which is going to look very much like a state diagram, but signal flow graph is a somewhat of a broader category, so we're talking about like okay, you take the signal flow graph, your state diagram techniques and you apply it to a finite impulse, impulse response, filter design or fir filter and there's a little bit of review of DSP algorithms and signal flow graph, and they talk about the cut set method.

A

If you're familiar with that already. This won't be too challenging if you're not then the month that we have to advertise this. We can probably get people up to speed and we're going to talk about parallel and serial fir filters, so filter design starts off on day two and there's a bunch here. This is probably a semester course, honestly, um because this stuff is, is non-trivial so day. Two second half you can see that, like oops sorry, uh two hours for signal flow and and the frr stuff, then we go right into multi-rate signal processing.

A

This isn't. These are filters, they're still filters, but they modify the rate this the sample rate. This is so when we talk about multi-rate we're talking about manipulating the sample rate of our signals. This is like the heart of what we do at Ori with respect to transponders.

A

So polyphase techniques are a big deal. Multi-Rate techniques are a big deal and, by day, two not very far into day, two we're getting right into it.

A

The Cascade integrator comb filters or hogenauer techniques are are mentioned here and those are that's one, big significant way to do it. You have up sampling and interpolation filter ship down, sampling and decimation filters.

A

We're going to assume that you know why polyphase and multi-rate Signal processing is good, I think what we'll do is we'll have an intro video about this or point you to an existing video. That can explain this.

A

So if you are wanting to take this class- and you want to do this sort of stuff- that we can get you up to speed in order to to get the full use out of actually implementing it, now talk about efficient, efficient arithmetic for IR filtering, so you notice, we've been talking about finite impulse response, so I IR means infinite impulse response filtering.

A

These are the two big classes of digital filters, you're going to like directly interface with both of them round out day, two with quartic techniques and channelizers, and this is a lot of people.

A

Think, oh, if you think cortic, then you think trig, but we there's a lot in there um and originally when, when we were working the the class uh list and the the content I said gosh most people are gonna already know about cortic, uh but there was some pushback on this like we have to include it, because it's so important and a lot of people will not know, know it.

A

So as of today, we have this two hour section in this class about the architectures for fpga implementations for cortic techniques, and this is broadly useful, um so I'd really like feedback on whether or not this is something we should skip over or still include day. Three is actual programming: we're actually going to implement stuff we're going to use the HDL workflow advisor. That is the it's really, it's really kind of nice. You start up this toolbox, you give it your.

A

You know constructed prepared, Matlab code and you walk that code through the workflow advisor telling it along the way you configure. You know it uh configure it either Matlab or simulink model. In this case, it was starting out with simulink, so we're using simulink to to kind of Goose up our Matlab model or Matlab code generate and build the HDL and C code.

A

So you have HDL for the pro for the uh for the program, programmable logic, Fabric and then there's C code for things like the embedded arm and zinc and you deploy it to a zinc platform. If you're not familiar with zinc is what we use at Ori and zinc means that you have both the fpga and the arm a hard arm and they're very well integrated. So you do have to co-develop like from the beginning with your arm and your programmable logic.

A

The fabric and the list is lots of big words like configuring, a subsystem for programmable logic. That's it's great. You know configuring the interface and peripherals and the IP core and integrating with the SDK um most people I think are probably familiar with taking IP like you develop a core, and then you walk it through the IP packager in save of Auto. You know you, but if you're not, then that's also something that that we can help.

A

You get more familiar with so that you know what they're talking about during the class next part is to model a communication system using simulink. So what we're going to do is uh go through. This is understanding the ad9361. The 9361 is what's in the Pluto, and how do you interface with that with simulink? If you interface with simulink, you can use HDL coder to deploy your particular algorithm directly to this device.

A

So our goal is to like remove the friction and get you talking to the devices as quickly as possible, using the best tools that we can come up with for the job and we're going to simulate a comms system that includes a transmitter, the transceiver on the Pluto, a channel yeah and the receiver. So this is the full RF test environment so not bad.

A

We implement this verify the operation of the transceiver, we're going to use real data. So what this probably entails. Is you as a remote learner logging into a system run by Matlab in an Ideal World? We would all meet in a classroom and everybody would get their own Pluto and you get to do this at a computer, we're not able to pull that off because of how distributed we are, but in the future these classes might be available.

A

We might be able to do this, especially if we're able to kind of put together, maybe a cruise or a longer term, in-person class.

A

So the next part is setting up the ADI Som as an RF front end and we're doing over-the-air signal capture. So that's again, this will be students logging into this and starting it. You know implementing it we'll do the best we possibly can we'll also have all of this equipment also available in our remote Labs. So you can do this over time.

A

It's not a One-Shot deal like we're going to take this class and we're going to make it possible for you to experiment and to learn this stuff by um working with our remote labs and the final thing, the verification, the verify algorithm performance. This is a big deal, because this is something that a lot of people don't do when you do a design going back and doing bit accurate verification and validation. We're we're very committed to this. This is something that the class includes and then the Prototype development- and this is hardware software code design.

A

So, like I said, the zinc has the hardware and the software cores on the same chip, and if you want to really utilize this amazing resource, you kind of have to know how to do something called co-design and that's what's talked about here. The final thing is to deploy a standalone system.

A

So all of this work when you're ordering your Hardware around is great but like how do you set it up in order to be by itself in space or on the top of a mountain or in a car or in your handheld, and we want open source?

A

You know really cool amateur radio stuff to be doing this, so this this class will give you the tools or give us the tools as a community to get started better on on achieving these goals. So it's not that we haven't done good designs and published good work, but having a sort of a unified top-down approach where it walks you through, and you can go from octave or Matlab code, all the way through to a standalone system with a little bit.

A

Less friction I think that we deserve this as a community, so the pro host class date is going to be early May, it's the second through the 4th of May or the second to the 5th that that day, three um in conversations with Matlab, it's probably best to split day three up into two days, to have a four hour session and a four hour session, because I think anybody that's ever actually done this in the lab knows that this is all a lot of hard work and especially, if you're trying to do it mobile or sorry remote.

A

um That's a lot to ask. So that's that's something that, as soon as we commit today um to to moving forward with this, that we'll have to settle um I'm in favor of of it being kind of split up over over two days. But I also understand that that most people that work on open source, experimental stuff may not have four full days to donate to the cause.

A

um It's already kind of a huge class I'd say it's it's easily a College semester worth of material, that's being summarized here and that will be, will have to do a lot to to ramp people up and support them in order to take full advantage of it. We can achieve it and matlab's very motivated, um but this is no doubt. This is a ambitious training project.

A

Okay, so that's the the outline of the course there's also I have lots of of information about the cost and I put that out on the list. uh So the floors open for feedback, opinions, guidance, suggestions, questions anything.

C

Well Rochelle: this is Mike uh kt7d I'll. Give you a a a couple of comments. One is this: uh there's a I, don't know whether you're familiar with Casper, uh Israeli astronomy, uh uh open source group.

A

Gosh that name sounds familiar.

C

Anyway, yes, if you go get on the web and look up, Berkeley uh and radio astronomy in in Casper, uh you'll you'll find us, but anyway, they're using almost or have used over the last I'm guessing 10 years. Almost the same set of tools that you're talking about, namely Matlab and the uh and the.

E

C

uh Products they would like to get away from that, but, of course, that that's more effort than than right now I'm, not sure how how much progress they're making you know in getting the open source truly open source tools, but uh I'm I I've found out about this stuff and you're interested in it and so I I. Would, uh uh if there's a role for me here, I might uh like to sit and be part of the taking it actually taking the course. So.

D

It uh depending.

C

On schedules and and stuff like that, but it's definitely a semester or two course. If you, if you, if you finish the final exam.

A

Yeah I agree. No thank you for the the um the pointer to Casper. That sounds that sounds familiar I'm almost completely sure I've come across Casper before yeah I I wrote a I wrote a message out to our list about the whole kind of the motivation for doing this, and and the motivation is twofold, one to be good at these tools like this is this is a workflow that keeps coming up over and over again. This is what the companies we interact with use and getting people up to speed with modern current. You know.

A

Industry standard stuff only helps us, as you know, as pretty as people that want to produce good designs for amateur radio and for both terrestrial and and space, but that's not the end of the story. So what we need to do is get some amount of competence across our community, like our group needs to get competent with these tools, which is hard because they're expensive.

A

We have an excellent opportunity here because there's a a really wonderful opportunity here, because we're a non-profit and we're open source. So we have access to these tools at a at a at a very nice um rate and we still pay for them. But you know the retail price of these is is out of our league, getting some people uh from our community getting a bigger diversity of people out there interested in this work.

A

Competent with these tools lets us go back to the open source tool makers, the people that are working on open source Frameworks for fpga and Asics, and it it lets us talk to them with a little bit of competence like authenticity like okay, we have trained with these tools, we've used them to produce work, here's what works!

A

Here's, the disadvantages and the advantages and the some of the tools that were that were looking at are things like amaranth, so amaranth is python related and it allows you to go from somewhat of python to HDL and our newsletter for April coming out in a couple of days, the the guest editor is Dr Daniel estefas, and he used amaranth to create a fpga, only Spectrum analyzer for the Pluto for the 9361.

A

and his experiences. He wrote us against us editorial and were that that'll be in our newsletter, so that's where we want to go. We want to make things like amaranth, bigger, better. You know we want to take on anything that we possibly can take on. You need to know what's being used in industry. You need to know what industry and proprietary tools are capable of in order to have a reasonable. You know good answer to it and we're in a unique position to be able to do this at this time.

A

So we are really focused as an organization on producing good designs. That's what we want to do. We want to produce the the designs human readable, completely open source. However, we get there doesn't matter whatever the best tool for the job is we're told HDL, coder and Matlab are excellent. Okay, we're going to go all in on that.

A

It doesn't mean it's an either or you know we're we're providing this to people and it's a broad benefit to individuals and to to us as Ori and to the Hobby of amateur radio and to open source Digital radio.

A

This sort of informing ourselves by like actually using the tools, we're going to use them to produce work, and then we can go and say: okay, we're in the open source Community. Is there an opportunity to to take on some of these functions and to provide tools that perform at least as good as what is? uh Is industry standard? So those are the two different things that we want to do.

A

We want to get to the end result, which is to produce the designs by whatever means necessary and to broaden the scope of of fpga Open Source designs, which really needs to be broadened. I think we can all agree, not a whole lot. There needs to be more.

A

um You know it's not in the same league as open source software and other languages and for their applications, and also to to be able to speak more intelligently to the tool makers that are working very, very hard to make open source fpga stuff possible.

A

So those are kind of the two two factors and I will go and track down: Casper and and educate myself that I'm familiar with the name and I'm, uh but not not the specifics and I think Rick had some comments and he I saw him talking, but earlier, but I did not see. I didn't hear any audio, so I'll turn it back over to you. Rick.

D

Well, I I turned off the audio so that I wouldn't bother.

E

D

But irrelevant conversation.

D

um I think the courses outstanding um it covers material um when I have not touched on in some years, since the development of the HP SDR and the courses that were given uh associated with that by Eric, Weidman and I was just trying to figure out how many years ago that was, it was too many. But.

D

Okay, that course was 10 days long I think.

A

10 days so was it was it two weeks and and a weekend in between.

D

I'm, not quite sure, no I think it was one day a week um and it may it wasn't like an eight-hour day either I think they were done in the evenings. I can't remember: okay,.

A

Yeah, that's that's more reasonable.

D

I did find on the openness, open, HP SDR site, um a list of all of the notes and the and the MP4 um video on of the training course, which is still as relevant today as it was then um yeah.

A

D

Of course, much of the source code, uh albeit in Vero one uh most most of my work now is, is not in very long but uh and it's not for silence Parts, as for all Terror Parts, but if you're looking at the broad issues, who cares, uh the important thing is the algorithms and how they're assembled into what became hbsdr and they're all the same algorithms.

D

Well, almost no, the same algorithms I'm talking about in this course, except it did not include any information on polyphase filters and, of course, and did not include thanks to you. I now have the second generation well, second generation of Fred Harris's book, which is daunting, the say the least, but.

A

D

E

A

Company I think, but you.

D

Know together together, we.

A

Can get through it no problem sure.

D

Yeah I read every page of the first iteration or the first edition, but now to go through the second edition and try to find out what got improved is a little difficult but I'll do it um so yeah the course sounds. Great um I have mixed feelings about.

D

When, when were you proposing having this.

A

This is second through the fourth or second through the fifth of May.

D

A

So about a month, that's what that's when that's when math works could fit it in for us there's. You won't be surprised about this I, don't think, um but there's one person at mat at Matlab that has the competence to teach it, um because this is advanced. Work in digital, Communications and and Matlab teaches teaches courses across the Spectrum, but they have one person that that can teach it and there's they're in uh they're in hot demand. So so our the offer to us is uh early May.

D

So what happens uh if you can't make it uh I, will it be recorded, so it can be played back.

A

Question in ninja classes from from math works or not and their their workshops, so they're pretty interactive and I've been to one before and as a recording. It's not a straight class, um however I'll ask because I have to get back with them today we're going to work over the next week to finalize everything. So what happens? If you can't make it that's a good question.

D

And in particular, I'm uh I've been out of plain April 30th for haystack Observatory um I have the first is um repairing equipment at the Westford telescope and the rest of the week is teaching a course at the technical operations Workshop, but I think that's exactly the same week. Yeah.

A

Yeah, of course, yeah there's, never a good time is.

D

E

Yeah, you meant.

A

You I think you mentioned this before your your.

D

Experiences and your.

A

Schedule yeah.

D

It's not scheduled, uh likes have been booked and hotel rooms or booked, and.

A

D

Right: okay, I'll.

A

I'll ask uh you know what do we do for people that are really super enthusiastic, but this date is not does not work for them at all. I'll ask about recordings I'm, pretty sure that the answer is they don't provide recordings, um but what we're also trying very hard to do is put together a a cruise with people to teach fpga design in the future.

A

um If we have a large waiting list for this class, then then I think we could offer it again if it turns out to be successful because it is, this is Advanced. This is uh generate. This is a class that is combined, it's four different, pretty substantial classes from MathWorks, and we we worked through the four different sets of of curriculum to try to come up with one yeah.

D

A

D

In in Columbia Maryland, which is where I obtained the first of my Pluto's, uh the class was relatively inexpensive because I think uh Analog Devices sponsored it. It was pretty much the cost of the Pluto for the class, but it was only one day and in focused it. It focused on that one central part of this class, which is build a radio and put it in a Pluto and I believe it was with math math works and simulink.

A

Oh yeah, that's undoubtedly the software-defined radio class that they have because that's exactly what they do and a lot of this there's there is material in our class, that's drawn from that one. So it sounds pretty pretty similar, so I mean my my sort of ambition here is that we we pull off this class. We get a lot of feedback. We look hard at what we're trying to do.

A

We get some good results with with using the tools we get more people with their hands, dirty and and feeling like they can control this stuff that, yes, this is understandable, yeah, it's hard, but ordinary people like you and me, and Mike and James, and all of the people. Listening to this, you can do this. You know you might have to hit it a couple of times really hard with the tire iron, but it's gonna, it's you're gonna get it!

A

You know it's uh a lot of times the stuff's made to be out made out to be probably more complicated than that really is like a practitioner, a practical person that just wants to get it to work. You can. You can figure this stuff out. You can get it to work. You can get an appreciation for the math, you can understand. What's going on and once you have an intuitive grasp, then a lot of the the more Ticky parts- hey.

A

You know there might be a library call for that or you know you can get somebody to help you. You know so I'm an optimist here and and I'm looking for this to really kind of open things up on the the digital experimenter side, uh for a community that really kind of needs it, and- and we can do this again- you know we can take this fine tune it. We can take what we learn. We can. We can teach you ourselves, you know we can.

A

We can spread it around uh and and just get it out there to the to the community a little better than it is so. That's kind of the goal have.

D

You chosen a family of Arts or even a specific part or you're ongoing design efforts.

A

Yeah, the Baseline Parts, the 9371, that's what we've been using in remote labs and that's the target. It sits on top of the zc706, which is a 7000 series, zincs, so 7000 series series, zincs, zinc, part with a 9371 is what we've been gravitating towards for the that's for the trans, with the space transponder, space and terrestrial transponder, or what we used to call Five and Dime.

A

The uh ofdm um drone slash deep space work, uh that is on a zcu 106, which is an ultra scale plus part, and we have a board ready to go in our lab. We just got a adrb 9002, that's a TR, the sort of the Next Generation transceiver chip, so the 9002 part will go with the ultrascale Plus board for that particular ofdm project. So there's two different Hardware platforms we have available. We also have the the Pluto. If you want to use a 9361, it's awesome. It's great, there's that's available too!

A

So we we try. We we have paths forward for all three of these uh chips, so yeah, I, I, understand completely and I have all empathy for uh schedule, challenges and there's you know, there's only you know you look at it. It's like what 52 weeks in a year that means only 52 weekends. Out of a year, it's impossible to schedule a conference, a symposium, a meeting, a class. It's really hard and I will do all that I can to make it possible to get something out of this.

A

Even if you can't get to it and your feedback about the material is deeply appreciated. We're trying to get it right.

A

And I know that we're going to have to provide support to people that are almost there, but not quite there. We don't want them to show up and get blown out of the water. You know in this class, but in order for that to happen, they've got to to get some of the stuff under their fingers, and you know it's doable.

C

Yeah you know, one comment uh is I realized that it probably doesn't work schedule wise, but uh Rick's, comment about this course she attended. It was like one day a week over weeks, rather than uh Bunch it all together.

C

We would be really in a way better, because that way, the the people that that really want to get into this have time to to learn the things that they they need to learn in between days right yeah. Maybe that's not practical.

A

Well, I I am in strong agreement, I mean I'd, be a fantastic model um and what I'll do is I'll I'll go back and see what they think Matlab and, and these courses are really set up for, like eight hours a day with breaks. You know the typical corporate training and they're, either one or two or three day classes for all the stuff they offer, and so that's how they're used to delivering it.

A

But I'm gonna talk to them and tell them you know and and they're, obviously familiar with corporate training that you know all of these people are Pros. They all know and I'm sure that they will be able to immediately. Tell me why they choose to do it this way, but I'm going to ask. Is there any way that we can offer this course over time, so that people see what they're going to be doing and then can ramp up and get ready for it, because there's an awful lot of free training?

A

That's really excellent from Matlab about digital signal processing, image, processing, artificial intelligence, machine learning just go to their website their little Matlab Academy. It's great I've. Already taken a ton of their courses, I spent lots of hours for free, getting good at Matlab and simulink and learning like ordinary stuff about DSP.

A

So if he had something that was spread out in time, it'd be a whole lot easier for those of us that have day jobs and still work to work it into our schedule. uh You know, or going to school and and also you'd, be able to kind of backfill.

A

If there's anything that you needed to know so, I'll ask for that for May, we're probably going to get a fire hose, uh but but I view this as sort of a you know, an ongoing challenge for educating people in digital Communications, experimentation for open source and amateur radio and whatever it takes to get it working good.

A

uh You know Count Me, In, so I will I will raise this issue today in the meeting with with Matlab and see what we can do so I definitely appreciate the feedback, I I think you're on the right track: cool. Okay, um let's see what other we do. Let's see we have posted a uh unboxing video of the adrv 9002. So we we open it up, we show you what's in it, uh that's posted on all our social media and I think it went out to the list.

A

So it's exciting to see, and then what I'll do is bring it up in the lab and do all of the um oh, like hello worlds with this uh transceiver chip and the the baseboard, um if not today, then very soon over the next week, because we we have a team that wants to use it so they're they're on me, man, they're, they're, asking for it so so I, better, better, get it up and running and the way that you access this is through our virtual machines.

A

So you log into our uh or our big computer in remote, Labs, West and.

D

A

Into the particular virtual machine where this particular Hardware is attached to, and so that's how we that's, where we provide the access um and in the show notes for this particular video I'll, put in the the how to get how to get access to uh virtual machines. They'll put the page in so that, if anybody listening is super interested in again getting involved and they know how to apply for an account.

A

And yeah that's pretty much it. We did have an unscheduled power outage at remote, lab West. um We had an earthquake here, so the power outage wasn't specifically linked to the earthquake, but it happened in short order. uh So so who knows? uh We still don't have an explanation, but but it's uh it's always exciting.

A

When you have an unscheduled power outage, because you have to bring everything back up and- and we did find some things that needed some- some maintenance so uh kudos for power, outages, uh the the flooding and the storms that have affected um most of California have not really affected San Diego. Very much so we're we're okay.

A

Here, um it's way more rain than we usually get, but we're not um we're not suffering a lot here and fortunately, remote Labs South, uh as James explained, um was surrounded by some extremely powerful tornadoes that have caused a lot of damage and have made my day job interesting. But we uh we did not suffer any physical plant damage at remote, lab, South and everything is, uh is working well, so we're looking forward to a good summer and to demonstrations at uh IMS 2023.

A

That's an IEEE conference, International microwave Society or International microwave Symposium 2023 in San Diego. So we should have some demonstrations there in June. We'll have a big demonstration at Defcon in August and we're a big part of the show at qso. Today, ham, Expo in September it'll be 9th and 10th of September 2023 and we're going to sponsor and help organize and get things going for another great Show online.

A

There is also in September. There is a uh IEEE iWork conference, that's in Little, Rock Arkansas and it focuses on chips, act, money and remote lab South will be a big part of that, so lots of stuff happening over the next year or so, and everybody's invited to to participate and we're working really hard to make. All of this. uh This effort turn into good things for the open source, Digital, radio and amateur radio communities all right any last comments or questions before we close.

C

I, have it off topic the question if you have any time after this is over? Oh.

A

Sure, no, why don't I go ahead and and close out the meeting the recorded meeting and then uh then we'll have office hours? Thank you, everybody and see you again next week.

E