Role for FPGA or CPLD with Raspberry Pi

Hi all,

Thanks for pointing me at this Morgaine.

I think there are a few point to consider here I want to throw in. However I want to be clear that I'm very pro getting schools, collages and Uni's more hands on and involved with things like Raspi and FPGA/CPLDs.

First off from a education point of view we have to consider what you guys are talking about and the target audiance. As someone who has teached kids in electroincs clubs and takes part in school engineering contests (as a judge) I had written this blog on my first views of the Raspi in educations. This was followed up by a head of ICT who poested this on Deaign Spark too.

My point is however that education is just not geared up for the Raspi and certanlly not for FPGAs. I ahve spoken to a teacher only yeasterday about this and she confirmed that they have no idea what they are doing. So throwing FPGAs at them will scary them off. The key point she picked up from my blog above is that unless a package is supplied to schools that is clear and simple to use then it will not get past the teachers desk. This sounds harsh, but its so true. Its take two years for the Arduino to apear in school contests.

Moving on to haveing a CPLD or FPGA attached to a Raspi I think is a good idea. The Raspi is limit in GPIO and anything that can help expand this will be good. As a first off I would use somthing like a simple CPLD as a direct access to the GPIO as this will hold its configuration at power up, allow simple patcheing and basic logic circuits to be run from it.

The XAPP058 is a great starting point for someone to generate a JTAP programmer for the Raspi. This can then use XSVF files to program either the CPLD or another FPGA if attached.

At this basic level I think you could then start offering not only programming but some simple VHDL / Verilog to schools and collages. However as pointed out you can't run the tools on the Raspi so will HAVE to have another machine for this. You then have to ask if this is protical. What is the Raspi really doing in this project that you can't do with a PC or some of the specially designed education kits like ther terasic DE2. Find that answer and argue it and then you can push Raspi and FPGA tech togetter into schools. However in my view, use pre complied XSVF files to teach logic and basic electroincs via the Raspi.

Moving up as I said you can then add larger FPGAs like the Spartan-3, good price point and enough flex for entry levels. However I think what your offering is a high end education kit or a really nice development kit.

Anyway just some of my ideas tio throw in - truned intoa long answer so sorry about that but would be intrested to see anyones developments and hear your views too.

Paul

(@monpjc)

Hi all,

Thanks for pointing me at this Morgaine.

I think there are a few point to consider here I want to throw in. However I want to be clear that I'm very pro getting schools, collages and Uni's more hands on and involved with things like Raspi and FPGA/CPLDs.

First off from a education point of view we have to consider what you guys are talking about and the target audiance. As someone who has teached kids in electroincs clubs and takes part in school engineering contests (as a judge) I had written this blog on my first views of the Raspi in educations. This was followed up by a head of ICT who poested this on Deaign Spark too.

My point is however that education is just not geared up for the Raspi and certanlly not for FPGAs. I ahve spoken to a teacher only yeasterday about this and she confirmed that they have no idea what they are doing. So throwing FPGAs at them will scary them off. The key point she picked up from my blog above is that unless a package is supplied to schools that is clear and simple to use then it will not get past the teachers desk. This sounds harsh, but its so true. Its take two years for the Arduino to apear in school contests.

Moving on to haveing a CPLD or FPGA attached to a Raspi I think is a good idea. The Raspi is limit in GPIO and anything that can help expand this will be good. As a first off I would use somthing like a simple CPLD as a direct access to the GPIO as this will hold its configuration at power up, allow simple patcheing and basic logic circuits to be run from it.

The XAPP058 is a great starting point for someone to generate a JTAP programmer for the Raspi. This can then use XSVF files to program either the CPLD or another FPGA if attached.

At this basic level I think you could then start offering not only programming but some simple VHDL / Verilog to schools and collages. However as pointed out you can't run the tools on the Raspi so will HAVE to have another machine for this. You then have to ask if this is protical. What is the Raspi really doing in this project that you can't do with a PC or some of the specially designed education kits like ther terasic DE2. Find that answer and argue it and then you can push Raspi and FPGA tech togetter into schools. However in my view, use pre complied XSVF files to teach logic and basic electroincs via the Raspi.

Moving up as I said you can then add larger FPGAs like the Spartan-3, good price point and enough flex for entry levels. However I think what your offering is a high end education kit or a really nice development kit.

Anyway just some of my ideas tio throw in - truned intoa long answer so sorry about that but would be intrested to see anyones developments and hear your views too.

Paul

(@monpjc)

I fear that Paul is right about the difficulty of the Pi in schools, and even more so when it comes to stuff like FPGAs. But I do think that their are plenty of people who could help out  - why can't schools get practising engineers in from industry or Unis to help. The schools my kids go to have always been keen to ask for money or fund raising help but never seem to show any desire to tap into parents/supporters brains. If on the back of the Pi wave some kind of support network could be built that we be a good thing. But it would need schools to accept the principle of outsiders doing 'teaching' .

Michael Kellett

One thing that would put me off visiting a school would be the requirement to get a 'working with children' card and its equivalents in other countries, and all the drama of insurance and saying approved things. But (never begin a sentence with 'BUT' ), my views seem to be stuck in the 60 and 70's when I was at school.

Hi Paul, nice to see you here.

I agree with everything you say about the problems in education.  I come from a very academic family, having been a university lecturer myself in EE and Computing before I went into industry, and half of my relatives being teachers in maths or science, so I've seen the difficulties personally both first-hand and second-hand.

But all is not lost.  While the schools system as a whole is just not geared up for working with the Pi because of lack of curriculum, staffing, skills, time and money, and even less geared up for teaching programmable logic, the Pi is so inexpensive that official sanction and government-backed investment are in many cases not necessary.  I have little doubt that many teachers will be picking the Pi up personally and just "making it available" to their keener youngsters, even when there is no scheduled time for class-wide involvement, and in some cases even when it's beyond their personal area of competence.  You don't need a government curriculum nor well-equiped teaching labs to spark interest in technology.  Inquisitive minds will be drawn automatically to whatever is  available.

I can't help but draw on personal experience of this.  At an age of 14 or 15 I was reading books on fine structure of stellar spectra and the stability of halide complexes.  My topics of interest were so far outside the school curriculum and the personal knowledge of my physics and chemistry teachers that my questions were more disruptive than helpful in class, yet the teachers all encouraged my inquiring deeper and deeper into what fascinated me.  What's more, labs and reagent cupboards were "made available" most definitely without official sanction (the school governers would have been horrified), and as a result lunch periods were a time of much fun and of very intensive personal learning and oiled the path towards a very technical career.  You just can't keep youngsters from learning if they want to.

The Pi is much safer than playing with chemicals, fortunately, but the same rules apply.  Teachers can easily encourage IT learning by making cheap computer systems available, curriculum or no curriculum, and the very low price of Raspberry Pi makes that possible.  It doesn't need to be taught class-wide at all, and it would in any case be the height of boredom for the majority,  The one or two natural geeks out of a class of a few dozen pupils is all that it takes to create a new generation of future engineers and scientists, and they will gravitate towards interesting technical things irresistibly if they are made available.

On the specific issue of programmable logic, you might notice that earlier in the thread I was trying to bring CPLDs into the picture, not just FPGAs, and this was directly in response to the worry that the more complex a system is, the less valuable it is for education.  Even CPLDs are too complex for educating kids about the principles of programmable logic, with far too much "and magic happens here" being done by industrial-strength power tools.  Education is very different to vocational training, and exposing youngsters to proprietary tools with mind-boggling interfaces has deeply problematic issues of principle as well as hiding the mechanisms involved.

Personally I think that even simpler programmable logic devices with basic open source tools would provide a better foundation for teaching the elements of this subject than any of the big name tools used for commercial work.  What's more, this approach would require nothing but the Pi itself, no additional powerful x86 PC to run the tools.  I know that all of us here are keen to use the most powerful devices and tools available, but that's because we are already experienced professionals, not pupils nor long-suffering school teachers.  There is a big disconnect occurring whenever someone suggests that programming a Spartan-3A might be featured in a school curriculum.  Yes it's possible, but it rather misses the point, and it adds major hurdles as well.

Morgaine.

I should also mention that while I was in academia, I experienced first-hand exactly the same thing as described repeatedly by Eben Upton of the Foundation, a very pronounced reduction in technical competency among our intake from UK schools over the years.  So many applicants lacked the elementary knowledge required to begin an EE/CSc degree that we had to offer remedial courses to bring them up to speed, or else we'd have 1st year intake consisting exclusively of extremely bright Chinese youngsters.

It really was depressing.  A major change is required in school education or we're heading towards disaster as a country.  It's a technical world, and the real currency is technical skills, not dollars.  You can't avoid providing your future citizens with a technical education if you expect to remain relevant in the world.

Morgaine.

"Logic!  Why don't they teach Logic at these schools?" asks the Professor in The Lion, the Witch, and the Wardrobe.  When I was a lad in USA schools, there was no instruction in Logic, and I imagine it's no different today and probably no different in the UK.  I had to learn Logic on my own.  My mother actually got me started by buying me a dice-based logic game called WFF 'n' Proof, probably because she didn't know what to get and the sales clerk said it would be "educational".  WFF 'n' Proof was pretty useless as a game since there wasn't anyone to play it with, but the instruction manual was a terrific introduction to symbolic logic and got me hooked on the idea of solving logic problems automatically.  The Madison, Wisconsin public library had a terrific collection of computer and electronics books, and before long I was making RTL circuits using schematics from the GE Transistor Handbook.  Then 7400 TTL showed up at an impressionable age and suddenly I could easily make really good logic circuits.  Bliss!  CPLDs and FPGAs brought it up to the next level, where you can do far more and not have to worry about the details of logic minimization and having to re-wire to fix errors.  Real chip design on the cheap, no masks, no minimum quantities, and no NRE.  Double Bliss!

My point here is:  So what if it's impractical to teach logic design in elementary/high school classes?  The kids who will run with it will do so on their own if the tools are available and cheap and there is good documentation.  Kids who don't want to bother with logic -- or thinking in general -- aren't going to be sold no matter how good the curriculum is.  To quote Jean-Jacques Rousseau (Émile, from memory): "A child should not learn by Rote what he cannot understand.  He must learn by Experience what he can understand."

@John:  My answer:  Exposure.

Kids can't get enthused about things that they've never heard about.  A Pi sitting in the corner blinking its LEDs during science class is a very powerful attractor for youngsters with the right mental disposition.  And if you can work it into science experiments as a tool to turn things on and off, you're halfway towards giving future engineers a lifetime of enthralling fun that just happens to be very valuable to society as well.

The process is self-selecting, but it needs a spark.  And in this new Pi era, sparks are cheap to provide.

None of this argues against your point that many kids will find out about technology on their own, which is very cool.  But not everyone will, and given the price of a Pi, providing additional exposure in schools has an astronomic benefit/cost ratio.

Morgaine.