Modern Construction

[Wendymott]

Is it an "age" thing. I still attempt to construct radio's and transceivers that are reasonably modern, but not "cutting edge" ..... me "cutting edge" thats funny.
It keeps the grey matter churning over and me off the streets. However I think we / I are getting to the limits of what is now possible. When I was in the TV industry, there were lots of parts available to make.. even Nicam decoders or Colour decoders with relative ease, until the Dreaded I2C controlled devices started appearing, thus unless you had the knowledge to programme your own controllers, television construction was confined to "old Tech".
It appears that Amateur radio has gone the same way, yes its progress, but the final nail in the coffin of home construction. All the modern "Black boxes" have massive component integration, making it impossible for us "plebs" to make anything like. I am stuck in the era of MC1496's, MC3357 and the like. Thank goodness there are a few about.
LSI "large scale integration" is now the thing, yes it keeps costs down, but reclaim or reworking is now a thing of the past.
I am fortunate to have a friend that writes good software for the PIC microcontrollers, but thats it....without that I would still be plugging in FT243 Xtals.

[kellys_eye]

I would agree Wendy but I don't really worry too much. What it's done to me is encourage me to explore even older technology to the extent I am now 'in to' valved receivers and transmitters. Ok, I still enjoy the basic semiconductor units but not to the extent this would take me into LSI or 'hard to obtain' devices with limited longevity.

The push 'forward' has obvious advantages - no point in denying the versatility of SDR for example - but when it comes to the ability to build a 'simple' valve tx/rx it gives me a much greater sense of satisfaction than hooking up a black slab that does it all.

I'm an even bigger luddite than you!

As an aside, I can 'croon' over a nice tetrode but feel nothing towards a DIP package!

[Aub]

Kelly's, I agree completely.

Aub

[G6Tanuki]

I think the modern side of the hobby has sort of moved over from the chassis bashing and soldering aspects to the code-cutting types. Whereas in the past we did calculations for L and C for a pi tank these days the technology interested hams are compiling their C++ DSP code and running it on their raspberry pi .

[Radio Wrangler]

Also, some of the large asics have no published data available to amateur constructors, only to companies with NDAs and contracts.

Similarly software can be controlled and compilers or general information may not be made public. One example is the cpu at the core of the raspberry pi.

Note that copyright lasts an awful lot longer than patents.

David

[Andrew2]

I agree, Wendy. I'm currently mucking with some FM TV on the amateur bands (I last did this in the early 90's using satellite LNB's and homebrew receivers) and finding components to do the same thing these days is not easy. In the past I've used NE564 PLL demod chips but these are difficult to find now. As for the more 'modern' (!) SL1452 as used in the old analogue satellite receivers, forget it unless you want to pay through the nose on Ebay. Like you I'm still diddling with 30 year old tech and I cannot get my head round most of the more recent stuff as without a clock generator and an some kind of data it won't do anything! Gah! Luckily I have a couple of complete satellite front end/IF/demod boards kicking about for now. We're doomed I tell ye!

[emeritus]

It's ironic that a private individual can perfectly legally make something protected by a patent for their own private, non-commercial use while it is in force, but cannot legally reproduce, by copying, something protected by copyright for at least 70 years after the author's death (the period keeps getting extended by international agreements and I have not kept up-to-date since retiring). On the other hand, copying must be proved, which is why it is good practice for the creator to include redundant code or features to provide evidence that something was actually copied and not created independently.

[Wendymott]

Not just me then.. I wonder how many of the members are over 60...this means that every year, there will be less, constructors and enthusiasts. Without the tech books, in 50 years, the "kids" will be asking.. What Valves were, and how did they work.
I did buy a Raspberry Pi when I retired 8 years ago, but I couldnt get my head around the code writing..so I donated it to the DDARS....I understand assembler..... and I wrote my own code for my TV pattern generator, but I realised there were a lot more clever people out there than I, so I stuck to what I know.

[Craig Sawyers]

There are people out there who use the raspberry pi as the core of a music server. The mind boggles.

Craig

[duncanlowe]

A raspberry pi is quite hard to learn to use if you aren't familiar with certain aspects of computing. I've been programming some things since I was at school, albeit I was one of the very first. I was taught to program in C, not C++ or C# just C.

I found that PIC controllers had a free C compiler so it was relatively easy to start programming these. You can get them to do something with very little understanding needed and learn on the go. Of course these are just controllers and you need to build a circuit around them DIL versions were still available last time I needed some.

Then, along came Arduino. It's a bit like a halfway house between PIC controllers and Pi. Ready built and ready to go, and you can do simple stuff very easily.

I've now started to understand Pi a little, but it's certainly way harder.

along came Arduino

And it's very good development environment, all the tricky stuff is done for you (port pins etc. which are particularly long winded for ARM devices). Lots of other micros have been ported to this tool. I was skeptical to start with "hobby stuff" I thought, but no, you get code running quickly without all the hassle of setting up the various bits of the chip. It is free and lots of libraries are available for all sorts of sensors, displays, the list goes on and on. Hardware is cheap too.

Back on topic, in my 50 years or so of playing with electronics I have learnt to design, use and construct SMT stuff, I have no problem hand soldering 0.5mm pitch ICs or even 0201 components. Mind you for work we have to use unleaded solder, lots of flux is your friend for that.

[kellys_eye]

I've built a standalone SDR receiver (32-bit development board, 320x240 LCD etc) that used 08 sized SMD and also many pieces of test equipment based on the Arduino but still feel more at home with the valved stuff - despite the fact that I grew up at the 'end' of the valve era and rarely dealt with them during my industrial days - the only genuine valves I came across were magnetrons and 4CX250Bs in marine equipment - most stuff moved to solid state as my career developed.

The older I get the more I regress! I'm in the over-60 bracket and like to keep abreast of all types of electronics delvelopment but the software side has certainly overtaken me. Like Wendy I used to code in assembly and have done some PICBasic and occasional Arduino stuff but the comlexities of it aren't, I feel, worth progressing with. There's still a LOT to be done with simple hardware.

At the near end of sixty I find ratting off a bit of code to do something quite easy and enjoyable, the interface to the real world is the best part. The fun is getting it working using available "libraries" done by someone else, I don't cut my own screws as they are sold in packets at a reasonable price. I will, however, cut a thread for something that isn't "out there", same with software.

[Andrew2]

I'm 68 now and started in the TV trade in 1969. Everything was hardware and no surface mount stuff. Needless to say, most of my homebrew projects were made from TV components, including UHF tuners etc.
I got into computers (ZX81 then BBC-B) in the early 80's and became reasonably proficient in BASIC and wrote many bits of software for controlling external gadgets, but I never got into assembler or anything like that. A few years ago a friend persuaded me to go down the PIC route, but I found myself completely boggled-out by it and it gathers dust now, virtually unused.
I think I'll be fiddling about with the old stuff 'til my time is up.

[Malcolm G6ANZ]

I agree with Andy (#6) It seems that everything has to have a processor of some sort in it. If you can write code and spend time debugging it then great. But there are those of us who never learnt to write complex code be it in basic or C++. So to get a processor to do anything useful you have to rely on the program already being written.
I have thought about going back to transmitting fast scan television, but have been put off by the cost of a digital encoder and transmitter.There is no way I could write code to digitise a video stream, so I would be buying a black box and plugging it in not really learning anything. I could probably generate FM TV on 23cms but this mode seems to be out of flavour and 23cms appears to be dominated with repeaters. Having got a 23cms receiver working and tuned to the local repeater I watched it over several weeks and saw no activity from any station. Just the test card.
I'll just keep playing with my PAL CCTV and not bother with transmitting.

Malcolm G8KGH (EX G6ANZ/T)

[TonyDuell]

I am happiest designing with lots of simple components -- transistors, op-amps, simple logic ICs. I dislike 'design' by sticking modules together, such modules include many LSI ICs and software libraries. There is the tendency to make the problem fit whatever solution you have, rather than to make the solution fit the problem. This to me is bad design.

I am not a programmer. Yes, I can program, but I don't enjoy it. I am much happier ith a soldering iron. I tried an Arduino module in a project last year, I got it to work. But I didn't really enjoy it. I'll probably use them again, but only if I have to.

I got a Raspberry Pi some years back and I found only 3 things wrong with it. The hardware, the software, and the documentation. It's odd, I like unix, I am happy with a shell prompt. But it's not what I want on an embedded system. It's not a real-time operating system and shouldn't be used as one by assuming the processor is fast enough to avoid problems.

A lot of modern electronics seems over-complex for no good reason. That Arduino project I mentioned was to link a keyboard to an old Microcolour graphics terminal, which in turn will be linked to a VAX11/730. When I looked at keyboard interfaces it appears that to use a USB keyboard you essentially need a Raspbrerry Pi or ARM-based Arduino. So let me see if I understand this :

I have a VAX, clocking at 5MHz, supporting half a dozen users

I have a graphics terminal based round a 68B09 processor

But to link up a keyboard I need an ARM clocking at around 80MHz.

No thanks... In the end I used an older keyboard with a more sensible interface,

Alas this forum is not immune to such 'design'. I remember some years back asking about the auto-start adapter for a Uher cassette recorder. It had to pulse a line low when the thing powered up. One suggestion was a PIC processor along with a voltage regulator to power it and level shifters on the inputs and outputs. My solution? A couple of diodes and a capacitor.

[Maarten]

The fancy solution for your pulse would have been to use a voltage detector IC (3 pins)...

The USB protocol is needlessly complicated if all you ever want to do is connect a keyboard. Even the AT protocol is overkill. An XT keyboard should do just fine and could probably be transcoded using a PIC (much like the role it was originally intended for: controlling peripherals).