Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

No Analog electronics in UK?

Status
Not open for further replies.

Flyback

Well-Known Member
Hello,
In 2002, I was in the final year of an electronics degree at a UK university. All the various lecturers were coming to speak to us about their subject, as they wished us to do a masters and choose their subject.
The Microwave Engineering lecturer came in, and announced to us all that "Microwave engineering is the last bastion of analog electronics in the UK".

Does this mean that there are active political plans to get rid of analog electronics in the UK? I regard SMPS as analog electronics. Do the UK politicians wish to get rid of SMPS work in the UK?
It makes you wonder because below the age of 16 years old, no single UK school teaches electronics. By that I mean they don't teach about eg stuff like Transistors, opamps, 555 oscillators, linear regulators etc etc.

The only semiconductor that's discussed in UK under 16 education is the rectifier diode. That's it.
 
I left school in 1999 (in the UK) and I did hardly any electronics at all, I mean we didn't even learn IT, most of what I know is self taught, I did go on to college and pursued a further education in electronics and computer engineering, but by that time I knew most of it and was just obtaining the bits of paper to say I passed.

Universities I think will only teach the basics and give you a good lesson on how to research subjects.

Electronics is a vast field, a lot of suppliers are going down the single chip solution and give you baby steps on how to implement it, i.e. specifying exact capacitor and inductor values etc so people are not having to think as much about the theory behind it all.

I don't think analog electronics theory and the like is as relevant any more unless you are designing specific products, these days it's all digital and micro controllers. I don't have a particularly strong analog background, but I know enough to get by and do what I need to!
 
Everything is analog when speeds are fast. Most of the 'black magic' of digital is dealing with analog signal effects. So if you don't have a good hunk of analog and RF under your belt you're working with one eye closed.
 
switch mode power supplies are analog electronics, and they are about the most ubiquitous circuit in the world. The majority of smps's are controlled by controller ic's and very very very few are software controlled...that is how it will be for at least the next two decades.
 
switch mode power supplies are analog electronics, and they are about the most ubiquitous circuit in the world. The majority of smps's are controlled by controller ic's and very very very few are software controlled...that is how it will be for at least the next two decades.
I don't think analog control for SMPS's will last anywhere near two decades. The trend is to do the control loop of an SMPS with a digital processor, either emulating PID control or using Fuzzy Logic control. Such a circuit generally requires fewer parts and can be made more robust with respect to component tolerances and drift since the feedback parameters are determined in software, not in the value of the feedback components. It's also simple for the processor to perform functions such as a slow startup, or have adjustable voltage or current limits under software control, and report on the status of the supply through a simple bi-directional serial bus. This can be particularly important for power supplies in large systems. So analog may not be dead but it is certainly getting squeezed out of many tasks that it used to do.
 
Im sorry but that is a myth.
Vast majority of smps's are mains connected and have an opto in the feedback path......what good is it to have software compensators when the parameters of the opto are highly variable and extremely complex code is needed to account for that...there just aren't enough software people in the world to do this.......do you remember when they computerised the NHS records in the UK?.....It cost billions because there weren't enough software engineers and those that were there charged a fortune.....software engineers don't have any desire to teach their trade to others who will then become a competitor in the jobs market.

At the end of the day, for an smps, the only way to ensure a stable feedback loop with adequate gain and phase margins is to measure it with say an AP300........so whats the point in doing the feedback compensation in software when you still have to verify it with the frequency analyser.

The DSPs required to do software control of SMPS is far far more expensive than the cheap analog controller ICs that abound. -the software engineers that can understand the control algorithms are even more expensive.

Don't get me wrong, a company called Gardasoft do software controlled smps, but that is because of the particular performance requirments of the smps's that they design there...these type of performance requirements just don't exist in most SMPS's.

Please Read the word on software controlled smps from Dr ridley...attached
 

Attachments

  • _digi smps revolutiopn.pdf
    1.4 MB · Views: 379
In the end as things get more complex and more integrated digital will probably prevail. I can remember people saying disk drive servos would always be analog loops, but now they are all DSPs. Not so much because DSPs are better or cheaper but because they can perform multiple other tasks at the same time they are doing servo. I can, for example, see a DSP in a TV set managing a whole lot of "stuff" and doing power supply management in it's spare time.
 
Im sorry but that is a myth.
....................................
Well, time will tell who's right. But don't think it will take 20 years to find out.
 
digital electronics is just a form of analog electronics. Software makes use of digital electronics.

How can a digital system ever be quicker than analog, when all digital systems need the analog signal to go through an ADC and anti aliasing filter first......and those two are analog stages anyway.
The digital system is always going to be more expensive...and need expensively payed software engineers.
 
Heaven forbid SMPS goes Digital...

Remember, SMPS is something where if you still have the slightest thing wrong after a rebuild, switch on and BANG. A Variac barely helps these days anymore.

I shudder in my boots. They are already hard to repair....with Digital control I believe nigh impossible :nailbiting::nailbiting:

Regards,
tvtech
 
Heaven forbid SMPS goes Digital...

It already is, and always has been, the clue is even in it's title - SWITCHING - digital electronics switches ON and OFF (like a SMPSU) and analogue electronics works in a linear fashion.

Processor based electronics is only a minor part of digital electronics, and as already suggested there's little reason to want to use a processor based solution for a SMPSU when existing digital SMPS chips are simple, cheap, reliable and easy to use.
 
Also, Dr Ridley , in post#7 above, also says why current non processor based chips are actually already "digital" .

There is a reason why motor drives have gone to processor based drivers, but I forget what it is...I think its because the switching frequency is much lower.
 
It already is, and always has been, the clue is even in it's title - SWITCHING - digital electronics switches ON and OFF (like a SMPSU) and analogue electronics works in a linear fashion.

Processor based electronics is only a minor part of digital electronics, and as already suggested there's little reason to want to use a processor based solution for a SMPSU when existing digital SMPS chips are simple, cheap, reliable and easy to use.

Hi Nigel

You have made me think....can you believe I have never regarded SMPS as Digital? Seriously.
When I think Digital I see +5v =0n..0V =off...

And nothing in between. Maybe it is because with SMPS you are dealing with Mains where everything can go wrong in the blink of an eye....
You will not find me with a Scope scratching around in the Primary of a SMPS....in fact you will not even find me with my YEW there unless I am checking to see if the supply startup Voltage is there....still steady hands though...slip and all is history.

Crazy crap.

I still cant see SMPS that I deal with daily as being Digital. Sure they switch....does that make them Digital though????

Regards,
tvtech
 
I think they say they are "digital" because the output of the pwm comparator is either high or low (fet on or off).
I don't know if that really makes them "digital " though.
The situation is that of feeding the output voltage divider signal into an adc and operating on it with software, in order to control the smps.........there is a very good reason why this is probably never likely to happen in massive amounts, but I cant explain why.....it just seems obvious.

The ridley article is good on this (#7 above)
 
OK.

When a SMPS Primary blows up....sometimes the Secondary regulation is damaged too....the ONLY natural way that can happen (besides a lightning strike or other act of God....or if it is a TV LOPTX arcing over to the Primary) is if somehow the Opto has let the team down.....and compromised it's main job.....to keep isolation between Primary and Secondary and provide feedback to Regulate the supply...

The damage to the Secondary regulation side can be substantial....blown transistors (faces missing) etc. Tedelex 51CM (every thing else good) from here are legends with this....fixable and worthy because discrete components are used in the Secondary regulation side....

Now picture this: Your SMPS has Digital control/feedback. Isolation barrier/feedback is compromised.....as above....now you have a Microprocessor controlling that simple task.....

Expect nothing but an impossible repair.

Regards,
tvtech
 
I tend to agree, and of course , as you know, theres more than that...for example, a switch mode led driver using a zxld1360 ic is not software based, and is extremely cheap and simple, wouldn't it just be complete and utter overkill to have this done with a software controlled control algorithm.?

fig 1 page 2........
**broken link removed**

Goodbye Software controlled SMPS......you are the figment of the imaginations of overzealous DSP sales people.

Yes there will always be those niche applications.....but even then, we know its just people doing pie in the sky research generally....don't get me wrong , sometimes pie in the sky comes good.

We have several decades of analog smps controllers ahead of us at least. As long as we don't run out of oil/gas/coal first.
 
I think they say they are "digital" because the output of the pwm comparator is either high or low (fet on or off).
I don't know if that really makes them "digital " though.

Of course it does - as I explained earlier, the difference between analogue and digital is dead simple - and SMPS's are digital - they switch ON and OFF.

Likewise a class D audio amplifier is digital (hence why they are called digital amplifiers) because they do nothing except switch ON and OFF - very much like SMPS's really.
 
ok, thanks, I concede that, (not that I devoutly disagreed on it), anyway..The concensus appears to be that software controlled smps feedback loops are pretty much pie in the sky, and will be for many, many decades.
I remember being at a radar company, and we bought in these DSP controlled SMPS demo boards from microchip....they actually came with instructions and bits of sample code so that people could program the dsp and control the 5W buck converter demo board.
I had a go at programming it and failed dismally.

But that company's principal software engineer also was not able to do it.
I think that says it all about software control of smps feedback loops...."pie in the sky"

Not only that, but the DSP board actually came with some code that could just be loaded in and it would work (control the 5w smps) ...but even though it was just a 5w buck converter, the dsp ran so hot that you couldn't touch it!!!......indeed software controlled smps feedback loops are pie in the sky..do you agree?
 
^^^^
Nigel deserves a Rep from you. The very first time he has replied to one of your posts/threads Flyback....you should feel honored :cool:

Regards,
tvtech
 
Status
Not open for further replies.

Latest threads

Back
Top