Making 60Hz from 50Hz

[ppppenguin]

There is an occasional need to make 60Hz mains from 50Hz or vice versa. This site gives a simple approach:
http://sound.whsites.net/clocks/freq-changer.html



Mod edit: Updated link: https://sound-au.com/clocks/freq-changer.html

[Nick]

You could do that all with one 8-pin 50pence  chip and a half dozen external passives... 


Plus a few lines of code. Shape the input AC 50/60hz (or 100/120hz), bung it into a low-end avr running off internal clock, determine if we're looking at 50 or 60 Hz in, then use an input edge to drive one output at 50hz and one at 60hz (simple maths)..

Whilst all that traditional MSI stuff is great, it's not how things are normally done today...

[ppppenguin]

The design I referenced does give approximate sine waves rather than square waves. The modern approach to this would be digital waveform synthesis and a simple DAC. Possibly as simple as a few resistors. Then a power amp as before.

For most purposes there's not much point in referencing the output to the input frequency. A crystal is more than good enough for most purposes.

[Kalee20]

A really good article! Interesting that the frequency requirements are highlighted in the preamble, that an ovened crystal oscillator is barely good enough and the AC mains is much more frequency-stable, long-term.

The idea of linking the generated 60Hz to our mains is super. Clocks etc will have exactly the sort of short-term gain or loss, while long-term will behave exactly as they should.

I wish I was interested in American electric clocks, then I'd make up a unit!

[Refugee]

I would just use a phase shift oscillator and strobe it to a neon lamp with a strobe mat for the local mains frequency.
Most reasonably posh turntables have a strobe lamp on them anyway.

[Terry]

... strobe it to a neon lamp with a strobe mat for the local mains frequency ...

But surely, it isn't the local frequency that you want to lock it to?

Here's a 'dual standard' version you could print:

   

[Nick]

In the UK, power companies are olblighted to stick to 50Hz +/- 1%, i.e. 0.5Hz variation is permitted. There are strict rules governing permitted excursions outside these limits, see http://www2.nationalgrid.com/uk/services...-response/ for the simple view...

I'm not entirely sure whether long term frequency stability is still guaranteed - it used to be the case that the generating authorities would correct any daily deviations during the night to keep any synchronous clocks relying on accurate long-term 50Hz for their accuracy correct.

Years ago, nixie clocks, station clocks, time recorders etc. used to rely on the mains for their frequency standard, but that largely changed with Droitwich, MSF Rugby, DCF77 and others. Now of course, it's pretty common to use GPS-disciplined time.

The truth is that there is pretty much no reason to play with the supply/demand numbers for the National Grid in order to maintain long-term 50Hz stability - no-one of any import cares, and it is an expensive exercise.

See http://www.dynamicdemand.co.uk/grid.htm for a bit of fun

[jimmc101]

There are some interesting ideas on Roman Black's website using a PIC, in particular '50Hz to 60Hz sinewave converting inverter!' about 3/4 the way down the page.

Jim

[Nick]

I think I first used Bresenham's Algorithm in about 1983 when working for DEC (Digital Equipment Corporation) in the USA & Canada (the job was a custom system for BP) - I read about it in Foley & van Dam's classic work, Fundamentals of Interactive Computer Graphics - a classic in it's day and still highly popular - I was designing bespoke ECL (bit slice) graphics accelerators (*) for the visualization of geophysical logging data. Fun times. It's a successive approximate scheme that works as well on a pixilated screen as it did on the original plotter it was designed for back in the '60s.

Since it's inception, Bresenham's algorithm has been applied to a whole host of applications - it's simple, fast and flexible.

(*) My thesis was on the design & implementation of bit-sliced processors based around the AMD 2900 series of chips - these were also used by DEC in the PDP 11 series of highly popular minicomputers, hence them offering me a job. I was also offered a job by IBM as I had won that year's UK Undergraduate prize, but I couldn't stand the white shirts and blue suits that everyone wore at the IBM Hursley research site... DEC was for more relaxed...

[Terry]

I'm not entirely sure whether long term frequency stability is still guaranteed - it used to be the case that the generating authorities would correct any daily deviations during the night to keep any synchronous clocks relying on accurate long-term 50Hz for their accuracy correct.

Currently, the grid is running at 49.967Hz, which is pretty close to 50Hz and can be seen to move above and below 50Hz throughout the day at http://www.gridwatch.templar.co.uk/

In fact, while I've been typing this, it has gone up to 50.042Hz, so an average of 50.0045Hz between just those two readings.

Unfortunately, I don't have any old electric clocks to put the long term stability to the test.