A/C frequency for a Clock Signal - Sine Wave vs. PWM Square Wave

[QualityGuy]

Greetings Fellow Members;

I've tried to find the answer to this without success, although I know it's out there somewhere...

I have a small, spare back-up UPS that generates a PWM / simulated sinewave.  I'm going to connect several vintage digital clocks that use the A/C line to keep time to this UPS since the clocks don't have battery back-up. 

Do clocks that keep time based on the A/C frequency require a 'pure' sinewave, or can they run OK on a PWM / simulated sine wave as the clock signal?

Thanks,

William

[Kleinstein]

It depends on the circuit. Good ones should run OK from the UPS, some might be sensitive to noise and higher frequencies from the UPS.

[QualityGuy]

Well then, it sounds like a trial and error situation!  Most of the time, when we have a power glitch, it's only for a second or two.   

[CraigHB]

UPS units put out a pretty clean looking sine wave.  It's the AC inverters you buy off the shelf that can put out those stepped sine waves.  Even so you can find a pure sine wave option with those at about twice the cost.  I've run into trouble powering a few things with a stepped sine wave inverter.  They're good for most stuff, but they don't work on everything.

[ericloewe]

UPS units put out a pretty clean looking sine wave.  It's the AC inverters you buy off the shelf that can put out those stepped sine waves.  Even so you can find a pure sine wave option with those at about twice the cost.  I've run into trouble powering a few things with a stepped sine wave inverter.  They're good for most stuff, but they don't work on everything.

Cheap UPSes definitely do not output anything close to a sine wave:
http://www.tomshardware.com/picturestory/732-apc-bn650m1-ca-ups-tear-down.html#s40

[QualityGuy]

UPS units put out a pretty clean looking sine wave.  It's the AC inverters you buy off the shelf that can put out those stepped sine waves.  Even so you can find a pure sine wave option with those at about twice the cost.  I've run into trouble powering a few things with a stepped sine wave inverter.  They're good for most stuff, but they don't work on everything.

Cheap UPSes definitely do not output anything close to a sine wave:
http://www.tomshardware.com/picturestory/732-apc-bn650m1-ca-ups-tear-down.html#s40

And that's my concern; I have vintage, mostly 30+ year old digital clocks that have some internal parts that are in the unobtainium category, or very expensive. 

It looks like a "basic" better (pure sine wave) UPS isn't a whole lot more money: https://www.amazon.com/CyberPower-CP850PFCLCD-Sinewave-Compatible-Mini-Tower/dp/B00429N18S

Thanks to all for your comments!  They are much appreciated.

[Kalvin]

Just make sure that the UPS is producing accurate and stable sine wave at 50Hz/60Hz. For example, if the frequency is 1Hz off, your clocks will run too fast or too slow.

[Kalvin]

If the sine wave quality is not good enough, possibly a simple, passive 50Hz/60Hz bandpass filter is sufficient to clean the sine wave signal..

[QualityGuy]

If the sine wave quality is not good enough, possibly a simple, passive 50Hz/60Hz bandpass filter is sufficient to clean the sine wave signal..

Will consider it, and thanks.

[danadak]

If the client side is a digital clock input then the slow moving risetime of
a sine based 50/60 Hz input can cause a lot of clocking issues. Think
about the input hanging right at the threshold of the client side logic, and
ordinary system noise causing multiple counts as it is transitioning thru
this region.

Best solution is a comparator with hysteresis to negate the effect of noise.


Regards, Dana.

[Melt-O-Tronic]

If you're concerned enough about the clocks (and I would be if it's a question of running motors in valuable vintage appliances on dirty power), consider a modest investment in a true sinewave UPS.

CyberPower UPS on Amazon

I have no affiliation.  I know about this because it's the model that Rigol recommends for powering their spectrum analyzers in the field.

[SeanB]

Quick brute force approach is to use a pair of identical transformers with the secondaries connected together, the primary of one getting the input sine, and the primary of the other ( now the secondary) giving the sine wave output. A non polarised capacitor of a few uF on the secondary, and another 1uF on the output should provide enough low pass filtering to get a reasonable sine wave even from a near square wave input, from the LC action of the cascaded filters.