Implications of Unstable Frequency from Inverter

[Cwhitla]

I have a battery powered "generator" from Lion Energy that was advertised as having a "modified Sine Wave" inverter. I purchased it about 3 years ago in relative ignorance of what I was really getting. I recently tested the inverted output with a small hand-held oscilloscope and found the waveform actually looked pretty decent with the appearance of a pure sine wave but the frequency varies between 58-70 Hertz. I don't have much of an electronics background but I'm concerned about what I can realistically power with an inverter that produces such unstable frequency. Obviously no clocks but what else? Will a refrigerator be damaged by the fluctuation? I realize I might be able to power a separate external inverter with the DC battery 12V power output but that is limited to 20 amps max which seems a little feeble. Any real world advice would be appreciated.
Thanks

 

[Nigel Goodwin]

It should be perfectly fine, very little is frequency dependent.

Your 'mistake' was in testing it's output

 

[Diver300]

The "Modified Sine Wave" inverters are usually a sort of square wave, with alternate on-periods negative. The sequence is off, +ve, off, -ve which repeats at 60 Hz (or 50 Hz if that's the local mains frequency). It's therefore surprising that you are getting anything that looks like a sine wave.

How are you measuring the frequency? Oscilloscopes that have a frequency measurement function are often fooled in lots of conditions. If you are getting a waveform that looks good, if the frequency is varying from 58 - 70 Hz, the on-screen width of one cycle will vary by more than 10%, so some of the waveform will look blurry?

Can you post a photo of the oscilloscope screen when the frequency is being measured? How quickly does that frequency vary?

I agree with Nigel that very few appliances will mind the frequency variation, even if it is as bad as what you measured.

(I once made a frequency stabiliser where the 50 Hz UK mains wasn't accurate enough, and 49.98 Hz, not 50 Hz was required. To say it was a specialised requirement is understating the situation)

 

[Cwhitla]

Thanks for the comments so far. Obviously, I'm very new to Oscilloscopes. I attached a JPG file. I also took a MOV file but it looks like I can't load that to the forum. While watching the video I saw the frequency fluctuate into the high 70's so it's even worse at times than I initially reported. The frequency fluctuates several times per second so it's pretty hard to convey the actual readings without a video.

 

[Diver300]

Ignoring the number of Hz displayed, was the waveform steady on screen?

When I started using oscilloscopes, frequency was measured by counting how many division on the timebase per waveform, and calculating the frequency with a slide rule. The waveform had to be steady on the screen, because oscilloscopes were just CRT displays showing the waveform directly. There was no computer memory, so for a line to be visible the waveform had to be repeated very often. An unsteady waveform would just be a blur.

I can't find my slide rule, so I've had to use the calculator on a computer, but I can still count the divisions. I make it 10.1 divisions for 3 complete waveforms, and the timebase is 5 ms/division. That is therefore 10.1 * 5 = 50.5 ms for 3 cycles, or 16.833 ms/cycle or 59.4 Hz.

It's not a very accurate way to measure frequency, no better than ±1%. If the waveform shown is steady, my frequency estimate would seem to say that something has gone wrong with the frequency measurement shown on screen. It also shows that the inverter's frequency is fine.

I don't know what is happening with the frequency measurement that the oscilloscope is doing. I suspect that it is counting the number of times per second that the waveform is crossing the 0V point, and sometimes small amounts of noise mean that the a crossing is counted more than once.

 

[Nigel Goodwin]

(I once made a frequency stabiliser where the 50 Hz UK mains wasn't accurate enough, and 49.98 Hz, not 50 Hz was required. To say it was a specialised requirement is understating the situation)

Considering how accurate the UK mains frequency is, you must have had a serious requirement

However, to be fair, it's the long term accuracy that's extremely good, there 'may' be minor variations depending on load during the day.

 

[Diver300]

Considering how accurate the UK mains frequency is, you must have had a serious requirement

However, to be fair, it's the long term accuracy that's extremely good, there 'may' be minor variations depending on load during the day.

At the risk of thread drift, it was a serious requirement. It was to power the synchronous motor that kept a 36" telescope pointing in the right direction as the earth rotates. It only ran at night, so long-term accuracy made no difference. At night the 50 Hz of the mains is often higher as the grid catches up.

Also the telescope's design from 50 years before hadn't allowed for the lens affect of the atmosphere, which was why 49.98 Hz was wanted. I provided 1 mHz steps of adjustment and independence from the incoming mains frequency.

And it was 3 phase, and I used a custom transformer.

 

[Nigel Goodwin]

At the risk of thread drift, it was a serious requirement. It was to power the synchronous motor that kept a 36" telescope pointing in the right direction as the earth rotates. It only ran at night, so long-term accuracy made no difference. At night the 50 Hz of the mains is often higher as the grid catches up.

Also the telescope's design from 50 years before hadn't allowed for the lens affect of the atmosphere, which was why 49.98 Hz was wanted. I provided 1 mHz steps of adjustment and independence from the incoming mains frequency.

And it was 3 phase, and I used a custom transformer.

Very nice, and nice to know the exact reason.

 

[Cwhitla]

Diver300,
Yes, the waveform looks very stable, at least to me. Incidentally, I have tested my grid power as well as a pure sine wave inverter and the Hz variation were both at 59-60. Thank you for the calculation as that appears helpful. I am wondering if they labeled the inverter as "modified sine wave" since it has some issue with frequency stabilization rather than an altered waveform? In any event, my concern was the viability of running something like a refrigerator or freezer as well as my furnace blower without damage to the units while in a multi-day power outage. It seems the consensus is that I won't have a problem and while I always want to find out the truth (bad meter or settings, user error, what have you) I feel somewhat more comfortable with everyone's advice. Thanks.

 

[Diver300]

Can we have a photo of the oscilloscope trace while it is measuring the mains or the pure sine inverter?

I really don't think that the modified sinewave inverter would be described as that because it had a frequency stabilisation issue. Also I don't think yours does have an issue with it's frequency. I'm fairly sure it's a measurement issue.

It's quite common for an oscilloscope to measure frequency wrong. I've seen other examples of oscilloscope frequency measurements being taken as correct when they were way off. In one, I did exactly the same sort of estimate as I did for your waveform to show that the number was wrong.

 

[Cwhitla]

Attached is a photo of the grid power measurement. Just in case it might be pertinent, I do have to hit "Auto" to get the display to synch up properly - both main and inverter.

 

[Diver300]

Well the two photos you posted, the inverter output and the grid power measurement, are both on 5 ms/division, and the width of the waveforms are virtually identical.

That means that they are the same frequency, give or take a percent or two at most.

Exactly what happens to the measurement isn't clear.

The "Auto" button can be for automatic level setting of the trigger level, or it can mean that the display will trigger after a short time even if the waveform doesn't trigger it. It's difficult to know what is the best way to get a steady waveform, without having an oscilloscope of that type to play with.

 

[Cwhitla]

Thanks for your observations and comments. I guess Nigel was spot on.