Inverter generator vs. AVR generator

[alwaink]

Hi,

This is kind of a "spin-off" thread from the other one that I posted before.

http://www.eng-tips.com/viewthread.cfm?qid=188842&page=3

I have a 2kW Sterling cooler load that needs a backup system, and the performance drops off sharply outside 60Hz +/- 0.5Hz.

Besides UPS, "Itsmoked" suggested Honda inverter generator. There are other brands of inverter generators, but none of them shows the frequency tolarence in their specs. It was a tremendous hassle to find the answer from their customer service as they wouldn't let me talk to their engineering staff because I'm not a dealer.

Honda's non-technical person told me their inverter generator has freq tolerance of +/-1% (0.6Hz) the first time I called and +/-5% (3Hz) the second time I called. Yamaha told me theirs is +/-0.1% (0.06Hz). I find it hard to believe the Yamaha one has better tolerance (+/-0.1%) than a UPS (+/-0.5%)

Could someone explain to me how an inverter generator works compared to an AVR generator? And what kind of frequency tolerance I should expect in both inverter and AVR?

Thanks.

 

[itsmoked]

In an inverter generator the "generator" produces - one way or another - 180VDC. Then an embedded controller takes that 180VDC and synthesises 60Hz 120VAC. The AC current is monitored by the controller so the engine speed can be adjusted to always provide the required "power" needed to match the power being provided out to the load.

Because of this setup a quartz crystal is the time base. Like a watch. This means the synthesized AC should be very stable and very accurate. I'm surprised they aren't all like the Yamaha.

I am still rather stunned that there exists a load that is so lame as to require this kind of accuracy though....

What is this Sterling cooler again?


Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[alwaink]

Thanks again, itsmoked. With regard to this cooler...as far as I understand, there's a linear motor that drives a magnet ring. This magnet ring makes the piston oscillate back and forth with respect to the displacer in the cooler, and there's a certain degree of phase shift between the piston and the displacer to create pressure difference, which in return exchanges heat. The magnet ring is specifically tuned to 60Hz to resonate and keep the phase between piston and displacer constant for optimal performance. Outside of the +/-0.5Hz resonance range that the magnet ring is designed for, the performance drops significantly because the piston and the displacer will be out of phase.

That is why +/-0.5Hz is very critical in this case when I'm looking for a backup power source.

As of now I'm considering both UPS and inverter generator. UPS requires less maintenance but the battery backup time is limited and takes time to recharge. My greatest concern with a generator was its frequency stability. But looks like an inverter generator should have stable frequency. It just bugged me that I wasn't assured by a technical person from the manufacturer. Honda gave me 2 different frequency tolerances both times I called.

 

[itsmoked]

Thanks for the description.. Strange load indeed.

Might I suggest that you don't give a dang about the efficiency during generator running? I presume this would be only occasionally or during power outages? So you run with many percent less efficiency for a few hours... In the grand scheme it won't cost much more since it is only occasionally.

Otherwise if this 60Hz is so important to good operation maybe the unit should convert any power it gets to DC then invert it to precisely 60Hz itself.

Keith Cress
Flamin Systems, Inc.-

http://www.flaminsystems.com

 

[waross]

I am trying to imagine a piston oscillating at 60 Hz. and having enough travel to do anything useful.
I wonder if the machine is converting to DC and using a 60 Hz. based clock to time the DC pulses. If so it may be more productive to use a small frequency stable source to drive just the controls.
respectfully

 

[alwaink]

Sorry, I don't know any more details regarding the specifics on how the cooler runs.

Yes, in reality I think it's alright if the cooler runs less efficiently for a few hours if the backup source doesn't fit that +/-0.5Hz. But my boss would rather know the different options that allow it to run most efficiently. Too bad I'm not the one controlling the budget

Thanks for the replies.

 

[ScottyUK]

I can't imagine an inverter using anything other than either a crystal or ceramic resonator as the reference oscillator. You will need some expensive instruments to detect the frequency error introduced by either of those devices. I work at a power station and I guarantee that the grid frequency is many orders of magnitude less stable than a crystal. Your quartz watch uses the same technology as its reference. I guess you are just not getting your question answered by someone who actually knows the answer. 0.5% error is likely a value for an engine-driven set with a basic governor.


----------------------------------

Sometimes I only open my mouth to swap feet...