Mains converter to power Aussie scientific machine in the USA.

[CaptDon]

If you intend a world market you will have to solve this issue indeed. You are spot on about controlling the work the compressor does. You could use a 50/60hz rated compressor that has 'just enough' capacity at 50hz and will have somewhat excessive capacity at 60hz. You seem to be in that situation now to some degree. A proportioning valve or 'unloader' in conjunction with an expansion valve would be the ticket here and even more so if they can be modulated, either electrically with P.I.D. self learning control loops, or maybe pre-programmed  curves based on detected input frequency. Best wishes!! In many large air conditioning systems unloaders are common. The system must be big enough for the worst heat of summer, yet run partially unloaded in the spring and fall. The evaporator has a temperature sensing expansion valve to keep the evaporator from becoming excessively cold and freezing up in mild weather, yet it can still be a heavy workhorse in summer. This spring/summer control idea applies in kind to your 50hz/60hz scenario.

[Zero999]

Have you searched for 60Hz to 50Hz converter?

The old fashioned way is with motor generator set, which has the advantage of being resistant to overloads, and low electrical noise, but is bulky and subject to mechanical wear.
https://www.piller.com/en-GB/203/motor-generator-set-50-60hz-from-12kva-up-to-1500kva

There are also inverters specifically designed to convert 60Hz to 50Hz and some also do the voltage conversion.
https://www.kccscientific.com/50hz-60hz-guide-frequency-converter-power/
http://www.gohz.com/how-to-convert-60hz-to-50hz

What is it? Is it a new kind of oxygen generator of ventilator for COVID-19 patients?

The machine is used in research. An NDA forbids me to say what it is. But your GoHz may fit the bill nicely. A big thank you. I have contacted them with some specs and we will see what they come back with. The compressors can be switched on at different times to reduce inrush current.
I will need to run some tests or see some data to verify the load claims of the compressors. I am also interested to know why they cannot use a feedback loop to control the work of the compressors if they run at 60Hz.

GoHz has a 5kVA model that might be suitable.

In any case, if the GoHz solves our problem, I owe you a beer  .

Hopefully that'll help you at the moment. You might find the inverter can cope with some surge, so 5kVA might be a bit overkill.

It's a shame you can't say what it's for. It sounds interesting.

The device may be sensitive but are you sure it is sensitive to mains voltage and frequency?


Can't you do a one off (partial) conversion?
I.e. change whatever won't run on 110V/60Hz for something appropriate.

Presumably if you're going to demonstrate it in the US you're going to want to sell it in the US.
Then you would make a better impression with something that doesn't need a big expensive converter.


The gasoline generator can probably be rented for just the demonstration.

Sorry, I should have been clearer. The demonstration is clinical trials. Eventually they will use an alternative compressor arrangement (eg: a variable speed compressor).

That sounds like the best solution in the long term. The compressors should be 3 phase 220V to 240V, that way a VFD can be used to generate 3 phase, from a single phase input, of any frequency. If you choose the correct motor, perhaps one rated to 60Hz, or more and a with a bit more power than you need, then it should work from 200V to 250V, making it suitable for use in most countries.

[VK3DRB]

I have asked the client for the actual specs on the compressors. I only heard from a conference call, so it is early days. Also, I need to know what power is available at the test facility, so I have asked him.

A single phase mains voltage of 230V 60Hz with high current will be fine to drive this... http://www.gohz.com/10kva-frequency-converter
It's "bigger than Texas", weighing 130kg.

At least there looks like a solution. Cost is not a big concern, which is a welcome change.

The machine will not be coming near patients by the way, so leakage currents etc wont be the big issue as it is in IEC-60601.

I big thanks to all for the responses; they have been very helpful  .

[VK3DRB]

Never trust verbals. I managed to get the datasheets for the compressors; and a proper spec. The bigger one has a 33A locked rotor current (ie: starting current) for <500ms  . Plus they need FOUR systems, not one. And they want WiFi monitoring.

The 10kW unit might still do the power stuff, but at 130kg and its size, I think this might be a better, smaller, cheaper and lighter solution:

Use TWO 90-264 VAC to 24VDC 60A battery chargers (in parallel?) to supply a 24V LiFePO4 100Ah battery which is connected directly to the input of a pure sinewave inverter. The LiFePO4 battery acts as a huge capacitor, and that and the chargers provide current to one of these: https://inverter.com.au/10a000-watt-surge-24-volt-pure-sine-wave-inverter-ps5000a24. Therefore the battery acts as a capacitor for the inrush current to handle the surge. Cost is $3500 USD plus enclosure plus battery. If it works here, we buy another battery in the USA due to shipping issues.

I assume as long as the average current drawn from the battery is less than that available to supply it from the chargers will mean the system will work and hopefully be reliable.

For the WiFi monitoring, I will find something like an iMonnit that measures current from a ferrite clamp, but probably cheaper.

[bdunham7]

Never trust verbals. I managed to get the datasheets for the compressors; and a proper spec. The bigger one has a 33A locked rotor current (ie: starting current) for <500ms  . Plus they need FOUR systems, not one. And they want WiFi monitoring.

The plan of using a large converter so that the device thinks it is at home has the problem that it is difficult to duplicate the very low source impedance of an actual mains connection.  If LRA = 33, then you actually are still within the 10kVA rating of the converter you were considering, so wouldn't just starting the compressors in sequence still work? 

The single-phase VFD that I posted would have the advantage of allowing you to soft start the compressors individually (1 VFD per motor, on the motor only) provided that they can be soft started. 

[mikeselectricstuff]

Never trust verbals. I managed to get the datasheets for the compressors; and a proper spec. The bigger one has a 33A locked rotor current (ie: starting current) for <500ms  . Plus they need FOUR systems, not one. And they want WiFi monitoring.

is it really not possible to get a 60Hz versions of the compressors ?

[VK3DRB]

Never trust verbals. I managed to get the datasheets for the compressors; and a proper spec. The bigger one has a 33A locked rotor current (ie: starting current) for <500ms  . Plus they need FOUR systems, not one. And they want WiFi monitoring.

is it really not possible to get a 60Hz versions of the compressors ?

Apparently not. The client has already investigated this.... they realise they have to spend thousands to find a solution. These are unique compressors. However down the track the compressor manufacturer will be producing variable frequency versions of the compressors but according the client this will occur much after the testing in the USA. I have a feeling this product (patent pending) being tested in the USA was a bit of an afterthought because they left it late.

All being said, they have not actually MEASURED the currents with a clamp meter. Typically, published LRA is worse case with the published load. I have asked the client to measure the actual currents. The fact they they thought is was 20A LRA when if fact it was published at 33A, says they have not done enough work upfront IMO. Whatever the currents measured, I have to add overhead for realistic worst case.

I built a low cost 2kHz bandwidth 50A clamp meter about 10 years ago used for high power SHG laser's current analysis so I could develop adaptive PID algorithm for closed loop power control. It worked a treat. I actually thought of commercialising the tool, but never got around to it. It would have been handy right now. I suggested the client to hire a Dranetz or something similar from a rental company for a week.

[VK3DRB]

I used two universal voltage input Victron AC to DC chargers (a single one of higher amperage does not have a universal voltage input), two Pylontech batteries and a Victron Sinewave inverter, built into a small rack on castors. The client is very happy. The whole thing works very well with his system's load, and there are no glitches. 

For the intermittent high current drain (rotor locking current), the batteries easily supply the current and recover quickly. We have done a lot of measurements using the Victron CerboGX system and my calculations were verified in practice as being very close to the mark .

The system only needs to be used for a year, and it has worked for a month without issues. It is all stable and robust.

I used Victron, because they are top shelf quality, built for reliability. And available off-the-shelf here in Australia.

Thanks you for your comments here. They certainly helped. 

Awesome!