Author Topic: AC booster for startup surge (similar to grid-tie)  (Read 2555 times)

0 Members and 1 Guest are viewing this topic.

Offline jonrogerTopic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
AC booster for startup surge (similar to grid-tie)
« on: March 17, 2019, 12:07:22 am »
I have a 2000W 120 VAC inverter generator that can't produce enough short term power to start some loads (eg motors).    So I'd like to build something that provides a short (say a few cycles, 50 msec) AC boost.  This is similar to a grid-tie inverter, but there is no utility grid involved.  Let's assume I understand all the safety issues.  And that I can't reduce the startup surge.

I'm looking for feedback on the design.

Small transformer based (ie, isolated) power supply creates 250VDC and stores it in a large capacitor.

Buck converter

H-bridge to invert the single ended output from the buck converter.

A micro-controller would measure the existing AC waveform to learn frequency, timing and typical voltage (at every point in the AC waveform).  Whenever (as in a few  microseconds) the AC waveform voltage drops a little below normal, the buck converter would charge the inductor and then discharge it into the H-bridge.   Repeat as needed.    If boost is provided for more than 50 msec, shut down (this helps with heat dissipation).

Feasible?   I know some $1000+ hybrid inverters have generator/power assist.  Any lower-cost off-the-shelf solutions I have missed?
« Last Edit: March 17, 2019, 12:45:16 pm by jonroger »
I am available for custom hardware/firmware development.
 

Offline soldar

  • Super Contributor
  • ***
  • Posts: 3595
  • Country: es
Re: AC booster for startup surge (similar to grid-tie)
« Reply #1 on: March 17, 2019, 11:20:45 am »
To be honest I do not think you will find a practical general solution of the type you are thinking. In practical terms the solution would be to get a more powerful inverter.

On the other hand, it depends a lot on the load. Maybe some loads can be soft-started and gradually brought up to speed. Maybe some motors could be fitted with a manual starter, like a lawnmower, so you could get it turning by hand and then apply electrical power.

So, it depends on the inverter and on the load.
All my posts are made with 100% recycled electrons and bare traces of grey matter.
 

Offline unitedatoms

  • Frequent Contributor
  • **
  • !
  • Posts: 324
  • Country: us
Re: AC booster for startup surge (similar to grid-tie)
« Reply #2 on: March 17, 2019, 11:31:22 am »
There should be some low cost mechanical solution. Like a motor paired to generator across large flywheel. Or just a flywheel mod to existing generator.
Interested in all design related projects no matter how simple, or complicated, slow going or fast, failures or successes
 

Offline soldar

  • Super Contributor
  • ***
  • Posts: 3595
  • Country: es
Re: AC booster for startup surge (similar to grid-tie)
« Reply #3 on: March 17, 2019, 11:42:56 am »
If the generator-inverter is of those small types like the Honda 2000 I can't see any practical way of mechanical boost. Possible maybe yes, but not practical.

And if the limitation is in the electronic inverter then it would not work anyway and you just need a more powerful model.
All my posts are made with 100% recycled electrons and bare traces of grey matter.
 

Online coromonadalix

  • Super Contributor
  • ***
  • Posts: 7001
  • Country: ca
Re: AC booster for startup surge (similar to grid-tie)
« Reply #4 on: March 17, 2019, 11:45:56 am »
You need to find a suitable inverter who accept startup motor spikes .... i know they are out there but easier said than done, i had one in my old company truck, it was a 3kw   but I never knew the brand :(
 

Offline jonrogerTopic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC booster for startup surge (similar to grid-tie)
« Reply #5 on: March 17, 2019, 12:35:50 pm »
Any thoughts on the big picture circuit design?   And if the concept (real-time correction of AC waveforms as they sag) will work?

A more typical AC inverter (or VFD) design doesn't use a buck converter and PWMs the H-bridge (with an output inductor for smoothing).   I believe that when the H-bridge is off, such designs cause the output inductor to push current through the body diodes on the H-bridge mosfets - not good for heat dissipation.

Below is an example of a similar product (but way overbuilt for super low duty cycles and it needs a battery).   It measures load current,  but I think that voltage sag is a convenient proxy for load current.

See example 2.4, but replace shore power with generator power:
https://www.victronenergy.com/upload/documents/White-paper-Achieving-the-impossible-EN.pdf
« Last Edit: March 17, 2019, 12:49:17 pm by jonroger »
I am available for custom hardware/firmware development.
 

Online NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9319
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: AC booster for startup surge (similar to grid-tie)
« Reply #6 on: March 17, 2019, 01:29:51 pm »
Is there an easy way to hack it to allow V/Hz ramp up?
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline jbb

  • Super Contributor
  • ***
  • Posts: 1265
  • Country: nz
Re: AC booster for startup surge (similar to grid-tie)
« Reply #7 on: March 17, 2019, 07:04:17 pm »
+1 for soft-starting the load if you can.

In principle, you can make a power electronics box to supplement the d. for load peaks.  This would pretty much be a STATCOM device with a nice big energy storage capacitor.  To get fast response, you would likely use a simple inductor output filter (not LCL) with a high switching frequency (probably > 20 kHz) and some cunning control design.

However, this is quite complicated to design and may interact with the inverter control systems, especially if using voltage-only control.  For example, the STATCOM and inverter could fight about what the appropriate voltage level is, or they could oscillate against each other until something goes bang.

If you measured the load currents you could implement a current compensation scheme, which could help with control design.
 

Offline jonrogerTopic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC booster for startup surge (similar to grid-tie)
« Reply #8 on: March 17, 2019, 08:12:26 pm »
Thanks for the STATCOM word.   

I think that by recording the waveform from the primary source (the generator) and only injecting charge when voltage is sagging, I can avoid pushing the line voltage above what the generator expects to see.   So hopefully it remains happy.

I suppose that with good external diodes in the H-bridge, there is little reason not to use a design with the inductor after the H-bridge (ie no buck converter stage before the H-bridge).

I agree about current measurement - knowing both the instantaneous inductor current and the load current (these are different things since there are two independent power sources in parallel) would be helpful for control.

25Khz sounds about right.  Every 4 usec it would sample the line voltage and decide if an additional pulse is needed.  Doable on a 180Mhz micro-controller.

I do know that Honda Eu2000i inverter generators do something similar.   They can be run in parallel with no control lines between them (the second one syncs to the first and they share current output).     I just don't want to buy and run a second one just for 50 msec here and there.

"Stackable" inverters are also similar (but most seem to use control lines).
« Last Edit: March 18, 2019, 03:25:46 am by jonroger »
I am available for custom hardware/firmware development.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: AC booster for startup surge (similar to grid-tie)
« Reply #9 on: March 18, 2019, 12:10:20 am »
There should be some low cost mechanical solution. Like a motor paired to generator across large flywheel. Or just a flywheel mod to existing generator.

Maybe a big synchronous motor would work but it would hardly be inexpensive.
 

Offline jbb

  • Super Contributor
  • ***
  • Posts: 1265
  • Country: nz
Re: AC booster for startup surge (similar to grid-tie)
« Reply #10 on: March 18, 2019, 12:11:45 am »
I’m assuming this is single phase?

Regarding the H Bridge:
- yes, the diodes will be critical because reverse recovery could be really high. Standard or superjunction MOSFETs will probably not be acceptable.
- the classic topology for this sort of system is a big DC bus capacitor and then the power switches (H Bridge for single phase) and then a filter inductor. This is referred to as a Voltage Source Converter (VSC).
- switching frequency will depend on the acceptable current ripple, the desired current ramp rate and the required closed loop bandwidth
- with care, the VSC can draw a little active power from the main AC circuit to top up its capacitor bank

Regarding generator paralleling without extra wires: this can be done using droop controls. Frequency droop control is used to share the active power and voltage droop control is used to share the reactive power.  If your generator has these droop controls built in it could be much easier to make it play nice with a STATCOM.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: AC booster for startup surge (similar to grid-tie)
« Reply #11 on: March 18, 2019, 12:17:09 am »
Thinking about it, I could design a power stage and control system which operates in parallel with an inverter to stiffen the output but that is hardly an economical solution.
 

Offline jonrogerTopic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC booster for startup surge (similar to grid-tie)
« Reply #12 on: March 18, 2019, 03:06:39 am »
No doubt that a one-off build isn't cost effective.   But cheap grid-tie inverters (a similar device) are < $300 on ebay.   That's much less than $1000 (plus size, weight, noise, gas) to add a second generator.
I am available for custom hardware/firmware development.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: AC booster for startup surge (similar to grid-tie)
« Reply #13 on: March 18, 2019, 03:53:56 am »
The problem comes down to a solution better than just using a larger inverter and I do not think there is one.

A synchronous motor is not quite the right solution for this but a doubly fed motor could do it.  More practical would be a motor-alternator and it should be possible to put one together from surplus parts for a reasonable price.  With a large rotating mass, the motor-alternator isolates much of the surge from the inverter.  Of course now the inverter has to be able to start the motor-alternator.

Easier would be to install a variable frequency drive on the motor configured to ramp it up as quickly as needed.  These have become very economical.
 

Offline jonrogerTopic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC booster for startup surge (similar to grid-tie)
« Reply #14 on: March 18, 2019, 01:11:49 pm »
I checked and evidently it's hard to find VFDs that will take in 120VAC and output single phase 120VAC.   Do such things work well anyway (or do I need to switch to a 3 phase motor (difficult))?
I am available for custom hardware/firmware development.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: AC booster for startup surge (similar to grid-tie)
« Reply #15 on: March 18, 2019, 06:02:40 pm »
I checked and evidently it's hard to find VFDs that will take in 120VAC and output single phase 120VAC.   Do such things work well anyway (or do I need to switch to a 3 phase motor (difficult))?

I have seen an awful lot of them used with PSC (permanent split capacitor) motors on small machine tools and they seem to work well enough.
 

Online NiHaoMike

  • Super Contributor
  • ***
  • Posts: 9319
  • Country: us
  • "Don't turn it on - Take it apart!"
    • Facebook Page
Re: AC booster for startup surge (similar to grid-tie)
« Reply #16 on: March 18, 2019, 11:24:09 pm »
Some high end inverters have V/Hz ramp up. It's not hard to do and there used to be a cheap brand of inverters (Vector) with that feature, but apparently they ran into patent issues. (That was 15 years ago or so so the relevant patents probably would have already expired by now.)

If it's easy enough to get at the inverter board in your generator, take a picture and that will give clues how easy/difficult it would be to hack it.
Cryptocurrency has taught me to love math and at the same time be baffled by it.

Cryptocurrency lesson 0: Altcoins and Bitcoin are not the same thing.
 

Offline jonrogerTopic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC booster for startup surge (similar to grid-tie)
« Reply #17 on: March 19, 2019, 03:20:15 pm »
Does anyone have a link to a 120VAC input, 120VAC output VFD?   I can't find a single one.  Lots of them for 240V and they say that some single phase motors can just use two outputs of a 3 phase  VFD.

Independent of practical advice, I think that a small STATCOM or "boost only GTI" is a very interesting design challenge.  Doable, but tricky - and evidently hasn't been done by anyone.
I am available for custom hardware/firmware development.
 

Offline maxwell3e10

  • Frequent Contributor
  • **
  • Posts: 911
  • Country: us
Re: AC booster for startup surge (similar to grid-tie)
« Reply #18 on: March 19, 2019, 06:22:11 pm »
Interesting challenge. Could one just build a 60 Hz resonant LC circuit to store enough energy? I had a similar problem (starting a furnace blower from an inverter) and ended up using a centrifugal V-belt clutch.
 

Offline jonrogerTopic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC booster for startup surge (similar to grid-tie)
« Reply #19 on: March 20, 2019, 01:02:13 pm »
I look at a design a bit more and if operated in discontinuous mode (wait for inductor current to drop to zero), control appears to be easy.
I am available for custom hardware/firmware development.
 

Offline David Hess

  • Super Contributor
  • ***
  • Posts: 17427
  • Country: us
  • DavidH
Re: AC booster for startup surge (similar to grid-tie)
« Reply #20 on: March 20, 2019, 01:54:29 pm »
Interesting challenge. Could one just build a 60 Hz resonant LC circuit to store enough energy? I had a similar problem (starting a furnace blower from an inverter) and ended up using a centrifugal V-belt clutch.

You could install a ferroresonant transformer.  Or a normal transformer which is undersized for the peak power would also limit the current.  I do not know how well either would work.

 

Offline jonrogerTopic starter

  • Regular Contributor
  • *
  • Posts: 72
  • Country: us
Re: AC booster for startup surge (similar to grid-tie)
« Reply #21 on: March 22, 2019, 04:02:24 pm »
I did some more investigation/modeling on controls and it looks to me like micro-processor control would be fine in both continuous and discontinuous modes.    But it would require knowing output voltage, inductor current and load current - to properly calculate the amount of energy to transfer without using a slow adaptive algorithm.     Low frequency (eg, 5Khz) provides adequate time without too large of an inductor.
I am available for custom hardware/firmware development.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf