Brand new inverter generator dies with suddenly applied full load, runs it fine if loaded gradually 3

[LMF5000]

I bought a brand-new Zipper ZI-STE1200IV on eBay (from Germany, no idea who the OEM of the unit is, but it's identical to the Parkside PGI-1200A1 generator sold in the UK).

The specs claim 1200W for 5 seconds and 1000W continuous output. I've been testing it with one of those halogen heaters that have 3x 400W heating elements. The brightness of the heating elements gives a good indication of the output voltage (faster than any digital meter can measure). I also have an inline power meter attached and a tachometer that takes its signal from a sense wire wound around the spark plug cable.

Here's the story:
1. If I turn on a 400W heating element, the revs fall, the voltage droops, then it picks up and delivers full voltage no problem
2. If I turn on the second 400W heating element, the revs fall slightly voltage drops slightly, then it picks up and delivers 800W. I even turned on the third 400W element and it managed to supply all 1200W.
3. If it's idling and I turn on both 400W elements at once (800W sudden load), most of the time the generator struggles, the revs drop, and the engine dies of low revs. Looking at what the throttle servo is doing during this time, it doesn't move when the load is first switched on (throttle remains in "idling" position, around half-open). As the engine starts bogging down over the next 2 seconds, the throttle barely opens. When the engine is really struggling it opens the throttle fully, at which point it finally dies. The engine is impossible to hand-start at this point (throttle full open) without full choke.

In addition to the above, it hunts a little (revving up and down) when cold at no load, and it over-revs for a few seconds every time a load is removed.

Any ideas what the problem could be?

 

[BrianPetersen]

It's not a Honda or Yamaha.

My hobby is motorcycle roadracing, and that means tire-warmers, and that means portable generators. The Honda EU2000i and the Yamaha EF2000i are the two prime choices. Champion are hit or miss. Anything else, you're taking your chances. I know lots of people who bought a cheap off-brand, had no end of troubles, and then ended up buying a better one. The non-inverter generators are cheaper, but are too noisy.

The only times I've had trouble starting the Yamaha have generally involved me forgetting to turn the fuel valve on, or something of the sort.

All of the inverter-generators will have an output drop if you immediately request full load from idle. The engine is idling and is not capable of making full output power ... so it doesn't. But the good generators will allow the revs to increase and pick up the load in a second or so without coughing and stumbling and stalling.

On the one you have, it sounds like the governor isn't working properly. If you have a money-back guarantee, send it back. If you have a warranty, get them to fix it. There is a fair chance you will get fed up with this unit. I've told you what you should buy instead ...

 

[jraef]

Halogen lamps have a very high PTC (Positive Temperature Coefficient) of resistance, meaning the resistance changes drastically from hot to cold, rising with the temperature. That therefore means when you first energize a cold lamp, the resistance is low and current goes very high for a moment. How long that moment is varies by a number of other factors, but some meters will not be able to record it.

Add to that the fact that some inverters are designed to "soft start" into rapid load changes, while others are designed with a "crowbar" circuit that shuts it down if the change in current is too rapid and you end up in your situation. Yours obviously has the crowbar circuit.

If you don't want to or can't change the inverter, you could rig up your own "pre-charge" circuit. Get a resistor that will limit the current to the rating if your inverter (maybe a tad lower) and wire it in series with the load. Then wire a contactor in parallel with that resistor as a bypass, and control that contactor with a timer, 1 second is likely enough. So when you turn in the lights, the resistor is in series and limits the inrush current to what the inverter can handle but gives power to the lamps so they warm up, then a second later the contactor shorts across the resistor to give the lamps full voltage. An alternative to this is to use an SSR or SCR controller with a Ramp function feeding the lamps, but finding one that will work reliably on an inverter output may be tricky. If you already have an SSR or SCR controller for the lamps, my guess is that it doesn't work on the inverter output.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[LMF5000]

BrianPetersen - This unit was purchased for €160 including shipping. No 4-stroke inverter generator on the market seems to come anywhere close to the price. To give you an idea, here's the spreadsheet I drew up of all the options available to me as part of my due diligence exercise before buying:

(the "Volume" column is simply length x width x height as a rough measurement of the "size" of the unit). Sending it back could prove expensive and the problem could be intrinsic to the design in which case the replacement unit would suffer the same issue so I'm trying to do my own troubleshooting first.

jraef - any idea what the inrush power could be for my situation? I've seen it variously quoted as being 3x to 10x the steady-state power (so 1200 to 4000W for my 400W element) but my meter was unable to measure anything like that. I'd like to power things like computers, monitors, fridges and power tools with this generator - I suppose the resistor would damage the computer PSUs for example?

From the answers so far it seems I am asking too much of the unit. Does anyone suspect a fuelling problem (dirt in the carb from manufacturing) or is this just the nature of small cheap generators?

 

[BrianPetersen]

160 euros = "too cheap to be good" which is a close relative of "too good to be true".

Maybe the carburetion is off. Maybe the governor is not working right. Maybe the way the inverter handles a sudden load change isn't right (see jraef's post). But you shouldn't have to mess around with stuff on a new unit to make it work right.

I know there are much cheaper units than the Yamaha. But the Yamaha starts on the first pull (when I remember to turn the fuel on!) and stays running. If you overload it, it will pop the circuit breaker, of course. But being able to run two sets of tire warmers and not have to worry about it is good enough for me.

 

[Skogsgurra]

I would guess that the Cold resistance is somewhere near one tenth of nominal (hot) resistance. When you switch such a low resistance to the generator, voltage drops and since there is no voltage (or very Little), you don't get the heating needed and that's why the lamps never come to Life. The gradual (Group-wise) Connection is probably the best you can hope for.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[ScottyUK]

LMF5000,

Probably not a fueling problem, more likely the electronics are protecting themselves from overload. 20% overload for 5 seconds is pitiful, but a bigger output stage costs more money and at that price you can't expect much overload capability. As others have said - you get what you pay for, and in this case you've either got quite a lot for next to nothing, or something which is a poor imitation of its competitors. I guess it depends on your perspective!

 

[LittleInch]

Think about it for even a moment.

You have a 55cc petrol engine running at some revs with no load ( say a small moped waiting at the lights)
You then get a large person on the said moped, drop the clutch and THEN open the throttle. What happens - nothing.
Why do you expect a generator to act any differently, especially one that costs < £140.
I've had 1 MW generators do the same thing when I connect in 400kW of DOL motor load when only initially running at 10% load, so expecting miracles doesn't happen.

Also it's curious as to why the next model in their range is twice as expensive, but less power....



Remember - More details = better answers
Also: If you get a response it's polite to respond to it.

 

[iop95]

It's a problem in loop control design/implementation/tunning.
Too slow response followed by overshoot.
Looking for control board if there are some pot for adjust response control loop.

 

[LMF5000]

Some more info I didn't include in the original post: it "idles" at 4000rpm. It will run 800W of halogen heating at 5000rpm. If I give it half choke it doesn't die (even though the control circuitry is still throttling up at the last possible second and the engine still almost dies).

Also, crucially, the halogen lamp glows brightly a second after powering it on. The generator revs fall and the engine dies 2-3 seconds later. So inverter and flywheel momentum seem to comfortably cope with inrush current. It's an engine or control issue, I think. I'm about to take a look at the carb and throttle mechanism.

I paid €160 for an inverter generator that starts 2nd pull every time and runs an 800W load when it's in a good mood. So far not so bad. This was mainly bought to learn about small petrol engines - the generator aspect is just a bonus that will be used infrequently in my case, but the engineer in me wants to find the root cause of this issue, if only for the educational benefit.

 

[jraef]

The hot to cold resistance ratio for typical halogen lamps is roughly 16:1 IIRC. So if this is the only load, your initial current when cold could be 16x the current when hot.

This is an "inverter" generator, implying that it is a DC generator that feeds an on-board inverter. Most inverters will have a thermal "overload capacity" that is rated in terms of a percentage over a long time so is software based, such as 30 or 60 seconds, and a higher "hardware limit" that is in the order of 1-2 seconds based upon not causing damage to the transistors from too rapid of a dI/dt (change (delta) in current / change in time). So from the sounds of it, I speculate that you are hitting the "hardware OC" limit and tripping the inverter off, which then probably kills the generator because you don't want to keep energizing the inverter if it cannot dissipate the energy. If this is just a test bench, turn the lamps on one at a time.


"You measure the size of the accomplishment by the obstacles you had to overcome to reach your goals" -- Booker T. Washington

 

[Skogsgurra]

I missed the "inverter" part of it. Then, I am surprised that it can take even two halogen lamps. You have a very good piece of equipment there. Do not try to send it back. It only complicates things and doesn't help.

I would also strongly recommend that you don't try any fixes. First, there are no pots to change and second, a commercially designed unit does not have any spare capacity.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[LMF5000]

LittleInch - the generators in the spreadsheet were the ones available to me at the time of purchase (mostly eBay prices - except the Kipor ones which are from a local dealer). The price difference reflects the different brands.

I'd like to stress that this is a halogen heater, not a halogen lamp. Probably less inrush current since the operating temperature of the filament is somewhat lower than that of a lamp.
Here's what the one I was using looks like (3 elements of 400 W each):

Update:
[ul]
[li]I took off the air filter and checked the nuts holding the carb - they were only tightened slightly, so I tightened them quite hard (probably gave them another quarter turn).[/li]
[li]I checked the spark plug, it was just starting to turn light orange. Same with the top of the piston. The engine has almost 1 hour of running time on it so far (from me testing it)[/li]
[li]I took off the servomotor and noticed the throttle moves nice and smoothly, no binding or excessive play. When putting the servo back, I discovered that if I tighten the servomotor's self-tapping screws fully, the throttle binds a little near the idle position; loosening one of them a quarter turn lets it move freely so I left it in this setting.[/li]
[li]I sprayed carb cleaner into the main jet, drained the float bowl with the drain screw and opened the fuel stopcock to fill it with new fuel. I will leave it like this until tomorrow morning when I can fire it up and test it. I don't have the halogen heater with me but I can load it with a convection heater that draws 1060 watts on the 235V mains voltage (hopefully under 1000W at the end of a long extension cord).[/li]
[/ul]

If you're curious as to what the innards look like, here are some pics I took during the disassembly:

Inverter:

Top view (fuel tank removed):

Side view of carb:

Throttle servo:

Carb without servo:

Full album is here:

http://imgur.com/a/Xds2S

 

[itsmoked]

You may have found it in the binding. I'd solve the binding issue some other way than a loose screw.

I bet that's a stepper motor with that many wires.

What are the two copper wire gizmos on the right side top pic?

Keith Cress
kcress -

http://www.flaminsystems.com

 

[LMF5000]

itsmoked - if you mean the inverter view, I think those are inductors for the inverter board. Probably part of the power conversion (I'm a mechanical engineer so can't elaborate on their electrical function, I just know they don't seem to be connected to the engine in any way).

Started up the generator this morning, it started (smokily) on the second pull with choke, idled nicely without choke (though it was hunting when cold and I could hear an occasional misfire when warmed up), and ran a 500W power drill at full speed without problems.

Hopefully I will be able to test it with some heaters later.

 

[Skogsgurra]

A 500 W drill. Was it fully loaded? If not, you probably didn't test with more than 100 W.

Please understand that 1200 W cold halogen lamps shall trip the generator or inverter. It is the only thing you should expect. Anything else is self-deception.

Gunnar Englund

http://www.gke.org

--------------------------------------
Half full - Half empty? I don't mind. It's what in it that counts.

 

[LMF5000]

The drill draws about 350 watts at full throttle (unloaded).

Tested the generator properly today. Ran perfectly starting the heater with two elements on (800W load). Even picked up and ran with the third 400W element (1200W total) though the servo pushed it to wide open throttle to do that and I didn't leave it for more than 10 seconds. Looks like reassembling the servo and/or cleaning the carb did the trick. Or perhaps the engine is getting better since it's still running-in (only 1.2 hours on the clock).

I took a video of the running and heater cold start-up, will see how it looks and maybe post the link eventually.

 

[LMF5000]

Well, finally found a resistive element fan heater that draws almost exactly 1000W on the "low" setting.

Tested the generator today and it worked perfectly picking up the 1000W load.

Here's a 40-second video of it running on and off load, with an inline power meter to measure the output:

https://youtu.be/W_qsRlJUlIM

Thanks to everyone for your suggestions and advice.

 

[BrianPetersen]

So, it was a simple problem related to dodgy build quality. Congrats for figuring it out. Most users would have just sent it back (and for something under warranty, I would be one of those).

 

[DanEE]

quoting Skogsgura Please understand that 1200 W cold halogen lamps shall trip the generator or inverter. It is the only thing you should expect. Anything else is self-deception.

Yep.. the very low cold resistance of halogen lamps made them a good additional final check tool in working with high energy DC power plants.. e.g. for initial charge up of the large value caps on the DC input side in UPS systems and as a part of redundancy checking using different methods before closing high energy connections.

Never wanted to be a part of whats shown here...

https://www.youtube.com/watch?v=DpQeDcEpEn0

makes for a bad day