Specific Genset Problem

Thread Starter


Joined Feb 12, 2010
Some interesting information about electricity supply in Indonesia.

To cope with the economics of peak hour demands, the local solution is to ration. In dire circumstances the power may be off between 5 pm and 9 pm every second night. More usually it is off every fourth or fifth night, but with no discernable program, meaning that you can’t organise life around the “jadwal”, or schedule.

If you run a fridge and freezer, and have people needing good lighting, for example to study, a genset is a must.

This is not a “backup” system made ready for emergencies. It is a necessary part of almost continuous electricity supply. If you want uninterrupted power supply, that will involve a battery system and UPS device in addition to a genset.

In Indonesia, the “low-start current” fridge is much advertised. When an electric motor switches on, the initial current draw can be very high and enough to kill a small genset or trip a fuse. The “low start current” model is supposed to alleviate this problem.

I began by hastily buying a small 2-stroke generator at around 1 million rupiah (about 100 dollars). It bore the hopeful code of 1300, but the small print said a rated 700 watts with a higher short term capacity of 750 watts. Honest enough, thought I. With lighting using less than 100 watts there was enough, I thought, to start up the fridge, rated at 75 watts.

Indeed there was, but when the fridge came on it seemed as if the genset had been kicked in the guts, rpm dropping dramatically before recovering. But when the fridge, ostensibly drawing 75 watts, switched itself off, the genset would race, blowing the occasional bulb before the governor got things under control. The addition of a freezer to the household at around 100 watts brought things to a head. I began alternately plugging in the two refrigerator systems. It was impossible to leave the house during the power cut. Gensets in Indonesia are highly over-rated….

A popular device in Indonesia is the “STAVOL”, a unit which is supposed to regulate AC voltage and protect appliances. I bought a 1500 watt unit. It clearly couldn’t cope with voltage surges, and the time delays in corrections were of the order of governor response in the genset.

I reasoned that the flywheel effect of a heavier genset might get us over the sudden load changes, and bought a single-cylinder 8 HP diesel to couple with a 3 KW generator. Such couplings in Indonesia can be dreadful, and I paid attention to belt tightness and alignment. The flywheel and crank is deliciously heavy, but engine experts know that the power delivered by an engine with one cylinder, and its rotational speed, fluctuates wildly. Even a four cylinder diesel has this problem, and a coil-spring infested coupling (incorporated in every car’s clutch plate) is used to connect the engine less wearingly to a constant load. Obviously, a v-belt coupling can absorb the fluctuations of a single-cylinder engine and the fluctuating load of a single-phase dynamo.

Out of the box, the engine gets rid of its heat by boiling water in a small tank – enough to last about a half hour. An operator must top up this water very regularly if the unit is to run the four jadwal hours. I soon bought a bolt-on plate with two pipes welded in to do the water circulation thing. Locals hook it up to a 200 litre drum of water, using what should be a thermo-siphon. There was a question of whether to drain the engine’s tank after every use. Hooking up a 20 litre bucket above the engine, discarding the hot water, gave me about 30 minutes respite. I needed a larger reservoir. Meanwhile, a great idea arose. We had been heating bath water on a not inexpensive wood fire. The first ten litres of water from the engine were a bit rusty, but after that there was a good supply of hot water which was clean enough for everyone to luxuriate in a free hot bath! Tending the water supply became a family bath routine. Great!

I also bought a DC HF welder, “knowing” that it would not operate on the 1300 watt limit of the power supply in our town, which lies at 1200 above sea level in Indonesia. Yes! This means it gets cold here! Love that hot bath at night!

With a 3000 watt generator I felt that I could do light stick welding to my heart’s content. I just had to weld the frame for the genset first…. A trial run using 2.6 mm electrodes on the house supply (1300 watts, remember) with fridges, iron, rice cooker and kettle turned off proved successful, and the genset was completed. With relish I fired it up to do some serious welding.

I was shocked. The genset response to the wildly fluctuating load of a welder was hopeless. I could not successfully burn the small rods. It was much better to use the house supply, drawing perhaps 1000 watts that the genset could not deliver.

Chastened, I next discovered that when the genset was running both fridges ran with a worrying low-frequency vibration superimposed over the normal slight vibration. Was I heading for fridge failure? The lights also flicker.

My first checks were frequency and voltage. Here there is a problem – Indonesians do not measure the frequency of their genset output. The small commercial genset units are factory preset to 50 herz, but what about the popular diesel thumpers? This was resolved by borrowing a Herz meter from the power company – none could be bought in town. I found that when the output voltage was around 220-230 volts the frequency was close enough to 50 Herz. The vibration remained. It also showed, slightly less, when using the small genset.

Searching the net, I found a few reports of flickering lights and one or two technical treatises on “interharmonics”. My lay mind sees this as the potential problem: fluctuations in engine-dynamo speed set up a “beat” effect, the combination of cycles of different frequency resulting in something like spring tide “in-phase” voltages interspersed with neap tide wave interference, explaining the flickering lights, the vibrating fridges, and the high-tech high frequency welder tormented beyond all reason.

One website suggested that there is a commercial unit which overcomes this problem. Little else.

Is there a neat explanation of this phenomenon, and a simple and cheap solution?


Joined Jul 17, 2007
Very interesting post.

The first thing I'd suggest would be to put your country in your profile. While you have stated it in your initial post (this is always appreciated), adding your general location to your profile actually helps considerably with replies. For instance, it would likely be worthless for me to suggest parts that you might only order in the USA; quite frankly I would have no idea where to suggest for you to obtain parts.

Secondly, 1 HP = 746 Watts; it seems initially that you have purchased an engine with 50% overcapacity, assuming that the generator and drive components are 100% efficient. Unfortunately, mechanical drive components are nowhere near that efficient, and neither are generators. You are very likely actually operating at near the peak HP of the engine; assuming 75% efficiency in the generator, and 75% or better in the V-belt drive; numbers scientifically picked out of a random hat.

Diesels tend to be lower RPM engines, requiring very heavy flywheels to keep their RPM's in the ballpark of "stable" - or to even keep running. 16:1 compression ratio is about the minimum to sustain ignition; ratios up to 22:1 or more are often used. The compression stroke is the equivalent of slamming on the brakes in an ordinary automobile (as far as crank speed is concerned) and the power stroke is the rocket blast-off.

While V-belts will absorb some shock, that is generally via slippage. Slippage = loss of power.

A couple of thoughts come to mind; one mechanical, one electronic. The first is reasonably economical, so let's pursue that for the moment.

The engine has a heavy flywheel, but it's likely that the generator itself does not. Were the generator fitted with a heavy flywheel, and the motor connected to the flywheel with a torsion bar drive, I believe that the generator speed would be much more consistent.

However, making a flywheel that is both statically and dynamically balanced by hand would be a challenge for most, even with a well-equipped machine shop.

A torsion driveshaft would also be a challenge; it would have to be tough, yet springy. The harmonics of the engine and the generator would have to be taken into consideration, along with the material engineering, and frankly my math skills aren't up to that.

The electronic solution would involve taking the output of your existing 3kw generator, and building an inverter to generate a reasonably pure 50Hz 220v/230v sine wave. This would not be easy nor cheap, and would require a good bit of investment in time and materials.

Sorry I don't have any "cheap" or "easy" solutions for you, but I feel they are in the realm of realism.


Joined Dec 5, 2009
The governing of the generator motor to keep up with electrical draw is what is giving you the peak hits. That has already been concluded. I think Sgt.Wookies flywheel on the dynamo would be a good idea if you had no governing on the motor. The high torque causing slippage is giving you the problems. If you can keep the motor at a constant, you will minimize these problems. I have an old generator at work that allows the motor to run at top speed regardless of draw. It bogs down due to magnetic loading on the dynamo but doesn't change speed in relation to draw for fuel conservation. On other units, while using a droplight and a radial arm saw at the same time, will blow lower rated bulbs when stopping the saw.

Staying at full rpm opposed to a fuel-savings mode, seems to 'regulate' the power better.

If fuel isn't your main concern, keeping the motor from changing speed in relation to demand (which is a slow process) by staying at max throttle could help. See if you have an option for 'auto idle'

on this site, they are talking about honda generators that continually produce 120 60Hz. Engine speed doesn't effect freq.
They also say almost all other generator manufacturers change freq, and voltage constantly, which can be bad on large loads.
Last edited:


Joined Mar 24, 2008
I'd put a UPS between the voltage sensitive components, such as the light bulbs or low current appliances. A good UPS regulates voltage almost as a side product.


Joined Jul 31, 2009
Sorry I can't help much technically but I appreciate your problem. Up until last summer I had lived in Nigeria for a couple years (oil job). The power there goes out at least once a day, usually three or four times, and I lived in one of the good areas. We had a big generator in our compound, 22 townhouses, but made no attempt to have a UPS. When the grid power went out, we would all sit in the dark (if it was night) and listen to the door slam as the guy on duty went out back and fired up the big diesel.

When it was up to speed then everything would come back on. When the grid power would come back on, he would wait a few minutes to be sure it would stick and then shutdown the generator and put us back on the grid. I actually preferred being on the generator because the frequency and voltage were a lot more stable. The grid power would fluctuate like crazy including blowing up light bulbs occasionally. It was common for people to buy these "stabilisers" to protect computers and TVs. I never bothered because those things were just junk anyway and never had any trouble. Modern electronics have pretty good power supplies of their own.

Sorry to drift off topic, but we do have it really good powerwise. I mean when it goes out here in California it's reported on the news! Can you imagine? I once figured that if you took the whole countries electrical generating capacity and divided it up amongst the 150 million Nigerians, each would get about 50 watts from the grid.

Thread Starter


Joined Feb 12, 2010
Thanks. I'm going to look up UPS threads first, before going into this further.

UPS is another topic of interest.

I thought of changing the pulley sizes and letting the engine loaf at higher revs and flattening out the fluctuations.


Joined Dec 20, 2007
I can't remember when was the last time the electricity here failed. Maybe a couple of years ago. The authorities planned in advance very well.

The electricity consumption hit the highest peak in winter due to Christmas incandescent light bulbs instead of the usual peak in summer due to air conditioning. So the authorities traded new LED Christmas tree lights for each string of old incandescent lights. Now everybody has LED Christmas lights and the electricity consumption is lower.


Joined Oct 2, 2009
When the fridge/freezer shuts off, you are experience what is known in the trade as a "load-dump". In most generators (or alternators) the output current is proportional to the instantaneous field current. The field has a lot of inductance, and even though the voltage regulator reduces the field current instantaneously, if takes a while for the field current to decrease just due to the inductance of the windings, which in turn causes the generator to put out excessive current (for the reduced load), which causes the voltage to spike up.

I deal with this in aircraft and automotive alternator systems.

Just off the top of my head, it would seem that this could be controlled with an add-on shunt regulator. As the thermostatically-controlled load shuts off, the shunt regulator sees the resulting voltage increase and switches in a temporary load equivalent to the fridge, and then slowly begins turning the added load off, allowing the main regulator to catch up...