Getting a generator to sync to 50Hz

[large_ghostman]

Hi
I am having problems getting a generator to stay at 50Hz, I have found that there is no sweet spot regarding the revs and the RPM to get 50Hz!!

Its driving me mad, I can get 55Hz and I can get 43Hz, but for some reason I cant get the revs bang on for 50Hz. So I give up and admit defeat and ask for ideas on how you get the generator to stay in sync. I thought about grid tying it but so many rules etc I dont think this would be a good idea

 

[crutschow]

What is the generator and how are you trying to adjust it to 50Hz?

 

[Nigel Goodwin]

Why does it matter?.

 

[large_ghostman]

its a diesel generator, I was trying to alter the throttle depending on load, but even at idle loads I cant get the revs right. Its a 3 cylinder lister running a 32KVA generator. As why it matters, there are short periods where there might be a crossover between gen and grid being on.

I am not convinced it does matter too much within reason, but I dont want to risk some of the equipment just because the frequency is off. So maybe the first question should be what kind of ley way is there in mains frequency for things like oscilloscopes etc?

My main issue is when the mains comes back on, here its common for it too switch on and off a few times in a matter of seconds when the power gets restored after a cut. The gen is currently set to stay on until the mains is stable for 2 mins, also sometimes I switch it on when the mains is there but have to let it warm up for 5-10 mins so the engine settles.

I dont know maybe I need to step away from this project for a week or two, I have battled this bloody generator for a couple of weeks and now I cant see the wood for the trees. Its not my generator and I got it going for someone else, but its proving a ***** to get the revs right. I think 50Hz is around 1200 RPM but I wouldnt trust the rev counter on it, I also think the turbo is a bit of a problem as it seems to kick in around 1400rpm (according to gauge).

So at the moment I am sort of trying to balance it between the cut in of the turbo, sorry I have gone brain dead with this one! Its been nothing but agro from day 1.

Nice engine but I dont think the gen is matched to it very well

 

[Nigel Goodwin]

I am not convinced it does matter too much within reason, but I dont want to risk some of the equipment just because the frequency is off. So maybe the first question should be what kind of ley way is there in mains frequency for things like oscilloscopes etc?

Frequency is pretty well completely irrelevant for the vast majority of gear, and certainly so for a scope.

I've even done discos on generators (cheap ones), as long as you don't have synchronous motors on the decks it's fine - what you don't want is synchronous motors and frequency variations, the ear is VERY sensitive to changes in pitch (wow and flutter).

 

[tcmtech]

It sounds like the engine has a sloppy governor that is worn out in the 50 Hz light load holding range.

For a 50Hz output your engine RPM's should be either 1000, 1500 or 3000 assuming it's a direct drive 6, 4 or 2 pole system of course.


Also are you sure your fuel filters are clean and there are no air leaks letting little bubbles into the injection pump? Those simple oversights can make diesels hunt all over the place in load ranges that they should be stable at.

I've got two Allis chalmers diesel tractors (8030 and 7050) that have mystery air leaks in the fuel systems that are driving me nuts for that very reason. They won't drip fuel anywhere when the systems are pressurized but they loose their prime and surge and hunt at light loads which is obviously a air leak problem.

 

[large_ghostman]

There isnt a governor as such, the throttle is controlled by a servo I rigged up, the problem now is the arm dosnt have fine enough control of movement. So every time it moves just a single step it alters the revs by too much or too little, the turbo is also a bit of PITA with a bit too much float in it.

I am pretty sure the engine never had a generator on it originally, I think it was put on later on. I might try and use a stepper motor or something and try and get finer control, its not mine and to be honest I dont think they are willing to spend what it needs.

I have spent alot of time trying to bodge and coax it into life when really a few £ would have sorted it. The fuel pump needs a overhaul as well and probably new seal insides, I was told it has sat for 20 years but I am inclined to think thats a conservative guess lol.

I offered to buy it but they want too much for it

 

[tcmtech]

There isnt a governor as such, the throttle is controlled by a servo I rigged up, the problem now is the arm dosnt have fine enough control of movement. So every time it moves just a single step it alters the revs by too much or too little, the turbo is also a bit of PITA with a bit too much float in it.

Well there's your problem and you already know it.

What exactly are we supposed to do about poor design on your part?

 

[mbarazeen]

If the frequency is flexuating There must be a way to adjust it mechanically to control the gain of the feedback loop. Or the governer may be having a mechanical issue. You can possibly think of upgrading the control. Also check the changes in the load, that may influence on RPM.

 

[large_ghostman]

Well there's your problem and you already know it.

What exactly are we supposed to do about poor design on your part?

Yeah I know I know......... I had hoped for some way around it, the engine didnt have a governor on it so I had to bodge it. Maybe I can use a stepper in 1/2 step mode and put a micro on to count the revs, I was also told that tying it to the mains would force it into sync. The problem I have with that is if the power fails and they try and work on the lines!!

I might take the turbo off altogether and that will get rid of one problem. I was first told £200 max for parts but slowly thats creeping up, I would give u but then I dont get paid!! On the other hand its getting to the point its not worth the trouble, I think originally they assumed because they have access to waste veg oil that the electric would be free.

I was told the plan was too use the oil just filtered, this didnt work well at all so now its turned into 'real' bio diesel with methanol etc. They have 3 takeaway shops and a small farm, but really tight with money. They think you can make the fuel bi monthly and be done with it, I tried to explain how hygroscopic it is and how they are better making what they need every couple of weeks. Might as well talk to the engine!

It would be a nice generator if it had some money spent on it! I also sugestead they put a heat exchange on the exhaust system and water cooling, but they thought £100 was too much for the bits! When I get my generator I think I will go that route.
I love the Lister engine though, it was nice to work on except the old imperial sizes for spanners lol. Also finding a couple of taps for stripped threads was a PITA, I love the thump thump noise it makes

 

[dr pepper]

Marine gens sync or sort of sync using a pid and throttle control, older ones still have some slight error and there is a spike at switch over, sync on some gens can be done by the gen acting as a motor if sync is well off forcing the 2 to an extent to sync, or blow a fuse.
A more reliable less mechanical way would be to use an inverter generator, where the gen itself gens dc, then its just a grid tie inverter to the mains, possibly done ghetto style with a car alty.
Read up first on the reg's, you dont want to end up electrocuting some poor lectric board worker that thinks he's shut down your supply.

 

[tcmtech]

What was that engine originally used for and how were they normally governed when used in stationary constant RPM applications?

 

[Tony Stewart]

THe rate of change of RPM under no load and full load is very asymmetrical in gain and thus stability of the loop must be defined or characterized first before an optimal PLL can lock properly. THis may require a dummy load under no load conditions or a bigger flywheel.

In any case the transfer functions must be known for each stage for each polarity ( increasing and decreasing speed to define to proper feedback loop under all load conditions.

Phase frequency type mixers used in CD4046 CMOS devices can be used if you have a rotary encoder to have a frequency multiplication effect and thus higher bandwidth of the mixer filter for faster time response to step loads. Phase shift per second is frequency error and at line frequency, if you dont have enough inertia ( flywheel effect) to prevent frequency changes faster than you can control it, it will be unstable.

It appears you have a hysteresis effect on locked frequency, which indicates poor speed detection methods, so you need to examine your precise dynamic feedback of frequency error and your reference frequency stability that controls the rate of power generated and consider a frequency multiplying rotary encoder. THen use a similar PLL with a digital counter in the feedback to lock at a multiple of line rate. A type I mixer would not be effective as it can be harmonically locked. It must be a phase/frequency type mixer as found in CD4046. Any Xtal can generate the required reference frequency of line rate * N with a suitable divider counter.

The frequency error will depend on having adequate feedback gain, bandwidth and power and perhaps dummy load to control the generator RPM without going unstable hunting for speed. Any stiction in the power control will cause hysteresis hunting effects, as accelerating power may be more than braking or load power or visa versa. i.e. asymmetric gain in the feedback loop is needed for special filtering on PID. ie differentiating speed changes increases bandwidth and phase margin to an extent but gain of P I & D must be derived from solid control system principles and a good knowledge of the transfer functions of each stage. i.e. a better definition of the problem.

 

[tcmtech]

If it was me I would just put a old fashioned belt driven mechanical governor unit on it and call it good.

 

[Tony Stewart]

I remember a V8 driven rope tow on ski hill in FLin Flon, when I was in Grade 2. They were trying to fix the belt drive governor when it went out of control on full throttle and I was going up the hill faster than down.

The last thing I remember was "don't let go" ( till the top)

 

[tcmtech]

Drive belt failure is always possible with the old externally mounted belt driven mechanical governors but when compared to homemade electronic control they are still extremely reliable and are still used on many 100's of millions of power unit engines to this day. Many of them now multiple decades old and still working just fine. I should know being we have a few machines on the farm that are 40 - 60+ years old and still using them without problems.

I don't know what LG has for salvage or scrap yards in his area but around here I have always pulled them off junk equipment for $10 - $15 apiece and now have a few five gallon buckets full of assorted versions of them for 'just in case' I ever find a reason to use one or in case someone like LG comes along and needs one for a real project.

 

[large_ghostman]

I have no idea what it was used for, but I would guess its spent time in a fishing boat. I might try and get a governor but its reaching the point I am going to walk away.

I asked for £11 for a gasket kit and got asked if I could make them instead! Yeah I can make them but why spend hours of my time doing that for some else when you can buy them cheap.

Offered £300 for the engine and generator and said I was going to call it a day after the weekend, I also dont like the situation of the mains going down and the generator still being connected, I mentioned this a few times and no one seems that bothered about it. IF something did go wrong and the guy working on the line got a belt then its me and not the owner thats going to get trouble.

They have a tiny solar installation thats grid connected and I think they want to connect it to the meter and get the feed in tariff, its been denied but I dont want anything to do with that. I will look at mechanical governors for when I get my own generator, I am also after a lister

 

[Nigel Goodwin]

Offered £300 for the engine and generator and said I was going to call it a day after the weekend, I also dont like the situation of the mains going down and the generator still being connected, I mentioned this a few times and no one seems that bothered about it. IF something did go wrong and the guy working on the line got a belt then its me and not the owner thats going to get trouble.

Assuming it's just a standby generator?, then that's not a problem as you use either a manual switch, or an automatic relay, to disconnect the mains when you use the generator (a double pole double throw switch/relay disconnects both live and neutral). For my small (2KW) generator I used a separate fusebox that feeds the house lighting, and the downstairs sockets - so I could use the TV and Microwave in the kitchen - just have to remember NOT to plug the kettle in I fed this fusebox via a DPDT relay, with a 240V AC coil, the common contacts feed the new fusebox, the NC contacts go to the incoming mains, and the NO contacts go to the generator. The relay coil is wired to the NO contacts - so when you start the generator the relay automatically switched from mains to generator.

There seems little point in trying to do mains feed in from a generator, as they cost far too much to run to make it viable, and there's VERY strict controls on what you can connect in that way - which presumably doesn't allow home made equipment?. One obvious requirement of the professional equipment is that it doesn't allow feeding back down the mains when the mains is down.

 

[tomizett]

I agree with Nigel there - I can't see the need to grid-tie, and it's just adding a whole host of other problems trying to sync the engine to the mains when it's already tricky enough getting it to produce a stable output. The scheme described above seems absolutely like the right thing. With the generator running I'd imagine you could flip the relay to and fro and not disturb any appliances, no need to sync to the grid. I suspect your friends just want to make a few bob from the feed-in tariffs?

 

[Nigel Goodwin]

I suspect your friends just want to make a few bob from the feed-in tariffs?

It's hardly 'making a few bob' when it's costing you considerably more to make than you get paid for it