Low voltge from ST 15 Gen Head

[Horsepoor]

Fact pattern:
Found the perfect spring for my 16 HP listeroid so it will hold 60 hertz.
Got the fuel rack set just right so it can handle anything from 100 watts to sustained 10,000 watts loads with only a 1 to 2 herz drop.
But my voltage drops to 102 volts with a heavy load which should not be a problem of a ST 15 head running at only 10 kW.
I've looked in the blue dog house, but know very little about what I am seeing, request detailed advice on how to bring voltage up - please.

[cujet]

Hi Bruce,

Disconnect your AVR. It cannot raise the voltage above what the ST head will naturally provide.

Raise the RPM a touch. Voltage climbs rapidly with RPM. Don't worry about going all the way up to a smokin' 62HZ. It's OK.

We have considered a capacitor to raise field voltage and smooth out the ST sine wave. I think this is worth strong consideration.

Chris

[MacGyver]

Yes, a little capacitance across the field winding helped greatly on my ST7.5.

I ended up with a 220uF @100V on mine. It brought the voltage up about 10 volts *and* made the waveform look better.
I tried larger capacitors, but they made little difference in voltage or waveform beyond what the 220uF did, and the larger caps slowed the response during changes in output current. (Large caps made the voltage overshoot quite a bit for a second when a large load was removed)

[ronmar]

Cujet, He has an AVR?  The ST's don't usually, and it dosn't sound like it to me.  If he had a properly functioning AVR I would suspect, he could set the voltage anywhere in the vicinity of 115-120V(and probably higher) that he wishes, within the load limits of the genhead of course.  I know I can when I hook one up to my ST-5...

The capacitance on the rectifyer output/field input, should help boost the output as mentioned.  When I did mine, I treated it like a DC power supply. I followed the DC power supply rule of thumb of 2000uF of electrolytic capacitance per amp of current flowing thru the rectifyer.  This gave me a very pure DC to the field with ripple measured in millivolts...  This also allowed me to run a hour meter connected directly to the rectifyer output, so it counts time whenever the generator is making power.

And speaking of rectifyers, check your rectifyer connections, or better still replace the stock rectifyer with a domestic solid state model.  If you have a bad diode, or bad connection on the rectifyer, you could be loosing harmonic current that needs to get to the field to keep the output voltage up...  It is good to have a spare anyway, and they are not very expensive. Not sure what your gen specs are, but one of these should do the trick  http://surpluscenter.com/item.asp?UID=2008081711305814&item=22-1206&catname=electric  I also noticed a little output waveform improvement when I replaced the stock screw together rectifyer with one like in the link.  But I also soldered all the poor factory electrical crimp connections in conjunction with the rectifyer swap, so that may also have had a hand in the improved output waveform.   

[MacGyver]

The capacitance on the rectifyer output/field input, should help boost the output as mentioned.  When I did mine, I treated it like a DC power supply. I followed the DC power supply rule of thumb of 2000uF of electrolytic capacitance per amp of current flowing thru the rectifyer.  This gave me a very pure DC to the field with ripple measured in millivolts...   

I'm not sure that the rule of 2000uF per amp applies here. My (extremely limited) experience shows that too much capacitance leads to voltage overshoot problems.

I started with 11,000uF and worked my way down. 4700uF, 2200uF, 1000uF, 470uF, 220uF, and 100uF.  With 100uF the waveform was obviously not as good as with the higher values, but the difference between 220uF and any of the larger caps was almost nothing on the scope.

What I did notice, especially with the 3 largest caps was that there was a substantial voltage overshoot when a large load was dropped. I didn't like seeing that output go to 135v for a second when I suddenly switched off 2000 watts of load.
I ended up with a 220uF cap,  and a 2 ohm resistor in series with the field as being the combination that gave the best waveform and most stable voltage for my particular ST.

I'll get some more scope photos to post of the final configuration once I get it all put back together. I've got the cooling system and generator head off of the frame right now because it's all being moved to it's new shed...

Here's the link to the few photos I took with no capacitor and with 11,000uF
http://www.weirdstuffwemake.com/sweetwatergems/geek/st_head.html

[Horsepoor]

Thank you to each of you for your posts. Yesterday was generator work day with Tropical Storm / Hurricane FAY threatening. Having dodged this bullet, I intent to work on gen head for the next storm in the pipeline. I just want to be sure I have to fix whatever I break during my experiments. Need some feed back please. If the 220uF worked well on the ST 7.5 gen head, so would it be reasonable to a 470 uF on a ST 15 ? I am pulling as much as 10 KW during peak demands. Also, thank you for the links, reading the web sites and vendor info is very helpful. I look forward read your / any comments.

[oliver90owner]

Am I missing something?

Is it running at 10kW from 16HP?

If'n it is down to 100V from 120 then the output is likely down by about 15%?


Sounds to me like the limit might just be engine horses and the voltage is drooping until it can just keep up?

Is it smoking at 10kW?

Regards, RAB

[ronmar]

No, you are not missing anything RAB, but I sure did.  That is a very good question...  Bruce, how are you maintaining 60HZ at 10KW with only 16HP?  Or worded differently, are you SURE you are maintaining 60HZ, or at least the 58 "with only a 1 to 2 herz drop" you mentioned at that 10KW load?    If the problem was available engine HP, then the frequency would fall off under load, which in turn would let the voltage droop excessively in a harmonically excited head.  As mentioned, a few HZ can make a large difference in output voltage.  I got the impression from your original post, that you were maintaining around 60HZ, which means to me your load was not exceeding available HP.  But I missed the 16HP available you mentioned at the start.  With losses involved in belting, driving an oversized head, and power generation in general, that exceeds the 2HP per KW which is generally accepted as the sustainable power generation figure...

Bruce, because of the nature of this type centrifugal governor that is directly coupled to the throttle, they have to vary(decrease) a little in frequency/RPM to give a steady throttle increase with a load increase.    The spring is trying to drive it toward full throttle,and the expanding weights are trying to drive it to no throttle with RPM increase.  Since the throttle position is dependent on flyweight position and that in turn is dependent on engine RPM, there MUST be a steady decrease in RPM as the load increases, otherwise there will be NO throttle increase to counter the load.  A good way to set one up is to set for 62HZ with no electrical load.  Your maximum sustainable load will be reached as the RPM decreases under load to provide around 58HZ.  In your setup, I would guess that to be closer to 8KW than to 10KW.   

[Horsepoor]

Thank you on your posts - each comment was proven to be correct, although I did not get so far as to disconnect the AVR. The low voltage associated with a heavy load was also associated with a 58 Hz or 57 Hz frequenncy reading. Ok - next to the HP figures with some verifed measurements but first I have to say that I actually have is a GTC rated at 20 HP at 1000 RPM. I down rated this to about 850 RPM + or - 10 RPM as calculated from the pulley on the gen head to produce 60 Hz. Here are my most recent full rack
Measurements:
Initial setting 62 Hz @ no load, increased load slowly until I was pulling 47.3 amps at 230 volts, slowing bouncing between 58 to 57 hz. I calculate 10,879 watts at this load which I held for more than 20 minutes while I measured temperatures on the exhaust system and radiator. Maximum heat on the exhaust system was 655 F right side and 635 on the left. Fuel used - waste jet fuel. Black smoke very noticable. The cooling system had a maximum of 215 F at the outgoing flow with some visible boiling in the 50% coolant mix as it entered the flexible tubing in a 5 psi pressure system. The return was 170 F for a consistent 40 F temperturn drop. This 40 F drop was even present at lighter loads.

Conclusions: Could have pushed for a little more power out of the engine but I would have had to cheat by opening the fuel racks to 100% by hand. I can truthful state 11 kw is possible but I really dont want to run it this hard for very long. The system lives a very happy life at 25 to 35 amps at around 230 volts for 6,000 to 9,000 watts. It runs both of my central air-conditioners (one 2 ton down stairs and one 2 1/2 ton upstairs), assorted lights, 5000 BTU window air conditioner, frig(s), well pump, etc. I am extremely pleased.

It has taken me almost two years to reach this point starting with cleaning all that casting sand out of the engine, re-machining parts, building the stand, refining the cooling system, finally finding the correct spring, breaking it in (about 20 hrs) so the rings are pretty well seated, rotating tappets, etc. I am so please with the Lister engine design, this must be the adult replacement for the little pill steam engine I never received as a child. I post very little but enjoy reading this forum and your comments with great interest. Thank you for your help.

[RJ]

The capacitance on the rectifyer output/field input, should help boost the output as mentioned.  When I did mine, I treated it like a DC power supply. I followed the DC power supply rule of thumb of 2000uF of electrolytic capacitance per amp of current flowing thru the rectifyer.  This gave me a very pure DC to the field with ripple measured in millivolts...   

I'm not sure that the rule of 2000uF per amp applies here. My (extremely limited) experience shows that too much capacitance leads to voltage overshoot problems.

I started with 11,000uF and worked my way down. 4700uF, 2200uF, 1000uF, 470uF, 220uF, and 100uF.  With 100uF the waveform was obviously not as good as with the higher values, but the difference between 220uF and any of the larger caps was almost nothing on the scope.

What I did notice, especially with the 3 largest caps was that there was a substantial voltage overshoot when a large load was dropped. I didn't like seeing that output go to 135v for a second when I suddenly switched off 2000 watts of load.
I ended up with a 220uF cap,  and a 2 ohm resistor in series with the field as being the combination that gave the best waveform and most stable voltage for my particular ST.

I'll get some more scope photos to post of the final configuration once I get it all put back together. I've got the cooling system and generator head off of the frame right now because it's all being moved to it's new shed...

Here's the link to the few photos I took with no capacitor and with 11,000uF
http://www.weirdstuffwemake.com/sweetwatergems/geek/st_head.html

How many watt resistor did you have to use? I also ran across this recently.

http://www.rentaitrade.com/ProductShow.asp?ArticleID=137


[/img]http://www.rentaitrade.com/UploadFiles/2008515104225201.jpg[/img]

INPUT Voltage : 70-140V AC(PMG,or auxiliary winding,single,three phases)

Reaction is measured:90-480 VAC

MAX.OUTPUT:DC 10A 125V

Function:Built-in low-speed,over excitation LAM,

REGULATION:+/-0.5%

Thoughts?

-RJ

[MacGyver]

How many watt resistor did you have to use? I also ran across this recently.

http://www.rentaitrade.com/ProductShow.asp?ArticleID=137

The resistor I used was 100 watts... because I had one. If I had to replace it I'd buy a 50w.

The actual power dissipated in my case is about 16 watts max.
I will eventually replace mine with a 5 ohm 50w adjustable resistor like this so I can set my voltage just a little lower.
http://search.digikey.com/scripts/DkSearch/dksus.dll?Detail?name=AVT50-5.0-ND

I wouldn't touch a cheap chinese AVR with a 10 foot pole, unless I could afford to buy 3 of them and stash 2 away as spares.

If I decide I need an AVR, I'll build my own so I know it's built "Cockroach Tough". And yes, I'll still build at least 2 extras to stash away as spares...