New-to-me-803A!

[Fungal]

My wife and I live on an off-grid farm... something I've been working on getting operational for about seven years. We sold our primary residence last year, and moved out to our farm full-time. We have about 20kW of solar panels in five sub-arrays, and approx 90kWh of battery storage. We rarely run out of power, but sometimes mother nature doesn't cooperate, and we use generator power to charge batteries (if we have too many rainy days in a row).

We live in hurricane country, and I've historically been a devotee of Kohler generators, mostly NG or Propane-fired. I've owned four, and they've mostly been reliable units,. I had been pretty satisfied.

Had.

My lastest Kohler generator was a 12RES model. That's a 12kW, 3600 RPM, air-cooled set. It puts out 240VAC, which is what my main inverter requires. It self-destructed at 311 running hours, and only 14mo of operation. I reached out to Kohler about warrantying it (their off-grid warranty is 18mo, or 1000 hours, whichever comes first). They declined.. apparently their warranty requires the generator to be installed-and-registered within 12m of purchase... and it took 19mo to get this one installed.



Yeah. Great. Thanks for nothing, Kohler.

It had been sitting on its own pre-cast concrete pad, so I removed the generator, and the pad. We added some paver base to level it out, packed it down, and added a second precast pad to accommodate a larger generator.







I already had a MEP-831 that I've been playing with (very finicky generator... it runs, but I'm still working on it), and elected to replace the Kohler with a MEP-803A. Ziggy was kind enough to sell me one of his, and it was promptly delivered on a truck, in great shape:





(to be continued)

 

[Fungal]

What's interesting about this generator is that it doesn't seem to want to put out power at its rated capacity. These are supposedly rated at 10kW, with the ability to go 10-20% higher if needed.

This one doesn't.

Here we are with a roughly 7kW load:




Here is the power graph from my inverter, which is charging batteries with 4.5kW, and passing 2.4kW on regular AC load to the generator, for a total load of 6.9kW.



I put a clamp-on ammeter on each leg of the generator to see what's happening. It's pretty well-balance between the two legs:







Any thoughts on why we're seeing this? The generator's panel-mounted meter claims 125% or so of rated load... but a 7kW load should only be about 70% of rated capacity.

What am I missing?

 

[ZiggyO]

I reached out to some of the experts here on this-- I will let them chime in here.

For background info, I load tested the genset using an essex a427a load bank, but I did so on 120/208 3ph as that is how the load bank is primarily configured. It passed with flying colors. I have since learned that the 120/240 1ph configuration monitors load differently than the 120/208 3ph configuration. I am interested in figuring out what is going on here as well.

 

[Toolslinger]

I recall reading in the past where someone had the percent output setup from an 802a rather than 803a. That would be about right on your draw vs meter reading I believe.

 

[Fungal]

I recall reading in the past where someone had the percent output setup from an 802a rather than 803a. That would be about right on your draw vs meter reading I believe.

I had thought about that... and it would make perfect sense. That would be about 125% (or a little more) of the rated load for an 802A, but this is definitely an 803A, confirmed by data plate, and the fact that the engine is a four-cylinder. The 802's are only a two-cylinder.

I wonder if the wrong electronics got put into this one, and so it THINKS it's an 802 instead of an 803?

 

[Toolslinger]

Yes, that's what I meant. Most of the electronic side is the same parts. I haven't had to deal with the percent rated output portion yet, so I'm a bit fuzzy, but there is a way to convert one setup to the other. It has been discussed, you just have to find the right thread.

 

[kloppk]

The electronics are exactly the same.
The only difference is number of turn of the gen head output leads thru the CTs.
803 use half as many turns as an 802.
Turn count is listed on the schematics in the -24 TM.

802A



803A

 

[Fungal]

So this one might have an 802A generator head, bolted onto an 803A engine?

 

[Toolslinger]

No. It likely has the electronics setup for the 802a. It is a matter of how many loops of wire are running through the current transformer in the back electronics enclosure. (thanks Kloppk, I knew I remembered this...)
I can't believe I don't have a picture handy of that area, but you would have to take the top panel off the electronics enclosure to see what is going on.

 

[ZiggyO]

Kloppk gave me some more insights on how the monitoring works. Here is what he explained to me:

"There is a fault that can cause a 803A to pass a 208 3 phase load test but fail with an Overload fault when running 120/240 single phase. Namely in S8 contacts 13 to 14. In 120/240 S8 13 14 close placing burden resistors R13 in parallel with burden resistors R12. If those S8 contacts don't close properly R13 dosen't get connected in parallel to R12. The result is that load current sensing circuit will falsely send a load signal to K8 that is twice the actual load. That will cause K8 to trigger a false overload.
Treating S8 with Deoxit and exercising S8 usually resolves the issue.
Only do that with the genset stopped.
It's a moderately common issue with S8."

As I mentioned earlier, in 120/208 3ph it fully loaded up-- if my understanding is correct, would it only load to half capacity if the ct loops were wired as if it were an 802?

 

[kloppk]

- if my understanding is correct, would it only load to half capacity if the ct loops were wired as if it were an 802?

Correct. Doubling of the loops thru the CT's would make the current sensing system think there is twice the load.

 

[ZiggyO]

Ok, so that means that there should be the correct number of loops since it loaded fully in the 120/208 3ph mode. Assuming that is the case, then the likely culprit are the two burden resistors not going into parallel via S8?

 

[kloppk]

In 120/240 Single Phase and AM-VM set to L3-L1 the %Load meter M8 is driven by the current sensed by CT2.

I'd suggest measuring the AC Voltage at K8 terminal pairs:
1-4
2-4
3-4
The AC Voltage across each pair should be almost the same if everything is working properly.
5.6 volts AC represents 100% Load
7.5 volts AC represents 133% Load

Here is a easy to read schematic of the wiring

 

[Fungal]

OK. It's a little late in the day today, but let me see if I can load it up tomorrow, and get some voltage numbers from those terminals.

 

[Tinstar]

This is from my 803A upon initial inspection.
(yes, it was cleaned after pic was taken)
I think this is what you guys are referring to.

 

[kloppk]

An 802 CT & loops for comparison


For the curious, this is what's inside the CT assembly

 

[Fungal]

OK... so I have some voltage readings at a 7-8kW load:

The 3-4 reading is the standout:

1-4 is 7.22 VAC
2-4 is 7.13 VAC
3-4 is 3.96 VAC

 

[kloppk]

Interesting results. I have a theory. Letting morning coffee soak in first.

 

[kloppk]

In 120/240 mode the gen head windings get configured like this by S8.


CT2 is what drives the %Load meter in this case. CT2 has stator windings output currents 8-11 & 2-5 passing through it.


In 120/240 mode the AC voltages generated by the CT's & burden resistors going to K8 should be virtually identical.
However, the AC Voltage you measured K8 3-4 was about half of the other two.
My theory is if one of the windings either 8-11 or 2-5 aren't carrying any current then CT2 would only generate half the current and burden resistor R11 would only generate half the voltage going to K8 3-4.
If so, it may be due to a few issues. They could be...
- An open stator winding
- Faulty S8
- Wiring error
- Incorrect number of loops through the CTs

I'd suggest performing the Stator Resistance check in the -24 TM to verify all 6 of the stator windings are good, they are isolated from each other and chassis ground.

While you have the 12 stator wires disconnected from TB3, I'd check to see if S8 is making the following contact closures while in 120/240 position.
I would check from the terminals on TB3 for the ones that come directly from TB-3 to S8 terminals.
Also check continuity from the TB3 terminals to be sure they are wired correctly to S8s terminals.


That's my theory after two cups of

 

[Fungal]

This model of generator is new to me. What do I have to open-up/access to start checking all of that? Is that all accessible from the control cube?


*** edit ***

And I clearly need to get a printed TM