Posted by Mr Wizzard on December 21, 2005, 7:45 am
Got a new Honda eu2000i to run my home furnace here in the NorthWest.
No go. Furnace is a York Diamond 80 (small modern garden variety deal)
Furnace starts its cycle, ventor motor runs, hot-surface igniter fires, gas
valve open, furnace fires, igniter shuts off and starts its cool-down, and
after about 5-6 seconds, the gas valve closes, and flame goes out. Then it
immediately goes back into the hot-surface igniter glowing cycle, and
Put it on my overly expensive 200 Mhz Tektronix hand-held digital scope,
and shit, the sine wave is cleaner than anything you'd see in a printed text
book. Expanded the heck out of the trace, and see very slight high frequency
hash riding on the sine wave, but its _so_ minute, that I didn't give it a
thought (more impressed in how utterly perfect the sice wave looks in
At first I thought it was a voltage surge after the high-current igniter
but realized that the gas valve kicks out about 5 seconds _after_ the
shuts off. Frequency is dead-nuts on. I tried a large Variac, and other
Decided to start putting shit (inductors etc) in series with the Honda
and the furnace, and started rooting thru my junk box. Found a real nice
ferrite pot core inductor with what looks like to be 16 gague wire. About
an inch and a half in diameter, probably an inch high. (nice looking thing)
Anyways, well-A! Furnace starts and runs fine. Worried about possible
too thin wire, I put the scope "accross" the inductor and let the furnace
run thru its cycles. Man!, talk about hash? Whoa. Some shit getting
caught in that choke, *thats* for sure. Anyways, getting about 5v PP
accross the choke (which by the way looks like it is a component of
a high frequency switching power supply ro something). Anyways,
5v peak-to-peak accross the inductor is not enough loss to worry about,
and the inductor is not getting warm, so the heck with it, I'll put it in a
little project box with extension cord ends and make that portable
So there you have it. My conclusion is that the hash (running around
the electronic control circuit board) was screwing up the "flame sensor"
(of all things) which is probably an ultra sensitive analogu thing like a
thermistor. Wish I could find my WaveTek meter that measures inductance.
So maybe you ran into this situation, maybe you didn't, but thought I'd
pass this on.
Posted by nicksanspam on December 21, 2005, 1:26 pm
Got 2 Variacs? How about plugging the EU2000 into a wall socket?
Posted by Mr Wizzard on December 22, 2005, 6:49 am
Got tons of Variacs. (you should see this 15-Amp job I got)
How about plugging the EU2000 into a wall socket?
What would that do ?
Posted by nicksanspam on December 22, 2005, 11:24 am
I used to think 2 EU2000s sent special signals to each other in order to
operate in parallel, but it seems they don't. The AC outputs are just tied
together and the inverters figure out how to sync and share the load, so
it seems very likely that a single EU2000 can act as a grid-tie inverter :-)
I don't own one, but it seems safe enough to try this with a variac hooked up
as an autotransformer and 2 light bulbs or another Variac in series. If it
syncs, the bulbs should be dark. Viewed in a fixed font:
| | C
| Honda | C 120 VAC wall socket
| | C
Moving the variac slider s downwards should make the bulbs light again.
Then short out one bulb, then the other, put a Kill-a-Watt meter into the
Honda socket, and run the exhaust into the top of a $00 gas water heater,
then out a window, cooler, to preheat water for showers, so most of fuel's
heating value ends up in the house. You are trying to heat the house with
the Honda powering the furnace, no? The house needs a CO detector.
With a 125K Btu/gallon high heating value and a 10,000 hour engine lifetime,
Honda EU2000 cogen looks marginally economical: if we burn 1.08 gallons of gas
with a fuel value of 135K Btu in 4 hours at the 1600 W rated load and make
6.4 kWh (21.8K Btu) of electricity and 113.2K Btu of heat (another 33.2 kWh,
ie 39.6 kWh total) and the heat replaces electric resistance and the Honda
costs $99 (mayberrys.com) and wear adds $.09/h, ie 5.6 cents/kWh, this kind
of cogen seems to make sense at an off-road gas price of $.75/gallon if
electricity costs more than 100x$.75/39.6+5.6 = 10.7 cents/kWh.
Posted by philkryder on December 23, 2005, 5:24 am
do you think that those hondas will give 10,000 hours of service at
1600w continuous load?
Does this also imply that more than 2 of the Hondas could be hooked in
If so, then, with NG, this might make a very attractive n+1 off grid
power source - with both high reliability and low cost.