Swapping the leads should not make a difference. There is nothing special about a light bulb. Since you did test it under load I would try a different meter. Maybe an analog meter.
Thanks for all the interest from the more "electrical savvy" than me.
Here is the results of my testing last night:
- Light bulb test showed no visible sign of voltage drop. Testing under load did not change the voltage reading as either. The readings with or without load are the same. 72V (+/-) and 172V on (reverse polarity). The average of the two being 122V.
- The neutral to ground test showed 0 V. The "hot" to ground test also showed 0. (A possible grounding problem?)
- I reviewed the wiring diagram for the 3500ti unit and it does show leads going from the inverter unit to both the + and - contacts. I checked the Honda EU3000 wiring diagram and it also shows the same.
- I then tested a portable 175W inverter (cigarette lighter style) in my truck and it also shows Voltage coming from both the + and -. At vehicle idle the portable inverter showed 90V between the + and neutral and 90V between the - and neutral. Which seems to confirm what "TEB1272" posted regarding each leg of the receptacle making up 1/2 of the 120 voltage. It appears that maybe the inverter unit is not putting out the proper voltage on each lead?
- I ran through the trouble shooting section of the Kipor service manual (kiporguydotcom) and it seems to point to the inverter as well. I have 330V upstream of the inverter while the generator is running, and > 30V & > 1V while pulling the start cord on the black-black / white-white test.
I hope someone can prove me wrong because if I have to ditch this Kipor I'll likely be likely going to a Champion "camper" model for fear of losing this kind of money again!!
With all of the information here I think the only explanation left standing is John's speculation of a DC level superimposed on the genset's AC output. I was thinking along those lines when I made my first post but my brain couldn't put the scenario together then. But I'd still be hesitant to declare the genset broken without seeing the output on an oscilloscope, because of my fear that there could be another explanation. I can't think of an easy way you could separate the AC and DC levels to make a more confident measurement of a DC level. Maybe somebody else has an idea for that. Separating the AC from the DC is fairly easy, but separating the DC by itself isn't.
The Kipor's AC output is floating with no intentional reference to the chassis ground or the outlet's ground connector. A reading from a hot line on the Kipor to the ground has no meaning. The output of the small inverter, however, is most likely referenced to ground and will show a reliable reading (not 120) from either output to the chassis ground.
I was curious how some of my meters would handle a DC level when set to read AC. With a 14 volt DC input, analog meter #1 read 33 volts one way and had a small minus reading with the leads reversed. Analog meter #2 showed 25 volts one way and 5 volts with the leads reversed. Both digital meters refused to display a stable reading, but if there was a superimposed AC component they might have. Based on this my opinion is that you can't tell for sure what the meter is telling you, but it's hinting that there might be a DC level in the Kipor's output, pending further verification.
The 330 volt reading (I assume AC) from the 3 alternator phases is in the right ballpark, similar to my Honda inverter generator. You should have this same reading from all 3 pin combinations mentioned in the troubleshooting info.
* This post was
edited 07/24/08 09:00pm by Wayne Dohnal *
Several of the posts indicated possible problems with the testing equipment. I wouldn't give up on it without getting a qualified electrician with really good equipment to check it out. Also, all kinds of odd things can appear to happen when improper grounding exists. Some of these guys can probably give you better advice than I can about checking grounds, but I don't think you should get 0 volts between both hot and ground and neutral and ground-especially if this thing works like I think it does and should show 60 volts in each case.
Light bulbs of the incandescent type are among the most sensitive things to overvoltage, and if you really have 170 volts a light bulb should last mere seconds. I would also look into the idea mentioned earlier of having both DC and AC voltages present-again the need for a really good piece of test equipment. The fact that the light bulb works OK, and that you keep getting readings that average about 120 volts, makes me think you are seeing an artifact of test equipment that cannot handle the situation this generator has.
Or maybe you could find someone with a Honda or Yamaha inverter type unit and see what readings your test equipment give on them. If it is about the same, it might be that your generator is OK.
Sine wave inverters produce significant high frequency ripple (this is a result of the PWM switching used to synthesize the sine wave).
Internal to your meter, the common terminal is connected to the common ground for the whole electronic circuit. The "hot" input is only connected to a voltage divider input. If the common is connected to a very noisy signal, ever little trace on the circuit board becomes an antenna and currents start flowing in and out them, messing up the measurement process. When you reverse the leads, there is a different noise pattern and the readings mess up the other way.
Yes, I've seen this phenomenon before. It doesn't mean you meter is cheap either, I have seen $700 lab fluke meters demonstrate it. I usually figured that this meant I needed to clean up the filters in my design some!
You can try plugging in a surge protected power strip and measuring on the far side. The extra filtering may be enough to stop the error.
Quote: - The neutral to ground test showed 0 V. The "hot" to ground test also showed 0. (A possible grounding problem?)
Nope, it means the generator is ground isolated, and the insulation is really good. If there was a little leakage, you'd get some voltage reading from each measurement.
Matthew_B wrote: Sine wave inverters produce significant high frequency ripple (this is a result of the PWM switching used to synthesize the sine wave).
Internal to your meter, the common terminal is connected to the common ground for the whole electronic circuit. The "hot" input is only connected to a voltage divider input. If the common is connected to a very noisy signal, ever little trace on the circuit board becomes an antenna and currents start flowing in and out them, messing up the measurement process. When you reverse the leads, there is a different noise pattern and the readings mess up the other way.
Yes, I've seen this phenomenon before. It doesn't mean you meter is cheap either, I have seen $700 lab fluke meters demonstrate it. I usually figured that this meant I needed to clean up the filters in my design some!
You can try plugging in a surge protected power strip and measuring on the far side. The extra filtering may be enough to stop the error.
Quote: - The neutral to ground test showed 0 V. The "hot" to ground test also showed 0. (A possible grounding problem?)
Nope, it means the generator is ground isolated, and the insulation is really good. If there was a little leakage, you'd get some voltage reading from each measurement.
Shouldn't the ground be tied to neutral at the source (generator in this case)? This is a code requirement, I think.
Sorry I forgot to post the make of the meter. It is a Equus 4320 Digital Meter. It is by no means a top of the line meter. I checked the website "iequusdotcom" and the model is not even listed anymore.
Lots of info to digest from all the posts here.
One more test I tried was to plug in a 1500W unit heater with fan, into one of the 20amp outlets. The heater kicks on and the fan attempts to run but stops and the heater coil goes red hot and the overload in the fan shuts it off. Yet the overload indicator on the generator does not go on? But it seems to run an 1800W hair dryer with no problem?
Is is safe to do a test run on the RV microwave or A/C unit?
Quote: - The neutral to ground test showed 0 V. The "hot" to ground test also showed 0. (A possible grounding problem?)
Nope, it means the generator is ground isolated, and the insulation is really good. If there was a little leakage, you'd get some voltage reading from each measurement.
Shouldn't the ground be tied to neutral at the source (generator in this case)? This is a code requirement, I think.
.
Either way is legal on a portable generator.
If you sink a ground rod it's safest to ground the neutral. That way, if there is a ground anywhere, it will blow a breaker.
If you don't sink a ground rod when you use a portable generator, floating output is the safest. If the hot line earths but the generator isn't grounded, the case of the generator goes to 120V. Someone walking up and touching it would get zapped.
BTW, the generators that produce "60V" on each output aren't center grounded outputs. They have balanced leakage from each leg to ground. Very likely, the leakage to ground is dominated by capacitors connecting from each leg to ground for filtering the high frequency from the inverter (known as "Y" caps). The fact that the Kipor doesn't show this behavior means that the Y caps are very small or not even there.
stacaz822 wrote: Also, check out this waveform page.
I've seen the link before but it's worth repeating here for others. The EU2000 inverter is impressive. I wonder what the Kipor output looks like?
Offline
Online
