Charging from a generator, 2 red blinks

[TORQU3]

It charges ok on mine gen set, it's kind of a beast though. Frequency and voltage are very steady under load. It's run through the house with a 100A reverse service plug and a 200A whole house transfer switch so it has proper grounding and neutral bonding at the main panel.

 

[SSedan]

Being in a place that sees -15f or worse every year and 100% electric I should consider whole house backup. Hopefully I don't jinx myself but sofar in 8 years the meaningful outages have been summer.

 

[KenC]

I have a 12k wholehouse standby generator, 5% THD, that runs on LP, and one time just after I had returned home and plugged in my Chevy Volt, the power went out. I forgot that my Volt was charging so I didn't unplug it. It charged up fine, but was a complete waste of LP since the Volt has a far more efficient generator on-board, already.

I imagine the Model 3 would be able to charge on a wholehouse standby generator, but unless you have an LP or NG hookup that's relatively cheap, it can become quite expensive and use up an LP tank quite quickly.

 

[TORQU3]

Mine is diesel and while it's a lot more cost efficient than a propane unit, it's still much more expensive than grid power. I would only have it charge my car if needed of when I want to run the gen set overnight but I need a base load to stop the gen from wet stacking. In that case I would set it to a 10A charge or so.

 

[ZOMGVTEK]

Charging off an EU7000iS with gasoline is about 24-26 MPG. Not great, but not terrible.

 

[darth_vad3r]

I am not sure a non-inverter generator would not work just fine.

Huh? Almost a triple-negative with the “non-“ too

I had to read that 4 times. I think you are saying you think a non-inverter generator may work just fine?

 

[darth_vad3r]

I think the concept of electronics needing good power supplies more applies to things that don't have batteries. Since the rectifiers are just providing power to batteries in the car it is not really the end of the world if the power is not perfect. I think typically the issue is if you don't provide consistent DC power to things like computer chips they freak out, but in charging a car (which also has many microcontrollers running) those chips are buffered by the battery.

Do we actually know this to be the case? Can the car route “shore power” directly to accessories? (well not directly directly, but after AC to DC conversion.

e.g. If you have 6 kW AC input power and then turn ~1 kW of accessories on (AC + high fan?) is the car (a) sending all 6 kW to the battery and taking 1 kW out of the battery, or (b) sending only 5 kW to the battery and 1 kW to accessories?

The charge display certainly drops to show the charge speed in the lower mi/hr rate anyways ... this itself doesn’t prove anything, but it made me think only X% of the AC was being added to the battery, and the rest was diverted to accessories (I didn’t think it meant the net add amount after all power went to the battery and then some came out for accessories, but it certainly could mean that I suppose).

EDIT:
I re-posted this as its own thread with a poll.
Does “shore power” go through the battery?

This topic came up in the battery degradation scientifically explained thread as well.

Perhaps we can gather discussion about this in the new thread.

 

[ZOMGVTEK]

You absolutely don't need an inverter generator. Making good AC from DC using electronics as it turns out is quite difficult. Not all inverters are created equal. The difference between one 'pure sine wave' inverter and another can be vast. How they respond to non linear loads can be very different. The THD ratings can be very misleading and the real THD can be massive in real service. Especially if you backfeed a house with 50 small SMPS devices without PFC all running unloaded or lightly loaded. A cheap 'sine wave' inverter most likely makes worse power quality than a cheap traditional generator. The real benefit to inverter generators for the average person is they tend to allow for varying engine speed while maintaining frequency, so they can be quieter and save on fuel.

The power coming into your house is generated by a generator much the same as the ones in cheap generators. However, they're very large, so you mitigate a great deal of the issues with that design. With a traditional generator, so long as you way oversize it, it should produce fairly good quality power. But the voltage regulation of traditional generators also has huge vairance. Some have HUGE swings as loads turn on and off. With a soft ramp, this shouldn't matter much though. But if you want to backfeed your house with a small cheap generator, it can be risky if the AVR is bad, which it probably is. A sensible rule of thumb is to never exceed half the rated power of the generator with a single load. But even then you can have significant issues in some situations with large inductive loads like A/C compressors.

 

[darth_vad3r]

Do we actually know this to be the case? Can the car route “shore power” directly to accessories? (well not directly directly, but after AC to DC conversion.

e.g. If you have 6 kW AC input power and then turn ~1 kW of accessories on (AC + high fan?) is the car (a) sending all 6 kW to the battery and taking 1 kW out of the battery, or (b) sending only 5 kW to the battery and 1 kW to accessories?

The charge display certainly drops to show the charge speed in the lower mi/hr rate anyways ... this itself doesn’t prove anything, but it made me think only X% of the AC was being added to the battery, and the rest was diverted to accessories (I didn’t think it meant the net add amount after all power went to the battery and then some came out for accessories, but it certainly could mean that I suppose).

I re-posted this as its own thread with a poll.
Does “shore power” go through the battery?

This topic came up in the battery degradation scientifically explained thread as well.

 

[ZOMGVTEK]

Do we actually know this to be the case? Can the car route “shore power” directly to accessories? (well not directly directly, but after AC to DC conversion.

e.g. If you have 6 kW AC input power and then turn ~1 kW of accessories on (AC + high fan?) is the car (a) sending all 6 kW to the battery and taking 1 kW out of the battery, or (b) sending only 5 kW to the battery and 1 kW to accessories?

The charge display certainly drops to show the charge speed in the lower mi/hr rate anyways ... this itself doesn’t prove anything, but it made me think only X% of the AC was being added to the battery, and the rest was diverted to accessories (I didn’t think it meant the net add amount after all power went to the battery and then some came out for accessories, but it certainly could mean that I suppose).

This is a tricky question. Lithium batteries are quite happy charging off appalling power quality. You can rectify AC and dump it into the pack without filtering, so long as the peaks are within limits. The battery WILL filter it to some extent, but relying on a low enough ESR pack to ensure no issues with other electronics inst terribly sensible in production long term. It's quite complex and depends on what else is on the battery.

The charger always dumps 100% of its output into the battery. It has no mechanic or intelligence to route power anywhere. Now, the power will preferentially go effectively directly into a load as that would be a lower resistance path than intercalation. The pack has some non zero internal resistance, so a voltage rise is required to force current into it. This means the chargers output leads will be higher voltage than the battery while it is charging, making the battery practically invisible for most low frequency light loads.

The SMPS that charges the battery typically won't have much of an objection to high THD. It may have some impact to efficiency, and as a result very high THD on the input can lead to significant impact to heating in some cases. But within sensible limits, bad incoming power inst very important. It's likely the PCS is protected against horrific power quality, and it absolutely has thermal protection. So while it is technically possible to design a charger that works in remarkably bad conditions with absurd swings, Tesla would be the one to know how their design performs and what the software allows. I wouldn't really say the SMPS is buffered by the battery though, thats kinda true, but misleading. You really don't need to be overly concerned with power quality though, just as long as voltage is within range. Even if its for 20ms or something, high voltage might be an issue depending on how high, and how much load is on the charger at the time. So you still need to be careful running a generator near its limit. If the charger stops suddenly, voltage can spike massively. I suspect in practice the risk of damage in this case is low, but non zero.

 

[Evoforce]

Thinking about wading into this and wondering the best 240 V unit to buy...

 

[SSedan]

What kind of hardware is needed to check output power quality?

Far as what a battery can charge on, with something like a Tesla what is OK for the pack is basically immaterial, what Tesla allows is what matters. Might be no good reason for it but if Tesla programmed the car to stop charging due to dirty power then dirty power is a problem.

 

[miimura]

What kind of hardware is needed to check output power quality?

Far as what a battery can charge on, with something like a Tesla what is OK for the pack is basically immaterial, what Tesla allows is what matters. Might be no good reason for it but if Tesla programmed the car to stop charging due to dirty power then dirty power is a problem.

I was just writing basically the same thing.

The question about whether you need an inverter generator or not is dependent on Tesla's judgement of what is OK or not. In the early days, people said that Model S and Roadster were very picky and if the frequency was changing more than 1/2Hz or so, it would stop. Same goes for voltage. It is well known that Tesla is using voltage sag as a proxy for dangerous wiring and will at least cut down the amperage if it is varying or sags too much under load.

 

[darth_vad3r]

This is a tricky question. Lithium batteries are quite happy charging off appalling power quality. You can rectify AC and dump it into the pack without filtering, so long as the peaks are within limits. The battery WILL filter it to some extent, but relying on a low enough ESR pack to ensure no issues with other electronics inst terribly sensible in production long term. It's quite complex and depends on what else is on the battery.

The charger always dumps 100% of its output into the battery. It has no mechanic or intelligence to route power anywhere. Now, the power will preferentially go effectively directly into a load as that would be a lower resistance path than intercalation. The pack has some non zero internal resistance, so a voltage rise is required to force current into it. This means the chargers output leads will be higher voltage than the battery while it is charging, making the battery practically invisible for most low frequency light loads.

The SMPS that charges the battery typically won't have much of an objection to high THD. It may have some impact to efficiency, and as a result very high THD on the input can lead to significant impact to heating in some cases. But within sensible limits, bad incoming power inst very important. It's likely the PCS is protected against horrific power quality, and it absolutely has thermal protection. So while it is technically possible to design a charger that works in remarkably bad conditions with absurd swings, Tesla would be the one to know how their design performs and what the software allows. I wouldn't really say the SMPS is buffered by the battery though, thats kinda true, but misleading. You really don't need to be overly concerned with power quality though, just as long as voltage is within range. Even if its for 20ms or something, high voltage might be an issue depending on how high, and how much load is on the charger at the time. So you still need to be careful running a generator near its limit. If the charger stops suddenly, voltage can spike massively. I suspect in practice the risk of damage in this case is low, but non zero.

Thanks, could I interest you in re-posting and voting in the poll?
Does “shore power” go through the battery?

 

[eprosenx]

My main goal is backup home power, fridge/freezer cycling, maybe microwave so the kids can make some food themselves.
This should be quiet, which will be welcome. After the damage Saturday we had to listen to a neighborhood full of loud generators, we we're lucky at just a 34hour outage, others had it worse.
Maybe every other year we have an outage long enough to risk food freshness.

Maybe this particular unit proves a poor choice, I will admit it if it does.

FWIW, I totally was not suggesting it is the wrong unit for your use case. Your primary use case sounds like home backup, and yeah, I don’t want my electronics damaged by a poor quality generator. That is why I paid the stupid money for the Honda inverter generator for my house.

 

[ZOMGVTEK]

If you ask me, the 'best' unit would be a Honda EU7000iS, or a whole home backup unit that's at least 20kW, and don't charge above say 6-8kW. Over sizing 2x is kinda a minimum, 3x is much more comfortable. About the only sensible way to measure power quality is an oscilloscope. Some very high end meters are basically cheapie oscilloscopes and will tell you THD with reasonable accuracy as well, but they're not super cheap. And super cheap test gear tends to not inspire much confidence, especially if what you're looking for is high frequency switching noise from an inverter.

What Tesla allows is very much what matters, but they're not going to say, and in practice you can see brief swings outside of range very frequently that may or may not be a problem. I suspect about the only way to check is to just plug it in and try. I'd say its exceedingly likely they were very well aware people would try to charge off generators. I'm assuming the design is excellent and it should be modestly tolerant, but you don't know until you know.

 

[ewoodrick]

You should be able to make this work. I actually don't understand why Tesla says not to run it off a generator (probably just to avoid support calls). The Tesla is just a rectifier and it ramps into the load slowly (unlike an electric motor) and so it should be pretty soft on the generator. I would the think the Tesla should be pretty tolerant to crappy voltage/frequency as it is just converting it to DC anyway...
.

The Tesla ISN'T just a rectifier. It's boost switching supply. It's got to take the 120/240 VAC to battery voltage.

It's also closely watching the supply voltage and current. If it sees things starting to deviate, it will start to crank the current down or abort the charging altogether.

The Tesla's charging control rules are close to that which your standard household or business UPS uses. Did you realize that most UPSs refuse to work on a generator? Too much voltage and/or fluctuations and the UPS just says no.

Again it's not "just converting to DC anyway"

 

[ZOMGVTEK]

 

[eprosenx]

Nice!

Running that in a shed is a horrible idea carbon monoxide wise...

 

[ZOMGVTEK]

The exhaust passes through the wall, the shed is ventilated, and I wouldn't spend more than a minute or two refueling. The exhaust has a distinct odor, so its quite likely it would be noticeable if there was a leak. It would have to be a substantial leak and at high load for there to be appreciable CO in there. Having an atmospherically vented gas water heater in the house is way more dangerous if you ask me.

But yeah, probably a good idea to not spend much time in there with the generator running.