Charging with a generator

[ajdelange]

Also, the load does not have to be resistive nor, for that matter linear. It only starts to distort the voltage from the generator if the peak, non sinusoidal, current spike is large enough to produce appreciable voltage drop across the generator source impedance.

Sponsored

 

[ajdelange]

I talked with my friend and he thought that charging from any appropriately sized generator would work but that it would not be a guaranteed sort of thing. It also could be that a fine enough waveform comes out of the generator normally but hook some weird load with a bunch of power switching to it and then it goes south.

So what your friend told you is that charging from most any generator should work but that if you screw something up in connecting it might not. IOW he didn't tell you anything that common sense wouldn't have told you anyway.

How much risk to personnel and or property did he specify for "going south"?

 

[NY_Rob]

....Now for perspective I'll note that I just installed an off grid Powerwall system which is backed up by a Kohler RES20CL (20 kW) generator which does NOT have an inverter but which DOES have brushes (yes, in 2022). I did not choose this generator. Tesla told me that this is one of two generators they accept for Powerwall backup.

Did you put your scope on it yet?
We have a quite old 20kW Kohler at my job, it was probably new when the first Space Shuttles launched. We use filtered power at work, but this generator powers the UPS/conditioner if the mains go out. We loaded it up good one day and after it hit operating temp I put a portable scope on it and it's voltage waveform under 50% load was really nicely formed, clean, no clipping, etc... not bad from a 35+ year old generator.

 

[Davethadog]

First case, even with ICE, boils down to either poor planning or pushing your luck on how far your energy (gas/diesel/kWh/whatever) will get you. I'm still gonna call you dumb as I hand you my spare gas can. ESPECIALLY if your 90k smart truck TELLS YOU HOW FAR YOU CAN GO WITH YOUR CHARGE. You definitely get the "way to go, bub!" honk.

In a power outage I have the ability to do the same, but from my propane genset on the house, not a portable generator. I'm not going to blow through all my propane to charge my truck, that's poor resource allocation. I'll run down to the EA station at Fred Meyer (Kroger) to fill up.

Let me be clear, I just think we need to think on things a little before we do them. And also lean on our friends/neighbors for help. I may call you dumb but I'm still buying you a beer ?

Guess you’ve never been dumb enough to try something fun.

 

[Zoidz]

Think you may be confusing active rectification with inversion. In active rectification you turn the switch on when you want it on and off when you want it off. In the simplest case you would not use active rectification but rather just diodes (as evidently Tesla does in this application). Whenever the voltage on L1 > L2 by more the stored voltage on the output cap diode A and D conduct. When the line reverses so L2 < L1 by more than the voltage on the output cap then B and C conduct. The duty cycle is determined by how fast C is being discharged i.e. by the load imposed on the circuit by the rest of the rectifier. It is this which primarily determines the bridge duty cycle. Looking back at all those crappy waveforms posted a few numbers back, take note of the ones with the incredible 2nd harmonic. In those if the rectifier is lightly loaded the cap voltage will be high and the diodes may turn off when the voltage takes that dip in the middle of each cresting part of the waveform, In that case instead of two nice smooth half sinewave shaped current pulses you would have 4. How does that increase dissipation in the diodes?

Now you keep talking transistors (or other active devices) here. One certainly can control the output of a rectifier by gating the input devices and input devices are often switched in 3ø systems (Vienna rectifier) for power factor (including harmonic content) control in the current drawn. But Tesla (can't speak for Rivian) evidently does not do it that way rather gating the high frequency inverter transistors. This pretty much renders the input semi conductors immune to input wave shape so I am afraid your argument remains jejune.

Uhhh, you are talking old school technology using only diodes. Modern EVs such as Rivian use MOSFET devices for the PFC and DC-DC stages to boost to the 400+ VDC necessary to charge the batteries.. In simpler terms it's a switching power supply that boosts the voltage from 120/240 VAC to traction battery charging voltage.

Here's a modern EV on-board charger reference design.

ST Microelectronics - "The reference design embeds two sections: an interleaved totem pole PFC with SiC and a dual galvanic isolated full bridge LLC DC-DC ZVS resonant converter, based on MDmesh DM6 super-junction power MOSFETs."

This is the type of charger I have been referring to all along when mentioning active devices and duty cycle.

I'll summarize my initial thought behind my posts - We simply do not know what Rivian implemented in the filtering/PFC stage, and therefore we don't know what damage dirty power could cause. Been there, seen it in Allen Bradley VFDs. Not happening to my Rivian.

You're taking what I intended as a simple conceptual example about not risking dirty generator power possibly damaging a Rivian on board charger, and repeatedly morphing down a rabbit hole.

As they say on Shark Tank - "For that reason, I'm out."

Cheers!

 

[ajdelange]

Did you put your scope on it yet?

No and when I went out there to put a THD meter on it this morning I found that the Kohler tech had locked it and I can't find my key so I'll have to waith until new keys get here before I can. This guy, BTW was really, really good. When the machine was intalled it ran three minutes and quit because they get tach from an auxilliary stator winding and that circuit opened. I, of course, feared a bad stator but asked the installer to check the wiring too. They didn't do that. So I got another company out (a Kohler dealer) and they spent a day coming to the same conclusion but when I asked them to check the cabling they didn't either. Finally I get a Kohler tech and he did check the wiring and, voila. The cable was routed such that it touched the flywheel.

Anyway I fully expect THD on this machine to be less than 3%. That's probably too much for our nervous nellies and those who apparently don't understand how car chargers work but it would be fine as far as I am concerned (not that this maching would charge a car - it would only charge the Powerwalls if we get a long string of sunless days.

 

[ajdelange]

Uhhh, you are talking old school technology using only diodes.

Uhh no, I have been talking about the Tesla power supply architecture because that has been revealed. I don't know how close or distant to that architecture Rivian's may be.

Modern EVs such as Rivian use MOSFET devices for the PFC and DC-DC stages to boost to the 400+ VDC necessary to charge the batteries..

I consider Tesla modern and they don't do that (apparently). In any case your argument seems to be that if a semiconductor passes X joules as half a sine wave it dissipates less power than if it passes it as two pulses of higher frequency. I've asked you to explain the mechanism for this and you can't and the reason is that it is BS (other than the tiny amount of extra dissipation from the extra switching transients).

In simpler terms it's a switching power supply that boosts the voltage from 120/240 VAC to traction battery charging voltage.

In simple terms a Tesla charger is a filter, a rectifier, an inverter, a transformer and another rectifier. No DC/DC converter that I can see.

Here's a modern EV on-board charger reference design.

ST Microelectronics - "The reference design embeds two sections: an interleaved totem pole PFC with SiC and a dual galvanic isolated full bridge LLC DC-DC ZVS resonant converter, based on MDmesh DM6 super-junction power MOSFETs."

Well that's nice. But you know, I am having trouble finding the THD limitation on this device. Common sense, who is evidently not your friend, should tell you that if it isn't speced it isn't an issue.

This is the type of charger I have been referring to all along when mentioning active devices and duty cycle.

Still waiting for something other than hand waving with regard to the mechanism by which duty cycle effects efficiency (joules dissipated per joule transferred).

I'll summarize my initial thought behind my posts - We simply do not know what Rivian implemented in the filtering/PFC stage, and therefore we don't know what damage dirty power could cause. Been there, seen it in Allen Bradley VFDs. Not happening to my Rivian.

Well we are not talking about the VFD's in your truck. We are talking about the battery charger which has a dramatically different architecture (although they both contain inverters but no vector field control in the charger). No - it isn't going to happen in your Rivian if you never charge with a generator not that I am suggesting you should. I was really hoping to get you to a common sense understanding of how these things work.

You're taking what I intended as a simple conceptual example about not risking dirty generator power possibly damaging a Rivian on board charger, and repeatedly morphing down a rabbit hole.

There is a possibility of damage to your Rivian every time you take it out of the garage.

 

[ajdelange]

For the very, very risk averse here I feel I would be remiss if I did not mention that there is indeed a hazard associate with charging a BEV from a generator but it is a hazard to people rather than the vehicle. As has been mentioned EVSE will not pass power to the vehicle unless it measured 120V between the earth wire and L1. That is becaue in a normal building installation earth and neutral are bonded at the service entrance so that demonstration of 120V between the earth wire and one of the phases insures that the vehicle frame is earthed. When you install that little grounding plug you fool the EVSE into thinking the truck is grounded but it is not unless you ground it. This is done by driving a ground rod adjacent to the generator and wiring it to the generator frame. Interestingly enough in one of the many YouTube offerings on "Can you charge you XXX from a cheap generator" the ground rod is visible but not mentioned nor is it mentioned in any other that I know of. Funny that no one who has tried this has reported any damage - but you never know).

Now I have been laughed at for bringing this up elsewhere by people of opposite disposition to the Chicken Littles here. They argue that there is no need to protect against events whose probability is so low (thunder storm, St Elmo's fire...) whereas the latter group argues that you must not subject yourself to the risk of low probability events. You will have to make up your own minds.

 

[NY_Rob]

I fully expect THD on this machine to be less than 3%. That's probably too much for our nervous nellies and those who apparently don't understand how car chargers work but it would be fine as far as I am concerned ..

< 3% THD should be absolutely fine for charging any EV, anything <5% THD should be fine IMO.

When researching portable generators in the 8-10kW range I saw some with as much as 25% THD.. and that was a Westinghouse product of all companies. 10%+ THD seemed very common- but a no-go for sensitive electronics, and a very few portable non-inverter generators in the <5% THD range, but there are some:
https://www.northerntool.com/shop/tools/product_200686833_200686833


And a reasonably priced open frame inverter generator with <3%THD:
https://generatorbible.com/generators/champion/_201067/

 

[lostpacket]

Question from someone who thinks sinusoidal is a nasal congestion medicine:

How would one properly ground the truck instead of using a neutral ground bonding adapter plug?

 

[ajdelange]

You should use the plug and ground it as explained in No. 53.

 

[zipzag]

You should use the plug and ground it as explained in No. 53.

For portability I use a T handle 3' ground rod intended for an electric fence. I'm hoping its good enough for clay soils.

At home is connecting the generator ground to the ground of a 20A receptacle acceptable?(assuming my house is properly wired and grounded)

 

[ajdelange]

At home is connecting the generator ground to the ground of a 20A receptacle acceptable?(assuming my house is properly wired and grounded)

Yes, that should work.

 

[ajdelange]

< 3% THD should be absolutely fine for charging any EV, anything <5% THD should be fine IMO.

I found this

Champion portable generators will output an industry standard total harmonic distortion (THD) rating of about 12%-20% depending on load applied. They will produce a sine wave, not a modified or square wave. This is perfectly acceptable for running common commodities found in your home such as TVs, computers, your appliances, etc.

Champion sells to Home Depot (and elsewhere). When I needed a cheapie generator for a quick test I got one of their units and sure enough the THD is a bad as they say (never observed above 15%). Nonetheless it ran Mitsubishi heat pumps, "sensitive" electronics and charged my Tesla without problems (for short periods).

Champion further advises

Please consult your specific appliance manufacturer to determine if the appliance requires low THD (less than 5%) to operate properly.

and the bolding is theirs. I don't think I've ever seen a piece of equipment with a power source THD limit caveat on it. Has anyone else? Certainly not my Teslas or in the Rivian manual.
?
Again I can't help wondering if the assumption that sources (generators) need to be so clean hasn't drifted over from the fact that sinks need to be so clean. Even linear power supplies, for example, pull current in pulses resulting in high harmonics in the drawn current which has real, and actually damaging, consequences (negataive and zero sequence currents) in the supply system. Hence PFC in even the simplest of loads these days. We demonstrated in an earlier post how good the PFC is for the Tesla charger.

I guess I should finalize by saying that if I wanted a generator for use in home backup, camping ... I would not buy a Champion. The fact that the generator has such a lousy THD spec says it is poorly built otherwise too.

 

[ajdelange]

OK. I think I'm finally getting this. You MUST use a pure sine inverter generator because the "delicate" semiconductors in the rectifier part of the Rivian/Tesla chargers cannot withstand any harmonic content without being damaged. The only part I'm still a little confused about is that somehow the semiconductors in the rectifier part of the generator inverter circuit are not so effected.

Sponsored