Direct from Solar PV to EV charging, possible?

[MontyFloyd]

Direct from Solar PV to EV charging, possible?
The benefit is avoiding conversion loss, be it DC > AC > DC or even DC >DC conversion.

I read here there is some complexities in the charge system, but IMHO it seems over analyzed (and can become quite complex with a Powerwall in equation , and no, there is no PV to EV in a PW setup. Even PW does not seem to take PV DC directly)

There are electrical devices that limit min-max volt and current, and of course a disconnect (relay), so why not have a PV > EV option?
( I can see good reasons both way, which is better?)

Thoughts?

 

[RTPEV]

Direct from Solar PV to EV charging, possible?
The benefit is avoiding conversion loss, be it DC > AC > DC or even DC >DC conversion.

I read here there is some complexities in the charge system, but IMHO it seems over analyzed (and can become quite complex with a Powerwall in equation , and no, there is no PV to EV in a PW setup. Even PW does not seem to take PV DC directly)

There are electrical devices that limit min-max volt and current, and of course a disconnect (relay), so why not have a PV > EV option?
( I can see good reasons both way, which is better?)

Thoughts?

This is not going to be possible.

The battery management system on the EV, when charging the battery (whether from AC or DC) is going to call for either constant current or constant voltage during the charge cycle. Since a solar array is inherently not going to supply a constant anything, at the very minimum you are going to be going through a DC/DC converter to provide the required output to the battery.

As for trying to eliminate the DC->AC->DC conversion, I suppose in theory this is possible (this is essentially what would happen at a V3 Supercharger with solar panels attached to the shared DC bus), the additional equipment involved would make some a scheme economically impractical compared to the relatively minor savings you would get from avoiding the conversion. It's simply not worth it.

 

[MontyFloyd]

This is not going to be possible.

The battery management system on the EV, when charging the battery (whether from AC or DC) is going to call for either constant current or constant voltage during the charge cycle. Since a solar array is inherently not going to supply a constant anything, at the very minimum you are going to be going through a DC/DC converter to provide the required output to the battery.

As I said, there are solid state devices that will permit only the acceptable volt / current values. These devices been around for decades.

As for trying to eliminate the DC->AC->DC conversion, I suppose in theory this is possible (this is essentially what would happen at a V3 Supercharger with solar panels attached to the shared DC bus), the additional equipment involved would make some a scheme economically impractical compared to the relatively minor savings you would get from avoiding the conversion. It's simply not worth it.

The underline is the major factor to consider. Still, even a 1% loss over time and/or load will add to more than the cost.


So it seems the answer to my question, such devices for large battery packs (EV or otherwise) are rare, if used at all.

 

[RTPEV]

As I said, there are solid state devices that will permit only the acceptable volt / current values. These devices been around for decades.

Yes, I'm not saying they don't exist, but the type of device that will be able to provide the voltage the car (BMS) is requesting) is the DC/DC converter you were looking to eliminate.

 

[MontyFloyd]

Yes, I'm not saying they don't exist, but the type of device that will be able to provide the voltage the car (BMS) is requesting) is the DC/DC converter you were looking to eliminate.

There is more than one way to control electricity.
At the very simplest is a voltage controlled relay, simply turn off if V is exceeded. With enough switch load capacity, loss will be practically non existent.
Alternately is a Zener diode, that will allow V up to a point.

The thing about DC-DC is they are a pseudo AC system, creating one half sine wave with a coil (actually, somewhat similar to a magneto). It is a functional but somewhat brute force way to regulate DC.

There is absolute zero reason to require a DC-DC control system, so what are the factors to make it a preferred method?
At a guess, abundance of low cost and easy to configure devices. All because it is low cost does not mean it is best.

 

[miimura]

Yes, I'm not saying they don't exist, but the type of device that will be able to provide the voltage the car (BMS) is requesting) is the DC/DC converter you were looking to eliminate.

From an electrical engineering standpoint, a system to DC charge an EV through a well known interface like CHAdeMO directly from a string of solar panels is not that difficult. The problem is that it's not worth the time and expense to engineer, manufacture and market such a device with the required safety protections. The addressable market is too limited to make it a profitable endeavor so you can market the small improvement in efficiency that it will deliver.

 

[MontyFloyd]

required safety protections

Of course! How did that slip my mind (and I am involved with such things). That would be a major and expensive item to clear

 

[Earl]

The thing about DC-DC is they are a pseudo AC system

As pointed out above: Most DC-DC converters essentially convert from DC to AC, then back to DC anyway. Also, keep in mind that solar panels have an optimum current that they can draw for a particular voltage for any solar condition. Therefore they have MPPT (Maximum Peak Power Trackers) which are essentially DC-DC converters controlled by a computer monitoring the solar output. Once this is done, it establishes a particular Voltage and Current available but that would have to be changed to the Voltage and current that the car's battery needs for charging.
There is a company, www.dcbel.energy , that is trying to develop and produce a similar device as described above. I don't know how much progress they have made though.

 

[gnuarm]

There is more than one way to control electricity.
At the very simplest is a voltage controlled relay, simply turn off if V is exceeded. With enough switch load capacity, loss will be practically non existent.
Alternately is a Zener diode, that will allow V up to a point.

Not trying to be offensive, but you just showed you do not understand electronics enough to even discuss this issue. A Zener diode is literally the last thing anyone would consider using to limit the voltage output from a PV array.
What you are describing above is a DC/DC converter. You seem to think by saying "semiconductors" you get around the losses. In a DC/DC converter much of the losses are in the semiconductors!

The thing about DC-DC is they are a pseudo AC system, creating one half sine wave with a coil (actually, somewhat similar to a magneto). It is a functional but somewhat brute force way to regulate DC.

I think you are trying to describe a switching converter. Yes, the transistors are switched on and off. The current does not switch on and off, rather it is routed in different directions. In a boost converter the current is routed to ground for a portion to build up current in the coil, then the coil is connected to the output capacitor to add charge to it. Due to the nature of an inductor the output voltage of the coil varies to suit the output voltage through the current rate of change. Switch quickly enough and the ripple in the voltage is minimal.
The losses are largely in the transistor switches. When switching they can not turn off immediately, rather they pass through a linear region during which high power is dissipated. If the switching rate is kept low the total loss is minimal, but large coils and capacitors are required.

There is absolute zero reason to require a DC-DC control system, so what are the factors to make it a preferred method?
At a guess, abundance of low cost and easy to configure devices. All because it is low cost does not mean it is best.

There is only one reason to use a DC-DC converter. That is to match the variable output voltage and current from the PV panels to the constant voltage of the battery.

I don't know anything about the details of how the BMS decides how much current can be drawn by the battery. Obviously it should not exceed the capability of the EVSE, so there is a control from the EVSE to indicate this. If this can be adjusted in real time it might be possible to let that control the current from the PV maintaining a proper voltage to the car. But I doubt it can be adjusted in real time.

So a DC/DC converter is likely required.

 

[gnuarm]

As pointed out above: Most DC-DC converters essentially convert from DC to AC, then back to DC anyway.

I don't know what you mean by that. There's no real AC power flow in a DC/DC converter. The current or voltage in any one point may be switched, but it's never really AC. Every component sees either a relatively steady voltage or current if not both.

Also, keep in mind that solar panels have an optimum current that they can draw for a particular voltage for any solar condition. Therefore they have MPPT (Maximum Peak Power Trackers) which are essentially DC-DC converters controlled by a computer monitoring the solar output. Once this is done, it establishes a particular Voltage and Current available but that would have to be changed to the Voltage and current that the car's battery needs for charging.

When the output is a constant voltage finding the MPPT is a simple matter of optimizing the output current. EVSE normally tells the car the maximum current available (at least in level 2 charging), and the car simply takes what current it wants up to that amount. I believe the DC voltage is fixed in a level 3 charger, so the power is controlled by the BMS unless it can respond in real time to a signal from the EVSE. If not, there is no point to MPPT since the power transferred to the car can't adapt.

There is a company, www.dcbel.energy , that is trying to develop and produce a similar device as described above. I don't know how much progress they have made though.

DCbel seems to be a company all about image... "so you can live a life without compromise." One of those companies that don't care if you can actually read the web page, but more concerned about their "image", gray font against a gray background. Worse than this page. They also don't say what they are selling other than the name, "dcbel r16". Some sort of magic EVSE I suppose.

I would like to have a low power PV system that can be run in the car to mitigate vampire drain. I'm not sure of the J1772 protcol details, so I'll need to dig that up again. If the EVSE current indication can be modulated in real time it's a cinch. If not, a significant battery will need to be added. bummer

 

[MontyFloyd]

Not trying to be offensive, but you just showed you do not understand electronics enough to even discuss this issue.

I do not know, that is why I posted this question.
I know enough to wonder if alternate designs are viable.

A Zener diode is literally the last thing anyone would consider using to limit the voltage output from a PV array.
What you are describing above is a DC/DC converter. You seem to think by saying "semiconductors" you get around the losses. In a DC/DC converter much of the losses are in the semiconductors!

Just tossing an idea.

I think you are trying to describe a switching converter.

Yes

There is only one reason to use a DC-DC converter. That is to match the variable output voltage and current from the PV panels to the constant voltage of the battery.

Another very good reason.

So a DC/DC converter is likely required.

Or at the very least is best option for a safe conversion of a variable input source, any loss is acceptable for benefits.
I now see the limits / flaws of other ideas I posted.

Thanks.

 

[Earl]

I don't know what you mean by that. There's no real AC power flow in a DC/DC converter. The current or voltage in any one point may be switched, but it's never really AC. Every component sees either a relatively steady voltage or current if not both.

This is true, however, as someone on this forum pointed out:

In a DC/DC converter much of the losses are in the semiconductors!

The losses come mostly from the switching from a smooth DC to a pulsed signal. Whether it inverts it to AC is only part of the process.

Ideally a solar to DC battery charger would take the DC from a PV panels and convert it to DC at the correct voltage to provide the current that the battery wants for charging.

Having an MPPT convert a PV's DC into AC at optimal power, then having an AC-DC battery charger convert it back to the right DC current and voltage for optimal for battery charging really isn't much more lossy than DC to DC conversion. Essentially the same switching has to be done.

While their website does a very poor job at describing it:

They also don't say what they are selling other than the name, "dcbel r16".

From more careful looks at today's website and from their previous Ossiaco website and emails I exchanged with their principals in the past; It is a bi-directional CHAdeMO EV charger, an household AC inverter, a J-1772 EVSE, an AC transfer switch, and a PhotoVoltaic MPPT controller, all in one box. Note that they lament that the Tesla CHAdeMO adapter does not support bi-directional use. They aren't available and haven't ever sold anything so, for now, they are vaporware.
Its not exactly:

I would like to have a low power PV system that can be run in the car to mitigate vampire drain.

but, if they ever produce it, it would be more of a high power PV system that can fully charge your car and power your home, either from the PV or from your car's battery. I thought about hanging onto my Leaf for use as a rolling Powerwall equivalent if they ever come out with the product.

 

[gnuarm]

This is true, however, as someone on this forum pointed out:

The losses come mostly from the switching from a smooth DC to a pulsed signal. Whether it inverts it to AC is only part of the process.

If I can be a bit pedantic, there is no pulsing of the DC "signal". The current is switched between two different routes. In a buck converter the output current is switched between the input or through ground. In a boost converter the input current is switched between ground and the output. It's hard to describe the voltages, but the important one, the output, is relatively constant in both configurations.

I guess my point is there is no benefit from desribing what is happening in a DC/DC converter as any sort of AC signal.

Ideally a solar to DC battery charger would take the DC from a PV panels and convert it to DC at the correct voltage to provide the current that the battery wants for charging.

Having an MPPT convert a PV's DC into AC at optimal power, then having an AC-DC battery charger convert it back to the right DC current and voltage for optimal for battery charging really isn't much more lossy than DC to DC conversion. Essentially the same switching has to be done.

There's the sticky wicket. An MPPT is going to control the IV (current/voltage) operating point of the PV panel to obtain maximum power. But then where does that power go? I don't know the details of EV charging, but I am aware that in level 2 charging the EVSE only indicates the maximum current available and the Battery Management System (BMS) draws as much as it wants up to the EVSE indicated maximum. So if the MPPT adjusts the operating point of the PV to produce more power (and therefore current) and the car does not use it, where does the power go? I expect it will be necessary to have a battery or a very large capacitor in the EVSE to buffer the fluctuations in PV supplied power to match the BMS drawn power.

This is something I'm sure has been worked out, I just don't know how they do it.

Converting to AC is pointless and is less efficient than direct DC/DC conversion. It's actually hard to produce a sinewave AC power output. Often a modified sinewave is used. But if it is not useful (which in this case it is not) it absolutely is less efficient than direct DC/DC conversion.

While their website does a very poor job at describing it:

From more careful looks at today's website and from their previous Ossiaco website and emails I exchanged with their principals in the past; It is a bi-directional CHAdeMO EV charger, an household AC inverter, a J-1772 EVSE, an AC transfer switch, and a PhotoVoltaic MPPT controller, all in one box. Note that they lament that the Tesla CHAdeMO adapter does not support bi-directional use. They aren't available and haven't ever sold anything so, for now, they are vaporware.

AC transfer switch? So they are combining this with a connection to the AC line, but I guess that's just a mechanical switch and no power is supplied to the AC line. That's why they are producing AC, so it can work off the PV panel OR the AC line using a simple swtich and all other components off the shelf. If this were designed at the board level they could use a simple DC/DC converter and add an AC input, so either AC or DC to DC. That is much more simple than the arrangement they are using.

I wonder how they get the PV power matched to the power input to the EV?

Its not exactly:

but, if they ever produce it, it would be more of a high power PV system that can fully charge your car and power your home, either from the PV or from your car's battery. I thought about hanging onto my Leaf for use as a rolling Powerwall equivalent if they ever come out with the product.

I have no interest in powering anything from my EV other than the EV. The most expensive part of the car is the battery and I want to preserve it as much as possible. I'm not going to use an expensive car to replace a $500 generator.

I would like to explore wind power which is claimed to be cheaper than PV solar. But that's at the commercial level and a small windmill seems very pricey for what you get.

 

[Earl]

PV power matched to the power input to the EV

I'm assuming that with the J-1772, they either take a lowest-common denominator current available at 240v and just cut off charging if there isn't enough PV production. Or, they may adjust the J-1772 pilot signal to roughly line up the current with the I/V curve for the PV as the sun moves across the sky (cosine curve).
With the CHAdeMO connector, they have a direct DC connection to the battery so they can truly do a DC-DC connection between the PV and the battery, using traditional MPPT to adjust the current going into the battery.

use an expensive car to replace a $500 generator

Yes, I agree. Note that I said Leaf. An old 2011 Leaf with a 20 KWh battery and 110K miles on it isn't "an expensive car". Its book value is less than $3K and it is a backup grocery-getter in addition to a battery on wheels.
Then there's the emergency power outage use: I'm ok with using my Tesla battery for a few days on the rare occasion (if ever) that I need it instead of having to keep a generator in running condition with fresh fuel on hand. I certainly wouldn't use my Tesla for peak shaving to save a few $$ per day in peak electricity costs.

 

[Mr T]

My Bluetti AC 500 directly accepts variable PV voltage to charge its BC300 battery. That is what lead me to this thread thinking, "Couldn't an EV have the same similar port? Would be way easy to string together several panels in series and plug em in (which is what I did for the Bluetti.) Is the electronics inside the AC500 that expensive and complicated? Reading above leads me understand I know oh so little about what is involved.

 

[gnuarm]

I'm assuming that with the J-1772, they either take a lowest-common denominator current available at 240v and just cut off charging if there isn't enough PV production. Or, they may adjust the J-1772 pilot signal to roughly line up the current with the I/V curve for the PV as the sun moves across the sky (cosine curve).

I've never looked into what they do. The PV max power is a continuous function. I don't recall if the EVSE current is continuously adjustable, or if it is stepped. But it has reasonable granularity if it is stepped, so the PV can be operated a bit under optimal and not much is lost.

With the CHAdeMO connector, they have a direct DC connection to the battery so they can truly do a DC-DC connection between the PV and the battery, using traditional MPPT to adjust the current going into the battery.

Not sure what you mean by "DC-DC connection". You can't just connect the PV panels directly to the battery. The DC voltage has to be adjusted by someone to match the battery. I've always assumed the idea is for the EVSE to provide a DC level that is at least as high as the highest battery voltage, and there is a buck converter in the car to match the battery voltage. Needing to do both buck and boost is more complex and uses more parts like inductors, which are large compared to the rest of the circuits. I've never met anyone who can confirm or deny this is a correct assumption.

An MPPT is designed to get the maximum power out of the PV panel. When charging a battery, since the output voltage is nearly constant, you can optimize the output current and the power is automatically optimized.

Yes, I agree. Note that I said Leaf. An old 2011 Leaf with a 20 KWh battery and 110K miles on it isn't "an expensive car". Its book value is less than $3K and it is a backup grocery-getter in addition to a battery on wheels.

Doesn't mean much to me. When you've chewed up the battery, you still have to either fork out a bunch for a new one, or find another beater and toss this one out. Not many people drive "cheap" BEVs. That's a bit of a self-contradiction, actually.

Besides, you can't do much backing up with a 20 kW battery anyway. When the air conditioning or the heat pump is running, that 20 kW will spin down pretty quickly.

Then there's the emergency power outage use: I'm ok with using my Tesla battery for a few days on the rare occasion (if ever) that I need it instead of having to keep a generator in running condition with fresh fuel on hand. I certainly wouldn't use my Tesla for peak shaving to save a few $$ per day in peak electricity costs.

To me this is like needing a gun for home protection. I don't feel a need for either. I've lived many decades without worrying about being without power for a spell. Only once, in that entire time did I need something like this. Even then, the car would have been drained flat in a couple/three days and I'd have no way out. Sounds like a very bad idea to use your means of escape for powering your home in such conditions. Better to just escape and be safe.

 

[gnuarm]

My Bluetti AC 500 directly accepts variable PV voltage to charge its BC300 battery. That is what lead me to this thread thinking, "Couldn't an EV have the same similar port? Would be way easy to string together several panels in series and plug em in (which is what I did for the Bluetti.) Is the electronics inside the AC500 that expensive and complicated? Reading above leads me understand I know oh so little about what is involved.

It would need a PV charging mode, just for this application. PV charging a battery is pretty simple. The required voltage is controlled by the MPPT to match what the battery needs. Actually, it would ignore the battery voltage and simply adjust the current for MPPT. The battery would manage the voltage automatically. The only way to change the battery voltage is to run the charging current up or down. As long as the battery is not charged too fast, there's no problem.

But the charging port on BEVs I've seen, are not designed for that. At least, that's how I understand them.

 

[Earl]

Not sure what you mean by "DC-DC connection"

I mean that with DCFC, there is no circuitry (eg. AC-DC 'charger') between the charging port and the battery. This can easily allow an appropriately designed PV MPPT to feed DC to the battery at the available power level for the PV and the appropriate voltage for the battery.

 

[RTPEV]

The DC voltage has to be adjusted by someone to match the battery.

Yes, true, although also add that when in constant current mode, it's not the voltage that is adjusted, but the current. When in this phase of charging, the voltage is high enough above the battery voltage such that the current flowing into the battery obeys Ohm's law. But essentially yes, the voltage/current is being adjusted by something.

I've always assumed the idea is for the EVSE to provide a DC level that is at least as high as the highest battery voltage.

For DC charging, yes, the EVSE provides the proper (requested) DC voltage/current. When in constant current (CC) mode, the EVSE will provide as much current as it is capable of up to the limit of what the car is requesting. When it switches over to constant voltage (CV) mode, the car requests a specific voltage a set amount above the package voltage such that current continues to flow into the pack, and the charger complies with the request, holding the output voltage at that set value.

and there is a buck converter in the car to match the battery voltage.

While there is a converter in the car (to support AC charging--see below), this is bypassed in DC charging. The car just requests a given voltage that is appropriate to it's pack's requirements, and the DC charging station complies. There is no voltage conversion happening in the car during DC charging.

Needing to do both buck and boost is more complex and uses more parts like inductors, which are large compared to the rest of the circuits. I've never met anyone who can confirm or deny this is a correct assumption.

No, this is not correct.

First of all, the car does have buck/boost converter (or at least a boost converter) on board. This is the car's onboard charger which is a scaled down version of the external DC charger, except it obviously is designed for lower power operation, meaning it is going to operate in the ~10kW regime (25A or so max current). But it definitely has the ability to boost the input voltage of 120V or 240V up to the 400V or so that the car's pack needs to charge.

But additionally, buck converters, boost converters, and buck/boost converters all require inductors, diodes, capacitors, and switches. The difference is mainly the topology of the circuit.

 

[gnuarm]

I mean that with DCFC, there is no circuitry (eg. AC-DC 'charger') between the charging port and the battery. This can easily allow an appropriately designed PV MPPT to feed DC to the battery at the available power level for the PV and the appropriate voltage for the battery.

I'm not sure what you think you are describing. There is no DC to DC connection without circuitry. The MPPT is the circuitry. But that MPPT would have to be designed for the job. The car would also have to be designed for controlling the input voltage.

I can't say I've ever reverse engineered the connection between the car and a DC charger, but in a Tesla, there are only two pins for control. I've never heard anyone say the DC voltage is controlled over those two pins. If not, then there has to be a DC/DC converter in the car, because you can't dictate the voltage to the battery. The applied DC has to match the battery voltage at the supplied current. Does anyone know the protocol on the control pins for the Tesla? The J1772 protocols are simple enough, but what about for DC fast charging?