DC to DC Converters / Optimizers

[Grouch1980]

Hey guys!
Back with some other questions.

1. Are DC to DC Converters (located at each PV module) another way of saying Optimizers? Are they the same device?

2. Can DC to DC Converters decrease the total voltage in a series string of PV modules, in order to connect to the input voltage rating of inverter? Let's say you have 15 PV modules, each at 50 Voc. So the total Voc of the string is 750 Volts. (leaving out the temperature correction factor). Can the DC to DC Converter, located at each PV module, clip the total string voltage so that it falls below a 600 volt input rated inverter? If so, how does one go about calculating or choosing the Converter... is this something that's done by the Converter vendor, or are there any formulas / graphs one can use to calculate this?

3. What's a more typical installation being installed out there? Strings with DC to DC Converters? Or strings without them, and having to use combiners?

Thanks again!

 

[jaggedben]

Answers to 1 and 2 are yes and yes. Optimizers can also increase the string voltage.

Plain old strings don't meet rapid shutdown requirements after the 2014 NEC. So unless the jurisdiction lags behind on code adoption, you won't see plain old strings anymore on building rooftops, only optimizers or micros. You could still see strings on groundmounts and carports and such. Utility scale as far as I know never uses optimizers because of the cost, preferring string level monitoring and/or scheduled human maintenance checks.

 

[Carultch]

1. Yes. These two terms are interchangeable, when talking about a module-level device. DC-to-DC converter is the more general term the NEC uses. Optimizer is the term manufacturers call them, to highlight the benefit. What is "optimized" is that these products enable the modules to operate independently of all of their neighbors, so that shading and mismatch losses don't hinder the performance of other modules in the string. The MPPT tracking happens at the level of the individual optimizer, rather than in the inverter.

2. It depends on the specifics of the optimizer, as not all brands necessarily work the same. An optimizer works like the DC equivalent of a transformer, that can dynamically adjust its turns ratio. The product of Iin*Vin is roughly equal to Iout*Vout, just like it is for ordinary AC transformers. The optimizer allows Vin to track the MPPT of the module, and determines Iout and Vout as is needed to "cooperate" with the other optimizers in the circuit.

For SolarEdge, the way it works is the inverter sets the total voltage, and the optimizers work with each other to achieve it. The optimizers solve an algebra problem, so that output current is the same in each optimizer so they can combine in series, and output voltage adds up to the master value set by the inverter. The output voltage of each optimizer, will be proportional to the power available to each one For their 208V and single phase 240V inverters, a typical value of operating voltage is 400V with a maximum of 600V. For their 480V inverters, a typical operating voltage is 850V, with a maximum of 1000V. SolarEdge optimizers are a proprietary product for use only with SolarEdge inverters, and vice versa. You cannot mix either their inverters or their optimizers, with any other brand.

I'm not aware of whether or not you can control this to decide that you want the voltage to not exceed 600V on the 480Vac units. The inverter decides this from its factory settings. This technology does permit you to generally use longer strings than you otherwise would use with connecting uncontrolled modules in series.

For Tigo, they have the "-O" models of MLPE units and the "-L" models as well that are optimizers. The "-F", "-M" and "-S" models are MLPE devices, but are not optimizers, as they don't control the mix of voltage and current. "-F" and "-S" units participate in rapid shutdown, but not optimization. The "-O" units still call for string sizing the same as with ordinary module strings. The "-L" units allow for long strings. Unlike Solaregde, Tigo is meant for using just about any other brand of inverter.

3. Standard stings are no longer allowed on buildings anymore, unless you are still on 2014 or earlier. 2017 and later effectively requires module level shutdown. It doesn't say it that way directly, but reading between the lines, it specifies module-level rapid shutdown. In theory, you could cleverly get around this rule with sub-30V strings, or with array level shutdown on sub-80V strings, but in practice, it will mean module-level power electronics in some form or another, to comply with rapid shutdown. Could be micronverters, could be optimizers, could be MLPE devices that only perform the shutdown purpose.

It is only non-buildings where you can get away without rapid shutdown, and use standard strings without it.

The use of combiners has nothing to do with this issue, unless you are talking about contactor/AFCI combiners specifically, that once were the method of meeting rapid shutdown when 2014 applied.

 

[Grouch1980]

Answers to 1 and 2 are yes and yes. Optimizers can also increase the string voltage.

Plain old strings don't meet rapid shutdown requirements after the 2014 NEC. So unless the jurisdiction lags behind on code adoption, you won't see plain old strings anymore on building rooftops, only optimizers or micros. You could still see strings on groundmounts and carports and such. Utility scale as far as I know never uses optimizers because of the cost, preferring string level monitoring and/or scheduled human maintenance checks.

Great, thanks! Regarding calculations though... calculating the total string voltage / separating things out with combiners you follow the math and formulas, and you size the wires and OCPD devices appropriately. How do you know though by how much an optimizer clips the total voltage? or increases it? How would I know how many PV modules can be connected in series?

 

[Carultch]

Great, thanks! Regarding calculations though... calculating the total string voltage / separating things out with combiners you follow the math and formulas, and you size the wires and OCPD devices appropriately. How do you know though by how much an optimizer clips the total voltage? or increases it? How would I know how many PV modules can be connected in series?

The answer is manufacturer specific. Read the manufacturer's installation guides and other documentation to determine the calculations you need to do.

The short answer for SolarEdge is you multiply the nominal operating voltage of the inverter, with the maximum operating current of the optimizers. This produces the maximum amount of power at STC, that you can connect to the string. Voltage is fixed by the inverter, and more power on a string means more current, rather than more voltage. As an example, suppose you have 400V operating voltage, and 15A P400 optimizers with 350W modules. 400V*15A = 6000W. 6000W/350W = 17 optimizers max. There are other factors that come in to play, so this answer isn't complete for all examples.

 

[wwhitney]

For SolarEdge, the way it works is the inverter sets the total voltage, and the optimizers work with each other to achieve it. The optimizers solve an algebra problem, so that output current is the same in each optimizer so they can combine in series, and output voltage adds up to the master value set by the inverter.

Do you know the details of how the algebra problem is solved? I.e. centrally (each optimizer tells the inverter what power it has available, and the inverter then tells the optimizers how much current to provide) or in some distributed fashion?

Also, in the Solaredge architecture, is some of the functionality of a normal string inverter offloaded to the optimizers? I.e. theoretically the inverter itself only has to deal with a more limited range of DC input voltages, so one of its stages might be smaller/less capable than a typical string inverter, due to the DC-DC capabilities of the optimizers.

Cheers, Wayne

 

[Grouch1980]

SolarEdge optimizers are a proprietary product for use only with SolarEdge inverters, and vice versa. You cannot mix either their inverters or their optimizers, with any other brand.

Interesting. Is this true for most brands, where the optimizers and inverter should be the same manufacturer? it would make sense, since they have to talk to each other.

3. Standard stings are no longer allowed on buildings anymore, unless you are still on 2014 or earlier. 2017 and later effectively requires module level shutdown. It doesn't say it that way directly, but reading between the lines, it specifies module-level rapid shutdown. In theory, you could cleverly get around this rule with sub-30V strings, or with array level shutdown on sub-80V strings, but in practice, it will mean module-level power electronics in some form or another, to comply with rapid shutdown. Could be micronverters, could be optimizers, could be MLPE devices that only perform the shutdown purpose.

In NYC where I work, we're still in the 2008 code. It looks like Optimizers won't be mandatory for awhile. I came across the Optimizers in someone's design for a building here, which brought about my questions.

 

[Carultch]

Interesting. Is this true for most brands, where the optimizers and inverter should be the same manufacturer? it would make sense, since they have to talk to each other.

It's difficult to use the adjective "most", because I only know of two brands of module-level optimizers in the first place, and the answer is exactly the opposite for both of them. SolarEdge requires a proprietary combination of products, while Tigo doesn't even make inverters, and makes their product for use with conventional inverters.

You don't really get a different string sizing calculation with Tigo "-O" units, than you get with ordinary strings. The advantage to using them, over the simpler "-F" and "-S" Tigo units, is that they allow for selective deployment to stop mismatch losses from propagating. Like shading or different orientations, or a mix of different modules. Another advantage of Tigo optimizers is (also applies for SolarEdge) that they allow for non-uniform strings, which is helpful to have when your desired module total count is a prime number.

In NYC where I work, we're still in the 2008 code. It looks like Optimizers won't be mandatory for awhile. I came across the Optimizers in someone's design for a building here, which brought about my questions.

This gets in to a tricky situation with your AHJ, because the technology has changed a lot, since 2008. In 2008, 600V was the industry norm for the maximum system voltage, regardless of location. It was also the industry norm for inverters to have an isolation transformer built-in, with one polarity grounded through a GFCI device. A negative-grounded inverter would produce an AC waveform with only positive voltage, and then use a transformer to remove the DC offset, in order for its output waveform to be symmetric about ground and compatible with the grid.

Now, both of these design factors are no longer industry norms, and it is hard to find equipment that is still built that way. 1000V is the industry norm for DC voltage on anything that on a building with a 277/480V grid. Most inverters are also transformerless, with the two polarities both ungrounded, at equal and opposite voltages to ground. The inverter does this, so it can directly produce an AC waveform that is symmetric about ground and compatible with the grid. Many of these factors weren't accounted for in the 2008 NEC. Your AHJ will have discretion over what scope of later editions of the code will have to apply, in order to use equipment that relies on provisions in a later edition of the NEC.

2017 calls both of these types of systems "functionally grounded". I like the terminology for the first kind, but I dislike that it also applies to a completely different kind of system (the second kind) as well. I'd recommend the terms "GFCI grounded" and "Symmetry Grounded".

 

[ggunn]

Great, thanks! Regarding calculations though... calculating the total string voltage / separating things out with combiners you follow the math and formulas, and you size the wires and OCPD devices appropriately. How do you know though by how much an optimizer clips the total voltage? or increases it? How would I know how many PV modules can be connected in series?

With SolarEdge inverters and optimizers the optimizers trade voltage for current to limit the current to 15A and keep the string voltage at the operating limit of the inverter. They don't clip the voltage at the module; they exchange it for current.

The string length is variable within a range set by the optimizer and the inverter it is connected to; there is a table on the optimizer data sheet that will provide that information.

 

[ggunn]

It's difficult to use the adjective "most", because I only know of two brands of module-level optimizers in the first place, and the answer is exactly the opposite for both of them. SolarEdge requires a proprietary combination of products, while Tigo doesn't even make inverters, and makes their product for use with conventional inverters.

All Tigo boxes and SunSpec equipped inverters are not created equal. For simple rapid shutdown functionality SMA inverters will communicate only with the TS4-R-F, which is the only Tigo device that Solectria and Chint are not certified to work with. They will work with the TS4-F and TS4-A-F. Before you match Tigo boxes with inverters be sure to verify that they will play nice with one another.

 

[Carultch]

All Tigo boxes and SunSpec equipped inverters are not created equal. For simple rapid shutdown functionality SMA inverters will communicate only with the TS4-R-F, which is the only Tigo device that Solectria and Chint are not certified to work with. They will work with the TS4-F and TS4-A-F. Before you match Tigo boxes with inverters be sure to verify that they will play nice with one another.

Some inverters have the power-line-carrier functionality built-in, some inverters don't. For the ones that don't, you need the RSS transmitter to provide that signal, if you plan on using the Tigo "-F" products. It works like a current transformer in reverse, where the energy source comes from the leads of the "CT", and induces an AC waveform overlaid on the DC power line to carry the signal. This is specific to the "-F" Tigo devices, where the F stands for fire safety, and its only function is rapid shutdown.

All the other Tigo devices (-S, -O, and -L) don't depend on the inverter, or the DC power line, to carry the "keep alive" signal. Instead, they use radio waves to carry the signal between devices, and the TAP and CCA system as the source/receiver of those waves. The "-S" Tigo devices are therefore inverter agnostic, since the monitoring signals and "keep alive" signals are independent of the inverter. The "-O" and "-L" devices need to be confirmed to be compatible with the inverter, so that the optimization algorithms are verified to be compatible, with the inverter's own MPPT algorithm.

 

[synchro]

For SolarEdge, the way it works is the inverter sets the total voltage, and the optimizers work with each other to achieve it. The optimizers solve an algebra problem, so that output current is the same in each optimizer so they can combine in series, and output voltage adds up to the master value set by the inverter.

Do you know the details of how the algebra problem is solved? I.e. centrally (each optimizer tells the inverter what power it has available, and the inverter then tells the optimizers how much current to provide) or in some distributed fashion?

The following patent from Solaredge describes an approach where the optimizers adjust their operation on a distributed basis. This is one of the earlier patents of their 174 issued patents. And so it may or may not reflect what they have in their products.

https://pdfpiw.uspto.gov/.piw?PageN...8,587,151.PN.&OS=pn/8,587,151&RS=PN/8,587,151

 

[ggunn]

Some inverters have the power-line-carrier functionality built-in, some inverters don't. For the ones that don't, you need the RSS transmitter to provide that signal, if you plan on using the Tigo "-F" products. It works like a current transformer in reverse, where the energy source comes from the leads of the "CT", and induces an AC waveform overlaid on the DC power line to carry the signal. This is specific to the "-F" Tigo devices, where the F stands for fire safety, and its only function is rapid shutdown.

All the other Tigo devices (-S, -O, and -L) don't depend on the inverter, or the DC power line, to carry the "keep alive" signal. Instead, they use radio waves to carry the signal between devices, and the TAP and CCA system as the source/receiver of those waves. The "-S" Tigo devices are therefore inverter agnostic, since the monitoring signals and "keep alive" signals are independent of the inverter. The "-O" and "-L" devices need to be confirmed to be compatible with the inverter, so that the optimization algorithms are verified to be compatible, with the inverter's own MPPT algorithm.

Be that as it may, among the inverters that do have the powerline transmitter, for the F type devices SMA will only talk to the TS4-R-F, and Solectria and Chint will only talk to the TS4-A-F and the TS4-F. I have spoken with tech support for all three inverter companies.

 

[Grouch1980]

The short answer for SolarEdge is you multiply the nominal operating voltage of the inverter, with the maximum operating current of the optimizers. This produces the maximum amount of power at STC, that you can connect to the string. Voltage is fixed by the inverter, and more power on a string means more current, rather than more voltage. As an example, suppose you have 400V operating voltage, and 15A P400 optimizers with 350W modules. 400V*15A = 6000W. 6000W/350W = 17 optimizers max. There are other factors that come in to play, so this answer isn't complete for all examples.

Basic question: it's one optimizer per panel correct? so where you say 17 optimizers max... that's 17 optimizers connected to 17 solar modules correct?

Regular question: So each optimizer is generating 15 amps DC... How do they work in conjunction with the solar modules? Each solar module has, for example, an Isc of 5 amps (just throwing a number). Is the module current increased to the 15 amps, or does the module current stay at the Isc value of 5 amps? and it's only the optimizer producing the 15 amps? I'm assuming the module current stays at 5 amps, since thats its max rating.

 

[Grouch1980]

The string length is variable within a range set by the optimizer and the inverter it is connected to; there is a table on the optimizer data sheet that will provide that information.

Thanks! I came across a table i think for the P400, that Carultch mentioned. I'll go through it.

 

[Carultch]

Be that as it may, among the inverters that do have the powerline transmitter, for the F type devices SMA will only talk to the TS4-R-F, and Solectria and Chint will only talk to the TS4-A-F and the TS4-F. I have spoken with tech support for all three inverter companies.

They may say that because that is the only mix of products they tested, and that the inverter manufacturer is either ignorant of the essential difference, or can't make a statement outside the scope of their testing for policy reasons. However, I don't agree with the inference that this means there is an incompatibility for other combinations.

The only difference between TS4-F, TS4-R-R, and TS4-A-F, is the mounting interface, that has nothing to do with inverter compatibility. The rest of the product is identical, and has exactly the same design. Tigo addresses this in their FAQ, that the only difference is the mounting interface, and product compatibility with inverters and transmitters should remain the same.

TS4-F means it is built-in to the module as a factory-integrated device.

TS4-R-F was the first generation of the mounting interface for field-installing the MLPE. R stands for retrofit, and it was built for mounting to the two frames at the corner at once. You need a module that has "accessory friendly" frames on both a short edge and a long edge to use the "-R-" Tigo products. That means that the frames have the lower lip, on which you can install wire management clips and MLPE. Some modules don't have a lower lip on the short frame. The "-R-" units have been phased out, in favor of the "-A-" units.

TS4-A-F is the next generation of mounting interface. A stands for add-on. This Tigo device is meant for mounting anywhere on the module frame, and only requires one frame to mount it. This makes it more versatile, so that it can be mounted on modules that only have "accessory friendly" long frames, and so that it can be mounted in more locations than just the corner.

 

[ggunn]

They may say that because that is the only mix of products they tested, and that the inverter manufacturer is either ignorant of the essential difference, or [they] can't make a statement outside the scope of their testing for policy reasons. However, I don't agree with the inference that this means there is an incompatibility for other combinations.

That's exactly it; sorry I wasn't explicit. Solectria and Chint both told me not to use the TS4-R-F because they haven't certified it, so if something doesn't work I can't go to them for a solution because they told me not to use that box. It might work but I'm not going to risk it.

SMA told me the mirror image; the TS4-R-F is the only F series device they have certified.

 

[Carultch]

Basic question: it's one optimizer per panel correct? so where you say 17 optimizers max... that's 17 optimizers connected to 17 solar modules correct?

Regular question: So each optimizer is generating 15 amps DC... How do they work in conjunction with the solar modules? Each solar module has, for example, an Isc of 5 amps (just throwing a number). Is the module current increased to the 15 amps, or does the module current stay at the Isc value of 5 amps? and it's only the optimizer producing the 15 amps? I'm assuming the module current stays at 5 amps, since thats its max rating.

First question: in my original example, yes, but this isn't universally the case. There are 2:1 optimizers in their catalog, built for grouping the modules in pairs on each optimizer, and are most commonly used with the 277/480V inverters. In some designs, not only do you get to use the 2:1 setup, you *have* to do that. Such as systems that require more than 30 modules in series, where the shutdown/standby voltage would otherwise exceed the 30V limit. You are permitted only one instance with a single module connected, in the event you need odd module qty on a string with 2:1 optimizers.

Second question:
It would be the Imp and Vmp that would get converted, rather than the Isc and Voc. The Isc and Voc are never concurrent, and they are the extreme limits of the IV curve that produce zero power. At Voc, the module is an open circuit. It is energized with a voltage across its terminals, but nowhere for the current to go. When the optimizer output circuit is open, each optimizer defaults to 1 Volt per optimizer, for testing and restart purposes, so your string's open circuit voltage by strict definition, is only equal to the number of optimizers in series.

To follow through with your example, we'll consider modules with an Imp of 5A, and a Vmp of 50V, on an inverter with 400V as the operating input voltage. This would make it a 250W module at this condition of irradiance and temperature. Given 17 single module optimizers in series, that are all uniformly performing at 250W, the total power on this string is 250W * 17 = 4250W. In order to satisfy P=I*V, that means the output current would have to be 10.625A. The input to each optimizer would be 5A and 50V, and the optimizer would transform this to 10.625A and 23.53V.

The optimizer works analogously to a transformer, but since an ordinary transformer can only work with AC, it is a different topology of components. See the animation below for the theory of operation of a simplified buck/boost DC converter:

 

[Grouch1980]

It would be the Imp and Vmp that would get converted, rather than the Isc and Voc. The Isc and Voc are never concurrent, and they are the extreme limits of the IV curve that produce zero power.

Right right... yes the Isc and Voc are only the upper limits, my mistake, I should've said Imp. I'll go through everything else you wrote and get back.

 

[jaggedben]

Basic question: it's one optimizer per panel correct? so where you say 17 optimizers max... that's 17 optimizers connected to 17 solar modules correct?

Regular question: So each optimizer is generating 15 amps DC... How do they work in conjunction with the solar modules? Each solar module has, for example, an Isc of 5 amps (just throwing a number). Is the module current increased to the 15 amps, or does the module current stay at the Isc value of 5 amps? and it's only the optimizer producing the 15 amps? I'm assuming the module current stays at 5 amps, since thats its max rating.

I think the simplest way to answer the question is that the optimizers rarely if ever output 15A. That is just the maximum that they are rated to output and therefore what is relevant to conductor sizing.

I haven't read through the patent document, but my impression is that Solaredge inverters command the optimizers to go to a certain voltage, and then the amps depends on the sun on the array and thus the power available. Example: 12 modules in a string. The inverter wants the DC voltage to be around 380V. So the optimizers raise their output voltage to an average of ~31.67V each. The amps is then dependent on the power available. So if they are 300W modules in perfect sun then amps would be 9.47. (12x300W/380V) Amps must be equal through the string but voltage for each optimizer does not. So if an optimizer has more power to contribute it will raise its voltage above the others and if it has less it will lower its voltage, but together they still meet roughly the target the inverter wants.

The voltage and amps between each optimizer and panel is determined by the optimizer mppt and is likely to be significantly different from the optimizer's output.