Rogue MPT-3048 now available

[lorelec]

For those who've been waiting, just a heads-up that the 3048 is now available for ordering on the Rogue website.

http://www.roguepowertech.com/products/mpt3048.htm

If you've visited the page recently, you might have to refresh your browser to see the update.

[Salvo]

For us 12V guys, peak efficiency is at 70W. Regardless of battery system, peak efficiency is located at the low end of the charging current range, between 5 and 10A.

lorelec wrote:

And just to clarify, peak efficiency doesn't always happen at 70 watts. The 98.4% figure, for example, happens at 550 watts.

[Salvo]

Never heard that gain error is greatest at full scale. Can you provide a link? Are you saying the gain error is nonlinear? The gain is probably set by a couple of resistors connected to the error amp. These resistors are not perfect. They have a tolerance that more or less defines the measurement gain error.

I understand that errors can be divided into gain and offset errors. But you didn't provide a gain and offset error specification.

The whole object of the amp & volt measurement specification is to determine what's the possible error when reading let's say 5.0A. Are you saying it's impossible to determine this error without measuring it? What's the purpose of your spec? What's does 1% at 100A mean? Max current is only 30A.

I'm guessing your current measurement error is at best 3 to 10% of reading plus +/- 0.1A. 0.1A is LSD, not offset.

Sal

lorelec wrote:

Full scale error in terms of ADC output (which is what the display derives its measurements from) is the sum of offset error and gain error -- gain error being the greatest at full scale. It's a very common spec. Lots of references on it if you do a Google search.

Error at lower values has typically been observed at a couple tenths of an amp. Don't know if I have an exact spec for you at 5.0A. No one yet has gotten flustered over their ammeter being off by an extra tenth of an amp. Most charge controllers are off by much more that that.

[Almot]

Not every manufacturer lists efficiency in specs at different modes. When I look at some other brands, it says - by manufacturer, not a retailer's blurb - "very high efficiency". Nothing else. Then, if users bother to open the manual file (if there is a manual), it says in specs - "peak efficiency N%", no mentioning at what input and output. Then, if the user is really advanced, on the last few pages there are efficiency graphs for different modes. Many don't post any graphs at all.

[lorelec]

Salvo wrote:
Perhaps you want to add: "PEAK (at 70W)" when listing these efficiencies in the Feature section.

High efficiency (in/out): 98.4%(68v/54.3v), 98.2%(51v/40.8v), 98.1%(34v/27.2v), 97.5%(17v/13.6v)

And just to clarify, peak efficiency doesn't always happen at 70 watts. The 98.4% figure, for example, happens at 550 watts.

[lorelec]

Salvo wrote:
Perhaps you want to add: "PEAK (at 70W)" when listing these efficiencies in the Feature section.

High efficiency (in/out): 98.4%(68v/54.3v), 98.2%(51v/40.8v), 98.1%(34v/27.2v), 97.5%(17v/13.6v)

Perhaps a good idea, but to my knowledge not something that any manufacturer does. In reality, most people aren't THAT concerned about it. I don't recall anyone ever asking me whether the controller is 2.35% more efficient at "X" power level. I presume that people who are interested in the details are able to successfully glean the data they want from the graphs. Some manufacturers don't publish graphs, either.

The efficiency graphs are great! They should be added to the spec or added in the Features section, not relegated to the last pages of the manual.

Sal

Direct reference is made to the graphs in the section of the manual that discusses efficiency (find me another manufacturer that even discusses that in their manual). They are quite easily located in the back of the manual, directly following the Specification section.

[lorelec]

Salvo wrote:
Your accuracy methodology is new to industry standards. I did a search: "meter accuracy AT full scale", and Google corrected me: "meter accuracy OF full scale". What do you mean by 1% accuracy AT full scale (100A)? Do you mean 1% of reading? Why reference full scale then? What's the total error when displaying 5.0A?

Sal

Funny, Google didn't correct me when I use the exact same words.

"1.0% at full scale (100a)" is the same thing as "1.0% of 100a when the meter reads 100a".

Full scale error in terms of ADC output (which is what the display derives its measurements from) is the sum of offset error and gain error -- gain error being the greatest at full scale. It's a very common spec. Lots of references on it if you do a Google search.

Error at lower values has typically been observed at a couple tenths of an amp. Don't know if I have an exact spec for you at 5.0A. No one yet has gotten flustered over their ammeter being off by an extra tenth of an amp. Most charge controllers are off by much more that that.

[Salvo]

Perhaps you want to add: "PEAK (at 70W)" when listing these efficiencies in the Feature section.

High efficiency (in/out): 98.4%(68v/54.3v), 98.2%(51v/40.8v), 98.1%(34v/27.2v), 97.5%(17v/13.6v)

The efficiency graphs are great! They should be added to the spec or added in the Features section, not relegated to the last pages of the manual.

lorelec wrote:

Salvo wrote:
Specifications should be accurate. Car manufacturers can't get away with advertising fuel efficiency of 50 mpg and not mentioning speed is only 20 mph. That's crap! The charge controller manufacturers are counting on the average person not able to interpret the spec correctly.

Sal

All of the information you want is there in the efficiency graphs. When peak efficiencies are mentioned, that's exactly what they mean: PEAK. It's not an evil plot to deceive customers; it's an attempt at being concise and trying to give an example of the best the controller can do. Refer to graphs for further details. Most people are quite able to interpret a line graph.

[Salvo]

Your accuracy methodology is new to industry standards. I did a search: "meter accuracy AT full scale", and Google corrected me: "meter accuracy OF full scale". What do you mean by 1% accuracy AT full scale (100A)? Do you mean 1% of reading? Why reference full scale then? What's the total error when displaying 5.0A?

Sal

lorelec wrote:


You're not reading the spec correctly. It says 1% AT full scale, not 1% OF full scale.

[Almot]

Salvo wrote:
Specifications should be accurate. Car manufacturers can't get away with advertising fuel efficiency of 50 mpg and not mentioning speed is only 20 mph.

They can, in commercials, when they say "up to 50 mpg" and don't mention that this is on highway 🙂
In car specs - neither in concise list nor in a user's manual - there wouldn't be graphs showing the mileage VS different speeds, loads, and terrain. In a decent controller manual you will have the graphs.

Salvo wrote:

Ammeter Accuracy: 1.0% at full scale (100a)
Ammeter Resolution: 0.1 amp

If the display reads 12.5A, what error do we have?

Hard to tell. Note that it doesn't say "1A". It says 1.0% at 100A. Which is how my multimeter specs are written too, - average accuracy in % at the given range. The accuracy of the actual signal may vary across the scale (it is averaged in specs), and at 12.5A it can be more than 1% of or less, but probably not 22%.

[lorelec]

Salvo wrote:
Specifications should be accurate. Car manufacturers can't get away with advertising fuel efficiency of 50 mpg and not mentioning speed is only 20 mph. That's crap! The charge controller manufacturers are counting on the average person not able to interpret the spec correctly.

Sal

All of the information you want is there in the efficiency graphs. When peak efficiencies are mentioned, that's exactly what they mean: PEAK. It's not an evil plot to deceive customers; it's an attempt at being concise and trying to give an example of the best the controller can do. Refer to graphs for further details. Most people are quite able to interpret a line graph.

[lorelec]

Salvo wrote:


Ammeter Accuracy: 1.0% at full scale (100a)
Ammeter Resolution: 0.1 amp

If the display reads 12.5A, what error do we have?

The meter accuracy is 1% of 100A, or +/- 1A. Add the resolution error and total meter error is +/- 1.1A.

When the display reads 12.5A, the error is 1.1A/12.5A = 8.8%.
When the display reads 5.0A, the error is 1.1A/5A = 22%.

Sal

You're not reading the spec correctly. It says 1% AT full scale, not 1% OF full scale.

[Salvo]

Specifications should be accurate. Car manufacturers can't get away with advertising fuel efficiency of 50 mpg and not mentioning speed is only 20 mph. That's crap! The charge controller manufacturers are counting on the average person not able to interpret the spec correctly.

You just about need a degree in electronics to REALLY know what you're getting. Take for example the Rogue controller ampmeter accuracy.

Ammeter Accuracy: 1.0% at full scale (100a)
Ammeter Resolution: 0.1 amp

If the display reads 12.5A, what error do we have?

The meter accuracy is 1% of 100A, or +/- 1A. Add the resolution error and total meter error is +/- 1.1A.

When the display reads 12.5A, the error is 1.1A/12.5A = 8.8%.
When the display reads 5.0A, the error is 1.1A/5A = 22%.

Sal

[lorelec]

Salvo wrote:
Got some questions/comments on the spec.

1. Spec is 97.5% efficiency at (17V/13.6V). What's the power level?

2. More important, what's the efficiency at 480W, 68V/13.6V?

3. From spec: typical operating power consumption is 0.98W. Max power consumption is 1.5W. That doesn't make sense. Max PV power is 480W. At 97.5% efficiency, you're dropping 12W into the charge controller.

Sal

1) 97.5% efficiency with a 17vin and 13.6vout is at ~70w input.

2) You're unlikely to see 480w in with a 13.6v output. At maximum output (30a) and at 68vin (92.8% efficiency), (13.6 * 30) / 0.928 = 440w. Unless you have 40w of losses in your wiring, you won't see 480w.

3) Power consumption is what the controller uses for powering its electronics, and does not consider losses from power conversion.

[Almot]

I think self-consumption power and the efficiency are two different things. If you look at MS45 graphs, at 98V /13.6V is has same 92-93% conversion efficiency as the Rogue, but in specs MS45 says "self-consumption 1.3-2.7W". 92% = 8% loss, this is 38W from 480W panel, not 2.7W, right?

In summarized specs they all provide very basic information, and Rogue is playing by the same rules as others. In retailers sales pitch - that Rouge doesn't have since there are no retailers other than factory - you can find things like "the industry’s highest peak efficiency of 99%" (for MS45), but they forget to tell that this is only if you use 48V bank and charge it with 52V. In real life MS45 can have 92-93%, but for this info you have to read the graphs. Or things like "significant less power loss than other MPPT controllers" - since when 2.7W is less than 1.5W, I wonder...

[Salvo]

Looking at the Efficiency Graph on page 43, the specified 97.5% efficiency at (17V/13.6V) is actually peak or maximum efficiency. It occurs around 5 or 6A. As current increases (from the peak of 5A, efficiency drops down to 94% @ 30A.

Essentially, the controller is losing 3.5% power as a result of (I'm assuming) copper losses located primarily in the magnetics and some fet on-resistance losses. This is the best the controller can accomplish (MPPT = 17V). At MPPT = 85V and 30A, efficiency losses are 8%, or 38W when applying 480W.

Sal