Rogue MPT-3048 now available

Salvo wrote:
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.

Peak efficiency for most regulators (with the exception of some multi-phase designs) is normally at the lower end of the range. That's because resistive losses increase with the square of the current. Double the current (to 10a) and you're still at almost 97%, double it again (20a) at you're at 95.6%. Worst-case (30a) you're still at about 94%. You make it seem as though anything other than the sweet spot is going to be terrible -- which, in reality, is not true at all.

Salvo wrote:
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.

You're thinking of an op-amp, whereby gain is set is set via external resistors. Input bias current in that case certainly makes the response non-linear. With an ADC, the difference between offset error (measured with zero input) and gain error (measured at full scale) does normally cause a non-linear transfer function. I'll leave it up to you if you want to find a link for it.

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

No, the purpose (at least in my case) was to give a maximum figure, which is greatest at full-scale. The current sensors are good to 100a, so that's why 100a was chosen as the full scale figure.

0.1A would be the resolution. The LSD, by definition, could be any number from 0-9.

Sal, if you have any genuine critiques or questions, I'd be happy to entertain them. Nitpicking about non-issues doesn't qualify.

mena661 wrote:
I really don't understand what you're getting at Salvo. Are you trying to discredit Rogue or are you just asking questions?

Mena, this has been going on for months if not years. Salvo just doesn't like the Rogue guy. The Rogue guy knows it. Gets old. :(

Salvo has been taking his civilization pills for quite some time now, and has been very helpful on a number of topics. I am especially grateful for some help in particular.

No idea why the Rogue guy gets up Salvo's nose. Just glad it's not me anymore with my "measurement errors." :)

I really don't understand what you're getting at Salvo. Are you trying to discredit Rogue or are you just asking questions?

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.

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.

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.

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.

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.

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.