MPPT controller - voltage sensing without separate wires?

Nice! Can I limit the voltage sweep range by entering min and max Vmpp (hot panels in shade, cool panels in edge effect) if I know them? Or is it so fast now and the quickest interval between scans large enough that those few milliseconds don't really matter? Think edge case like solar roof panels charging batteries while driving past buildings and under bridges.

Jim

My reference was to - That stuff with sensing battery voltage during a MPPT tracking function (as Sal described in an earlier post). Additionally to the question Almot asked previous to my last post. I assume you are sensing the battery voltage pretty much all the time unrelated to the tracking function. Are there two processors in the new unit?

Can you give us any update as to when the data sheet will be available for your new baby?

Yep, the battery voltage measurement doesn't have anything to do with MPP tracking, and the sense wires are always active if they're connected. Mode transition points and voltage regulation in constant-voltage modes rely on them.

There are two processors in the new controller.

Should have a spec sheet on it in the next couple of weeks...it's getting closer to the top of the list!

What are the triggers and new user options on MPPT??

The MPPT on this controller is still based on the dynamic tracking that the 3024 used, but it's much faster now. It still uses an array sweep to locate the initial MPP when coming out of rest and between some modes, and then constantly tracks it from there on out. In addition, the user has the option of triggering an array sweep at an interval of their choosing, which can be helpful for arrays that are often shaded. The array sweep is on the order of tens of milliseconds now, so it's basically transparent. The controller also features a slower MPP sweep that is triggered by a fast-rising battery voltage, which will help avoid overshoot common with small (and AGM) batteries. Plus some other tidbits that work behind the scenes to make the tracking more effective and accurate.

I don't think it's that big a deal having voltage sense when current is less than 20A. Most people install charge controller close to batteries with good sized cables.

RV manufacturers on the other hand do a woefully inadequate job with both the location of the converter and size of cables. I had about 23 mohm line drop from converter to batteries. At rated current of 45A, I get a 1V drop across the cable. Charging was very slow. I opened the voltage feedback loop and added a sense line to battery. The charging performance of the 7345 was significantly improved. I added a diode to the circuit so that if sense line opens, the diode clamps max voltage to 0.7V above nominal.

Sal

Almot wrote:


......... Correct me if I'm wrong, this means same 1% voltage drop. So, 14.4V on the battery will read as 14.3V, reason enough to use separate wires.

My info is that battery voltage will be 'seen' by the controller as higher than it really is. So if the controller's set point is 14.4 it will start controlling at 14.4 while the battery has not got there yet at 14.3, leading to undercharging. Or it will mean taking longer to reach the same SOC, or not reach the highest SOC it would have in the daylight time available.

(Going way back, Salvo had something about voltage sensing being desirable in the 7355 converter that would have made it work better.)

HiTech wrote:

Measuring Voc requires a disconnect of the PVs from charging and is generally a sign of a crude MPPT strategy. Who still uses this?

I think Salvo meant that controller "can" measure battery voltage via power cables accurately when there is no charging current. Which is correct. The problem is that there is always some current.

Well - since users complained on "inconvenience" of accurate voltage sensing with separate cables, those who don't care about under-voltage, should be happy now :) ... If the battery is sitting next to controller, there is not much error, but then, there is not much inconvenience with extra sensing wires either. Whatever...

Correction to my previous post - it has to be ~6ft from battery to controller, to get 1% error at 30A. 12ft total run.

Salvo wrote:
That looks like a good approach. When controller disables charging current, the power cables look just like sense lines. The mppt algorithm may measure PV open circuit. At that time charging current is also disabled. That's a good time to sense battery voltage.

To make this work, a dedicated line to batteries is required that's not shared with other loads. That's usually the case anyways.

Iota converter does something similar. It senses battery voltage when ac power is off. When power gets turned on, converter knows exactly which mode is required.

Who builds 3048?

Sal

Measuring Voc requires a disconnect of the PVs from charging and is generally a sign of a crude MPPT strategy. Who still uses this?

Jim

What are the triggers and new user options on MPPT??

Jim

I guess asking about processors would be prying too much into design so new, considering piracy threat. Having a second processor won't help with measuring voltage accurately . If there is a voltage drop, you need separate wires that carry no current and therefore no measurement error.

There were good tables in the old 3024 manual. From those tables it follows that at 30A there will be ~1% post-controller "power loss" if you use 4ft of double-line cable #6, i.e. 8ft current run. Correct me if I'm wrong, this means same 1% voltage drop. So, 14.4V on the battery will read as 14.3V, reason enough to use separate wires.

Hi Marc

Kind of what I figured - Yes as you say all other proper controllers allow the use or non use of the voltage sense lines.

My reference was to - That stuff with sensing battery voltage during a MPPT tracking function (as Sal described in an earlier post). Additionally to the question Almot asked previous to my last post. I assume you are sensing the battery voltage pretty much all the time unrelated to the tracking function. Are there two processors in the new unit?

Can you give us any update as to when the data sheet will be available for your new baby?

The MPT-3048 still has the battery sense feature, just like the 3024 does. The difference is that the 3048 doesn't *require* a connection to those terminals, whereas the 3024 does. In other words, the 3024 won't function without a connection to them. Some people complained that it was inconvenient for them to use the sense terminals (because you have to run separate sense wires to the battery), so I made the connection optional this time around. If you don't connect the sense wires to the 3048, it will sense battery voltage from the BAT+/- terminals on the main terminal block. This is how every other controller measures battery voltage if it doesn't use remote sense wires. Yes, at higher currents there is a voltage drop between these terminals and the battery....so if that's a concern, use the sense wires. That's what they're there for :-) .