MX 60 data output collection

Re: MX 60 data output collection

tallgirl wrote: »

Does Morning Star have a published API for getting the data out via the web interface? I'm about to wrap up support for SMA inverters and if the MS interface is simple enough, I might add it next.

I forwarded your question to MorningStar, this is their response:

MorningStar wrote:

We do not have a formal document on how to do this just yet. Remote monitoring and data collection is becoming more important in all of our markets. Ideally we would partner up with a few monitoring companies and periodically send data out with whatever protocol they require. This gets around a lot of network hassles as well.

For the motivated few who want to write their own pages, there are a couple of ways:


They can write and serve their own pages on any web server. The issue is that the pages need to call a CGI script from the TSMPPT, which is another web server. Due to security issues (cross-site scripting), this isn’t always a palatable solution.
MS can provide a utility that can be executed on a web server that will periodically poll data from X TSMPPT controllers and store the data in a database. Then they can write web sites on top of the database. Since the product is new, we are waiting to see what customers want before we develop software solutions.

If I were to do this, I would use something like this http://jamod.sourceforge.net/ and then write a script to poll values using modbus over TCP. The advantage of this is all the modbus registers for the TriStars are documented and unlike the web component, modbus can talk with other meter-bus connected controllers thru a single Ethernet interface

Julie, if you have more detailed questions, I would suggest you contact MorningStar support http://www.morningstarcorp.com/en/form-support

Re: MX 60 data output collection

Dave Sparks wrote: »

SG, when the units are linked thru a hub will you only program one unit and the others are slaves as in the XWCC's? Thx.

The current firmware, no ...

Part of the development of the New MS charger we developed and were doing verification of the link feature. A concern that came up was to be UL compliant, it requires that if the linking fails, the slaves have to shutdown ... in essence you have a single point of failure ( cabling ) that could bring down the entire charger system as the meter, which is an independent device was the actual master polling controllers and handling fail-over of a charger drops offline.

So to keep the launch on track it was decided to deal with this as a later firmware upgrade with the proper testing to meet MorningStars quality level. Currently all chargers ship with BTS so the linking isn't essential to multi-charger installations but we do understand the benefits it can offer.

I have never had multi-XW chargers so I can't speak to what the Xantrex units do if the Xanbus faults when the units are sharing charging parameters or BTS values.

Re: MX 60 data output collection

On the MS units, shouldn't be an issue as the all timers and set points are user selectable, the bulk to absorb transitions should be close on units as they track the battery voltage same for absorb to float.

I've run a pair and they pretty much track states within a few minutes of each other even without linking

Re: MX 60 data output collection

Dave Sparks wrote: »

Very cool! Thanks SG! Were your arrays at similar directions? This problem arises for me when one array is pointing at the sunrise and array #2 is true south. Array # 2 does not start racking up absorb time (on some days...) the same as #1

I have done all kinds of things to solve this on the Outbacks. The XWCC just tells other units on the network to go to float and makes parrameter changes on the other units for you.

Is the Morningstar firmware is easy to upgrade in the field??

MS firmware is easy and field upgradable, easiest of the OB/Xantrex/MS as is just RS232 from a PC/Laptop to do the upgrade. Xantrex requires the 400 dollar XANCAN adapter OB you have to send into the factory only.

On Your OB issue, based on my testing I'd guess your Absorb/Float issue is related to the current sense resolution issues with the OB. At low currents and high voltages the OB lacks the resolution to measure current. This is seen different ways as night time shutdown needs typically a 1/4+ amp threshold.

Both the Xantrex and MS units use TI DSP processors and have 4X the resolution ( 12 vs 10 bit ADC ), also ripple current can influence measured readings, again the OB is effected by this significantly more than the other two.

MorningStar goal with the TrsStar Mppt was to have the best efficiency, convection cooled for reliability and most diverse low cost networking options ( RS232 and Ethernet ). Xantrex does offer better inter-device networking and integration as they have the entire XW line, which is best in class on the inverter side.

Lots of good options these days, I hope this information is helpful in your future decisions.

**Proper disclosure, I was a principal engineer in both the Xantrex and MorningStar programs.

Re: MX 60 data output collection

tallgirl wrote: »

Have you checked the voltage calibration on the two controllers? As soon as the batteries reach Vabsorb, the #2 controller should accumulate absorb time, even if it is only outputting a watt or three. If the fixed array is "below" the tracked array, my thoughts are it would get to absorb later, and if it were "above" the tracked array, it would get to absorb sooner.

Yes and even tried offsetting the absorb time lower on the fixed. In the end the XWCC made the Outback CC problem go away. It was a great CC back in 2004 but as SG said "there are lot's of good options these days"

The thing I like best about the new designs is that they are going back to eliminating mechanical components and "hopefully" building products that will last.
In the distant past charge controllers had very little to go "wrong" and did not require you to go to a forum for help.