George,

If you decide to go with the MX-60 and sell any of your SB3048s, let me know. I'm looking to buy one.

My system includes 2880 Wp of solar array currently split into two 1440 Wp arrays. I use two RVPP Solar Boost 3048 charge controllers to charge my NiFe battery array. The peak power I have seen from the controllers is 3300 W.

I've decided that I will be doubling the size of my solar array to support the additional load of my swimming pool pump and to eliminate my wintertime shortfall. I've been carefully watching OutBack power to see if the FX-60 might be a good replacement for the SB3048s to avoid additional wiring from the arrays to the house. I have three pairs of 1 AWG copper in place. Round-trip distance is about 340 feet. I also have a third SB3048, since I had planned for 50% expansion.

Here are my current thoughts on benefits:

Benefits of SB3048:

Paid for and operating fine.
No fan (quieter and less likely to fail).
Established reliability.

Benefits of MX-60:

Less expensive for 2880 Wp application.
Less wiring for 2880 Wp application.
Higher efficiency.
Automatic peak power adjustment.
Higher output voltage capability (although the inverter still limits me to 66 V).
Better display.
Remote control capability.

I particularly like the fact that the MX-60 achieves its peak efficiency at 40 amps, since my system typically operates at about that output current.

Unfortunately, if I double the current in my wires, I will increase their loss by a factor of 4 (obviously). The improved efficiency of the MX-60 would allow me to double the array size while mostly offsetting this efficiency hit in the wiring.

Anyway, it's great to have choices! Thanks OutBack! Please correct me if I have misunderstood anything about this comparison.

Regards,

George

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>Originally posted by George Estep:
Anyway, it's great to have choices! Thanks OutBack! <HR></BLOCKQUOTE>

Yes - we think so as well!

With the MX60 you can also increase the nominal PV array voltage to reduce the losses in the wiring from the PV array to the controller. We can handle up to 125 vDC open circuit - which will probably be 5 modules in series - or possibly 6 if the VOC is at the low end...

The fan on the MX60 is very small and nearly silent. We included it because the losses in the inductors for the converter was making the other parts in the box get hot. Hot electronics reduces their reliability - so the fan keeps all of the parts cooler to increase the reliability. The fan also increaes the efficiency of the inductors as well. Our testing has shown that the MX60 can operate without the fan under most conditions BTW.

In my opinion, based on the temperatures of the Solar Boost units I have tested, that they should have a small cooling fan as well to ensure reliable operation over the product's lifetime.
[img]images/smiles/icon_cool.gif[/img]

Thanks for the detailed reply, Christopher!

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>Originally posted by Christopher Freitas:
Yes - we think so as well!

...We can handle up to 125 vDC open circuit...
<HR></BLOCKQUOTE>

According to the specs for the RVPP 3048, max open circuit voltage is 127 VDC, which is slightly higher than for your unit. NEC requirements force the user to derate OC voltage somewhat (is it 25%?) to account for cold temperature operation. This pretty much precludes operation using a string of six. A string of five might fit within the NEC requirements, but I have 24 panels in my array, which is not divisible by 5.

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>Originally posted by Christopher Freitas:
In my opinion, based on the temperatures of the Solar Boost units I have tested, that they should have a small cooling fan as well to ensure reliable operation over the product's lifetime.
<HR></BLOCKQUOTE>

Regarding fans: If you are basing this on 30 A output, I agree with you. However, when sizing such a system, it seems that you must set the "normal" operating condition below about 25 A so that you can take advantage of the additional energy available from edge-of-cloud effects and low-battery-voltage operation. The result is that the SB3048 as well as your unit will tend to be operated very close to their peak efficiency points most of the time. (Either that's where the converter topology peaks out or you designed it to happen there, or both.) I don't think NEC derating of 156% apply to either of these converters since they both automatically limit their output current.

As a result of all of this, I doubt that your unit will operate above 50 A very often nor will the SB3048 operate above 25 A very often. In my application, the operating current is more like 21 amps due to the higher voltage of the NiFe batteries.

Further thoughts?

George

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>Originally posted by George Estep:
According to the specs for the RVPP 3048, max open circuit voltage is 127 VDC, which is slightly higher than for your unit. NEC requirements force the user to derate OC voltage somewhat (is it 25%?) to account for cold temperature operation. This pretty much precludes operation using a string of six. A string of five might fit within the NEC requirements, but I have 24 panels in my array, which is not divisible by 5.<HR></BLOCKQUOTE>

We limited the rating to 125 vdc because that is the maximum allowed by our breakers and ground fault protection devices. We have ran them with three of the Sharp modules in series with no problems.

You could add a 25th module - but that is rarely an option due to mounting/appearance issues. It also only works with 12vdc modules of course.

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>Originally posted by George Estep:
The result is that the SB3048 as well as your unit will tend to be operated very close to their peak efficiency points most of the time. (Either that's where the converter topology peaks out or you designed it to happen there, or both.) I don't think NEC derating of 156% apply to either of these converters since they both automatically limit their output current.

As a result of all of this, I doubt that your unit will operate above 50 A very often nor will the SB3048 operate above 25 A very often. In my application, the operating current is more like 21 amps due to the higher voltage of the NiFe batteries.

Further thoughts?

George<HR></BLOCKQUOTE>

I have not actually seen any curves of the RVPP Solar Boost efficiency so I can not comment on their performance. If you have a copy of their efficiency curve I would like to see it.

Our efficiency increases with power mostly - but is lower when the differential of the battery and PV voltage is large (but only slightly lower ~1 or 2%). The benefit of having a higher nominal PV voltage than the battery offsets the slightly lower conversion efficiency, IMHO.

We include the efficiency of our unit for both 24 and 48 vdc applications along with a variety of input voltages of the PV array. The latest PDF of the manual for the MX60 should be on our website soon - it also includes a "menu map" for the standard LCD display. Or let me know and I will email the PDFs directly.

The two 125% safety margins for the PV array size are not required with the MX60 controller - only one 125% margin is needed. I am not sure about the RVPP SB3048 though. OutBack also offers hydraulic-magnetic circuit breakers that are approved for use with the MX60 to allow operation continuously at the 60 amp output level.


[img]images/smiles/icon_cool.gif[/img]

[ December 03, 2002: Message edited by: Christopher Freitas ]

Hi Christopher,

Thanks for providing further information and the MX60 manual. I'm sorry. I do not have any efficiency curves for the SB3048. My comments are based on specific data points I have found in their literature.

I have read through your manual a couple of times in order to more fully understand the comparison between your product and the SB3048s that I am using. For me, further insight leads to further questions. I hope you are willing to indulge me with a few more of your excellent answers. ;-)

But first I would like to commend your team on this product. I find the efficiency curves in the manual to be nothing short of amazing. As a power electronics designer, I tip my hat to you and your team!

1) The efficiency curves in the MX60 manual found at http://www.outbackpower.com/MX60%20Manual%20Rev%203.pdf do not match the efficiency numbers listed in the MX60 datasheet at http://www.outbackpower.com/MX60.pdf. Which information is correct? I had based my previous statements on the numbers found in the datasheet.

2) On page 2 of the manual, you state "The MX60's setpoints are fully adjustable to allow use with virtually any battery type, chemistry and charging profile." Then on the very same page under "Important Safety Instructions" you state "This charge controller is intended to charge lead acid battery systems, sealed or vented with nominal voltages of 12, 24, 36, 48, 54 or 60 volts DC. For other battery chemistries, contact the battery manufacturer for specific charge control settings and methodology." My question is do you support alkaline battery chemistries such as NiCd or NiFe. I can see that most of the settings are adjustable, but I didn't see one to allow me to change the temperature compensation slope. Also, it appears that the settings might all be tied to calculations based on 6, 12, 16, 18, 24, 27 or 30 cells per string. Is this correct? If so, it may be challenging to get the settings right for chemistries other than lead-acid. Thoughts?

3) There is no mention of paralleling of the outputs of multiple MX60 units. Can this be accomplished without the MATE? If so, can the details of such arrangements be added to the manual or provided elsewhere? RVPP has, IMO, a rather nice method of handling this issue which is clearly documented:
http://www.rvpowerproducts.com/pdf/100202.pdf

4) Can the output of the MX60 be attached to the same battery bank as an RVPP SB3048? Your manual states on page 6 "The MX60 is designed for PV input only and will not function properly with other power sources or combinations of PV and other power sources or in parallel with other charge controllers." While this seems to answer the question, it really appears to be a discussion of input restrictions, not output restrictions.

5) Why have you limited the terminal sizes to #4 AWG? Is it because of the NEC requirement for disconnects at both inputs and outputs? I realize that you would like your customers to install their MX60s just like in Figure 1 on page 3, but some of your customers have already installed Brand X. In my case, my charge controllers are not adjacent to my DC Disconnect. What do you recommend for installation of the DC breakers when an MX60 may be the only OutBack product in the system (at least for now).

6) Can your converter be operated at 100% duty cycle? Assuming that it can, will it operate there if that is the peak power point? The reason I ask is this: I'm convinced that with some combinations of solar panels and battery systems this is often the proper operating condition. In my system, that is certainly the case in the summertime except during low-battery conditions. But with the SB3048, I must manually change the controller settings to get it to go there and then reset it to get it to operate in other modes. It's somewhat hard to describe, but basically, the SB3048's setpoint is not optimal if not adjusted seasonally in my application.

7) How often does the MX60 check it's peak power point? Does it have both major and minor searches? These are very important issues which are not discussed in the manual. Just as we have seen with the Brand X MPPT software problems, the rest of the specifications of the unit go out the window if this is done poorly. This is particularly true on partially cloudy days (like everyday in FL). The eight-second interval used in the SB3048 is one of the unit's strong points, IMO.

8) Why do the efficiency curves have the word "nominal" in the subtitles? I would assume these are the exact voltages used for the measurements.

9) What is the accuracy of the information displayed by the unit?

10) The hickup approach to startup and shutdown seems odd to me. The SB3048 has a clever (granted, it's patented) approach to startup and shutdown which results in a single relay cycle on a normal day. My charge controllers are near my living space, and I don't mind hearing the relay because it tells me when the units are waking up. I get the impression that your units may be annoying each morning and evening. Please correct me if I misunderstood.

11) Can you please provide efficiency curves at higher output voltages? In my case, I like to operate as close to the 66 volt inverter shutdown point as I can. The unit typically operates around 62 volts.

12) What are the provisions for upgrading the firmware in the MX60? Can the firmware be upgraded by the user? IMO, this kind of product has little chance of getting out the door without software issues, and I am not at all interested in installing a charge controller and then removing it to send it back to the factory just for firmware. The SB3048 does not have firmware, so this issue does not apply to that product.

13) On page 2 the manual says: "Your MX60 can also be connected to the OutBack MATE system controller and display to allow monitoring of up to eight MX60 controllers from a distant location - up to 300 feet away. The MATE also includes and opto-isolated RS232 port for connection to a PC computer for data logging and system monitoring." Is there an API available for your products? Can the MX60 be monitored without the MATE? How does the cable get from the MX60 to the MATE? Does it need to be in a conduit to comply with codes?

14) On page 5 the manual says: "Surge protection on both the input and output of the MX60 is required for extended warranty protection." I must say that requirement is likely unenforceable since OutBack does not specify the proper methods for surge protection of this device. Having designed surge protection circuits in the past, I can say that it is not trivial and is better not left up to the typical homeowner or installer. I have seen some pretty strange failure modes in the lab. Perhaps units which provide breakers (or other disconnects) and surge suppression at various voltages could be made available to your customers to assist them with meeting this requirement and the NEC disconnect requirements. Thoughts?

15) Is this product currently available to buy?

Thanks again for all the information you have provided so far. Any further insights you wish to share will be greatly appreciated!

Regards,

George

[ December 07, 2002: Message edited by: George Estep ]

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>Originally posted by George Estep:
I do not have any efficiency curves for the SB3048. My comments are based on specific data points I have found in their literature.<HR></BLOCKQUOTE>

I think providing several curves is necessary to represent the efficiency of an MPPT accurately since it varies with both the PV input and battery voltages involved. The curves we included in the manual are more accurate than the "composite" efficiency numbers we included on the product dataheet.

The controller is designed to be able to charge a variety of battery types and we are interested in adding any additional capabilites as required. From a UL standard listing standpoint - it was only approved for lead acid batteries - therefore the statement in the safety instructions.

The temp comp is set for -5 mV/C per cell of lead acid battery (6 per 12 vdc). When you select the nominal battery voltage we are setting the temp comp level. By selecting a different nominal battery voltage than you are actually using you can, in effect, adjust the temp comp value. This is a little ackward I agree but does make it simpler for most users.

Our charge control method allows for the output of multiple MX60s to be paralleled without problems. We have also allowed for mixed operation with other controllers such as RVPP and C40 controllers. Multiple MX60s do not require the MATE to be used to have acceptable operation.

We will in the future have means for a similar method of coordinating the operation of multiple MX60s further by using the MATE and a DC shunt on our OutBack communication "network". It will not be required however.

The statement in the manual about using the MX60 with PV only is only discussing the input side of the controller. We have had people already connect other DC sources (such as small wind turbines) together with their PV array to the input of the MX60. This makes it hard to MPPT and can cause other problems.

The #4 AWG wire limitation was due to the amount of wire bending space required by UL when larger terminals are included. Our goal was to make a product similar in size and layout to the C40/C60 controller - so we were space constrained. Since the NEC requires disconnects nearby to the controller - it is easy to transistion a large wire from the PV array location at the disconnect.

If you are not using the OutBack PSDC then standard fused disconnects would be another option.

The MX60 is not currently able to operate at 100% duty cycle (i.e. have the PV array at the same voltage as the battery). It is able to operate close to 100% however. This is a tradeoff in the design we used. I think many people will be looking at having a higher nominal voltage PV array than the battery nominal voltage (such as a 48 vdc array and a 24 vdc battery, or 60 vdc array and 48 vdc battery etc.) so this is not an issue for all installations.

For your application where you have NiFe (nickel iron) batteries - it would be an issue unless you could go with a higher nominal PV array (five 12vdc PV panels in series with a 48 vdc battery for example).

We felt that requiring the user to make seasonal adjustment to ensure high performance was not acceptable. That is why we do true maximum power point tracking (TMPPT?). Finding the max power point when it is close to the battery voltage can be difficult for many MPPT software schemes.

The MX60 includes both major and minor maximum power point sweep routines (we call them full and "mini-sweeps"). The timing depends upon a number of things - such as how stable the output current is. We sweep at least once a minute and more often under some conditions. These are critical issues for operation in places such as florida and are also something we consider to be hard won solutions. We are including the ability to set a number of these variables in the field with our assistance on the next software revision (currently being tested in the field).

The nominal in the title of the efficiency curves was a reference to the nominal system voltage (i.e. 24 vdc or 48 vdc). The numbers on the curves are the actual measured values during the tests.

The accurancy of the current meters on the MX60 are one of the design issues that we are struggling with at the moment. We are working this right now. This does not affect the MPPT function - but does impact the customers information about the PV array's perfomance.

Our wake-up/shut-down method normally works without any nuisance cycling of the relays. Under some conditions (such as a PV array that is shaded in the morning) it may restart once or twice until the PV array can produce reasonable power. I don't think it will be an issue on your system. The cycling is at a very low rate (about once every few minutes max).

I will try to get you an efficiency curve closer to your conditions - what is your PV array configuration. Most of the change in the efficiency is duw to the difference of the PV and battery voltage so you could extrapolate from our curves provided as well.

The firmware is technically upgradeable in the field - we are using a FLASH type processor on the MX60. It does raise some security issues for the software we use in the product. Having the code would make it a lot easier for someone to knock off our design. There are some clever options we are considering to allow field upgrading. Our mechanical design also allows replacement of the "guts" without removing the enclosure from the wall if required. This might be easier since then the conduit and wiring would not have to be disturbed.

The communication port on the MX60 to the MATE is not intended for direct connection to a PC computer. We are working with a number of people to develop software for use with the MATE and will provide the data format etc. to people who can write their own software. The cable used is standard CAT 5 cable with RJ45 connectors on the end of the wires (8 conductor twisted pair - non crossover). It should be acceptable to run it without conduit inside of buildings.

The surge protection clause will be hard to enforce we agree. We do want people to install something however to handle possible lightning conditions - anything should be better than nothing... We are working with a company who specializes in lightning protection for telecommunication systems on a future product.

The OutBack MX60s are shipping in volume at this time. We do have a significant backlog of orders and hope to fill most of them in the month of December. Check with your dealer/distributor to get the latest delivery schedule.

[ December 07, 2002: Message edited by: Christopher Freitas ]

Wow, Christopher!

Thanks! On a Saturday and everything!

It seems that these may work quite well for my application.

One followup:

Unfortunately, with the arrangement of 24 Solarex MSX-120s that I have, I don't think going to strings of five will be a viable option for me. What maximum duty cycle do you think the MX60 can achieve (approximately)?

Thanks again,

George

[ December 07, 2002: Message edited by: George Estep ]

<BLOCKQUOTE><font size="1" face="Verdana, Helvetica, sans-serif">quote:</font><HR>Originally posted by George Estep:
What maximum duty cycle do you think the MX60 can achieve (approximately)?
<HR></BLOCKQUOTE>

Its still early for us out here ! (I think there are 8 outbackers still here right now)... so today's Saturday huh? humm...

I should probably ask why you are so knowledgable about power supplies before I respond to more questions... [img]images/smiles/icon_smile.gif[/img] [img]images/smiles/icon_wink.gif[/img]

So - are you a retired NASA engineer? (that's my personal guess).

[ December 07, 2002: Message edited by: Christopher Freitas ]