Victron MPPT 100/50 vs DSSR50 efficiency.

[Dacian Todea (electrodacus)]

I just purchased a Victron BlueSolar MPPT 100/50

I will do some efficiency tests in different days (cloudy, sunny and partly sunny days) both with same number of panels 6x 255W just DSSR50 are all parallel two groups of 3 parallel panels and the MPPT will have 3 groups of two panels in series so voltage is large enough to do MPPT.

I just bench tested the MPPT and seems to work including the remote ON/OFF the MPPT will turn OFF charging if the pin 1(GND) and pin 2(RX) are shorted. I found a cable with 3 pin connector that fits perfectly in the VE.Direct port as I only need the two pins anyway and I will control this with the new DEXT.

A few numbers to compare dimensions but can also be observed in photo's

                           Victron MPPT           DSSR50

------------------------------------------------------------------------

Max current:          50A                            50A

Weight:                   1.5kg                        0.15kg

Volume:                  1.3L                          0.13L

Price                        227 CAD                  98 CAD

I wanted to open the MPPT to show internal photos but I was not able to and did not force it as I needed to test but made some google search and it seems it is glued internally with the same type of epoxy as the Battery Protect BP65 that I needed to cut (fully damage to see the internals).

But I fund a video you can watch if you want to see the internals  https://www.youtube.com/watch?v=nz_nRLYLizc

or this one https://www.youtube.com/watch?v=gC9EFUHKxhg

They use the potting in order to reduce the noise and vibration from the inductors and also help with inductor cooling.

In both cases is user error that lead to those 80A fuse fault as that can only be damaged if the battery was connected with wrong polarity so nothing to do with manufacturing defects. They have a fairly decent warranty of 5 years on this.

I also found a photo of the internal (some sort of copy of this as it has no potting)

I got the BlueSolar not the SmatSolar as it is a bit less expensive and I have no need for the Bluetooth functionality. It was on sale so decided to get it now.

Will try to find time soon to install it on my large battery remove all loads after discharging the battery close to empty and that should allow for a full day of charging where MPPT will be on the PV shunt and the DSSR50 on the battery shunt so then I can calculate what each of them produced while logging all data on a laptop.

I was not able to properly test the efficiency on the bench as the MPPT algorithm did not like the CC-CV bench power supply as PV input and was constantly trying to find the max power point but seems to be around 94 to 96% efficient so can be fairly good. Likely as good as any MPPT can get but this was at low current just 5A so 10% of max.

Fairly curious how it can compete with DSSR50 especially in cloudy days.

If MPPT algorithm is good I will expect fairly similar results for both DSSR50 and Victron MPPT.

Let me know if you want some specific tests as I will have this just as a temporary install for a few days maybe a week or two at most.

I will post the results here when they are available.

[Dacian Todea (electrodacus)]

Finally I got the Bluetooth dongle to set the battery as 24V.

It was latter in the day but managed to capture a few hours of sun for initial test and have everything prepared for tomorrow when it should be a full sunny day.

Bellow the graph for the last 3 days with some clouds except for today when it was mostly sunny. In total the DSSR50 connected to 6 panels (3x260W and 3x255W) has managed to generate just over 20kWh (I managed to keep the battery from getting to full charge on purpose to see all possible energy).

The last few hours at the end where when I connected the Victron MPPT 100/50 and the rest of the images will be with just that data.

Bellow just the data since the MPPT was turned ON.  You can see that for this short part of the day the DSSR50 has produced 1469.174Wh while MPPT a bit more 1519.057Wh

This means 1469.174Wh / 1519.057Wh = 0.967 so just about 3.3% less energy produced by the DSSR50 in average. There are periods where difference is over 6% but in average is just 3.3%. Will be curious to see how this will be different in a sunny day vs a cloudy day. 

The red line (ADC2) is the MPPT PV array voltage over 70V open circuit but around 55 to 60V at max power point.  The green line (ADC3) is an unused panel in the array to see the open circuit voltage of a panel.

You can also notice that it checks the max power point exactly every 10 minutes for about 30 seconds except at the end of day when it gets "crazy" :)

Some more data points

even more

This is a test to see that MPPT can be turned OFF by the SBMS0 by setting the EXT IO4 to type 0 then back to type 1. You see that DSSR50 starts immediately within one second from ON signal but MPPT takes quite a few seconds with 5s per division.

easier to see the mppt tracking every 10 minutes.

And the end of the day where MPPT will be less efficient than DSSR50 due to low power available and DC-DC requiring some minimum self power in order to work. The DSSR50 will be smooth to the end is just that PV measurements stops at around 300mA due to about 150mA offset and 150mA limit based on the selected current shunt to reduce the possibility of noisy readings.  The MPPT stops charging at 19:40:30 and based on that I estimate the DC-DC requires around 8W just to be ON and that will be normal for a such a large 1500W DC-DC converter. Or maybe there are two 750W DC-DC converters in parallel and maybe they can work with just one at lower power levels. Will try to see that with the thermal camera.

The DEXT can control ON/OFF nicely the MPPT with one of the normally closed SSR normally used for diversion. But care needs to be taken so wires can not become open circuit as that leaves the MPPT ON instead of safely off as is the case for DSSR50 or if you use the Victron dingle and glue that VE.Direct connector to the MPPT so it can not become disconnected.

[michael clark]

Good morning,

The  Victron MPPT is about 160 dollars in the USA so it is a lot more expensive to start with than the DSSR50....

how many solar PV panels can connect to the Victron at 24 volts...??? seems like 2 or 4 maybe??? (((1000 to 1400 watts total not sure on that part???)))

then, the next thought /question is how many Victron MPPTs will the SBMS0 be able to utilize....???

I plan to continue with the DSSR50 and SBMS0 regardless...!!!

for me, it seems the easy expandability of the DSSR50 at 4 panels each is still an economical safe way for the DIY system....

my system is stationary and I like the 24-volt system...

thanks for all your hard work and design of the SBMS0, DSSR50, and DEXT etc....

have a great day 😎

[michael clark]

https://www.victronenergy.com/media/pg/Manual_BlueSolar_100-30__100-50/en/technical-specifications.html

with an internet search >>> I did find the specifications above for the Victron MPPT 100/50 so they seem to match up to the DSSR50 some....

thanks

[Dacian Todea (electrodacus)]

Yes both DSSR50 and Victron MPPT 100/50 support the same amount of panels as they are both designed for 50A max charging.

In this test the DSSR50 is connected to 3x 255W panels and 3x 260W so a total of 1545W while the Victron MPPT 100/50 is connected to 6x 260W (pairs of two in series) total 1560W and this is about the limit of solar for both DSSR50 and Victron MPPT 100/50.

The SBMS0 can handle both the DSSR50 and any of the Victron MPPT and how many of them depends on the max current that SBMS0 can measure and that is 750A

The MPPT is quite a bit more expensive. I paid  227.22CAD on mine + 11% GST so with taxes 252.21CAD and it was on sale vs DSSR50 that is 98CAD so 2.5x more expensive in Canada.

But for panels that are far from battery there is some savings on cables as MPPT  requires half the PV cables.

In any case the test is more to see exactly what sort of gain can the MPPT provide in real world in exactly the same conditions.

In about a week or two I should have enough data to cover all types of conditions and I will also do shading test to see how that is handled by both systems. 

[sailingharry]

Dacian

Not sure what happens with the DSSR50 if you give it too much current, but that is a benefit of the MPPT.  If you overpanel the MPPT (ie, put in enough panels to meet your needs in the winter, and enough MPPT to meet your needs, in the summer you have 2-3 (or 4!) times as much solar as needed), the MPPT hits the peg and just pumps 50A without being damaged.  I suspect that in the case of the DSSR, you either manually reduce the number of panels connected, or (since they are so inexpensive) you install as many DSSR as needed to never exceed their capacity.

[Dacian Todea (electrodacus)]

Over-paneling the MPPT can be an option but it can not be more than about 2x based on spec and even doing that will likely reduce the usable life.

The spec for the Victron BlueSolar MPPT 100/50 is so that you can only connect two 60 cell panels in series the 72 cell panels will exceed open circuit voltage in cold days.

I use now 260W panels connected as 2s3p so a total of 6 panels and they can get up to 85V open circuit in winter and they can output easy 30A (3 panels in parallel) and since the limit for input is 60A that means no more than double the amount of panels I connected and this could cut it to close in some specific conditions so maybe to be safe 2s5p will be max thus not even 2x

You are correct that DSSR50 will not support over paneling but it can handle 6x 260W panels and in those few occasions where current gets to 60A it will be able to handle that and put all in the battery not limiting at 50A

Today I expected sunny but it was cloudy with some sun so I do not know how hot the MPPT will get when working at or close to 50A for longer periods.

Today it got to around 50C and that is acceptable. The DSSR50 is on the right side.

Below is the result from today's test and with warm weather panels got fairly hot thus shifting the max power point for the DSSR50 and still overall the DSSR50 has put in battery just around 5.2% less energy than the MPPT.

I added today the Efficiency calculation including efficiency graph (green) that has average efficiency in 4 minute interval (I selected 4 minutes in order to see the effect of the MPPT algorithm that checks the panels every 10 minutes exactly for about 20 to 30 seconds). It seems it connects the panels directly to battery to measure the short circuit current so during that test efficiency is about 50% and then it disconnects the panels to test the open circuit voltage of the panels.

Below is a zoom of MPPT tracking event in this case it takes 26 seconds and it is doing 3 open circuit and two connecting the PV array directly to battery.

[sailingharry]

Can you clarify "efficiency?" That number (~16%) almost sounds like a panel efficiency, which you would need some sort of solar measurement to calculate.  What efficiency are you measuring?

Also, what is your array voltage (Voc) and your battery voltage?  I wonder if MPPT would yield improvements at the ends if the MPPT array were much higher (double, tripple, etc) than the battery -- so that it can still operate at lower insolation.  You do see the MPPT giving clear benefits mid day, when your Vmp climbs higher, but that's where you would expect to see a benefit.

[Dacian Todea (electrodacus)]

That number of up to 16% represents how much less DSSR50 puts in to battery compared to the MPPT. But that is peak of that day for a few minutes the average for the entire day is just 5.18%

I measure and store the current going from DSSR50 in to battery and separately the current from the Victron MPPT going in to battery every second as well as the battery voltage.

That way I can calculate both Ah and Wh that each charger puts in to battery.

If you look at the first photo the cursor is set at 13:45:04  and at cursor power coming from DSSR50 in to battery is 1092W and from Victron MPPT is 1292W

So 1 - (1092W / 1292W) * 100 = 15.48% meaning that DSSR50 at that exact second pushes 15.48% less power in to battery with the average over 240 seconds (4minutes) of 16.32%

If you look at the efficiency graph you can see that efficiency is negative at the beginning of the day and end of the day meaning that MPPT outputs less than DSSR50 and usually that is when PV power is less than 100W.

So it could be that in a overcast dark day the DSSR50 will outperform the MPPT (not by much but could be 2 to 3% better witch means nothing).

At those sort of currents over 40A in to battery and about half over 20A trough the 10AWG PV cables with a length of about 40m (130ft) the voltage drop on cables will be about 6V even over * 20A = 120W loss on cables so for MPPT it will be 120W loss on cable but for DSSR50 it will be 2x that as there are two cables so 240W loss on cables a delta of 120W due to long cables alone and this large voltage drop 6V plus relatively high panel temperature means the DSSR50 in this conditions is on the other side of the max power point thus not quite efficient. If panel temperature was lower as it is the case in winter DSSR50 even with the long cable will still be in the max power point of the panel.

But even with this worse case scenario of long cables and hot panels the DSSR50 over a full day produced just 5.18% less than the MPPT  

5.118kWh DSSR50 vs 5.397kWh for MPPT for the day a total of over 10kWh put in to the 13kWh battery that I discharged it fairly deep the day before.

Normally battery will be full before noon so the MPPT will make zero difference. 

I will try to also measure the DC-DC efficiency of the Victron MPPT but is fairly difficult because it always changes the voltage at the input that also affects the input current see the red line in second photo representing the MPPT input voltage that was around 55V but fluctuates every second as it is tracking the max power point  and it is anywhere between 50V and over 60V as you see in that graph.

But approx max DC-DC conversion efficiency is around 95 to 96% with peak efficiency claimed by Victron of 97% so around 3 to 5% loss as heat due to DC-DC conversion and another 1 or 2% loss due to MPPT algorithm.

Thus a total loss of 4% to 7% for the Victron MPPT vs around 0.3% max for the DSSR50 but the DSSR50 has to deal with double the loss on the PV cables plus not always being very close to max power point due to panels temperature variation.

[sailingharry]

Thanks for this.  It's really interesting results.  For a use-case as you set up (and it's how you build  your systems, so it is a fair analysis), it is a surprising thumbs down for MPPT.  For the price of the MPPT, you could add a panel or two and be power ahead.

Three possible issues with your analysis (actually, not your analysis, but your conclusions).

1)  The DSSR requires panels matched to batteries.  I believe (but don't know) that Vmp has been rising in the supply chain.  For higher voltage battery banks, that's not a major shortcoming yet, but for mobile systems that are often 12V, it is already a challenge.

2)  For stationary systems, it is super easy (well, not really.....LOL) to get full uninterrupted sun.  For mobile systems with lots of spurious shadows that constantly change, it is much harder.  That could (does it?) make a case for Vmp higher than battery voltage to ensure a shadow doesn't bring Vmp below battery voltage.

3)   Mobile systems are almost always limited by space, and money is not a major design limitation.  Spending 50% more to get 10% more power makes sense, if you are already at max physical size.  I just bought panels at almost $1/W, rather expensive (800W for $600 US), because the size was an almost perfect fit for my maximum footprint -- I was barely even looking at prices as I shopped.  If I tripled my budget for solar panels, I'd still not be able to break 1kW!  So squeezing an extra 5% has appeal (except 5% is, practically speaking, zero, and not worth much cost/effort).  Or, conversely, getting panels at $0.50/W but at a shape that only allows 600W is not a good way to save money (I'd get 600W for $300 vs 800W for $600).

[Dacian Todea (electrodacus)]

1) Yes that is correct that DSSR requires panels matched to battery and typically that is 60 cell panels but it can depend on exact spec of the panel and what the application is (what climate and how far the panels are from battery).

2) I will perform shadow tests also to compare MPPT to DSSR50 and I do not think the MPPT will have a wind there based on theoretical analysis but I will measure that in real world in a sunny day so I have constant sun.

3) If price is not an issue of course you get what fits best in therms of dimensions. Of course 800W with better fit panels will make a difference compared to 600W panels and since is just hundreds of $ is not that relevant. For large fixed systems like mine difference is in the thousands of $ and there is no benefit in more efficient panels or even new panels as in my case I got used panels at almost half the price per Watt.

[jim]

I have zero hands on experience with mppt chargers, but have used my sbms120 for 2.5 years now (around 2.5kw of solar and no heatsink on the sbms120 and always stays rather cool).  It is always interesting to see it collecting 10-30w early in the morning and late in the evening... almost enough to power my inverter :) It took me a while to realize most mppt controllers would just be sitting by doing nothing at that point. It doesn't appear to add much overall, but it is interesting.

It also isn't a huge deal, but the dssr50 is operating with a 5w x 3 disadvantage due to the panels used if I read that correctly. Not exactly a fair fight.

I'm certainly interested in the overcast and shadow tests, but what you've already done is great.

Is there already a consensus on how well the Victron performs vs other mppt controllers? Victron seems to be well respected most places online. I hear of bad mppt or even "fake" mppt, but haven't seen many charts like this vs a real mppt.

I realize this is comparing apples to elephants, but I'm pretty sure how all this shakes out when compared to the realistic limitations of a typical grid-tie setup. They may have 10kw of solar panels, but then pair it with a string inverter capable of collecting only 7-8kw at any one time and some salesman handwaving about solar panel ratings. Installing a larger string inverter is often a huge jump in cost. People paying 5-20x as much for microinverters or string inverters vs a dssr50 seem to shrug off all those losses/limitations while pointing out how mppt is a requirement for solar or you are in the 1970s. The smbs0 can add almost as many dssr50 and panels as people could fit on their land, for very low additional cost. And then I see their solar output graphs and it is a truncated straight line most sunny days... I try to hide my shock when they tell me they paid $75,000 for it all. It makes solar somewhat depressing.

It really feels like the phrase I've heard over and over rings true; if you need more output "add one more panel." It seems this mostly only sucks if you are on a boat. How dare Dacian not design a solar system optimized for every conceivable permutation of solar generation (/sarcasm...).

Thanks for sharing, Dacian.

[michael clark]

I tend to agree... the Victron also requires an additional dongle....for more money and is limited to x number of solar panels plus extra fuses etc...

yes, Victron has many people who like to pay more money...(a very loyal crowd)

I do not have a Victron...

but I know the problems of the solar charge controller limits and the requirements for more solar charge controllers etc...

I think that the SBMS0 is a good method and like most >>> find out the battery is usually full early in the day..

I also bought second-hand solar panels as they will produce solar electricity for many years for less investment...

I am off-grid solar....a stationary build

when you add in combiner boxes and extra fuses the other brand's expenses will quickly add up vs the DSSR50 route with 2 or 4 panels at a time expansion....up to 750 amp charging is a lot of solar panels....

very few if any will have that amount of battery storage...

I am glad he is doing the real-life comparison as the big boys do not tell only half of the story....

their aim is to sell stuff and make money from the consumer....

Dacian has made products to fit his specific uses and they work with many other options also..

thanks for all the efforts

windy but sunny here today and my batteries are full....😎

[sailingharry]

But Dacian HAS designed a system optimized for all!  Yes, his full ecosystem is based on the "add one more panel" and also for the "just get lots of cheap ones" and it works brilliantly.  But here's the hidden pearl -- his system is so robust and full featured that it works well for other cases too!  On my boat, I'll be using Victron SmartSolar (so no dongle needed) MPPT controllers, directly controlled by the 'Dacus (as well as alternators and shore power chargers also directly controlled by the 'Dacus), and I am quite confident that his system is well suited for the task.

[Dacian Todea (electrodacus)]

Here is the result of the shading test I done yesterday.

The DSSR50 won in all 6 shading cases but it won by large margins on the last shade test.

It was sunny and panels were warm at noon thus the delta at the time of the test was around 18% less power for DSSR50 because of the low max power point voltage + the voltage drop on the long PV cables but this test was only about shading.

My array is designed for winter and in winter even at noon the DSSR50 will match the output of the MPPT.  Hope to get some colder days this week to show the effect of temperature on this particular setup.

Shade 1 test had the MPPT initialy confused and for a few minute selected a different max power point but then it did his typical scan that he is doing every 10 minute and he found the better max power point and that is what I used in the comparison.

Shade 2 Of course the DSSR50 will see no difference if one cell or more are covered on the same panel as that panel is just disabled but for MPPT there is a bit of difference as it has panes in series.

Shade 3 this is with two panels shaded. And both panels are in the same group either same parallel group in in case of DSSR50 or in same side on the MPPT case.

Shade 4 same as 3 but this time the two panels are in different groups.

 to be continued as it is to long for one post

[Dacian Todea (electrodacus)]

Shade 5 is where a 2x4" shaded a row of cells on two panels but is not fully shaded as cells are 6" so a bit wider than the lumber.

Shade 6 same as shade 5 test but this time the panels are in the same group and here the MPPT did very badly as expected and even managed to overall output less than DSSR50 not just as a percentage of unshaped panels.

The results have nothing to do with the Victron MPPT algorithm that works great (as good as it gets) but it is due to the physical configuration of the panels parallel only vs series/parallel panels.

I posted this last night but it was lost as rejected for being to long of a post. Thus the reason it is split in two parts.

[michael clark]

these are snapshots of the Victron mppt vs the DSSR550 efficiency....

what is the total production of the DSSR50 vs the Victron mppt for full days???

it is my understanding that the DSSR50 will produce slightly better in early morning and late in the day went there is less sun ... correct...???

My LiFePO4 battery is fully charged when there is a lot of sunshine available often early in the day, but in winter when the sunshine is lacking (((November through January or slightly longer)))...more electricity is sometimes needed.../wanted

It would be interesting to see the total accumulated solar electricity produced during full days...... one vs the other.... Victron vs DSSR50..

thanks for doing these comparison tests.

the Victron equipment is a lot more expensive than the DSSR50 and is not much of an issue for small systems but in large systems, the expense differential could be substantial...

I still like the DSSR50's modular approach as it can be easily upsized a few solar panels at a time....

thanks again,

[Dacian Todea (electrodacus)]

Yesterday for some reason the logging stopped before the end of the day so I do not have a complete day. Not quite sure for the reason maybe there is a file size limitation as it was almost 15MB (about 3 days). I will for now start a new log every day just in case.

It depends on the day how sunny or cloudy and how hot the panels get as the higher the power from panels the higher the voltage drop on the wire and that moves the voltage point at the panels and if above the max power point the current will start to decrease giving an advantage to MPPT. My array was designed for winter as I do not need even 10% of the array energy in summer.

The max power point voltage for my panels is just 30.6V (at 25C panel temperature) and battery is always in the 26.5 to 27V rage so as soon as panel is above 25C usually in the afternoon this days with +20C max ambient temperature and the current trough PV cable is high enough to create above 4V drop 27V +4V = 31V the DSSR50 will start to be on the other side of the max power point voltage.

I will try to do a diagram and explain all this and how to select the panels based on climate and length of cables (my cables are fairly long about 40m (130ft) ).

Here is the data from yesterday. The average was 3.3% meaning that DSSR50 produced 3.3% less energy than Victron MPPT. If I will have gotten all the data until the end of the day it will have been just below 3% as the panel temperature decreases and so is the current trough PV wires thus less advantage for MPPT more advantage for DSSR50.

You can see that at some specific point in time the DSSR50 produced 11.47% less than MPPT likely because PV panel got hotter at that point in time.

The efficiency is average over 4 minutes that is why it looks like a stair and the regular deeps in efficiency for MPPT have to do with the scan MPPT is doing exactly every 10 minutes to check the panels open circuit voltage and the short circuit current.

Below is with efficiency averaged over just 10 seconds and you can much more clearly see those scans MPPT is doing and while it may look bad overall (for the full day) the MPPT will only be losing about 1% of energy due to MPPT algorithm.

But it requires this checks in particular to deal with shading as you seen in the Shading test 1 that I did where he managed to find a different MPPT point that was better after one of this scans.

[Dacian Todea (electrodacus)]

There was a bit of snow especially yesterday so maybe data not as useful due to snow maybe not equally covering all panels.

This morning I cleaned the snow from the panels but some hours after the day has started. I cleaned the snow after I took the photo. The first 6 panels on the right side are on DSSR50 and next 6 are on the MPPT. The MPPT had a slight advantage before cleaning the snow but then with some sunny periods the DSSR50 had the advantage as the MPPT was limited to around 50A

Here is the graph for today with DSSR50 producing about 5% less than the MPPT

I was also working on a diagram to explain how everything works so feedback on that will be useful. So here is an example where MPPT was limited and below the diagram with this example data

All the green power labels are for power available at different points and then with red labels is the power lost as heat. In case of MPPT the 27.1W in this example is just the power loss assuming 98% DC-DC conversion efficiency and that will all end up as heat but there is another about 1% maybe slightly more that will be loss due to MPPT algorithm but that loss is not as heat.

While each set of 6 panels is about 1500W STC rating you can see peaks close to 2000W here there where even a few short peaks today slightly above 2000W when for a few seconds the DSSR50 was pushing 59A in to battery.

Here in this example there are 925.7W + 934W =1859.7W at the panels then 196.1 + 187.9 = 384.1W lost as heat on cables + 4.9W lost on the DSSR50 389W total loss thus in to battery 1470.7W was pushed (1859.7W - 389W).

Despite much lower losses on the MPPT due to slightly shorter cable and lower current due to 50A limit the power input in to battery was lower at 1327.4W due to this limit.

Feedback about how simple or not the diagram is to understand will be useful for me as I want to make a video and explain why we see this results.