Enhanced capacity system

131 views
Skip to first unread message

Ken R

unread,
Jan 17, 2020, 6:01:29 PM1/17/20
to electrodacus
This is a bit of speculative question so please treat it as such.

I'm looking a fully loaded DMPPT system with 44 panels but am concerned that it won't be enough.   I agree with your assertion that the cheapest route to more power is more panels.  Just curious about what would be the strategy of how to drop the extra modules when the output is threatening to overload the DMPPT inputs?  (Where/how would the overload detection be?, the DSSR looks like the solution for switching)

As a side note, I'm becoming amazed at how few solar people will even consider the merits of your strategy.  Entrenched interests rule the day.  Progress is not possible without deviance from the norm.

Ken


Dacian Todea

unread,
Jan 17, 2020, 7:52:40 PM1/17/20
to electrodacus
Ken,

Do you mean for both heating and battery charging or just for battery charging ?  If is just for battery charging you may be able to add a bit more panels but if is also heating then even 44 panels will be a challenge as you will need to keep the DMPPT below +50C and even me with just 29 panels installed I need to have a small fan in front of the DMPPT to keep it below +50C (as reported by the DMPPT450).
If you need a much larger system for both battery charging and heating you may want to consider the SBMS0 + DSSR20 as that will be able to handle up to 96 PV panels. The new DSSR20 that will be available in two or three months will have diversion for heating and while there is no digital MPPT for the heating part just the battery charging it will still be relatively efficient in sunny days.
I can purchase the DMPPT450 from you as there are other that will want that so I can resell that since I no longer produce it.

Ken R

unread,
Jan 23, 2020, 3:11:15 PM1/23/20
to electrodacus
I am in the design stage for the house.  Code is requiring me to have a conventional heating system and I will likely go with a propane fired hydronic system.  The long term goal is eliminate that use of fossil fuel and go fully solar electric.  I guess one could come up with a controller that adds heating elements until the voltage crashes and then back off one; would be a weak form of MPPT.  I appreciate your offer (I also have 2 SBMS120) but think I'll hold off and see how the whole DSSR20 thing works out.

On a side note, was talking to one solar guy in Edmonton and he looked at your system and stated he's had many people asking why they can't do solar electric heating; he may follow up.. This component approach makes scaling it so much easier.

Dacian Todea

unread,
Jan 23, 2020, 3:34:10 PM1/23/20
to electrodacus
Ken,

The DMPPT is not a weak form of MPPT is a Digital MPPT tho it can be argued that a DC-DC converter is also digital is not in the same way. The DMPPT can be around 97 to 98% efficient better than any DC-DC based MPPT that are typical around 92 to 95% efficient and the reason for lower efficiency on a DC-DC converter is the losses in the inductor's and loses in the electrolytic capacitors on top of the mosfets while DMPPT only has mosfet losses that are almost nothing less than 0.5% and the rest is because of the resolution of a digital MPPT still 98% is possible.
Even tho DMPPT is much less expensive than DC-DC conversion is still relatively expensive as PV panels continue to drop in price. For the past few weeks I test using PV panels as panel heaters. I have two large 60 cell panels inside the house that works as a 1000W heating element that works naturally at the maximum power point or very close to it without any electronics. I push about 500W in a 250W panel and 750W should also be possible as I only use 40 out of the 60 cells and it is always safe to touch due to large surface it gets warm but never hot. 

Ken R

unread,
Jan 23, 2020, 4:13:03 PM1/23/20
to electrodacus
I wasn't implying your DMPPT is a weak form.  It was my idea of a unit that would take in several modules (say 5-10) and selectively switch in heating elements to get MPPT functionality as weak; its only on the heat side.  I love the DMPPT because it does that for the input and output to get more precise tracking.  I've always thought of the D as being Discrete.

Using the modules as heating elements is an interesting strategy.  I will be following it closely and will have to think about how to integrate that into a house.  Wires in the floor a much easier to wrap my mind around and they have the thermal battery aspect.

Great work.

Dacian Todea

unread,
Jan 23, 2020, 4:30:14 PM1/23/20
to electrodacus
Ken,


That is what DMPPT is doing on the heating part switches ON heating elements until you get to the max power point of the array then stop based on array voltage. The multiple inputs are for battery charging only has nothing to do with the heating part and so there a number of PV panels are connected to battery in order to get to the set charger current.
So say it is sunny and each panel can provide 9A max power point current and say there is a total of 7 panels so max 63A available connected in 3 separate arrays 1 panel , 2 panels and 4 panels and say you have a small 40Ah battery and so you do not want to exceed 10A of charge current thus normally you could only have one panel connected and maybe based on your consumption that will be plenty in sunny days but in multiple consecutive cloudy days it will just not be sufficient.
Then DMPPT can switch first panel ON see that it provides 9A and just stops there but then later in the day when there is less sun say 5A/panel it can move to the two panels and even later or in a cloudy day with maybe just 1A per panel it can connect all 7 panels.
With unused panels it can do heating if that is needed.

This below simplified diagram shows how DMPPT450 is build inside so 6 inputs each will go trough an ideal diode and all of them add up in a single point from that point there are 6 outputs that can be switched ON individually to get the max power point for heating.
Then just before the ideal diode's there are switches that can redirect any of the 6 inputs to a battery and since battery will be less than 28V that will not interfere with the heating that works at around 29 to 32V normally depending on panel temperature so all current for the diverted input will flow to battery as it has the lowest potential and nothing will flow trough the ideal diode since on the other side it more than 29V and also of course the ideal diode will not allow current from that heating side.

dmppt450-diagram.png 

 
Reply all
Reply to author
Forward
0 new messages