Sun Power SPR-E19-320 to power element directly series/parallels issue

Estragon

The issue is not enough current. Sort of the opposite of the not enough voltage you had before with the 240v element.

Also, I hope you're being really careful with wiring. 160+v is dangerous.

jimmyaz

Estragon said:

The issue is not enough current. Sort of the opposite of the not enough voltage you had before with the 240v element.

???   Really?  please explain.  Personally I don't see that's the answer here.  I know that the 48v/1000watt element at 48v will pull 20amps.... I am only putting 6amps through it, but I don't see why the voltage would drop from 180 to 15volts because of not enough current.

mcgivor

Estragon said:

If you have panels in series, voltage combines while current in amps stays the same, so 6a would be doing pretty well. In parallel current combines. Put 4 in parallel you should get around 20a to heat the 48v element.

You need to match the voltage, not over volt the element, good thing you didn't have the grid feeding the 48v element, you would have smoked it. The panels are current limited, so yes in parallel not series.

jimmyaz

mcgivor said:

You need to match the voltage, not over volt the element, good thing you didn't have the grid feeding the 48v element, you would have smoked it. The panels are current limited, so yes in parallel not series.

Stated by Missouri Wind and Solar.  For example this element is rated for 24v/25amps=600watts.  It stated below that you can put 150volts to it at 4amps to get the 600watts.

"  You can run high voltage through the element as long as you don't exceed the rated wattage of the element.

For example:

1. A 600 watt element can handle a 600 watt max solar array output of 150 volts 4 amps
   
2. A 300 watt element can handle a max output of 300 watts 200 volt at 1.5 amps.  "
   

jimmyaz

I want to go high voltage DC, to reduce voltage drop over distance and to get max efficiency.  Also, I thought that you can go higher voltage less amps to achieve the same wattage output. 

I thought that if you have 1000watts panels, you can get a 800watts element and regardless of how you wired it, series or parallels you will get at least 800watts out.  But I am getting 100watts.

Something else is at work here...  It's somewhat acceptable if you say that the OHM is low that's causing too much short that the voltage drop to almost nothing... that I can accept. 

But if you say 150v X 5amps = 750 is different than 48v X 15amps... don't know how I can understand that.

mcgivor

jimmyaz said:

Estragon said:

The issue is not enough current. Sort of the opposite of the not enough voltage you had before with the 240v element.

???   Really?  please explain.  Personally I don't see that's the answer here.  I know that the 48v/1000watt element at 48v will pull 20amps.... I am only putting 6amps through it, but I don't see why the voltage would drop from 180 to 15volts because of not enough current.

It would appear the element is dragging the panel voltage down because there is not enough current to back up the voltage, this is an invaluable educational process is it not?

Estragon

@jimmyaz

Correct. The resistance is fixed on each of the 48v and 240v elements. Drive the at the design voltage and they will consume the design amps for the design wattage, if the design current is available. In your series config, you were limited to 6a by your panels.

McgIvor is right. Had grid power at 120v and 100s of amps been available you may now be looking at a puddle of molten metal. Probably just as well you only had 6a available.

If you need to go long distances and want to minimize wire size and losses you can either get voltage up to near your 240v element or up/down convert to target. If your panels are close to the heater, just get the appropriate size wire to keep losses reasonable.

animatt

If you have 4 panels.  For the 48v element 4 wire all in parellel. 
For the 4500w/240v element. You would do best with 2 in series and 2  of those in parellel.  This configuration would run at higher voltage then the 48v element. 


Missouri wind and sun I do not have direct experience.  But while they do sell some quality products they also sell some real junk. I would double check any info you get from them.

jimmyaz

So if get the 120v/1500watts AC element at Home Depot.

I series 2 panels, parallels two, then I would get 120volts and 12amps, then my voltage will not drop, maybe it will a little, but not significant.

Either that, or run all 4 panels in parallels to create 60VOC, 24amps to that 48V, 1000watts element.

No,unfortunately the panels is located on the opposite side of where the water heater is, the run is 100-120ft.  I only ran 12awg, but the good thing is I ran two set of cables, meaning I have 4 12awg wire.  I guess I can parallels them wire too.

Well, the bright side is 48VDC is less dangerous and probably less ARC to deal with.... and I don't have to find a high voltage digital power meter.    Darn.... just though high voltage DC would heat water much faster and more efficient.

BB.

You have to go through the basic electrical equations to "match" the solar array output to the heating element input.

For example, if you have a 120 VAC ~1,500 Watt element. The resistance would be:

• V=I*R; P=V*R; P=I^2*R; P=V^2/R

So, we have Watts (P) and Volts (V) and we want R resistance.

• P=V^2 / R
• R = V^2 / P = (120 Volts)^2 / 1,500 watts = 9.6 Ohms

We will assume Ohms is fixed--But it can (and usually will) vary with temperature (I don't have any information, so assume fixed resistance).

The heating element is 9.6 Ohms and can run 1,500 Watts maximum @ 120 VAC... The rated current will be:

• P=I*V
• I = P/V = 1,500 Watts = 120 VAC = 12.5 Amps

So--The first major numbers you need to worry about are the maximum current and maximum voltage. You cannot exceed both 120 VAC AND 9.6 amps or the heating element will overheat. If you feed it with less than 120 VAC OR less than 9.6 amps, then the element will not overheat and will not exceed 1,500 Watts.

With solar panels, they are Current Sources--They output (more or less) a fixed amount of current (of course, varies with amount of sunlight too). Say you have:

Solar Panel Specs: 
Nom: 320watts
VOC: 64.8 V
VMPP: 54.7
ISC: 6.24 A
Imp = 320 Watts / 54.7 volts Vmp = 5.85 Amps (missing from your first post)

If you put 5.85 amps into a 9.6 Ohm element:

• V=I*R = 5.85 amps * 9.6 ohms = 52.8 volts
• P=I*V= 5.85 amps * 52.8 volts = 309 Watts into heating element
  

One panel connected to the heating element is actually not too far from "optimum". Vmp tends to fall a bit when the panel gets hot, so it is possible that your Voperating will fall because Vmp falls, which will reduce operating current a bit.

If you have four panels, 2s x 2p that would give you Imp-array~11.7 amps and Vmp-array~109.4 volts. In this case, you are both under the maximum current and maximum operating voltage. Need to figure out the operating point.

• V=I*R = 11.7 amps array * 9.6 ohms = 112 volts operating point for element (over 109.4 volts Vmp-array)

So, in this case, you have enough current to run at 112 volts, but your Vmp-array is not high enough to supply the needed voltage for a perfect match between the array and the heating element (but is about as close as you can get with off the shelf parts)..

The operating point will now be the actual Vmp-array operating voltage:

• V=I*R
• I = V/R = 109.4 volts array / 9.6 Ohms heater = 11.4 amps
• P = V^2/R = (109.4 volts array)^2 / 9.6 Ohms = 1,247 Watts into heating element

Unless I made a stupid math error (always possible), that is how it should work with your array.

If you have two panels only, there is no really good way to get much more than 310-320 Watts or so... You can put two panels in series and have roughly a 50 volt drop (use small gauge wiring from array to heater). You use cheaper wire and let it get warm (have to check the numbers that you do not overheat the wiring)--But you are just "throwing away" the extra panel's worth of heat to save copper wiring costs.

I hope this makes sense--I am not sure I am being clear, but hopefully you can follow. If not, please feel free to ask for clarifications.

-Bill

jimmyaz

BB. said:

I hope this makes sense--I am not sure I am being clear, but hopefully you can follow. If not, please feel free to ask for clarifications.

-Bill

Thanks Bill

I think I get it now.  I think I am probably going to go with all parallels and run it on 50volts.

jimmyaz

Hey guys,

Just a quick report back.  Finally pick up the 4th panel today.  Hook it up as 2s2p, fairly overcast cloudy day.  If I only hook 2 panels in series, then I get 60v X 4 amps, but when hook them all together as 2s2p... voltage went up and amps almost doubles... 80volts X 7amps or so. 

Finally the sun peek out for a moment and I saw the meter said 104volts @ 8.2 amps, that mean 850watts.  I guess it isn't that bad for a 1280watts rated panels.  (this is on the 240v/4500w element)

Do you guys really think it's worth it to change to the 48v/1000w element and run all 4 panels in parallels?  Do you think I would get more out of it?

BB.

If you are happy with 4 panels--I would just keep them as is with the existing heating element and see how well it works.

In theory, 4 of your panels in parallel could overheat the 48 volt @ 1,000 watt heating element (Vmp > 48 volts, and Imp-array > 20.8 amp rated heating element). Unless the element is rated for > 1,000 Watts.

Realistically, you probably will not damage the element. But you will have to measure the voltage and current with your 4x parallel connection at noon and see how close you are to 1,000 watt and how much you are willing to push the heating element over 1,000 watts. The "normal" variation in voltage range for a 240 VAC element (~220 to 255 VAC "nominal voltage range") would suggest you can go upwards of 1,343 Watts or so (assuming your DC element has similar design margins).

-Bill

mcgivor

jimmyaz said:

Hey guys,

Just a quick report back.  Finally pick up the 4th panel today.  Hook it up as 2s2p, fairly overcast cloudy day.  If I only hook 2 panels in series, then I get 60v X 4 amps, but when hook them all together as 2s2p... voltage went up and amps almost doubles... 80volts X 7amps or so. 

Finally the sun peek out for a moment and I saw the meter said 104volts @ 8.2 amps, that mean 850watts.  I guess it isn't that bad for a 1280watts rated panels.  (this is on the 240v/4500w element)

Do you guys really think it's worth it to change to the 48v/1000w element and run all 4 panels in parallels?  Do you think I would get more out of it?

Good to hear you got it working, so remember the formula ?
4500*(104/240)^2=845W 

Not far off what you got.
So if you were to use parallel 
1000*(60/48)^2=1562W
Not sure I would try that, unless the voltage drops close to 48v under load, then you would get the full 1000W.

jimmyaz

Humm... but how can the 4x 320watts panels produce more than 1300watts? 1300 is the rated power and under optimal condition and almost never possible in real life.

Im thinking a max of 250watts per panels that will give me 1000w and perfect for the 48v/1000.

Im thinking that with the 1000w im closer to the effecient zone when the sun isnt full, i can still get decent power.

BB.

There are certain conditions (cloud configuration, cold temperatures, reflections from snow/sand/water, etc.) are all possible reasons for panels to output near (or even over) rated power (usually for relatively short periods of time).

You have to design for worst case conditions so the system remains safe. 1,300 Watts into a 1,000 watt rated heating element is getting close to exceeding safe/reliable long term operation (and note that some water conditions build up a deposit around the heating elements which may cause them to run hotter than normal).

Note that many available "DC" rated heating elements are just elements that have been derated for lower input voltage (and current).

A blown element is probably not going to hurt anyone--Just leave you needing to replace it.

So--For the sake of experimenting, sure, why not try it. Just monitor the input voltage/current and see how it does. In general, it is going to run under full rated power 99% of the time (most likely).

-Bill

Dave Angelini

Jimmy, are these panels still available? I have a client with 4 bad panels looking to replace. Thx!

jimmyaz

BB. said:

There are certain conditions (cloud configuration, cold temperatures, reflections from snow/sand/water, etc.) are all possible reasons for panels to output near (or even over) rated power (usually for relatively short periods of time).

You have to design for worst case conditions so the system remains safe. 1,300 Watts into a 1,000 watt rated heating element is getting close to exceeding safe/reliable long term operation (and note that some water conditions build up a deposit around the heating elements which may cause them to run hotter than normal).

Note that many available "DC" rated heating elements are just elements that have been derated for lower input voltage (and current).

A blown element is probably not going to hurt anyone--Just leave you needing to replace it.

So--For the sake of experimenting, sure, why not try it. Just monitor the input voltage/current and see how it does. In general, it is going to run under full rated power 99% of the time (most likely).

-Bill

Thanks Bill, since I already bought a bunch of paralles connector, I guess I will hook it up to the 48v in a 5gallon bucket and see what happen before I install it into the tank.  Trust me, I don't want to replace the element one bit... But if it can increase 200watts, it's probably worth it.

Dave Angelini said:

Jimmy, are these panels still available? I have a client with 4 bad panels looking to replace. Thx!

Hummm... yes the place where I bought these from still had at least 50more.  But they are NEW but isn't manufactured approved.... most got some type of flaws, bubbles that was rejected by the manufacture .... Some marked -10 on them, I assumed they mean -10watts upon testing...  The panels I got, I was able to pick out the best looking one.

Upon my testing when I hook it to a MPPT charge controller the other day... at noon, I was getting 240-250watts.  This was not at a optimal angle.  Suppose to have it at 57degress or so, panel was at 80degrees.

But for 45cents a watts and no shipping... hard to beat.

Dave Angelini

If these were the Sunpower E20 327watt panels you should see 6 amps short circuit as you tilt the panel in most all sun conditions.
Sounds like the rejects that were sold last winter. Thanks but....

mcgivor

mcgivor said:

jimmyaz said:

Hey guys,

Just a quick report back.  Finally pick up the 4th panel today.  Hook it up as 2s2p, fairly overcast cloudy day.  If I only hook 2 panels in series, then I get 60v X 4 amps, but when hook them all together as 2s2p... voltage went up and amps almost doubles... 80volts X 7amps or so. 

Finally the sun peek out for a moment and I saw the meter said 104volts @ 8.2 amps, that mean 850watts.  I guess it isn't that bad for a 1280watts rated panels.  (this is on the 240v/4500w element)

Do you guys really think it's worth it to change to the 48v/1000w element and run all 4 panels in parallels?  Do you think I would get more out of it?

Good to hear you got it working, so remember the formula ?
4500*(104/240)^2=845W 

Not far off what you got.
So if you were to use parallel 
1000*(60/48)^2=1562W
Not sure I would try that, unless the voltage drops close to 48v under load, then you would get the full 1000W.

jimmyaz said:

Humm... but how can the 4x 320watts panels produce more than 1300watts? 1300 is the rated power and under optimal condition and almost never possible in real life.

Im thinking a max of 250watts per panels that will give me 1000w and perfect for the 48v/1000.

Im thinking that with the 1000w im closer to the effecient zone when the sun isnt full, i can still get decent power.

True you couldn't actually achieve more than the panels maximum output, my mistake, but the formula would apply if the current was not limited and voltage stable.

Estragon

There are circumstances where you could see more than STC output. For example, cold high altitude location early morning, cloud edge effects, etc.

jimmyaz

Hey guys,

If my entire solar rack is with Unistrut and the solar get attached directly to it... Do I still need to run a bare ground wire to every single one of them... or can I just ground the Unistrut to the ground rod?  Wouldn't the unitrack be more conductive than the ground wire?

Estragon

You should ground each panel frame with a clip designed for purpose. Aluminum forms a layer of oxidation which reduces conductivity. The clips are designed to pierce the oxidation layer. At least that's the theory (and maybe code where you are). In the real world I have to wonder about AL/CU galvanic corrosion especially in coastal areas.

I'm a long way from any oceans so I used the clips on individual panels.

Estragon

Also, AL is less conductive than CU. You have to go up a size for a given current.

jimmyaz

Estragon said:

Also, AL is less conductive than CU. You have to go up a size for a given current.

So wouldn't the lighting hit the CU instead of the panels?

mike95490

jimmyaz said:

So wouldn't the lighting hit the CU instead of the panels?

Lightning generally goes for the corners & sharp points,  look at an old Franklin Rod

jimmyaz

Hey guys,

So solar is running, producing about 5kw on a good day and 3kw on a cloudy day.  But I had some extra time and been experimenting with the thermostat.... and unfortunately both the STOCK and the MissouriWind and Solar thermostat are UN-ABLE to shut off with 120VDC VOC (104underload).  They both arc and arc, get hot, but can not separate.

So now I am thinking relay is my only option, use a stock thermostat to control a DC relay... the question is, mechanical or solid state?  I can not find any cheap high VDC relay, except this one below.  But it's a solid state... bad idea?

http://www.ebay.com/itm/Soild-State-Relay-SSR-25-DD-DC-DC-25A-3-32VDC-5-200VDC-/351911944512?hash=item51ef962140:g:Se8AAOSwUKxYZb8X

BB.

Be very careful when getting cheap switching electronics:

http://canada.ul.com/safetyalerts/ul-warns-of-solid-state-relay-with-counterfeit-ul-recognition-mark-release-13pn-52/

Good chance the one you linked to is counterfeit and it will not hold up.

-Bill

BB.

http://www.instructables.com/id/The-inner-workings-of-Counterfeit-FOTEK-SSRs/

Another article.

-Bill

animatt

If I do my calculations correctly that would be like 17000 btus on a good day.
I can not remember water heater size. But would assume 40 gallons.
17000 bus would Raise water approximately 51.2f degrees. Neglecting heat loss.

You may have answered it above but I can remember. 

Why not just bypass thermostat and have element always connected to panels.
And have the appropriate pressure release valvE to drain to take care of events of over heating.

You may or may not have issues with having very hot water in water heater.  Depends on existing system.

Matthew