Author Topic: Water Heating Controlled by a Classic 150 (Read 45642 times)

ghurd · « **Reply #66 on:** November 20, 2012, 08:58:16 PM »

"The only thing I'm unsure of is if the thermostatic relays in the water heaters can handle switching 90 VDC power."

Why would you want to do it like that?
G-

ChrisOlson · « **Reply #67 on:** November 20, 2012, 09:06:06 PM »

Just to experiment with it and see if taking the extra power from the panels and sending it direct to the aux load is more efficient (get more hot water) than running it into the batteries and thru the inverter.

I've tried it before at battery bank voltage and it's a disaster because it's hard to find elements that will draw enough power and it takes big wiring. But with 90 VDC we're talking only ~12 amps to divert 1,100 watts of power. And I got higher DC voltage than that available from turbines - they run up 142 VDC. So excess could be diverted direct from those too in the form of high voltage DC.

It's just something I want to experiment with.
--
Chris

hydrosun · « **Reply #68 on:** November 20, 2012, 10:59:49 PM »

The relays in a normal hot water tank can't be used to switch dc. I've had them stick closed on 24 vdc.They can be used to control a dc rated relay.
I've looked into using the high voltage from the solar or hydro that normally goes into a Fm-80 to go directly into a heating element. One thing to watchout for is if the load is more than the output from the source it will try to move backwards through the controller from the battery. I've been told that might destroy the outback controller. Either contact midnite to see if their controller can handle that or make sure it can't happen. Either a large diode or a double pole switch before the input to the controller would do it.
Chris

ChrisOlson · « **Reply #69 on:** November 21, 2012, 12:57:51 AM »

The load can't be more than the output from the source because the Classic's AUX 2 with PWM drives the load, and it does it based on extra power available for the charge stage. If there's not enough extra beyond what the batteries require, then it won't even have the load turned on.

I wondered about those relay contacts in the thermostats, if the DC arc would weld them or not when they switch.
--
Chris

hydrosun · « **Reply #70 on:** November 21, 2012, 11:19:27 AM »

Even with pwm the momentary load is the full watts accross the heating element. So it could be coming from the battery through the controller if there isn't a diode that prevents it. I don't know if there is one in the classic but I was told not to do this with the outback controllers by outback engineers. It may work fine but it may not.
Chris

ghurd · « **Reply #71 on:** November 21, 2012, 08:55:18 PM »

Use the contacts to control something solid state.

Couple resistors, mosfet, and contacts to trigger 'something else'.
G-

ChrisOlson · « **Reply #72 on:** November 21, 2012, 09:35:58 PM »

The general consensus over on the MidNite forum on this, according to some of boB's comments, is don't do it. I think a diode on the input of the Classic would protect it and prevent it from backfeeding the load (assuming it can).

There is one guy over there already doing it and he says his Classic has suffered no ill effects. But he's got a lot smaller load than I do, and his panels are running at lower voltage than mine.
--
Chris

OperaHouse · « **Reply #73 on:** November 25, 2012, 11:35:08 AM »

Experiment indeed! Have only been able to find one data sheet that gives the speed of these
DC SSRY and that was .5ms turn on/off. They could get toasty over 200hz. I have been
wanting to buy one to see their real response.

I think the topic of the full load anytime when the DC SSR is on is interesting but not for the
reason previously mentioned. The load on the input of the Midnight can cause a dramatic
voltage change if the load current is close to source current. While others see a controller
as a black box, I am always thinking of the control structure they use inside. Any control,
no matter how good will have some set of conditions where it is less than stable.

An ideal situation would be to have a large capacitor bank to store power. This was mentioned
in another water heater post by GHURD. very seful in lower power situations. In that way a
5A power source could supply 10A pulses to a load. That large capacitor bank could cause
real problems for a controller. Any controller will have some capacitors inside to prevent
parasitic oscillations. This allows the current to be limited by the source. PV panels are
basically a constant current source and a mill is limited by coil resistance and inductance.
However, a large external capacitance can supply currents many times what the PV panels are
rated at. Potentially this can damage the switching devices of the controller. I built a
machine to magnetize and laminate plastic sheet magnets using a large capacitor bank. As
an example, that would produce 20,000 amps. The welding cables going to the magnet bar
would jump each time it fired. You should ask Midnight what they would think about you
placing 50,000 uF at the input of the controller.

One way around this is to isolate the capacitors from the controller with a diode or
inducter. The inductor does not isolate like the diode but limits the current for a period
of time protecting the controller electronics. A power transformer with a sufficient
current rating could be used as an inductor.

aleman83 · « **Reply #74 on:** December 26, 2012, 03:43:10 PM »

Hello Chris,
This is exactly what I have been looking for. I also have a midnite solar classic and would like to use my access power in water heating. I live in Honduras and my accessories and professional help are very limited.

The waterheater I could buy here would be a 110V 30 Gal 2000W one. Would this work with the same ssr? And then the even bigger question. I have no idea how to properly connect everything and can't seem to find any information about it that would make sense to me. What is it that I connect to the charge controller? What settings do I use on the charge controller? And where do I draw my power for the heater from? Does the power come straight off the same inverter I run everything else from? Maybe it just didnt make all that much sence to me because of the missing pictures.

I would really aprecciate your help.

Aleman

Eric L · « **Reply #75 on:** January 23, 2013, 01:49:02 PM »

A quick and hopefully final update on my experience with this since I made another large change a month ago.

I had grown concerned about the inverter transformer core saturation issue discussed above. The inverter was running hot, and I think the dirty AC power was stressing some items in my house, like the 'fridge compressor motor.

So, when I saw a good price on a 48 volt, rebuilt 120 volt AC inverter, I bought it, and ran an AC wire from it to the wh.

The obvious drawback of this is that it increased the cost of the project from a few tens of dollars to a few hundred. But there were some advantages. First, the voltage fluctuations and pulsing issues on the main inverter were solved, unsurprisingly. Second, because the new inverter runs at 120v, the load was smaller (1000 watts into a 4 KW 240 volt element). This turned out to be an advantage because the smaller inverter goes into steady, full-power mode with no pulsing and steady voltage more quickly, as long as sun conditions are good. Finally, my main inverter has less to deal with and more available capacity.

One thing I had to change was to put the offset in the Classic AUX settings up to 3.0 volts. At lower offset settings, the element wasn't getting enough power; basically if it's not getting the full 120 volts for much of the day, it won't make a significant contribution to the water heating. Setting the offset wide pretty much insures that it's receiving full voltage when the surplus power is available. Setting it too narrow did not do this, even when surplus was available, I think because my charge controllers could not respond quickly enough to the pwm-induced changes in load to keep the voltage from fluctuating.

In retrospect, I might have gotten a similar result by wiring the element at 120 volts just using my main, 240 v. inverter. I never tried this, but that would have lowered the load and might have kept it in a steady full-on state more often as well. But I don't mind that I did it this way; still plenty of hot water and now, plenty of spare capacity and no stress on the rest of my equipment.

gww · « **Reply #76 on:** January 23, 2013, 05:53:24 PM »

I don't know if this adds anything to this discussion but I found it interesting.

http://www.outbackpower.com/forum/viewtopic.php?f=29&t=2865

Thanks
gww

ChrisOlson · « **Reply #77 on:** January 23, 2013, 09:17:38 PM »

Quote from: Eric L on January 23, 2013, 01:49:02 PM

I had grown concerned about the inverter transformer core saturation issue discussed above. The inverter was running hot, and I think the dirty AC power was stressing some items in my house, like the 'fridge compressor motor.

Thanks for the update, Eric. I've never had any problems like that with our SW Plus 5548. The PSX-240 made some noise with the PWM width set to 1.0. But after I set that to 3.0 it works perfect and runs quiet. To set the offset, I kept increasing it (actually going more negative) until the absorb voltage wouldn't hold the proper voltage anymore at the bank. I ended up at -1.0 on the 48 volt system, and once it drops into float the element is going at just about full power.

I can only assume the Magnum does not have the heavy duty transformers in it that the big Xantrex has, and that's why it affects it like that. The Magnum inverter weighs only 1/3 of what the big Xantrex weighs, and all the weight in the Xantrex must put extra copper in there someplace where it's needed.

Since we last talked, we have some friends that live off-grid that bought a new MS4448PAE and when their washing machine is running on the agitate (wash or rinse) cycles all the lights in their house flicker like crazy. I took a little video to show him that while our washing machine does the same thing (wildly varying load when it's on agitate), our lights do not flicker like that - it is as stable as a rock.

--
Chris

Eric L · « **Reply #78 on:** January 23, 2013, 09:23:34 PM »

Quote

I can only assume the Magnum does not have the heavy duty transformers in it that the big Xantrex has, and that's why it affects it like that.

I think that's probably right. While the Magnum could do it, I think it would probably have reduced its lifetime and considering the options, putting the stress on an inexpensive rebuilt inverter seemed most sensible.

It will be interesting to see what my annual megawatt hour production will with your modification. At my location, I have good year-round sun, and right now one MWh/year per kilowatt of array looks to be quite obtainable, and it may well be higher since the panels are near full capacity most days.

OperaHouse · « **Reply #79 on:** January 24, 2013, 01:26:14 AM »

Chris, Have you tried your FET module yet running DC from the source? I found another module and it gave a response time of 10 ms which is pretty slow and could cause considerable heating of the module. If you look at FET driver chips they will often have capability to supply several amps todrive the FET gate. Once a FET is turned on the gate does not require any current. Even I like to believe that some black nbox module somehow bypasses these basic design needs. Most applications never intended to have these turned on and off constantly.

I question the slicing and dicing with the AC modules too. I think everyting would be happier if a one shot timer turned on the SSR for a one second period. This would eliminate transformer problems and give the controller time to evaluate the load. If the Midnight doesn't have the ability to do this now, it ought to in the future. .

bart · « **Reply #80 on:** January 24, 2013, 05:49:58 AM »

Relating to the Magnum MS4448PAE and the flickering light issue. Have the same issue with a 110v toaster oven. Just did a test with both analog and digital volt/amp meters. On both meters voltage is rock steady. On the analog amp meter, it does pulse with the lights. Not much, maybe a 1/4 amp or less.
Did email Magnum about this and they quit responding after their second response. No solution. Have a feeling they know this is a problem, just don't want to fess up to it.
Did find out later that when on utility power and have a light on the same circuit as the oven, the flickering is evident, but not as bad and does not effect all the lights as it does on the inverter.
Not a electronic guru at all, but my gut tells me that this has more to do with electronics/transformer than a load issue.

ChrisOlson · « **Reply #81 on:** January 24, 2013, 11:09:14 AM »

Quote from: bart on January 24, 2013, 05:49:58 AM

Did email Magnum about this and they quit responding after their second response. No solution. Have a feeling they know this is a problem, just don't want to fess up to it.

Our friends aren't real happy about it. They're wishing they would've spent the extra money on a XW4548 instead.
--
Chris

halfcrazy · « **Reply #82 on:** February 23, 2013, 08:57:04 AM »

Magnum has some REAL slow regulation. I complained early on in the AE series. Serial number 5 or something and they sped it up a touch but refuse to go any more. This can be an issue for a lot of people and drove me nuts living offgrid. I went from the SW's to the Magnum and was not very happy. I then went to a pair of outbacks and it is better but I sort of miss my SW's

Ryan

Treehouse · « **Reply #83 on:** February 26, 2013, 12:00:30 PM »

What about rectifying the 120v ac from the inverter to dc, then pwm to a opto isolated fet? or high speed ssr. Would this not evenly load the output transformer? Any ideas if this would work or be a better idea than doing it on the a/c side?

ChrisOlson · « **Reply #84 on:** February 26, 2013, 09:13:12 PM »

It seems to require a -3.0V offset to get full power at my heating elements once the voltage gets to the charge stage set point. Is that normal?
--
Chris

gww · « **Reply #85 on:** March 17, 2013, 07:20:46 PM »

I am assuming after watching chris's vidios that the ssr divertion mode only works on a resistance load. It would not work with a heat pump type water heater or air conditioner due to the cycleing and varible voltage/current.

Am I correct in this assumtion?
Thanks
gww

ChrisOlson · « **Reply #86 on:** March 17, 2013, 10:39:07 PM »

That would be a correct assumption. The SSR with PWM drive infinitely varies the voltage to the load. So you need to use a load that can handle variable voltage.
--
Chris

gww · « **Reply #87 on:** March 18, 2013, 05:06:23 AM »

As always "THANKS" Chris O.
gww

ruarangifarm · « **Reply #88 on:** September 23, 2015, 06:49:21 PM »

Hi OperaHouse

It has been some time, since this discussion.

Quote from: OperaHouse on October 03, 2012, 05:27:30 PM

.....
Just want to make people aware that PWM into a AC SSR can have some very strange results.

Maybe there has been some more research done on that?
Could you explain me more of what you know about it? 'PWM into a AC SSR can have some very strange results....'
I'm about to start my work to get my 'opportunity load' working to heat hot water...

Cheers
Ben

joestue · « **Reply #89 on:** September 23, 2015, 08:02:32 PM »

Quote from: ruarangifarm on September 23, 2015, 06:49:21 PM

Hi OperaHouse

It has been some time, since this discussion.

Quote from: OperaHouse on October 03, 2012, 05:27:30 PM
.....
Just want to make people aware that PWM into a AC SSR can have some very strange results.

Maybe there has been some more research done on that?
Could you explain me more of what you know about it? 'PWM into a AC SSR can have some very strange results....'
I'm about to start my work to get my 'opportunity load' working to heat hot water...

Cheers
Ben

most ac SSRs are just optically coupled and a circuit to trigger the triac.
there are optically triggered triacs but i don't know if they are making their way into ssrs yet.

so if you feed pwm into them, several things can happen, it may trigger randomly, because the on time duration is too short to trigger the scr completely.
if this is the case you've basically made a random noise generator.

it may trigger properly.

if it triggers properly, what the load ends up seeing is the interference pattern between the two frequencies.
the SSR turns off when the current collapses to zero. --so your minimum on time can range from 0 to 8.3 ms for a 60 hz waveform.

if it triggers properly, no matter what the pwm percentage is, the ssr's output will basically vary from full load to no load, according to the difference in frequency.

the duty cycle or pwm percentage will affect the output but without simulating it, i couldn't tell you what that looks like.

ruarangifarm · « **Reply #90 on:** September 23, 2015, 10:58:42 PM »

I understand those thoughts. Maybe not 100% but close.
I would like to outline the idea I'm following. SSR is not equal SSR, and so on.
I tend to use a 'phase angle controller / SSR'. The idea for that I picked up here:
http://scoraigwind.co.uk/?attachment_id=2485#comment-82016
I bought a MCPC2450C, this uses a analogue 0-10VDC input to regulate the phase angle of the load (element in hot water heater) This then should hopefully reduce the 'stress' posed on the inverter, and reduce 'noise' and flickering lights.
Now, we need somehow to 'smoothen' the PWM from the AUX output on the charge controller (in my case an FM80) to a more or less analogue signal for the SSR. The RC components shown in the picture in that link above, should do that.

Our background: off grid, 3kW PV, 3kW/240V Inverter, 620Ah at 48V, power system in shed, 'mains cable' from shed to house, Hot Water Cylinder with 2kW/240V elements, no DC cable buried between house and shed, although at the moment there is a phase spare in the 'mains cable'.

Would this work?
What to look for?
How could I improve this?

I'm a mechanical Engineer with some electronic knowledge and tools, and I probably could borrow a scope from a mate, but I'm not an electronic engineer as there are some posting in this forum.
Any comments and suggestions much appreciated!!
Cheers

joestue · « **Reply #91 on:** September 24, 2015, 12:21:19 AM »

Quote from: ruarangifarm on September 23, 2015, 10:58:42 PM

I bought a MCPC2450C, this uses a analogue 0-10VDC input to regulate the phase angle of the load (element in hot water heater)

I suppose I shouldn't be too surprised such products are available.. i guess next we'll see I2C SSRs next.

Whatever it takes to get the 0-10 volts from the pwm should be fairly straightforward. RC or level shifting may be required. if you need a zenar then it might not be linear.

dgd · « **Reply #92 on:** September 24, 2015, 12:36:58 AM »

Rua,

So is it because the HWC is too far away from the DC in shed that prevents you considering a DC element with Fet SSR pwm'ed by the controller.
A DC ssr would be much happier dealing with the pwm control

dgd

ruarangifarm · « **Reply #93 on:** September 24, 2015, 02:12:24 AM »

"I suppose I shouldn't be too surprised such products are available.. i guess next we'll see I2C SSRs next."
can I read this as something positive? Well it should reduce the strain on the inverter, no?
What is linear shifting? If not with a zenar what would you suggest? The output on the FM80 is 200Hz, 0-12VDC. That is why there is a zenar, I guess I would use a 10VZ, not a 9V as in the diagram.
I do understand the basics from it, but I couldn't just draw something up. If you could draw up a sketch, that would be awesome!

Yes, DC in shed too far away from HWC......
When you build all from scratch in a green paddock, you just can't think of everything in advance, you have to make a start somewhere, and take it from there.... that's what I'm doing now.

oztules · « **Reply #94 on:** September 24, 2015, 06:03:37 AM »

I think your scoraig link will work pretty much as you hope it will... may need to play with the 2kr or the 22uf or both.....

Assume you will modulate the extra wire in the cable for the heater input active.

With only 3kw of solar, it will be unlikely you will need to use the upper end of the 10v scale very often if at all, so if the dynamic range is not quite there, it may not be a practical difficulty.... actually a blessing for the little inverter.

If you have solar hot water, then this system may not be working a lot of the time in summer, and not enough in the winter..... air con is easy.with solar........ hot water is counter seasonal..... bummer really.

Unless you have high usage and small storage, I found that a simple timer between 10am and 4pm works fine. You don't want it turning on at night if you can help it. Here in the spring and beyond, by 9am, batteries are full ( 90%) from the night before, if it is clear, 12 if not.... ( I can dial up to 15kw of panels if I need to on a darker day... it helps...a lot...)

If you normally shower at night, you can leave it later for turn on if necessary... maybe 2pm - 5pm etc... every ones needs are different.

................oztules

OperaHouse · « **Reply #95 on:** September 24, 2015, 04:50:33 PM »

I must say that I have enjoyed all the hot water this year from PV solar diversion. I have a 36V string that will produce about 52V at power point. Frankly I don't know what the power point is. That value just stuck in my head from from some test I don't recall. It doesn't really matter much from a practical standpoint. All of my controllers operate at a fixed power points and that is how what goes where is decided. At some point I will track with panel temperature. With the camp falling apart and solar working beyond expectations, updating the software has been low priority. With the new Power Shed I am starting to clean up the controls.

How does this matter to you if you own a Classic? You can calculate what the power point should be or just measure panel voltage with a decent load. Instead of using the Classic diversion output let a micro monitor the panel voltage. When the panel voltage rises over the power point voltage divert power with a PWM to duty cycle a FET from the PV DC. I suggest a diode to isolate the capacitor bank. This also works with the older PWM controllers and the diode isolation is a must. PWM controllers rely on the current limiting properties of solar panels. So anytime a PWM controller is in the off duty cycle, that current can go to heating water.

ruarangifarm · « **Reply #96 on:** September 26, 2015, 05:25:08 AM »

Thank for those constructive comments.
Yes, I got some different R and uf values to play with.
PV: there is provision for another string of 750W of panels. Totals to 3.75kW is close to limit for the FM80@48V. There are actually 2 pairs of PV cables spare at the moment....so 1 pair could not be used anyway....may be I should use it to heat water only or have it available for cloudy days.
I got some spare cat 5e cables to the house, so I can either have the SSR set up in the power shed (and feed it through the spare phase) or have it right at the hot water cylinder.
Our hot water cylinder is 425L (112gal) rather tall. In winter its heated by the stove/cooker. I will hook up some hot water collectors ( I have to build them my self first) to it. That has best efficiency on lower water temps. Let's say it preheats the water to 50deg C (122degF) that doesn't need much for the PV to heat it to the save/desired hot water temp.
There are 3 elements in the cylinder (in different heights) with the first power the top element heats the top, then the next down element is turned on, when the top reaches desired temp.
Like that we should get enough hot water with only medium to little sun. When the whole boiler is full of hot water it will last us 3 days easy without any heat input.
Bit off the track, but it might help to understand our situation, and food for thoughts.

OperaHouse
In another post you wrote: "the cheapest way is with a UNO, a FET and some capacitors."
What is a UNO? I guess you are mentioning the same scenario here?
So that would be a PWM DC Voltage to heat hot water, right?
I don't have that cabling in the ground....don't really see me digging it all up again.

DamonHD · « **Reply #97 on:** September 26, 2015, 06:10:00 AM »

Arduino UNO: https://www.arduino.cc/en/Main/arduinoBoardUno

OperaHouse · « **Reply #98 on:** September 26, 2015, 03:58:07 PM »

My entire house system runs on a $6 UNO clone. I started programming in the 70's and until the UNO came along micro programming was pretty unfriendly. You are online so you have the computer. Just need $6 now and to download the free development system. For other things I do the UNO is too big and I use the stripped down version that is less than $3. It has 6 PWM output pins and 6 analog inputs. My system uses 2 for water heater elements, 2 for charge controllers, one for a good size relay ( It closes with full power and then drops the hold current to about 40%). Fancy code isn't needed for a really impressive system, just one line IF statements. I'm in the northern regions and the wife today was complaining about how hot the water is. Only 900W is used for heating water and that provides power for the fridge and everything else. 100% of what the panels can produce is always being used somewhere.

News:

Author Topic: Water Heating Controlled by a Classic 150 (Read 45642 times)