How about a solar car port on a budget?

[Muhendis]

1 hour ago, Crossy said:

There's certainly debate in the solar circles about whether it's actually worth bothering with. I probably wouldn't have done it if the Sofar inverter actually behaved as the manual says when set to "default" battery type rather than seemingly just being fixed as lead-acid.

Personally I see three separate, independent units which have no need to complicate things by communicating with each other.

The charge controller who's sole purpose is to try to put power into the batteries at a rate determined by the battery type.

The inverter which simply converts incoming dc to the required ac voltage.

The BMS which is there to give electronic protection to the batteries, balance them and report on individual cell state.

Each of these units can have the capability of letting you know if there is a failure. They don't need to control each other.

[SomchaiDIY]

3 hours ago, Muhendis said:

Personally I see three separate, independent units which have no need to complicate things by communicating with each other.

The charge controller who's sole purpose is to try to put power into the batteries at a rate determined by the battery type.

The inverter which simply converts incoming dc to the required ac voltage.

The BMS which is there to give electronic protection to the batteries, balance them and report on individual cell state.

Each of these units can have the capability of letting you know if there is a failure. They don't need to control each other.

Solar product market is the big mess all maker use different method  no standard for date transfer from system parts

 

 

Some reason for exchange data between bms and charging system

 

Temperature alarms

Limit current charge

Limit current discharge

Program logic in bms can control charge system direct overide charger logic

bms command battery type no need to set battery parameter at charger

 

I sometime read info at small private thailand electronic group they use popular mr pittaway diy bms system

 

The diy bms is good place for learning of how battery data share over system

The diy bms is esp32 base + now has the rs485 + canbus on custom control pcb

diy bms system is good for custom tune of second hand battery for people with c++ and the tech knowlege

 

https://github.com/stuartpittaway

 

[Muhendis]

16 minutes ago, SomchaiDIY said:

Temperature alarms

Limit current charge

Limit current discharge

Program logic in bms can control charge system direct overide charger logic

bms command battery type no need to set battery parameter at charger

Agreed but all these snippets of information are dealt with locally and alarms sounded.

There is no reason to share any of this between other units.

 

For example.

If the inverter starts to overheat it will control it's own problems and could shut itself down. The charge controller doesn't care about that and neither does the BMS.

Likewise any problem with over current from/to, state of charge of individual cells and over/under voltage of the batteries is controlled by the BMS. The inverter and the charge controller don't need to know. The inverter is only interested in whether there is enough voltage to run it's output.

Similarly the charge controller is only interested in filling the batteries. It doesn't need to know if one cell, for example, is not performing so well. That is the job of the BMS.

Communication between all functions might be a nice to have but it is really unnecessary.

[SomchaiDIY]

2 hours ago, Muhendis said:

Agreed but all these snippets of information are dealt with locally and alarms sounded.

There is no reason to share any of this between other units.

 

For example.

If the inverter starts to overheat it will control it's own problems and could shut itself down. The charge controller doesn't care about that and neither does the BMS.

Likewise any problem with over current from/to, state of charge of individual cells and over/under voltage of the batteries is controlled by the BMS. The inverter and the charge controller don't need to know. The inverter is only interested in whether there is enough voltage to run it's output.

Similarly the charge controller is only interested in filling the batteries. It doesn't need to know if one cell, for example, is not performing so well. That is the job of the BMS.

Communication between all functions might be a nice to have but it is really unnecessary.

Because limit of english some word here I copy from tech document translated to thai

If module has no data from next module this is ok because module can take care itself if has problem

 

Why share module data if this true

 

Some search for the charging algorithms of the lithium battery can explain why bms data is use on many system at charging and control

 

Charge discharge learn process can have more effects if more info available to the calculation over the time

 

Auto learning is big step in modern charge control system to be sure of good battery life and performances

 

Also all data on a share bus is advantage for system and final user viewing or transmit over other network

 

[Crossy]

I like to think of it as "communication" rather than "control".

 

The more information about other system elements each component has, the better it can regulate its own performance.

 

For example, I've noticed that when our weak cell triggers the "over voltage warning" for that cell (pack voltage is still within range) the inverter tempers its charge current to allow the balance system to catch up with the over volt cell whilst still adding some charge to the other cells.

 

Now just how much difference this information exchange makes overall is debatable. There are many systems in operation which do just fine with each component operating independantly.

 

But your solar setup is a "system" made up of component parts, it's (usually) better if it behaves as a system rather than a bunch of parts.

[Muhendis]

1 hour ago, Crossy said:

I like to think of it as "communication" rather than "control".

 

The more information about other system elements each component has, the better it can regulate its own performance.

 

For example, I've noticed that when our weak cell triggers the "over voltage warning" for that cell (pack voltage is still within range) the inverter tempers its charge current to allow the balance system to catch up with the over volt cell whilst still adding some charge to the other cells.

 

Now just how much difference this information exchange makes overall is debatable. There are many systems in operation which do just fine with each component operating independantly.

 

But your solar setup is a "system" made up of component parts, it's (usually) better if it behaves as a system rather than a bunch of parts.

I had a look at your Seplos balancer and this is what I found

https://www.seplos.com/sale-13671689-24-48v-8-16s-50a-smart-solar-lifepo4-bms-monitoring-system-lithium-ion-battery-management-system.html

Current Balance

Passive balance

150mA(Max)

 

Am I looking at the right thing?

If I am then it is small wonder the charge controller needs to be attenuated for the balance to do its job.

150mA is really not enough and passive? Ouch!

 

[Crossy]

Yeah, the built-in passive balancer is pretty carp, hence I have a 5A "flying capacitor" active balancer to help it along and that definitely helps in maximising the available capacity. My other pack which still has the original Daly BMS stays in balance pretty well without help.

 

Anyway, as I said, many systems work just fine with no comms, evidently yours is one of them ???? 

 

Sadly the Sofar inverter really needs to talk to the BMS (I think it gets lonely), maybe, if I'd known about the issue earlier I wouldn't have bought the Sofar.

 

The jury is still out, but the increased charge cut-off voltage and the more "sympathetic" interaction whilst balancing does appear to have squeezed a little more (about 10%) total capacity out of the pack with the BMS.

 

[BritManToo]

1 hour ago, Crossy said:

Yeah, the built-in passive balancer is pretty carp, hence I have a 5A "flying capacitor" active balancer to help it along and that definitely helps in maximising the available capacity. My other pack which still has the original Daly BMS stays in balance pretty well without help.

Same with my 2 battery packs, one works fine with just the Daly BMS, the other was hopeless until I put a 5A flying cap across 8 of the 16 cells.

[Muhendis]

2 hours ago, Crossy said:

Sadly the Sofar inverter really needs to talk to the BMS (I think it gets lonely), maybe, if I'd known about the issue earlier I wouldn't have bought the Sofar.

If I remember rightly, you got a good deal with Sofar so you are probably still ahead. 

Will the flying cap balancer work with the Seplos connected (just coms) or does the coms need a fully functioning Seplos balancer to be able to communicate. I am thinking that unless there is separate monitor wiring that may be an issue.

[Muhendis]

2 hours ago, Crossy said:

Anyway, as I said, many systems work just fine with no comms, evidently yours is one of them

????

My "system" is an antique with Gel batteries and is completely and totally dumb.

Maybe each unit sits there talking to itself but whatever, there's less things to go wrong and no bytes to worry about and believe me, at my time of life that's a plus.

[carlyai]

11 hours ago, Crossy said:

Yeah, the built-in passive balancer is pretty carp, hence I have a 5A "flying capacitor" active balancer to help it along and that definitely helps in maximising the available capacity. My other pack which still has the original Daly BMS stays in balance pretty well without help.

 

Anyway, as I said, many systems work just fine with no comms, evidently yours is one of them ???? 

 

Sadly the Sofar inverter really needs to talk to the BMS (I think it gets lonely), maybe, if I'd known about the issue earlier I wouldn't have bought the Sofar.

 

The jury is still out, but the increased charge cut-off voltage and the more "sympathetic" interaction whilst balancing does appear to have squeezed a little more (about 10%) total capacity out of the pack with the BMS.

 

What would you now buy instead?

[Crossy]

16 minutes ago, carlyai said:

What would you now buy instead?

 

In reality probably the Sofar again, getting a BMS that can talk to the beast isn't that much more expensive than getting a "smart" BMS that can't, I just didn't know about the inverter "bug" at the time. It's still a nice solid piece of kit and is one of the few hybrids on the PEA approval list.

 

My requirements were pretty strict, it has to be IP65 and fanless due to the location our inverters live in (being silent isn't actually an issue but it's nice to have).

 

If I was doing it all over again with a decent budget rather than going bitsa-expansion I might go for a big off-grid hybrid system, say 20kW (4 x 5kW inverters in parallel) and run as a whole-house UPS as well as solar. No export of course, but being DIY that's not officially available anyway.

[Crossy]

Simple 3mm acrylic cover installed, I feel rather more comfortable about the works now.

 

Toothpicks bottom left are to operate the LCD buttons through the small holes, I won't bother with the LCD if I get any more of these BMS's, the PC interface is pretty good.

 

Still debating whether to give it sides, the overall enclosure is rain-proof, but ...

 

[Crossy]

OK, top and sides added.

 

Notes to self:-

• Acrylic solvent adhesive allows some adjustment whilst it sets.
• Acrylic solvent adhesive does not allow adjustment if your carefully positioned parts move and then it sets.
• If you get acrylic solvent adhesive on your fingers it will leave fingerprints.
• Acrylic solvent adhesive fingerprints don't polish out.

Nowhere near perfect, but it will do the job and it's hidden 99.9% of the time anyway.

 

[Muhendis]

Does that Seplos BMS get hot?

If so, when will you be fitting the fan? 

[Crossy]

45 minutes ago, Muhendis said:

Does that Seplos BMS get hot?

If so, when will you be fitting the fan? 

 

Incredibly I did think of that!

 

Like most BMS's the Seplos FETs run as either hard on or hard off, both of which are low dissipation modes. The heatsink hardly gets above ambient but I will of course monitor the beast from the logs (it also looks after itself if things get too warm). The unit I have is the 100A version, the 200A has the same heatsink.

 

It's also worth noting that when this BMS is installed in their rack batteries it gets less airflow than my simple cover permits.

 

 

[BritManToo]

Only 44 units this month, 194bht.

[Crossy]

The numbers for June.

 

[Crossy]

The July numbers are in.

 

We didn't quite crack the 1MWh barrier this month but there have been a number of really yukky and wet days.

 

No developments on the hardware front, it just sits on the car-port roof and makes "free" power.

 

 

[Pink7]

On 6/21/2022 at 4:31 PM, Crossy said:

 

If I was doing it all over again with a decent budget rather than going bitsa-expansion I might go for a big off-grid hybrid system, say 20kW (4 x 5kW inverters in parallel) and run as a whole-house UPS as well as solar. No export of course, but being DIY that's not officially available anyway.

4 x5kw all in one? how much panel capacity you would connected to i?

 

Pink

[Crossy]

10 minutes ago, Pink7 said:

4 x5kw all in one? how much panel capacity you would connected to i?

Pink

 

I'd seriously consider the Growatt SPF 5000 ES, two,  maybe three of them in order to meet house peak load.

 

Going from what we use from PEA and what our current system generates then - Panels around 12kWP and 25kWh of storage.

[Pink7]

47 minutes ago, Crossy said:

 

I'd seriously consider the Growatt SPF 5000 ES, two,  maybe three of them in order to meet house peak load.

 

Going from what we use from PEA and what our current system generates then - Panels around 12kWP and 25kWh of storage.

Looks to be a good unit. I see its about the double price of the cheapest 5kv units, so i guess its a high quality product with some support and warranty/service etc?

 

Buying the cheapest stuff in some cases can be very expensive, Im all in to pay some more if the value, quality and life time is there.

 

 

[MJCM]

@Crossy

 

What is the angle of the roof on which your solar panels are mounted?

 

And can you change the Angle of the Panels?

 

We can have the panels mounted on our Water Storage shed, but the roof is at a Wrong angle (and much steeper then yours). If we put the Panels in a Horizontal line they would be more facing to the W, if we would put them in Vertical line they would be facing south but thus in an angle to the W.

 

My idea now is to have a structure build on top of the Shed to even out the angle and then on top of that a structure to adjust the panels. Or even have the panels lay flat but fully facing South.

 

[Crossy]

On 8/6/2022 at 2:36 PM, MJCM said:

What is the angle of the roof on which your solar panels are mounted?

 

Sorry MJCM, I totally forget about this post.

 

Some quick measurements and schoolboy trig. gives about 3.2^o of slope, panels are parallel with the roof and not adjustable.

 

We are about 14^o N so the panels are a bit flat for the classic "use your latitude" but they are working just fine.

[Pink7]

On 8/13/2022 at 2:57 PM, Crossy said:

 

Sorry MJCM, I totally forget about this post.

 

Some quick measurements and schoolboy trig. gives about 3.2^o of slope, panels are parallel with the roof and not adjustable.

 

We are about 14^o N so the panels are a bit flat for the classic "use your latitude" but they are working just fine.

What distance  between the roof and the panels you recommend? 

 

Pink

[Crossy]

7 minutes ago, Pink7 said:

What distance  between the roof and the panels you recommend? 

 

The "real" mounting kits put 2-3 inches (50-75mm) between the panels and the roof.

 

Some air flow is good, it helps keep the panels and roof cool, but too big a space could cause issues in a big wind.

[Pink7]

38 minutes ago, Crossy said:

 

The "real" mounting kits put 2-3 inches (50-75mm) between the panels and the roof.

 

Some air flow is good, it helps keep the panels and roof cool, but too big a space could cause issues in a big wind.

ok i suggested 20cm and my brother in law who will weld and mount this suggested 30cm. How much clearance you have?

 

Pink

[Crossy]

5 minutes ago, Pink7 said:

ok i suggested 20cm and my brother in law who will weld and mount this suggested 30cm. How much clearance you have?

 

It's about 40mm from the top of the roof "hump" to the underside of the panel surround.

 

This is my prototype mount. Panels sit directly on the top piece.

 

[007 RED]

@Pink7....  FYI..

I adopted a slightly different approach for mounting my panels.  I don't have access to a disc cutter, pedestal drilling machine  or welding equipment, so I had to think of an alternative way to mount the panels onto the corrugated cement tiles of the car port roof.

 

I opted to a 'bridge' type mounting solution as shown below.

 

The mounts are made from 1.2mm stainless steel which are bolted down to the metal structure of the car port roof.  They save a lot of weight.  Made for me by a local tinsmith from off cuts @ 50 THB each.

 

A small coat of clear silicon was applied to the underside of each bracket to help prevent any water ingression via the mounting bolts. 

 

Great advantage of stainless steel is that it won't rust and doesn't need painting = less maintenance.

Been in service for 10 months and no problems with wind.

[Pink7]

On 3/2/2022 at 4:32 AM, Crossy said:

Numbers for February, a short and variable month with a low daily production of 15.7kWh and a high of a very satisfactory 43.9kWh. 

 

Total solar for the month is 759kWh whilst we used 228kWh from PEA.

 

This is a average month? You have around 10kw panels? 

 

Pink