diy solar

diy solar

Sol-ark - eating their own dog food?

Llaves

New Member
Joined
Dec 27, 2021
Messages
204
Location
Llaves, NM
("Eat your own dog food" is an expression meaning for a company and/or it's employees to use their own products).

Maybe it's just me, but it seems like nobody at Sol-Ark, at least in a decision-making position, has direct experience with living with an off-grid system. I've had my 12K for about 6 months now and have missed some features that were present in my Outback system 15 years ago. The most notable is that the 12K does not generate any sort of alarm, fault, or warning if there is a gen-start failure. When I brought this up with tech support the response was along the lines of, "yeah, there's no warning". I submitted a feature request, which I would have thought would get a "oh god, how did we miss this" sort of response, but instead got a generic "thanks for your input". In a similar vein, with the Outback system I could start a generator charge cycle, put it in auto-mode, and the generator would shut down when the batteries reached the programmed SOC. The 12K does not have this feature. If you start the generator with Gen Force, it stays on until you turn it off. There's no way to set it to turn off once it reaches the programmed SOC for a normal generator cycle. (Or so I've been told by tech support.)

So, is it just me, or have other off-grid owners found Sol-Ark to be a bit clueless about life in an off-grid world? Seems like they need a user advisory panel to help them prioritize feature requests.

BTW - I want to be clear this is not a generalized complaint about their tech support staff. For the most part they are knowledgeable and timely in responding to questions. The problem exists at a higher level of the company in my view.
 
Is solark designed around to being off grid?or more of grid support?

As you found out each design has its strong points and weakness.
 
Is solark designed around to being off grid?or more of grid support?

I'm not sure any brand is more focused on off-grid than grid-tie. (Maybe someone reading this knows of one). But most brands seem to claim both grid-tie and off-grid capability. Clearly the grid-tie market is much larger.
 
("Eat your own dog food" is an expression meaning for a company and/or it's employees to use their own products).

Maybe it's just me, but it seems like nobody at Sol-Ark, at least in a decision-making position, has direct experience with living with an off-grid system. I've had my 12K for about 6 months now and have missed some features that were present in my Outback system 15 years ago. The most notable is that the 12K does not generate any sort of alarm, fault, or warning if there is a gen-start failure. When I brought this up with tech support the response was along the lines of, "yeah, there's no warning". I submitted a feature request, which I would have thought would get a "oh god, how did we miss this" sort of response, but instead got a generic "thanks for your input". In a similar vein, with the Outback system I could start a generator charge cycle, put it in auto-mode, and the generator would shut down when the batteries reached the programmed SOC. The 12K does not have this feature. If you start the generator with Gen Force, it stays on until you turn it off. There's no way to set it to turn off once it reaches the programmed SOC for a normal generator cycle. (Or so I've been told by tech support.)

So, is it just me, or have other off-grid owners found Sol-Ark to be a bit clueless about life in an off-grid world? Seems like they need a user advisory panel to help them prioritize feature requests.

BTW - I want to be clear this is not a generalized complaint about their tech support staff. For the most part they are knowledgeable and timely in responding to questions. The problem exists at a higher level of the company in my view.
Sol-Ark only had off-grid products until the 8K release. However, the 3K 24V/48V systems had limited generator connectivity.
Since ~spring 2020 TOU can be used to set Generator cutoff Voltage/% when using the two-wire start. Have all the charge boxes checked and set the stop %/V and wattage in the other columns. Recommend 70-85% for a good balance.
Prior to that, the inverter would only charge the battery until it would no longer accept 5% of the programmed Ah capacity.
GEN Force is for manual control/ testing of the dry contact relay (2-wire start).
There is a software update in the works across 5K-15K that updates generator setup and programming functions. Since it's lumping together several different but related generator requests it may not be until Q2-Q3 for its availability.
Regarding a generator start failure alarm, implementation is likely in the works but waiting until the server migration of our monitoring platform PowerView (PV Pro) is complete.
 
thanks for this update- sounds like some useful changes are in the works. Is this information available anywhere on the Sol-Ark website? I think it would be really useful if there was a page to post feature/enhancement requests, for users to vote on how important they are to them, and for Sol-Ark to list their prioritization.

While it's nice that the TOU can be used to set an upper limit on charging, it seems a rather baroque way to achieve a rather straightforward goal.
 
I'm not sure any brand is more focused on off-grid than grid-tie. (Maybe someone reading this knows of one). But most brands seem to claim both grid-tie and off-grid capability. Clearly the grid-tie market is much larger.

The SMA Sunny Islands are very much focused on off-grid.. as in "completely off grid on an island". (hence the name)

My Sunny Islands can be programmed when to start the generator (SOC) and what SOC to stop the generator, and if I wanted an alarm, I believe one of the aux relays could be programmed to sound a bell or whatever.
 
thanks for this update- sounds like some useful changes are in the works. Is this information available anywhere on the Sol-Ark website? I think it would be really useful if there was a page to post feature/enhancement requests, for users to vote on how important they are to them, and for Sol-Ark to list their prioritization.
I have also been told that they are limited in making App changes because the requests need to go through a third party.
Once the My Sol-Ark app is fully in use they can do anything they want.
While it's nice that the TOU can be used to set an upper limit on charging, it seems a rather baroque way to achieve a rather straightforward goal.
Not really. It seems that way if you have a small or just adequate battery pack but if your using a 100KWh rack of batteries you probably only want the generator to charge them up so that they have enough power to get you through the night. Who knows, the next day your PV luck may change.
Why run a generator to get to a huge bank up to 100% SOC and then find out in the morning that it's a bright sunny day and all that fuel was wasted.
 
Not really. It seems that way if you have a small or just adequate battery pack but if your using a 100KWh rack of batteries you probably only want the generator to charge them up so that they have enough power to get you through the night. Who knows, the next day your PV luck may change.
Why run a generator to get to a huge bank up to 100% SOC and then find out in the morning that it's a bright sunny day and all that fuel was wasted.
This is exactly the use case behind my request for the ability to change the charging limit. I'd like to set the start SoC to something like 25%, but set the end SoC (not currently possible) to 50%, which is probably enough to get me through the next 24 hours if the PV is a total bust. In a perfect world we'd have a more complex decision algorithm that might look at the time of day and the PV output so that if the PV is looking good and it's early in the day, it would just skip the generator altogether. Now add a hyper-local irradiation forecast to the mix and think of what you could do. But for now, I'll just settle for a Soc criterion for stopping rather than the fixed 95%.

I've never tried the TOU interface, but it seems like you should be able to achieve this goal using TOU, which allows setting start/stop Soc. Let's say we want 25% start, 50% stop. Then we set *every* time period to 25% start, 50% stop. While a pain in the butt, wouldn't that work the way we're asking for?
 
This is exactly the use case behind my request for the ability to change the charging limit. I'd like to set the start SoC to something like 25%, but set the end SoC (not currently possible) to 50%, which is probably enough to get me through the next 24 hours if the PV is a total bust. In a perfect world we'd have a more complex decision algorithm that might look at the time of day and the PV output so that if the PV is looking good and it's early in the day, it would just skip the generator altogether. Now add a hyper-local irradiation forecast to the mix and think of what you could do. But for now, I'll just settle for a Soc criterion for stopping rather than the fixed 95%.

I've never tried the TOU interface, but it seems like you should be able to achieve this goal using TOU, which allows setting start/stop Soc. Let's say we want 25% start, 50% stop. Then we set *every* time period to 25% start, 50% stop. While a pain in the butt, wouldn't that work the way we're asking for?
I don't have a generator hooked to my Sol-Ark so I have never actually used TOU with one.
It would seem to me that just using one TOU slot at for example 5pm and telling it to Gen charge to 50% will guarantee that if your PV has not bough the batteries up to 50% by 5pm the Generator will start and charge them to 50%.
 
I don't have a generator hooked to my Sol-Ark so I have never actually used TOU with one.
It would seem to me that just using one TOU slot at for example 5pm and telling it to Gen charge to 50% will guarantee that if your PV has not bough the batteries up to 50% by 5pm the Generator will start and charge them to 50%.
No, TOU works much like the more general Gen Charge parameter in that you specify the threshold at which charging starts - the SoC has to have dropped *below* the threshold, otherwise nothing happens. So for your example, the 5pm slot has two parameters - start and stop. You set the start to 25% just like you would the generic Gen Start SoC. But you also set a stop parameter, say 50%, which you can't do for the Gen Start setting. So, in your example, if start is 25% and end is 50%, then as long as PV has brought the charge above 25% when 5pm rolls around, the generator is not started, even if the battery is below 50%.
 
No, TOU works much like the more general Gen Charge parameter in that you specify the threshold at which charging starts - the SoC has to have dropped *below* the threshold, otherwise nothing happens. So for your example, the 5pm slot has two parameters - start and stop. You set the start to 25% just like you would the generic Gen Start SoC. But you also set a stop parameter, say 50%, which you can't do for the Gen Start setting. So, in your example, if start is 25% and end is 50%, then as long as PV has brought the charge above 25% when 5pm rolls around, the generator is not started, even if the battery is below 50%.
You have tried it and it failed?
I will take a look tomorrow.
 
Honest question: how realistic do you think it is for any manufacturer to be able to address all feature wish-lists, even in software?

I would kind of think you need external monitoring and control logic for much more than the basics-- the key is really that internally the system can report enough of the status information to find errors or failures. Just looking at the settings hell it seems like you need to be watching everything almost constantly, which is generally when supplemental monitoring comes into play.
 
Honest question: how realistic do you think it is for any manufacturer to be able to address all feature wish-lists, even in software?
Nobody is suggesting they implement everything every user asks for. I'm merely asking for more information about what is coming down the pike and suggesting they invite users to help prioritized possible new features. As the title of this thread suggests, how many Sol-Ark designers are living with an off-grid system? Without that day-to-day knowledge they're only guessing at what really matters to the end-user.

I would kind of think you need external monitoring and control logic for much more than the basics
In this particular case of a gen start failure alarm, I would consider this pretty basic.
 
My parents are off grid with 3x 12k inverters. The generator start control is handled by an independent system (runs on a PLC) that also acts as their battery monitor. I programmed in 3 types of manual operation modes - fixed runtime (minutes), charge to % level, and charge to full. There is also auto gen start options I programmed in, but they only leave the one enabled that keeps their lights from going off (starts at 46vdc and charges to 15%).

I could never imagine them using the built-in start logic on any inverter including the sol-ark because when living off-grid there are various situations where generator is needed. It has loss of phase and stall monitoring that sound alarms in a few places, and if startup fails, it retries 3 times before locking out and also sounding an alarm. It also increases the glowplug run time on the second retry in case it was exceptionally cold.

Fixed runtime gets used most often in winter if/when we are doing some fabrication like welding or plasma cutting and want to keep it running even though the batteries may be full. This is because we expect a continued heavy load and don't want to drag the batteries down just to bring them back up.

Charge to % is used most on winter mornings, where we know we may get some sun, but nowhere near enough to get the batteries full. Let's say we are at 10% at sunrise, but anticipate the solar can only generate enough to add 40% to the batteries - we would charge up to 60% with generator in the morning and let the solar finish the charging.

And then charge to full - this also used in a similar case as above, just in the evening time or of we got less power than anticipated... Let's say the generator didn't run in the morning - during the day we only got from 10% up to 50% - charge to full will bring them the rest of the way up so we never have to worry about the power going out at 3AM...

The power not going out at 3AM has a lot of peace of mind for me especially. One time when I was living on their property, I woke up at 3am because the power went out. I went outside to find one of the outbuildings engulfed in flames. When the main electrical feed was burning in this outbuilding, it shorted out the inverters causing them to shut down. The PTSD from waking up at night to this experience has made me a bit obsessive about making sure there is no way the power goes out at night, because the fear of waking up to another situation like this is pretty powerful.

I'm off topic now, but there are quite a few different situations in the off-grid lifestyle that require different amounts of generator intuition.
 
My parents are off grid with 3x 12k inverters. The generator start control is handled by an independent system (runs on a PLC) that also acts as their battery monitor. I programmed in 3 types of manual operation modes - fixed runtime (minutes), charge to % level, and charge to full...
What do you use for a PLC? Sounds like a slick system you built!
 
. The generator start control is handled by an independent system (runs on a PLC) that also acts as their battery monitor.
I considered an approach like this, but after 15 years with a system that only I could maintain, I decided that my new upgraded system should be as pure a commercial solution as possible. (I'm at that age where you take mortality into account.) Besides, since my Outback system had a generator start fault alarm over 15 years ago, it never occurred to me that such an obvious feature would be missing from a newer unit.
 
  • Like
Reactions: ENB
What do you use for a PLC? Sounds like a slick system you built!
They are the CLICK PLC from Automation Direct - however, the Productivity 1000 series is much more powerful in terms of features for slightly cheaper for analog IO. I use them for other things and they are very capable for the price, and can be expanded in the future.

They have become a bit harder to get with supply shortages, but the video tutorials on YouTube for these make it feasible to learn these as compared to something like an ABB or Siemens that requires what I call "proprietary training".

Also from Automation Direct, the C-More EA9 series HMI as a user interface. In the first iteration I used a bunch of buttons and indicator lights, but then I realized how powerful of a solution this was, realized that I would run out of front panel space in notime!

Some of this stuff is on my YouTube channel (HighTechLab)
 
Back
Top