Schneider SW4024 drop-outs with generator

Horsefly · August 2019

Since installing solar at our cabin in 2017, the generator gets run very seldom. In fact, my wife and I have never had reason to run the generator on our visits. However, both my brother and my sister and her husband have had to run the generator a couple of times. Both times were when the battery bank had gotten somewhat low. As a result the generator was both servicing loads in the cabin and charging the batteries.

My brother ran it in 2017, soon after the solar was up and running. He reported that periodically the lights seemed to flash briefly, and the SW4024 inverter/charger would make some clicking noise and humming would turn on and off. Fast forward to this past weekend, and my brother-in-law (sister's husband) reported something similar.

I've never seen any of this, so the specifics are a little sketchy. I've made a note to try and repeat the problem next time I am up there.

To me this sounds like the transfer switch relay is cycling, causing the power to drop while going between the generator and the inverter as a source.

I just noticed before I posted this a similar report on an old Xantrex SW4024 (https://forum.solar-electric.com/discussion/19912/xantrex-sw4024-drops-load-repeatedly-then-finally-holds). It sounded like the problem was a flakey disconnect on the DC side, and I am pretty certain ours is fine.

A couple of other notes:

We do have split phase 120VAC coming from the generator, and I know I checked it on both legs for both voltage and frequency before we hooked up the SW4024. Both looked fine without a load.
The generator is a rather cheap four year old Westinghouse rated for 7500W continuous, 9000W surge.
I've got the charger on the SW4024 set to a max charge current of 70% of the charger's max of 90A, so limiting the output of the charger to 63A. That only works out to something over 1800W.
I'm pretty sure I have disabled the load shaving and AC support capabilities of the SW4024, since we have no grid power.
I also am pretty sure I left the AC qual settings at the default, which says the SW4024 should accept between 95VAC and 135VAC, at between 55Hz and 65Hz.

I'd have to think the SW4024 is somehow deciding the generator power is no longer good enough, switching over to the inverter (and cutting of the charging), then after a few seconds it again qualifies the generator power, switches back again, and starts charging.

Has anyone seen anything like this, especially on a newer model Schneider?

Most of what I have for settings in the SW4024 are still at their defaults. Has anyone had to adjust the settings for use with a generator?

Estragon · August 2019

Any idea what, if any, AC loads the BIL was running in addition to battery charging? That they've needed the genny but you haven't suggests they may be running loads you don't?

Does the inverter do generator assist (adding inverted power to genny limit setting if needed)?

Horsefly · August 2019

Thanks @Estragon. Good questions.

Estragon said:

Any idea what, if any, AC loads the BIL was running in addition to battery charging? That they've needed the genny but you haven't suggests they may be running loads you don't?

There was really two things this time. There was really no direct sun while my BIL was up there this time, which he said ran down the batteries more than usual. However, he also had to run a compressor for a few minutes to test some new gas lines we are putting in. I get the impression (but need to explicitly ask) the drop outs happened with no real loads, but also when the compressor was running.

Estragon said:

Does the inverter do generator assist (adding inverted power to genny limit setting if needed)?

Yeah, that's what Schneider calls "load shaving" and "AC support". Load shaving limits the power taken from the grid / generator, and supplies anything greater from the inverter. AC support doesn't pull anything from the grid / generator as long as there is enough SoC in the batteries to supply the loads. I'm fairly certain that I disabled both of these features, but will check again when I go up there.

Horsefly · August 2019

So I did confirm with my BIL that the periodic dropouts were happening before, during, and after the air compressor was running. So it seems to not be related to having a big load on the system. The only other loads that were on were LED lights and a phone charger.

I guess I'm getting more suspicious of our generator. I need to check the voltages and Hz on both legs next time I'm at the cabin.

Dave Angelini · August 2019

Genset is the cause and if you have not adjusted the CSW settings, it is time.

Start adjust by lowering all settings, input breaker and charger percentage. Slowly raise them up.

Some gensets are pretty bad and if you have not used this one....

Horsefly · August 2019

Thanks @Dave Angelini. I'll do as you suggest.

We had a really great Honda generator for years. The engine blew a little over four years ago, right before my brother was taking his family up for a long weekend. This was not only before we had the solar, but before we had any idea we would put solar in. So I did the quick and easy solution, buying a cheap generator on Amazon and getting it two-day delivered to my sister's house (closest to the cabin) so it could get installed and be available for that weekend.

Whereas we all liked the old Honda generator, none of us like the replacement. It's much more noisy. The fact that we all hated the new generator was in fact one of the triggers for me to start looking into solar. So I guess it was a good thing!

Dave Angelini · August 2019

My brother in law can kill anything. Keep him away from it until it is adjusted. Great way to kill the CSW input relay

Horsefly · August 2019

Ok I have some updated info, but I'm still really confused.

I've been doing some remote diagnostics. Not the cool way with Conext Insight, but the old fashion way of my BIL trying stuff and texting me the results. To try and narrow down the problem, I had sent him several things to try. Without going into all of the details, here's what he found: Starting point is mid-day, with the Inverter up and providing power to some minimal loads in the cabin. The batteries are in float. The charge controller is up and running, but with not much sun, not much charging, and not many loads, there isn't much coming in from the CC.

BIL starts the generator, and watches the control panel readings for the AC L1 and L2 voltage and Hertz. They are not far off of 120VAC and 60Hz. The transfer switch kicks in and transfers loads to the generator, and the charger part of the inverter goes to float. However, after what my BIL says was 1-2 seconds, the transfer switch kicks back, so the inverter is running off the batteries again. After about 10 seconds, the AC In is qualified again, the transfer switch switches over to the generator, but again (after 1-2 seconds) it kicks out. As this was happening the voltages and Hz reported by the Schneider control panel never got out of the acceptable range. This cycle kept repeating, about every 10-15 seconds. In each cycle, the generator was not contributing to either the battery or the cabin loads except for 1-2 seconds out of every 15 seconds.

On a hunch, I asked him to turn off the breaker that provided PV power to the charge controller, and waited for the CC to shut down. He then started the generator and bingo! The inverter/charger qualified the AC In, the transfer switch connected the loads to the generator, and the charger part of the SW4024 started charging the batteries in float mode. This condition remained steady for as long as he was willing to leave it.

After several minutes, he turned on the breaker for the PV-to-CC, allowing the charge controller to re-enter the room. Immediately the SW4024 went back into the every 12-15 second cut out, leaving the generator really out of the equation.

So it appears that the generator is fine by the SW4024, as long as the charge controller is OFF. Once it is on, the SW4024 can't deal with the generator input, and keeps going back and forth between running off of batteries (10-12 seconds at a time) and re-trying to run off the generator (2 seconds at a time).

I'll have to check again when I get up there, but I am virtually certain that I had the bulk / absorption / float voltages set the same in the charge controller (Schneider 60/150) and the SW4024. I can't imagine that is the problem, especially with it all (and the system control panel) connected and communicating via the Xanbus.

For now, I guess if we need to give the batteries a boost and run the generator we just need to shut down the 60/150 charge controller. Obviously that isn't how it is supposed to work.

I did have my BIL check the firmware revision. The current one is 3.05 build 4 from August 2016. We bought it somewhere around December of 2016, so that makes sense. I see that in the next release of the firmware (3.07 build 3) one of the changes was "Improved generator qualification algorithm". With no more specifics, I have no idea if that could be related.

As far as firmware, we don't currently have a Schneider combox (or gateway), so we don't really have a way to update the firmware without buying some more components. Moreover, my EE experience is that you should never patch firmware unless you know it is fixing a known problem or vulnerability. The risk of turning the device into a brick is just too high. In this case, I can't tell.

I also don't know what firmware changes have happened in the 60/150 charge controller, although that device is simple enough I can't imagine it is the problem.

I know @mcgivor and I think @mike95490 have some Schneider components. Do you guys have any suggestions?

Thanks in advance for the expertise here!

mike95490 · August 2019

My XW, generator and other brand PV chargers, all work fine at the tail end of a rainstorm and the sun pops out, the generator unloads, but is still qualified and then when cloud cover comes back, generator lugs right down and starts grunting again.

Horsefly · August 2019

mike95490 said:

My XW, generator and other brand PV chargers, all work fine at the tail end of a rainstorm and the sun pops out, the generator unloads, but is still qualified and then when cloud cover comes back, generator lugs right down and starts grunting again.

Ugh. Thanks Mike. That's how I would expect things to work, just not how mine is working.

mcgivor · August 2019

My generator useage is limited but have never experienced what you described, the firmware of my unit is 3.01 so it's unlikely firmware is the cause. Based on the information it would seem as though there is a conflict between settings of the two components for the same parameter. Are there any fault codes on the SCP?

Horsefly · August 2019

mcgivor said:

My generator useage is limited but have never experienced what you described, the firmware of my unit is 3.01 so it's unlikely firmware is the cause. Based on the information it would seem as though there is a conflict between settings of the two components for the same parameter. Are there any fault codes on the SCP?

Well, knowing it can't be the firmware is a little bitter-sweet. Since I have no way right now to upgrade the firmware, it's just as well.

I'll have to check with my BIL about faults and warnings. He didn't mention any fault light coming on, but that doesn't eliminate the chance. Thanks for the idea.

mike95490 · August 2019

How is the generator loaded ? If one leg is loaded and the other not, that can cause an issue

Estragon · August 2019

Just to clarify, the CC turns OFF when the PV-CC breaker is opened? I would have thought it would stay on, but in some sort of resting state (same as at night).

mcgivor · August 2019

Horsefly said:

mcgivor said:

My generator useage is limited but have never experienced what you described, the firmware of my unit is 3.01 so it's unlikely firmware is the cause. Based on the information it would seem as though there is a conflict between settings of the two components for the same parameter. Are there any fault codes on the SCP?

Well, knowing it can't be the firmware is a little bitter-sweet. Since I have no way right now to upgrade the firmware, it's just as well.

I'll have to check with my BIL about faults and warnings. He didn't mention any fault light coming on, but that doesn't eliminate the chance. Thanks for the idea.

The fault codes are not a light but available in the menu of the SCP, if it weren't bedtime for me I would explain the process, it's not something I do on a regular basis so I would have to push a couple of buttons to find out

Dave Angelini · August 2019

The cc should stay on when solar input is off. There is no firmware upgrade needed.

This issue to me is a very basic problem. Back decades to when Trace first came out with an Inverter with a gen input.
I have not read the detail here so I could be wrong. I remember Steve, I think that when you first described this in a previous thread, there was no mention of the solar, correct?

If you had brought that in, it would have made this simple. You have to set the solar about .3V or so higher than the settings in the csw charger. This way the gen will back off and let the solar use its full output. Without doing this the scenario just keeps repeating like groundhog day.

There probably will not be any fault codes as the csw is doing what it is programmed to do. You just need to program it to be smart.

Horsefly · August 2019

mike95490 said:

How is the generator loaded ? If one leg is loaded and the other not, that can cause an issue

Both split phase legs from the generator go to the SW4024, and from the SW4024 to the split phase breaker box. There are lights and wall plugs on both legs from the breakers. We currently have both the kitchen plugs (Microwave oven) and the well pump on the same leg, which remains on my to-do list to fix. Aside from that, I think the loads are pretty balanced.

Estragon said:

Just to clarify, the CC turns OFF when the PV-CC breaker is opened? I would have thought it would stay on, but in some sort of resting state (same as at night).

Sorry to confuse. Yeah, that's what I guess I mean. I had him (BIL) turn off the breaker from the PV to the CC so that the CC would stop trying to charge the batteries. I specifically told him NOT to turn off the battery power to the CC.

Dave Angelini said:

The cc should stay on when solar input is off. There is no firmware upgrade needed.

This issue to me is a very basic problem. Back decades to when Trace first came out with an Inverter with a gen input.
I have not read the detail here so I could be wrong. I remember Steve, I think that when you first described this in a previous thread, there was no mention of the solar, correct?

If you had brought that in, it would have made this simple. You have to set the solar about .3V or so higher than the settings in the csw charger. This way the gen will back off and let the solar use its full output. Without doing this the scenario just keeps repeating like groundhog day.

There probably will not be any fault codes as the csw is doing what it is programmed to do. You just need to program it to be smart.

I'm not too sure what thread I might have brought this up before, but I'll trust that I did.

My brother reported (with very little detail) that there was some strange behavior from the inverter once before when he was running the generator, but I never saw it and didn't think much of it. I appreciate your expertise, and will try the 0.3V upper to the setting on the CC. I guess I'm surprised that nothing I've read from Schneider indicated that. From what I've seen, I would have just expected that the intended setting would be the same for both.

I'll post back here when I have some results, but I won't be able to get on-site until late Aug or early Sep. Thanks to you all for such great thoughts.

Dave Angelini · August 2019

As I said I do not remember the details here so forgive me if I said something that may be not accurate about the threads.
Schneider probably should do another service note like xantrex did and trace before them about this common problem.

I am sure Mike has done this with his setup, right? I do not use a genset here, but I help my clients do this right away after the installation. It is basic setup of an RE power system.

Since you are an engineer if you think about what is happening it makes sense right? Maybe a good place to start jokes about Engineering? Nothing personal just funny

Image: https://us.v-cdn.net/6024911/uploads/editor/c6/os8uo5udk1zy.png

Horsefly · August 2019

Horsefly · August 2019

I'm heading up to the cabin this coming week, and so was reviewing this thread and the manuals for the CSW4024 and the MPPT-60/150. I wanted to see if there was any guidance to setting the bulk / absorption / float settings between the two devices, as has been discussed here. In the manual for the MPPT-60/150, there is some discussion about the charge stages being synchronized automatically via communication on the Xanbus.

Synchronized charge states are active when more than one charging device (MPPT or Inv/Chgr) is connected in the system via the Xanbus network. i. The first unit to enter Bulk, causes all other chargers to enter Bulk. ii. The first unit to enter Absorption, causes all other units to enter Absorption. iii. The last unit that is ready to exit Absorption, triggers all units to exit Absorption and enter Float

This doesn't contradict the idea of setting the MPPT charge voltages slightly higher than the charger on the CSW, but it somewhat suggests that it should not need to be. I know my SCP is exchanging data across the Xanbus with the MPPT and CSW just fine, but I'm wondering if there is some miscommunication between the MPPT and CSW.

Still, I'm going to try @Dave Angelini suggestion. I'll set the three charge voltages on the MPPT to around 0.3V higher than those on the CSW.

Estragon · August 2019

I'm not sure the charge stage sync is the issue, more like giving a relay trigger value a bit of range to prevent chatter.

Horsefly · August 2019

Estragon said:

I'm not sure the charge stage sync is the issue, more like giving a relay trigger value a bit of range to prevent chatter.

Yeah, I get what you mean, but... Let's say that I set the CSW to voltages that are below that of the MPPT. I start the generator, and based on the battery voltage let's say it (the CSW) thinks it should bulk charge the battery. According to the manuals, the CSW would then go into bulk charging mode, and the MPPT would immediately go into bulk charging mode. So even though the battery voltage was not where the MPPT would want to be in bulk, it would go into bulk because the CSW did.

So my bottom line is, it sounds like it doesn't matter where the charge voltages are set between the two devices. The second device should simply follow the first device in charging.

Moreover, I can't see how this "synchronized operation" goes along with the inverter/charger dropping off-line every 10 seconds because the voltages are not set far enough apart.

Anyway, I'm going to give it a shot. @Dave Angelini has more experience with this Schneider equipment than I ever care to, and all those undocumented bugs / features are things I would never figure out.

Edit to add: Just in case, I am bringing along my WD40 and duct tape!

Estragon · August 2019

Do the devices share a common definition of voltage? It sounds like they just share a messaging protocol ("do this") and respond accordingly, bouncing between on and off states.

My assumption is the 0.3v suggestion is to prevent ambiguous voltages from creating this bouncing.

I'm not sure how that works in bulk though, as you'd think both the CC and genny would see < Vabs and both stay in bulk (with the genny still qualified).

Dave Angelini · August 2019

You are right Estragon. The system is doing what it is suppose to. We are talking very large systems that the engineers modeled. The chargers are synced in multiple XW system with multiple gensets. The solar voltage is causing the issue because it is changing with clouds. Large systems can use the AGS to automate and save fuel for folks in the great north.....

In Steve's case, a small system, the CSW is protecting the battery (an AGM) from over voltaging. The only way the CSW charger can back off charge in this simple system is to disconnect the genset. This is the cause of failed relays as many DIY users never even know it is happening. There are other ways that would cost some money and maybe have the same problem. This is the same with Outback. Back in the early days of solar people were killing agm's. Back before XW nothing was networked. A flooded would not care about spikes above the absorb voltage. I stopped using Outback because XW gear was all networked. Outback addressed that a bit later.

If you just used the genset alone it would be fine also. The solar and a variably speed genset is a great way to save fuel on a cloudy day.
A regular genset, not so much.

.3v is just what I have used for folks. The value can vary but it should be fine for most. Bill has written of this many time over the years also so it is nothing new.Not sure if it is FAQ? The 2 chargers are in competition. Let the solar win

Hope this is useful.

Horsefly · August 2019

Good news! The suggestion from @Dave Angelini worked exactly as advertised!

I turned up the Bulk, Absorption, and Float voltages on the MPPT-60/150 slightly, and turned the CSW charger settings for the same down slightly, leaving the CC about 0.3V higher on all stages than the charger on the CSW. I then tested it in every situation I could think of, running the generator while the batteries were in various stages of charging both with and without much solar. Everything worked as intended.

I never really doubted you Dave!

Thanks for the help!

Dave Angelini · August 2019

Jeeze I wish my wife had more faith.........................................

BB. · August 2019

Dave, your wife must have faith. She married you.

Take care,
-Bill

Dave Angelini · August 2019

There was a list for her when we married. It is longer now

The German part and my Italian are also a challenge at times.
You take care also!

Schneider SW4024 drop-outs with generator

Comments

Categories