Trace SW4024 Charging Issue

[Bmdale]

Hi to all;
We have a circa 1997 Trace SW4024 that recently started dropping off charging before charging is complete. We live off-grid and use a Honda eu3000 generator.
While watching the current flow to the batteries on the battery monitor, charging will proceed as expected (~60+A) and then suddenly go lower and lower until it is negative and the relay in the inverter will click. After a few seconds the relay will click again, charge current will increase to, maybe 12A, then drop to negative and the relay will click.
Bulk is set to 28V, absorb to 27. Charging stops before the battery voltage reaches 28V.
When it drops off seems to depend upon the charging current setting: at 22A it charges for about 7-8 minutes, at 18A maybe 12 minutes, at 15A probably ~90 minutes, but that timeframe may be becoming shorter - haven't had time to test.
The LEDs on the interface do not show any error. The Error Codes menu does not show any errors.
I've verified the generator is outputting acceptable power.
I have tested with all PV's turned off and all loads turned off. Same behaviour.
This problem became apparent last summer. We had switched from FLA to LiFePO4 batteries in the Spring. Pretty sure we had charged to completion via generator before the summer.
I found a later manual put out by Xantrex which had a trouble shooting section which mentioned a circuit breaker on the left side. I pressed that in and it seemed to stay in but no difference in problem.
While charging the other day, for the first time ever I heard the fan come on inside the inverter. I have not heard is since.
I've read that a dirty relay can cause this issue. Assuming there's one relay for AC1 and another for AC2, yesterday, I switched the generator hot from AC2 to AC1. The problem remains the same.
Any thoughts to share?

 

[Whinny]

What is your battery arrangement and type. It is possible you are hitting a BMS disconnect due to imbalance.

How was the little breeze and rain last night?

 

[RCinFLA]

The issue is probably when inverter changes charging from absorb to float voltage. SW's will try to pull battery down from absorb voltage to float voltage when it drops to float mode by pushing out AC power.

Usually it pushes the AC power out the ACout port to bring down battery voltage to float level if there is ACout loads.

Make sure you do not have Sell enabled that would allow a back feed into generator if there is not ACout loads.

 

[Bmdale]

What is your battery arrangement and type. It is possible you are hitting a BMS disconnect due to imbalance.

How was the little breeze and rain last night?

Whinny, thanks. Batteries are 4 x Simpliphi 24V (151Ah each).
Good point. Whether charging from 75% discharge or 10% discharge result is the same. PV panels with charge controller running at up to 66A doesn’t encounter this problem. I’m pretty certain the Trace charger worked earlier in the year. And, there’s no sign of BMS disconnect via the inverter and some other electrics I have connected directly to the batteries.
Thoughts on how to check the BMS’s?

 

[Bmdale]

The issue is probably when inverter changes charging from absorb to float voltage. SW's will try to pull battery down from absorb voltage to float voltage when it drops to float mode by pushing out AC power.

Usually it pushes the AC power out the ACout port to bring down battery voltage to float level if there is ACout loads.

Make sure you do not have Sell enabled that would allow a back feed into generator if there is not ACout loads.

Thanks Solar Wizard. This problem occurs when charging from anywhere between 75% discharge and 10% discharge, so I wouldn’t expect it to be switching out of bulk. Plus, when this shutdown occurs the battery voltage is always well below bulk setting of 28V.
Also, I should have mentioned, when this problem occurs, if I shut off the generator wait a minute then restart it the charger usually works fine for several minutes until it shuts down again.
It’s like something gets hot or an error develops and the charger can’t function.
I don’t recall any way to tell what stage the charger is in. Do you know?
Very good point re power out on AC1. I’ve already switched the generator hot back to AC2, though.

 

[RCinFLA]

Inverter will not immediately release from AC input unless there is load on AC input, like a grid collapse to draw back current.

An inverter-generator will tri-state (open circuit) its output when generator is shut down. A synchronous generator will absorb AC back from inverter causing inverter to release pass-through relay.

When pass through relay is engaged the inverter is slave to AC input freq/phase/voltage. If you just open AC input breaker the inverter pass-through relay will stay engaged and, lacking an AC input to maintain phase lock, inverter will gradually wander off frequency until it reaches its AC input frequency range limit then release pass-through relay. This can take a few seconds. Once pass-through relay releases the inverter reverts to being its own master and slowly pulls frequency/voltage back to inverter internal reference.

If you unplug a generator connected to inverter, then plug it back in a few moments later before pass-through relay has released there can be a mismatch in AC phasing causing a large surge current. There will also be inverter supplied AC voltage on the male plug until it opens pass-through relay.

Whenever you change or switch AC input you must ensure the green LED ACin shuts off indicating pass-through relay releases. The phase lock will no longer match unless pass-through relay releases and inverter goes through a new phase locking process to AC input.

SW inverters have two AC inputs. If you have generator on ACin2 and grid shows up on ACin1, it detects AC voltage on priority AC1 input. It releases pass-through relay to ACin2 generator, then runs on battery power as it readjusts its phase to AC1in before closing AC1in pass-through relay.

You need to set AC input current limit so it does not exceed generator capability. Charging power may exceed generator capability. You can leave heavy charging limit if you set AC input current limit. That will cause charging power to be limited to AC input current limit.

 

[Bmdale]

Inverter will not immediately release from AC input unless there is load on AC input, like a grid collapse to draw back current.

An inverter-generator will tri-state (open circuit) its output when generator is shut down. A synchronous generator will absorb AC back from inverter causing inverter to release pass-through relay.

When pass through relay is engaged the inverter is slave to AC input freq/phase/voltage. If you just open AC input breaker the inverter pass-through relay will stay engaged and, lacking an AC input to maintain phase lock, inverter will gradually wander off frequency until it reaches its AC input frequency range limit then release pass-through relay. This can take a few seconds. Once pass-through relay releases the inverter reverts to being its own master and slowly pulls frequency/voltage back to inverter internal reference.

If you unplug a generator connected to inverter, then plug it back in a few moments later before pass-through relay has released there can be a mismatch in AC phasing causing a large surge current. There will also be inverter supplied AC voltage on the male plug until it opens pass-through relay.

Whenever you change or switch AC input you must ensure the green LED ACin shuts off indicating pass-through relay releases. The phase lock will no longer match unless pass-through relay releases and inverter goes through a new phase locking process to AC input.

SW inverters have two AC inputs. If you have generator on ACin2 and grid shows up on ACin1, it detects AC voltage on priority AC1 input. It releases pass-through relay to ACin2 generator, then runs on battery power as it readjusts its phase to AC1in before closing AC1in pass-through relay.

You need to set AC input current limit so it does not exceed generator capability. Charging power may exceed generator capability. You can leave heavy charging limit if you set AC input current limit. That will cause charging power to be limited to AC input current limit.

Hi Solar Wizard. I think I understand. Correct me if I’m wrong: on the SW4024 no current is sent back through AC2? Thus protecting the generator.
We don’t have grid power here so normally nothing connected to AC1. Our generator is an inverter generator.
To be sure, in the scenario I described about restarting the generator the AC in Good LED is off when I restart the generator.
As described, I’ve tried setting the allowed draw on the generator lower and lower to get the Trace charger to run longer.
But sooner or later, it’s like the Trace heats up, although no detectable heat on case, and shuts down. Nothing out of the ordinary shows on the control panel.

 

[RCinFLA]

Criteria for SW to release from AC input is:

1) overcurrent on SW inverter through ACin. EU3000is should be limiting factor on this.
2) Too much wobble on ACin frequency. EU3000is should be stable on this with its electronic inverter controlling frequency.
3) Frequency out of range. Again, with EU3000is it should be accurate and stable.
4) AC voltage out of range.
5) Back push of AC current out of ACin port. (unless selling is enabled)

The EU3000is may tri-state (open circuit) its inverter if it detects back push AC current. This could cause SW to lose phase locking reference.
EU3000is also has paralleling capability. This is done by inverter-generator detecting AC present from an operating parallel connected generator when it first starts up. If AC voltage is detected on AC output of EU3000is when it is first started, EU3000is will put itself in slave mode and phase lock to other generator connected in parallel on AC output which is phase master.

This last feature of EU3000is may have something to do with your issue. If both the inverter generator and SW inverter are acting as slaves, neither is controlling the AC freq/phase so the system just wanders off AC frequency until one of them trips off due to out-of-range frequency.

Check AC frequency with a DVM with freq counter to see if it wanders off.

I run a Yamaha 3000is inverter-generator on my SW's with no issue but the Yamaha does not have paralleling capability.

Make sure you have an input neutral ground bonded as most inverter-generators have floating neutrals.

When inverter-generator is operating in ECO mode at reduced engine rpm, when a sudden AC load is applied the inverter-generator AC voltage peaks will be clipped until engine speeds up to supply additional AC load. I do not believe this should be an issue with SW as the inverter will just stop charging current during the inverter AC peak voltage dropouts. It might cause SW inverter to release pass-through relay though.

Other thing you can try is maintaining some ACout load on SW inverter while connected to generator. You can also try maintaining some AC load directly on inverter-generator output to see if it affects the disconnecting issue.

Check ACin current with a clamp-on amp meter to see if it is dropping near zero when you see the disconnect. Use menu 4 meters on SW to check ACin frequency, voltage, and current.

 

[Bmdale]

Criteria for SW to release from AC input is:

1) overcurrent on SW inverter through ACin. EU3000is should be limiting factor on this.
2) Too much wobble on ACin frequency. EU3000is should be stable on this with its electronic inverter controlling frequency.
3) Frequency out of range. Again, with EU3000is it should be accurate and stable.
4) AC voltage out of range.
5) Back push of AC current out of ACin port. (unless selling is enabled)

The EU3000is may tri-state (open circuit) its inverter if it detects back push AC current. This could cause SW to lose phase locking reference.
EU3000is also has paralleling capability. This is done by inverter-generator detecting AC present from an operating parallel connected generator when it first starts up. If AC voltage is detected on AC output of EU3000is when it is first started, EU3000is will put itself in slave mode and phase lock to other generator connected in parallel on AC output which is phase master.

This last feature of EU3000is may have something to do with your issue. If both the inverter generator and SW inverter are acting as slaves, neither is controlling the AC freq/phase so the system just wanders off AC frequency until one of them trips off due to out-of-range frequency.

Check AC frequency with a DVM with freq counter to see if it wanders off.

I run a Yamaha 3000is inverter-generator on my SW's with no issue but the Yamaha does not have paralleling capability.

Make sure you have an input neutral ground bonded as most inverter-generators have floating neutrals.

When inverter-generator is operating in ECO mode at reduced engine rpm, when a sudden AC load is applied the inverter-generator AC voltage peaks will be clipped until engine speeds up to supply additional AC load. I do not believe this should be an issue with SW as the inverter will just stop charging current during the inverter AC peak voltage dropouts. It might cause SW inverter to release pass-through relay though.

Other thing you can try is maintaining some ACout load on SW inverter while connected to generator. You can also try maintaining some AC load directly on inverter-generator output to see if it affects the disconnecting issue.

Check ACin current with a clamp-on amp meter to see if it is dropping near zero when you see the disconnect. Use menu 4 meters on SW to check ACin frequency, voltage, and current.

Hi Solar Wizard, all good thoughts.
We've had that generator for over a year and for sure the Trace never had a problem with the generator and SLA batteries. And, I'm pretty certain the Trace worked fine for several months after switching to LiFePO4.
When charging starts, the Trace takes a little look at the power from the generator and if it likes what it sees the relay closes and it slowly ramps up how much current it draws. I always assumed this was the charger 'latching' on to the generator frequency.
Curiously, there can sometimes be a little 'dance' between the inverter/charger and the inverter-generator: the charger asks for a bit more power, the generator responds by increasing throttle by one 'click' which makes more power than the charger asked for so it asks for less, then the generator slows down which decreases power, then the charger asks for more power... This dance never happens at full load, only at partial load.
I have tried with and without load on the inverter and the generator and behaviour is the same - at some point the charger just stops working.
Have verified that voltage and frequency from the generator never wavers, even while the charging issue happens.
Re grounding: I believe all neutrals are connected to ground at the circuit panel. Will check.
Good idea with the ammeter. Will give it a try.
Planning on lugging some old SLA batteries in (everything is in the house) and try charging those instead of the LiFePO4's.
Also, I believe the Bulk light on the inverter/charger stays on while the shutdown occurs, but will check.
Thanks again.

 

[RCinFLA]

Sounds like you are having feedback issues between inverter-generator ECO mode control and SW inverter load regulation. Sounds like the two feedback loops are fighting each other.

Does problem go away if you turn ECO mode off on inverter-generator, letting it run at full rpm's ?

SW will jump pretty hard on generator shortly after it connects if charging level or input AC current limit is set too high. SW will automatically back down charging draw to stay within AC input current limit setting if more ACout load is required. ACout has first priority on available generator power.

Sudden surge current causes pertubations to feedback control systems. The SW will over/undershoot AC voltage regulation. This, in turn, may cause ECO mode on generator to be perturbed. If SW overvoltages on surges it may cause inverter-generator to shut its inverter down. How long it does this is the question and if too long or permanent until generator is stopped and restarted the SW will wander off frequency and disconnect.

 

[Bmdale]

Hello All;
Seems I found the cause of this issue: I had the 'Absorb Time' set to 10 minutes. Curiously, the Absorb voltage was never reached but the charger went to Float anyway.
I'd read the manual previously, and just re-read it now, and the charger doesn't seem to behave the way I understand the manual to say. The confusion seems to be about when "0" for the timer happens.
In any case, I have bumped that setting up to 2 hours and the charger has run at 86A to full charge.
And, I'm guessing a confounding part is that when trying to diagnose the problem I lowered the charge rate, to say 10A, and the charger appeared to be charging properly but was actually only charging at Float current.
In case you're wondering why I'd set Absorb time to 10 minutes, the LiFePO4 battery manufacturer wants the batteries charged to absorb voltage then held at float voltage for 6 minutes and then all charging to end. The smallest non-zero 'Absorb Time' setting the Trace SW4024 will take is 10 minutes.
I guess now, it's kind of manual charging: I have to turn off the generator when the batteries are fully, or nearly fully, charged.
Thanks to every one for your help on this.