AEG SolarCube Pro error 170305110: what “Head unit undervoltage, level-3 fault” means
Error 170305110, Head unit undervoltage, level-3 fault, means the battery of the AEG SolarCube Pro is discharged too deeply and the battery management system has locked it. For days the app then says Battery Force Charging Mode while 2 to 6 W goes in. Our unit drained itself off the grid, sat locked for 4 days and 5 hours after it was plugged back in, and came out on its own. What each message means, why it drains while it looks switched off, and what does and does not help.
In this article11
- What error 170305110 means
- How it went on ours: from 47% to System hibernation shutdown and back
- Why it drains while it looks switched off
- How it came out of the lockout
- AEG SolarCube won't charge after 170305110: what helps and what does not
- Does this also apply to the HYXi Halo?
- Reported to HYXi and AEG
- Frequently asked questions
- Further reading
- Transparency
- Sources
TL;DR
Error 170305110, "Head unit undervoltage, level-3 fault", means on the AEG SolarCube Pro (AS-BBL08) that the battery is discharged too deeply and the battery management system (BMS) has locked it. It then does not charge, even though the home screen says "Battery Force Charging Mode". On our unit a second message came first in the alarm list: 170486302, "System hibernation", with the cause "System hibernation shutdown". Our unit came out of it on its own, with no technician and no intervention, 4 days and 5 hours after it was back on the grid. own measurementAEG StoragePRO 1.0.0 app and Modbus readout, cell voltage logged per minute · 30 September 2026
Everything below comes from one unit: the one we tested for our review of the AEG SolarCube Pro. Whether another unit behaves the same way, I do not know. What I do know is here with the date and the origin attached, so you can hold it up against your own app.
💡Tip
Is this code in your app right now? Leave the unit on the grid and look at the alarm list, not at the mode name on the home screen. On ours, no setting, write or restart made the recovery any faster. It took 4 days and 5 hours, and the unit came back with both packs.
What error 170305110 means
The message sits in the alarm list of the AEG StoragePRO app, labelled "Unprocessed". Tap it, and under "Fault Cause" it says:
The cell or battery pack voltage exceeds the limit range, thereby triggering the corresponding protection.
That sentence does not say in which direction the voltage left the range. The name of the message does: undervoltage, at level 3. own measurementscreenshot of the alarm list, AEG StoragePRO 1.0.0 app · 25 September 2026
"Head unit" refers to pack 1, the head of the system. On ours, pack 1 stood at 0%, with cells between 2,255 and 2,321 mV. Pack 2 no longer reported at all; the app said "Battery x1". Charge and discharge power were both 0 W. The inverter itself reported no fault: the lock sat in the battery. own measurementModbus readout and the app · 25 September 2026
If you read the unit over Modbus, the same state shows in input register 4979: 4 is connected to the grid, 7 is locked on battery undervoltage. own measurementModbus readout, read only · 25 September 2026 The other registers are on the AEG SolarCube register page.
How it went on ours: from 47% to System hibernation shutdown and back
After the test I took the unit off the grid and left it in standby. What happened next, with what the app showed at each moment:
| Moment | What happened | What the app showed |
|---|---|---|
| 11 and 12 September 2026 | the test on our techbench; then off the grid, in standby | |
| 15 September, 03:36 | last data point in the cloud: state of charge 47% | |
| 24 September, 22:49:12 | the unit goes into hibernation | alarm 170486302, "System hibernation", cause "System hibernation shutdown" |
| 25 September, 19:39 | grid reconnected. Pack 1 at 0%, cells 2,255 to 2,321 mV; pack 2 not reporting; charge and discharge power 0 W | error 170305110, "Head unit undervoltage, level-3 fault", "Unprocessed"; "Battery x1" on the home screen |
| 25 to 30 September | locked; 2 to 6 W goes in | home screen: "Battery Force Charging Mode", "Battery x1", 0%, Battery 0 W |
| 30 September, 00:51 | the BMS releases: the charge ceiling jumps from about 0.10 A to 750 W, the cells go from 2,500 to above 3,000 mV within minutes, pack 2 reports again | alarm list: "Head unit undervoltage, level-3 fault" set to "Restored" (00:50:45), then "Head battery hardware failure" (00:50:47) and "System hardware overcurrent" (00:50:51) |
| 30 September, 00:55 | charging again, with both packs | home screen: "Charge", "Battery Force Charging Mode", "Battery x2", 25%, "Battery Charge 763 W"; per pack: pack 1 at 2%, pack 2 at 47% |
| 30 September, 12:12 | almost full | pack 1 at 98%, pack 2 at 100% |
Battery Force Charging Mode does not say it is charging. The home screen shows the mode that is requested, not what the BMS allows. The state of charge stood at 0%, below the 5% at which forced charging starts, so the unit asked to charge. The BMS let 2 to 6 W through, for days, and the app showed "Battery 0 W" alongside. Nothing on the home screen mentions a lockout or how long recovery takes. The undervoltage alarms appear only in the alarm list. own measurementAEG StoragePRO 1.0.0 app and Modbus readout · 27 September 2026
On ours, Battery x1 did not mean pack 2 was broken. Pack 2 was invisible all those days and reported again at the release, healthy, at about 3,283 mV per cell. own measurementModbus readout after the recovery · 30 September 2026
The two hardware messages at the release. "Head battery hardware failure" and "System hardware overcurrent" arrived within six seconds of the undervoltage message being restored. What they mean is in no document I have. On ours nothing followed: it kept charging, and at 12:12 that day both packs were almost full. own measurementscreenshots of the alarm list and the pack overview · 30 September 2026
Why it drains while it looks switched off
This unit cannot really be switched off. Hold the power button for 5 seconds and the lights go out and the app says "device offline", but the socket still measures 9.3 W and Modbus keeps answering: only the lights and the cloud connection go off. Standby through the app measures 9.5 W. There is no isolator switch, and no button combination turns it off: the IoT button for 5 seconds restarts Wi-Fi and Bluetooth, the power button for 5 seconds is standby, and both together for 5 seconds is a full reset, after which it comes back on. Modbus does not do it either: the deep-off value in register 4021 falls back to standby. own measurementmetering socket and Modbus readout, after the recovery, battery almost full · 30 September 2026
9.3 W
draw with the power button off, at the socket; off the grid it comes out of the battery
Bron: own measurement, 30 September 2026
Off the grid, that draw comes out of the battery. In standby it then takes about 9.5 W from it, and nothing stops it before it is empty. own measurementModbus readout, standby off the grid, recorded in the bench report · 16 September 2026 On ours, hibernation only kicked in once pack 1 was already empty. derivedfrom the System hibernation message of 24 September and the state of charge of pack 1 at the reconnect on 25 September
How fast exactly it drains, we have not measured under control. From two end points I give no figure for it, and so none for how many days it can stand off the grid before it locks.
The drain was lopsided. Pack 1, the head, went to 0%, while pack 2 still stood at 47% at the release according to the app. Off the grid, the head carries the electronics that are always on: the BMS, the communication, and the light ring, which stayed on the whole time on ours. How much that ring accounts for is not measured. derivedfrom the state of charge of both packs at the reconnect on 25 September and at the release on 30 September
The limits in the app are states of charge, not voltages
None of the limits you set in the app protects the battery against the unit's own draw. This is what ours was set to:
| Setting | Ours | Why it did not help off the grid |
|---|---|---|
Anti-starvation protection (anti_starvation, holding 4121) | on | charges from the grid, and there was no grid |
Minimum discharge (discharge_min_soc, holding 4141) | 10% | governs discharging into the house, not what the unit itself draws |
| Forced charging, start and stop | 5% and 25% | charges from the grid; during the lockout it stood armed with no current flowing |
| Minimum self-consumption | 10% | a state of charge, not a voltage |
| Minimum off-grid | 30%, set by me on 26 September, after the drain | a state of charge, not a voltage |
The lock itself hangs on the voltage of the weakest cell, not on a percentage. A floor in percent stops the discharge into your house; the draw of the electronics carries on, at 0% too. derivedfrom the settings of 27 September, the per-minute cell voltages from 25 to 30 September and the BMS firmware in update package HYX-MS3000AC_V01.03.01.75
How it came out of the lockout
It came out on its own, on the trickle alone. On 30 September at 00:51, 4 days and 5 hours after the grid was back, the BMS let go. The charge ceiling jumped from about 0.10 A to 750 W, the cells shot from 2,500 to above 3,000 mV within minutes, all undervoltage alarms cleared, and pack 2 reported again. own measurementModbus readout, cell voltage logged per minute · 30 September 2026
4 days 5 hours
from reconnecting (25 September, 19:39) to release (30 September, 00:51), with no intervention
Bron: own measurement, Modbus log from 25 to 30 September 2026
What it was waiting for. The BMS waits until the weakest cell holds 2,500 mV, not until it touches it: every dip restarts the waiting time. For days the lowest cell touched 2,500 mV without anything happening, until it stayed above it for the whole waiting time. While it is locked, it charges at about 0.10 A at most. There is no mode, register or setting that speeds that up, not in the app and not over Modbus. A note on the origin: the mechanism comes from the firmware in the update package the app offers, not from the firmware that runs on this unit. What the unit did fits it exactly. derivedfrom the BMS firmware in update package HYX-MS3000AC_V01.03.01.75, and the per-minute cell voltages from 25 to 30 September
2,500 mV
what the weakest cell has to hold before the BMS lets go; a brief touch does not count
Bron: inferred from the BMS firmware and the cell voltage log from 25 to 30 September 2026
Why it can also go wrong. That trickle of about 0.10 A is also why recovery is not a given. It only works as long as it delivers more than the battery itself uses. If it delivers less, the weakest cell never holds those 2,500 mV, and then it does not come out without intervention. On ours it worked, once. Whether it works on another unit, I do not know. derivedfrom the precharge ceiling in the firmware and the recovery from 25 to 30 September
AEG SolarCube won't charge after 170305110: what helps and what does not
What worked on ours was doing nothing except leaving it on the grid. After 4 days and 5 hours the BMS let go, and after that it charged normally.
What did not help on ours:
- writes over Modbus: they were acknowledged and then read back unchanged, and forced charging stood armed with no current flowing;
- switching it off and on by hand, twice;
- another mode or setting in the app: the trickle during the lockout is fixed in the BMS;
- the anti-starvation protection and the discharge floor, for the reasons in the table above.
⚠️Warning
Never write to holding 140 to 179 to get around the lockout. That block does not change, appears in no document and looks like the battery's protection settings. The unit stores anything without checking it, and a lowered undervoltage limit is exactly how you force-charge an over-discharged cell. That these are protection settings is inferred from the values and not confirmed; that you do not write to them is settled. own measurementModbus readout, read only · 26 September 2026
Prevention. If you want to put the unit away, leave it on the grid. We know of no other way to stop it draining: the button does not switch it off, and none of the limits in the app is a voltage limit.
If it does not come out. How long to wait before giving up, I do not know. On ours it was 4 days and 5 hours, and I have not seen a second case. Note the error code and the times from the alarm list, and report it to your seller or to AEG's service address, storage-pro@aeg-solar.com. Do not count on a quick reply: two earlier mails to that address, about the firmware, got no answer in thirteen days, and after that only a ticket number. own measurementown mailbox · 25 September 2026
Does this also apply to the HYXi Halo?
HYXi builds the AEG SolarCube Pro. In its registers the unit calls itself HYX-MS3000B, and that type name also appears on the HYXi Halo datasheet. The register layout is the Halo's, but the meaning of some registers differs from HYXi's own document. own measurementModbus readout of the internal model name · 14 September 2026 manufacturerHYXi Halo datasheet V1.5 · 19 August 2026
We have not tested a Halo. Whether it gives the same error code, drains the same way and comes out the same way, I therefore do not know.
Reported to HYXi and AEG
I reported the defect in the lockout, that the trickle is so small that recovery takes days and is not a given, to HYXi and to AEG, together with what I found about the cloud. As of 1 October 2026, neither had replied.
Frequently asked questions
Veelgestelde vragen
What does error 170305110 mean on the AEG SolarCube Pro?+
Head unit undervoltage, level-3 fault: undervoltage on pack 1, the head of the system, at level 3. The battery is discharged too deeply and the battery management system locks it. The app gives as the cause: The cell or battery pack voltage exceeds the limit range, thereby triggering the corresponding protection. On ours, pack 1 stood at 0%, with cells between 2,255 and 2,321 mV, and pack 2 no longer reported.
Why won't it charge while the app says Battery Force Charging Mode?+
The home screen shows the requested mode, not what the battery management system allows. During the lockout it charges at about 0.10 A at most; on ours 2 to 6 W went in and Battery showed 0 W, for days. Look at the alarm list, not at the mode name.
The app says Battery x1 instead of x2. Is a battery broken?+
Not on ours. Pack 2 did not report during the lockout and came back at the release, healthy, at about 3,283 mV per cell. At 00:55 that night the app showed Battery x2 again.
How long does it take to come out of the lockout, and can I speed it up?+
On ours, 4 days and 5 hours on the grid, from 25 September 19:39 to 30 September 00:51, with no intervention. Speeding it up did not work: writes over Modbus were acknowledged and not carried out, switching it off and on twice changed nothing, and no mode or setting raises the trickle during the lockout. The battery management system waits until the weakest cell holds 2,500 mV. One unit, seen once.
What does System hibernation shutdown, alarm 170486302, mean?+
That the unit has gone into hibernation. On ours that message came on 24 September at 22:49, off the grid, when pack 1 was already empty. At the reconnect the next day, error 170305110 sat next to it.
How do I stop my AEG SolarCube from draining?+
Leave it on the grid. It cannot really be switched off: with the power button off it still draws 9.3 W, and off the grid that comes out of the battery. The anti-starvation protection charges from the grid and the discharge floor governs discharging into the house, so neither helps once the grid is gone. We know of no other way.
Does this also apply to the HYXi Halo?+
I do not know. The AEG SolarCube Pro is built by HYXi and reports itself in its registers as HYX-MS3000B, a type name that also appears on the HYXi Halo datasheet. We have not tested a Halo.
Further reading
- AEG SolarCube Pro tested for the full review: what comes out of the unit, its own consumption, and the rest of what happened after the test
- The Modbus registers of the AEG SolarCube for anyone who reads and controls the unit themselves over RS485
Transparency
APEX Power Supplies sent me this AEG SolarCube. No money was paid. I received the unit for free and may keep it. This article contains no affiliate links and no discount code.
Every figure carries its origin in the line: eigen meting (own measurement) with the instrument and the day, fabrikant (manufacturer) with the document, afgeleid (inferred) with what it follows from. What I have not measured is stated as something I do not know.
Sources
Bronnen
- [1]AEG SolarCube Pro tested — our reviewPublished 1 October 2026, in Dutch. The section Na de test: van het net af liep hij leeg (After the test: off the grid it drained) describes the same event; this page follows that text.
- [2]AEG StoragePRO 1.0.0 — the appScreenshots from 25, 27 and 30 September 2026: the messages 170305110 and 170486302 with their Fault Cause, the home screen with Battery Force Charging Mode, Battery x1 and Battery x2, the alarm list at the release, the 763 W and the pack overview.
- [3]our techbench — log of the undervoltage and the recovery, 25 to 30 September 2026Highest and lowest cell voltage of pack 1 per minute, the BMS state, the charge ceiling and the number of packs, from the reconnect on 25 September at 19:39 to after the release on 30 September at 00:51. Logged over Modbus, alongside the app.
- [4]our techbench — settings read on 27 September 2026Holding 4121 to 4141: anti-starvation protection, the floors for discharge, self-consumption and off-grid, and the start and stop of forced charging.
- [5]our techbench — standby and button-offStandby off the grid, recorded in the bench report of 16 September 2026. Button-off, standby through the app, the button combinations and register 4021, measured on 30 September 2026 at the metering socket and over Modbus.
- [6]BMS firmware, from update package HYX-MS3000AC_V01.03.01.75The package the app offers, which does not install on this unit. Read to understand the lockout mechanism: the 2,500 mV the weakest cell has to hold, the waiting time and the trickle of about 0.10 A. It is not the firmware running on the unit.
- [7]HYXi Halo — datasheet V1.5Archived on 19 August 2026 as content/productblad/hyxi-halo/2026-08-19.pdf. Names the type names HYX-MS3000AC and HYX-MS3000B.
