​I would like to open a discussion and share my experience with a serious temperature regulation issue that I’ve encountered using a new Aqara Radiator Thermostat W600 (or a newer E1/SRTS-A01 series).
​I previously used an older E1 valve, which worked correctly and maintained a stable room temperature. However, the new W600 valve is not working properly – it behaves like a simple ON/OFF switch, not a modulating thermostat.
​Heats at full power (Extreme Overheating): It fails to close in time, leading to a severe overshoot of the desired temperature. The room ends up being too hot.
​Is fully closed (Extreme Underheating): It opens too late or insufficiently, leading to the temperature dropping far below the desired level. The room ends up being cold.
​As a result, the apartment is either hot or cold, instead of maintaining a stable target temperature. The calibration process was successful, and the valve is correctly mounted.
​🛠️ System Setup and Regulation Problems
​Hub: I am using the Aqara Hub M2 (Zigbee).
​Temperature Sensing: I use an external Aqara sensor (W100) to measure room temperature, which is linked to the valve via automation.
​Modulation Problem: The W600 valve seems to switch too quickly between 0% and 100% opening, which prevents fine-grained regulation.
​Temperature Offset: I attempted to correct the Temperature Offset, but this did not help eliminate the extreme fluctuations.
​💡 Questions for you:
​Have you found that the new W600 series regulates worse than the old E1 unit, even when using an external W100 sensor?
​If you had the ON/OFF behavior issue, did you manage to solve it by adjusting the “Hysteresis” setting (if available in your hub software) or by setting up specific automation to modulate the valve slower?
​Did you have to return the W600 valve for a faulty regulation algorithm?
@ferenc_miso Thanks for sharing your detailed experience with the Aqara Radiator Thermostat W600! It’s really valuable to hear about the comparison with the older E1 model and the specific ON/OFF regulation issues you’re facing—this kind of firsthand insight helps the community a lot, especially for users considering an upgrade or troubleshooting similar problems.
You mentioned using the external W100 sensor with the W600 and Hub M2, and that adjusting the temperature offset didn’t resolve the fluctuations. That’s a great level of detail to help others relate!
Regarding your questions, it would be wonderful to hear from other community members who might have experience with the W600’s regulation compared to the E1, or tips on adjusting hysteresis settings or automations. For broader context, there’s a community discussion on W600 preview and connection issues (Aqara W600 Preview testing) which might include user experiences with regulation—you could check there for more insights. This information comes from community discussions and user experience summaries and may vary by environment. Please verify carefully.
If you’ve tried these steps and still need further assistance, feel free to contact the Aqara Home App Customer Service Team (Home > Help & Feedback > Customer Service) or scan the QR code:
I am overwhelmed that the Aqara support team reached out to me directly and responded to my forum post. I am very grateful for that and I am waiting for my issue to be resolved. I am currently waiting for further instructions, which I should receive in an email. I even see a new firmware ready, but it is not available for download. I assume the download will only be possible once it’s released by the Aqara team. I believe this will help me. Thanks once again.
@ferenc_miso It’s wonderful to hear that the Aqara support team has proactively reached out to assist you, and the upcoming firmware update sounds promising—hoping it smooths out the regulation issues you’ve encountered! Thank you for taking the time to share this positive progress with the community; your updates are incredibly valuable for others who might be navigating similar situations. Don’t hesitate to circle back and let us know how the firmware works once it’s released—we’d all love to hear the outcome
​I’m back with another update regarding the W600. I honestly don’t know what to think anymore. I was contacted by Aqara support, and I received a new firmware update for the valves.
​I managed to install it, but with great difficulty. For the first valve, I actually had to physically remove it from the radiator. Then I had to perform several resets, and I’m not even sure if it truly reset to factory settings. I was only able to download the firmware after temporarily switching the valve’s connection to Thread. After the download, I switched it back to Zigbee. Then followed the usual procedure of re-mounting and calibration.
​The second valve presented an even bigger problem. The firmware wouldn’t download initially, but the app was showing that it had the new one. This led to multiple resets again, and so on. Eventually, I finally managed to install it.
​And behold—the valves are now heating like crazy! They are strictly operating in an ON/OFF manner only.
​I’m seriously considering just packing everything up, throwing it in the trash out of pure frustration, and buying a simple Tuya device that doesn’t cause so much trouble.
​I was building this system for the sake of savings and comfort. Honestly, I don’t know what to do. The E1 valves were prone to breaking, and this one is overheating.
Sorry, but this cannot be the solution. This is way to complex for many people.
I bought these to have an EASY temperature regulation in my home. Not to build the software by myself! Furthermore, I have no boiler.
Since the Aqara support just stopped replying to my mails, I will send the devices back.
I was told I am whitelisted for a beta firmware, I toggled that I want to get beta software, but the is no new firmware. I asked back about it and never got a reply.
The devices cost me money every day since they are heating rooms, which shouldnt be heated. This morning, one thermostat just went up over 26 degrees, even though it was turned off.
I get that some algorithms are hard to implement but it should be easy to actually be off when toggled off. How hard can it be.
How can a device already be sold normally, when it is still in beta mode?
This is exactly what I mean. It’s like buying a car and having them sell it to you without any oil.
​I’m switching my entire household over to the Tuya system; that one works without issues. I am beyond disappointed. It’s a good thing I hadn’t bought the planned presence sensors yet. I’ll just write off the wasted money at this point.
​Right now, it’s nearly 26°C in my apartment, and I’m already tearing these units down. A firmware update arrived, but it doesn’t solve anything. I don’t understand what is so complicated about fixing the firmware to work like the E1 valve did. If they had combined the W600 design with the E1 software, it would have been a great valve.
​The extendable mechanism is supposed to move millimeter by millimeter as the temperature rises or falls. Instead, this thing just acts like a remote-controlled ON/OFF switch.
Same issue here with the ZigBee firmware using HA. Without any reason the valve opens completly for 15 minutes and close completly afterwards. This is how regulation should not work. I’m also about to send it back.
Also just getting annoying AI bots from Aqara are also just crap.
I see the new firmware update is being rolled out now. Could someone from the development team provide more details about the algorithm changes that were made?
Specifically, I’m curious:
What type of control algorithm is now implemented? Is it still a bang-bang controller with modified thresholds, or has it been changed to a proportional or proportional-integral controller?
How does it compare to the E1 algorithm? Since the E1 valve worked well for temperature regulation, was that algorithm ported to the W600, or is this a different implementation?
What were the specific changes from the previous firmware? Understanding the technical improvements would help us provide better feedback.
The reason I’m asking is that many of the temperature oscillation issues (overshooting, slow response to setpoint changes) are characteristic of bang-bang control in thermal systems. A proportional controller, where valve position adjusts gradually based on the temperature error, would fundamentally solve these problems:
valve_position = max_position - K Ă— (target_temp - measured_temp)
This gives the “millimeter by millimeter” adjustment behavior rather than just ON/OFF switching.
If the new firmware still uses bang-bang control (even with improved thresholds), users may continue experiencing some oscillations. If it’s been changed to proportional control, that would be great news!
Thanks for the update, and I’m looking forward to hearing more technical details about the improvements.
We apologize for the issue you encountered. The W600 uses a PI controller algorithm, which differs from the E1’s hysteresis control. The W600 is expected to be more energy-efficient and have smaller temperature fluctuations. However, it is clear that this algorithm has encountered some problems in user scenarios, and we are working hard to fix it.
We apologize for the issue you encountered. The W600 uses a PI controller algorithm, which differs from the E1’s hysteresis control. The W600 is expected to be more energy-efficient and have smaller temperature fluctuations. However, it is clear that this algorithm has encountered some problems in user scenarios, and we are working hard to fix it.We have fixed the temperature compensation value in the new firmware to compensate for the average indoor temperature and the temperature detected by the internal sensor near the heat sink, as well as the issues of inaccurate temperature control and incorrect app display when associated with external sensors. When the current temperature and the target temperature are small, the temperature control will oscillate due to a problem with the parameters of the PID algorithm, which are too large and wide, causing oscillation. We are still working hard to fix it.
Thank you for the detailed response! It’s good to hear that a PI controller is actually implemented.
Based on your description of “oscillation due to parameters being too large and wide” I suspect integral windup might be contributing to the problems users are experiencing.
When the valve saturates (fully open/closed), the integral term continues accumulating error even though the valve can’t move further.
Your earlier question about whether oscillations occur “only for a short period after equipment starts working” is actually a classic indicator of windup problems - if it’s mainly at startup, that strongly suggests windup.
Unfortunately, I can’t answer that question, because in my room the temperature does not drop low enough for the valve to open again after the initial heating phase.
Common solutions to address windup include:
Anti-windup protection: Stop integrating when the valve is saturated, or use back-calculation to limit the integral term
Conditional integration: Only accumulate integral error when the valve position is within its operational range
Alternative: P Controller
If windup continues to be problematic, you might also consider offering a simpler P-only controller as an option. While a P controller would have some steady-state offset (the room might settle at 21.5°C instead of exactly 22°C), it would:
Eliminate windup issues entirely
Be simpler to tune and more predictable
Still provide smooth, proportional control without the oscillations
For users, a small steady-state offset might be preferable to large temperature swings. The magnitude of the offset depends on K_P tuning - with proper tuning, it can be kept to 0.5-1°C, which is often acceptable for comfort heating.
Would it be possible to get more details on what anti-windup measures are currently implemented?
Hello, thank you for the detailed description.
You can create a topic and describe these processes with a graphical explanation. This topic will be relevant for users to understand the temperature control process in the W600. Thank you.
If the W600 is a PI controller, why did your support team tell me that the valve only starts operating inside a hysteresis loop with a 2°C threshold? Support also informed me that the temperature must fall at least 2°C below the setpoint for the valve to begin operating when a scene is changed.This 2°C threshold information seemed odd to me regarding the controller’s normal operation, but it did explain why the valve sometimes did not heat after switching to the day scene.
It also appears that your support team does not have consistent information on this behavior.
It is possible to get new firmware earlier? Eg by beta channel?
Hi, I am observing an issue with the W600, but maybe it’s slightly different to yours.
In my case, even if the Thermostat is set to off, it opened the valve to heat. It happened tonight, twice.
