Most reacted posts

Hi all, just sharing my recent experience working with a Growatt SPF 3500 ES and the included WiFi-F module. I have a few other solar systems already feeding their data into InfluxDB and Grafana, and wanted the new Growatt unit to do the same. Otti’s project on GitHub provided some inspiration...

As with all alleged battery breakthroughs, I'll have to see it to believe it. So many broken promises over the years has me irreversibly cynical. :LOL:

Not true. Some inverters float the panels at very high voltages with respect to ground. An unlucky person who touched a faulty panel frame or rack could have a nasty surprise/early death. This is why proper earthing of mounting systems is considered mandatory by practically all panel, racking...

Here's the list of working registers: Name Address Offset (from 0x1000) Type Multiplier Unit System Status 0 uint16_t 1 0: Standby; 1; (No Use); 2: Discharge; 3: Fault; 4: Flash; 5: PV charge; 6: AC charge; 7: Combine charge; 8: Combine charge and bypass; 9: PV charge and bypass...

I've taken a bit of a different approach to this issue and addressed most of it in software. I have a JK BMS, and have been using the amazing esphome-jk-bms software by syssi to monitor and send all the BMS statistics to MQTT. With a small modification, I've configured this to switch a relay on...

For anyone else who ends up here looking for answers, after a lot of searching and bit of trial and error, I've found that the plug type on the JBD BMS (for both the balance wires and the 4-pin UART/NTC connectors) is actually HY2.0. Molex PicoBlade, JST-PA, JST PH 2.0 and the others mentioned...

I completed the testing this morning: Conditions were ideal with a clear blue sky and 24°C temperature. After some experimenting with positions, I ended up using an east-west facing setup as this aligned the panels best with the sun. One panel was fully illuminated while the other was only...

Here you go. The switch in series with the 20R resistor is a momentary swtich for pre-charging the capacitors, the other is your high-current toggle switch. The momentary switch wouldn't need to be high current type, 5A would be sufficient.

Just an update on this - I think it's important to discuss our successes as well as failures, and this one was definitely a fail. I bought an old office style water cooler for spare change and stripped the compressor and refrigeration loop out of it. On its own it was able to cool the 2L water...

The information is relevant because competition within this market, especially in China, is brutal. Vendors will do just about anything to increase their margins or produce a product that is cheaper and more attractive than their counterparts. This includes ripping off and cheating their...

They'd be fine, they'll just produce a bit more heat. 0.75v @ 10.2A = 7.65W heat/wasted energy. 0.22v @ 10.2A = 2.24W heat/wasted energy. Worst case it's a 0.4% difference in the total maximum output of your array. Not worth worrying about.

Yes, I dumped the stock firmware before I started experimenting with my own, I've attached it for whoever might want a copy. I can also post all the MODBUS registers if it'd be helpful, through trial and error I found that the documentation for these systems has a couple mistakes and many...

In my setup the ESP queries the BMS for all its running statistics every few seconds, and then uploads the data to InfluxDB. It also operates the balancer relay if the necessary cell voltage conditions are met. While this was coded to work online/while connected to WiFi, it'd be trivial to...

A quick update: I've successfully programmed an ESP8266 to retreive data from my BMS (JBD/Overkill) and translate it into emulated Pylon CAN bus messages. The output looks identical to what the inverter would expect when configured with Option 5 - Li and Option 36 - L52. When I have a moment...

After reading this thread I'd be far less likely to buy any Victron gear. If I was to pay the premium for top quality equipment, I'd expect to be able to open, diagnose and repair it after the warranty expires. It's clear in this case that Victron intended these to be throw-away items, making...

I've always wanted to try this approach. Anyone who monitors their MPPT closely will see that the voltage it determines to be the maximum power point doesn't vary much throughout the day - about 10% even with changing angle of the sun, cloud cover, etc. So the approach of holding a fixed array...

It depends on a few things: - Will it be used outside for extended periods? The extension cord probably isn't manufactured to withstand the rigours of daily exposure to high levels of UV, moisture, heat and cold for years on end, as they're usually intended for short term, temporary use. PV...

You'd place the resistor in parallel with the disconnect switch, with its own momentary switch. That way while the big disconnect switch is in the "off" position, you can hold the momentary switch for a few seconds, completing the circuit through the resistor to pre-charge the inverter's...

No, it's in parallel with the disconnect switch, so it'd do nothing. I'll upload a circuit diagram a bit later to make it a bit more clear.

Just wanted to express my appreciation for those in this thread who put the miles in to diagnose and understand the issue of Growatt inverters ceasing solar production/charging. I was stuck on this same issue until coming across this thread. I made the recommended adjustments and now it's all...

After a bit more fiddling around I managed to resolve this - the latest version of JBDTools (V3.7-20220415) was able to write the settings successfully. This software is somewhat hidden on the Jiabaida website, so here's the link for anyone else experiencing this issue...

It's hard to tell from the photos if you have the necessary space, but you could wrap some damp cloth around the cell terminals to try to avoid heat conduction through them towards the cell, and use a metal or fiber cement heat shied to protect against the hot air stream. Check your insurance...

Extreme indeed, I'm amazed you didn't see a difference in that scenario. A great example of the value of real-world testing.

It seemed unanimous earlier in the thread that blocking diodes are unnecessary and generally a waste of time and energy. We then moved on to series vs parallel connections for panels facing different directions.

I had an unusual problem to solve recently - I have a cabin that's completely off grid and not connected to any public water supply, running on stored tank water only. This is in a fairly arid area, so water is precious and can't be wasted. One risk that's been sitting in the back of my mind is...

While this is true, I've managed to find peace with my Growatt AIO. Yes, the support is awful and the standby power can be high. But the unit does work really well once properly set up. I have mine running some heavy equipment in a workshop and it's never missed a beat. When I don't need AC...

I don't think you'll have any problems, the 390w figure just indicates how much power the Tracer can deliver according to the maximum current it can handle at a given voltage. Its firmware will self regulate and prevent an overcurrent situation. The total maximum PV power for the 3210AN at 12v...

This has been an interesting read and a failure mode I've not considered before. It would be trivial to make an independent device consisting of a voltage divider, micro controller and solenoid/relay to monitor voltage and force a disconnect if battery voltage exceeded a specific set point...

You only need to care about the voltage drop across the diode, which is where the inefficiency and heat comes from. The total voltage of the system doesn't really have any effect on this.

Try adjusting setting 12 (B2AC) to something above 48v. It doesn't have any functional effect in your configuration, but it might then allow changing 21 to your preferred value as the two are related. If you have no grid, you should probably also change 14 (charger source priority) to OSO...

That's correct, however my ESP32 connects to the BMS using Bluetooth. This is preferred as it has complete electrical isolation. The buck converter powering the ESP32 is currently downstream of the BMS power MOSFETs. This means monitoring will stop in the event of the BMS cutting off discharge...

My cells from Luyuan arrived this morning, total time in transit was 53 days. They look to be genuine EVE with the clear "B" marking on the QR labels, which is what I was expecting. It might be a while before I have a chance to test them properly, but I'll post the results of an initial...

Correct. It'd find the maximum power point close to where both strings are contributing as much as they can. It wouldn't be as optimised as an individual MPPT per string, but it wouldn't be far off, either. The MPPT voltage of a string of panels doesn't vary with lighting conditions as much as...

You might think so, but the resistance of the wires and terminal connections at high current make a non-negligible difference across the batteries. When under load, the wiring in first diagram pulls more current from the 4th cell due to lower resistance. This leads to an imbalance in...

I've added some extra functionality that provides an option to use MQTT to remotely switch the inverter's AC output in and out of standby mode. This is a function not available through the inverter's interface and can save quite a lot of energy by reducing idle power consumption when the AC...

I'm with you on this. I can understand why a vendor on razor thin margins might not want to deal with a customer who is upset about a <1% difference in tested capacity verses spec, particularly when that testing occurs using uncalibrated equipment, outside a laboratory environment and with no...

I'm probably a bit late to the party, however I did a somewhat related test last year, and found parallel is a more optimal arrangement for when whole panels are facing towards/away from the sun: https://diysolarforum.com/threads/blocking-diode-question-please.51248/post-653783 This would be...

I made the TechBuilder 1kW MPPT board. It worked well for some tests, but the MPPT algorithm is somewhat simple and can get stuck on a false maximum power point. The author seems to have abandoned the project too, so it's only worth picking it up if you're able to troubleshoot any issues...

In what order did you turn off the system? EPever controllers are known to fail if the battery is disconnected while solar is still available. BMS cell protection causing a sudden disconnect has also allegedly killed EPever controllers. They rely on the battery for maintaining stability.

Yep, my mistake. All these numbers are meaningless out of context, so let's tabulate them: Setting # Name Range Your configuration 12 Back to utility voltage 22.0V~25.6V 25v 13 Back to battery voltage 24.0V~29.0V 28.5v 19 C.V. charging voltage 24.0V~29.2V 28.8v 20 Floating charging...

Yes, this certainly contributes to the outcome and there are a lot of factors influencing the combined result. As always the real-world outcomes will differ from lab conditions and theory. Assuming some good weather I'll go ahead with the series/parallel shaded tests this weekend. I'm quite...

While I agree with all your points, the DIY market may be bigger than you think. I keep a close eye on eBay and FB Marketplace locally for second hand panels, and all reasonably priced offers sell very quickly. There are even a few solar installers selling the used panels they remove on a...

As this behaviour is "by design", there is no solution. Its just emphasises the need for normal safety measures, including but not limited to: properly earthing all your panels and mounting equipment, double-insulating everythig else, and never working on an energised system. This would apply to...

Yep, it sounds like your stick is stuck in a reboot loop because it can't connect to your Wi-Fi. You definitely don't need the <> characters in secrets.h, they're just placeholders. For example, if your SSID is wireless and your password is password123, the respective lines should look like...

Nice, thanks for the tip @Symbioquine. I've ordered one of these RS485 adaptors and will test with my setup. For my purposes I use a separate ESP8266 and RS485 adaptor for each charge controller. This way I don't have to connect them all together or manage device addresses. The ESP, level...

I bought a couple similar to this a few weeks ago, however didn't end up using them. There's a few variants of these ESP01 relay boards, it's important to use the type that exposes the UART interface (TX, RX, GND) so you can connect it to the BMS. I personally dislike working with the ESP01...

Another thing to keep in mind is that the current most active balancers can move is entirely dependent on the delta voltage between adjacent cells. A 5A balancer might be able to move 5A current under ideal conditions with a very large, unrealistic cell voltage delta (>1V), however if your cells...

yeah, I've learnt not to trust either my BMS or inverter for SOC calculation. Both aren't accurate enough to properly monitor consumption over long periods of time, particularly at low currents. I just watch for the accelerated voltage drop as the pack discharges or call it a day once the cells...

I've had a chance to expand my test battery and sort out the BMS issues that were slowing progress. Testing is nearly complete with most functions working as expected, most importantly the external relay for the active balancer. The shared UART connection still needs a bit of work, which will...

Wow, that's some crazy cell construction. You could try using a high power hot air gun to melt the solder, though I'd be worried about how much heat would transfer to the cells, both through the hot air stream and conduction through the terminals.

There's a lot of reasons the market has moved away from small panel sizes - mostly driven by cost. As average system sizes (kW) increase rapidly, larger panels mean less frame material, less mounting hardware and less electrical connections. This benefits the consumer with lower installation...

I'm the same - I've done the tests both on the ground and on the roof, and at least for my systems, parallel produced more power by quite a large margin. While it's an interesting thought experiment, there's too many variables at play to isolate a single factor and draw conclusions around the...

I think you'll find that a BMS is always powered from the batteries through their balance cables, so the direction of the battery pack current is not relevant to their operation. Because of this they can be active even during a low voltage disconnect and when a charger isn't attached or...

I agree with @timselectric, the bus voltage looks ok, though the BMS must be interacting poorly with the Growatt. Change setting 05 to USE, choose some conservative voltages for your constant voltage and float settings and then monitor. I use 55.2 for CV and 54.7 float with very consistently...

This sounds normal - once your cells are above 3.5v, they're already well above 95% SOC, possibly even higher if only 6A is trickling into them. Regardless of the cell "grade", they'll never be perfectly matched. At some point you'll have one cell reach 3.65v sooner than the others, and as long...

There are two fans in the SPF units, one for the inverter, the other for the MPPT. Are you able to tell which one is running fast? While the inverter fan will always run, it should be at low speed when there's not much load. Mine is audible, but comparable to an average PC when supplying <200w...

I'm not sure if Growatt have implemented a delay for their cutoff, it would be best practice (most BMS do this). Why not test and report back?

Yep, that's all correct. I did a bit of reading since your last post because I realised my lack of knowledge in the area and came to the same conclusion. It's strange that the AERL Earthguard seems to be the only device on the market for this purpose, or at least the only one I could find - does...

I have two panels to work with, I hope that's enough to prove a production difference one way or another. Fortunately I'm in the southern hemisphere, so now is nearly an ideal time for testing. I'll face the panels north/south with an approximate 30 degree tilt. This should be enough to make...

It looks like you've commented out line 31 in main.cpp by adding two slashes at the start of the line. This means all your settings in secrets.h won't be applied. Remove the slashes at the start of the line (or copy/paste this whole block again from GitHub to fix up the commenting), recompile...

Agreed, @mcart117 might be better off replacing the evaporative cooler with an old window air conditioner. They can be bought for less than $200 second hand and would do a far better job, with the added benefit of removing some humidity. It'd use a lot more power, though would only have to run...

Since my last post I've started using 55v for CV and 54.5v for float on USE. Despite only being 0.5v less than my previous settings, the actual charging profile has changed quite a bit, with the Growatt now only starting to charge once the battery drops below 52.5v. This means some days the...

- Higher rated output power (consider your needs into the future, not just now. These DIY solar systems tend to grow in uses over time) - Multiple units can run in parallel (probably not important for you) - Higher surge power (important for starting big loads like power tools, compressors...

That's interesting, I've also had someone over on GitHub suggest that CP2102 UART adaptors work for flashing without having to unsolder anything. It would seem that some UART interfaces probably have stronger drivers that are able to overcome whatever impedence the SIPEX chip causes. A Raspberry...

My other advice is that customer support from Growatt is just about non-existent, which is typical for cheap devices in this category. Once you understand and work out the quirks they're a pretty good unit, but you'll be relying on the community here for help if/when you need it.

Yes, one of the considerations when using an inductor is the flyback when current is removed which can cause massive voltage spikes. Have you considered a simple bypass circuit with a low value (1 ohm?) current limiting resistor and its own smaller MOSFET or relay? This should be enough to...

There is another theoretical advantage of a fixed voltage system. Under certain circumstances (like shading) a string of panels can develop more than one local maximum on their I-V output curves. Less sophisticated MPPT algorithms can get stuck on one of these lower local maximums, which reduces...

Yep, 56v to 5v continual for several weeks without a problem. I also use the 12v variant for running some LED lights. An important note for these units in particular is that their PCBs are all labelled as the 12v output version, you need to test with a DMM to be sure which one you have (i.e. 12v...

Remembering that your cells aren't ideal and perfectly identical, it's not surprising that charging to 14.5V is triggering the BMS, as that would mean each cell is at an average of 3.625v. You already know that you'll see up to 110mV deviation as your cells finish charging, so you'll need to...

All you really need to do is connect the two batteries in parallel with a current limiting device and a fuse for safety. There's plenty of voltage and current adjustable boost converters on Aliexpress that would do the job - you'd just set it to limit current to your desired maximum, and maybe...

This won't work, as the EPEver controllers aren't using Ethernet to communicate. They use MODBUS over RS485 which can support multiple connected devices electrically, however I'm not sure if the EPEver software can communicate with each device in this way as they all need a unique address. Can...

The STM32 would be well suited to the job by itself if you have no need for interaction with other systems. They're cheap, powerful and most importantly, 3.3v logic level to match the BMS.

That's for sure. There's also a big difference between theory and reality. Theory can help inform expectations, but I never assume a whole working system will exhibit the combined behavior of its isolated components, as tested in a lab.

https://szluyuan.en.alibaba.com/

Forget everything you read on that website, none of it is relevant for compiling and uploading the firmware for these adapters. At a high level you must: Download and install PlatformIO Create a new project for an ESP07 with the Arduino framework. Replace the contents of all the files in your...

The BMS wires are definitely on the thin side, but they should also be very short, which will help. The amount of resistance (and therefore voltage drop) in wires is proportional to their length and diameter. Assuming the BMS is of reasonable quality and that you make good connections between...

I have a similar problem to deal with. My batteries will be located in workshop that's exposed to pretty high temperatures. I've picked up a large aluminum toolbox that I'll line with fireproof blankets as thermal insulation and a safety measure. I'm hoping to pick up an office style water...

I think the reason this isn't seen more is the need (or demand) for MPPT. You'll lose quite a bit of efficiency when powering the inverter directly from solar by not maintaining the array at an optimal voltage independent of your load. I'd be interested to see how the actual voltage on the...

I don't think there is any firmware update that resolves the problem detailed in this thread. The issue is caused by the CV charging voltage and float voltage settings. I've set mine to 56.0 and 55.5 volts respectively, which has been running for a couple weeks now with good results. The...

I think it's fair to say that the various different lines of SPF inverters from Growatt have differing charging algorithms, so my advice below might not work for everyone. After experiencing the same problems described by apctjb, I've done a lot of experimentation with my SPF 3500 ES and have...

There's no special process that's specific to these modules, you can just follow the standard steps for flashing an ESP8266: Disconnect all power to the module. Short the GPIO-0 pin to ground. Apply power to the module. It will now start in serial bootloader mode. Connect your USB to UART...

I've seen the same thing - a couple batches of second hand panels I've tested (Sharp and GCL) outperformed their label rating. The Sharp panels were more than a decade old, but looked as good as new after a quick clean. I've seen (and avoided) some bad second hand panels too, though their...

The inverter uses a JST XH 2.54mm connector for the fans with the following pinout: Pin Function Wire Color 1 Control Blue 2 +12v DC Red 3 Sense Yellow 4 GND Black If you're using PC fans as your replacement, you'll quickly notice that the plug type doesn't match, and the...

I've received a CP2102 UART adaptor for testing and confirmed that it will flash the ESP without any modification to the board. This is a far better option than removing the SIPEX chip. They can be bought for a few dollars - I'll be keeping mine handy!

I've uploaded my code to Github: https://github.com/octal-ip/ESP07_Growatt_SPF_3500-5000_ES_Monitor I think I might have nailed the issue with the inverter's MOBUS interface hanging after a while too. I suspect that it doesn't like receiving too many unrecognised commands through the RS232...

RS485 is an interface type, not a protocol, so equipment having an RS485 port doesn't indicate any kind of compatibility. As far as I know the only BMS that can communicate with Growatt inverters natively are the Seplos range. They're quite expensive. You're right that the Growatt charging...

Andy at the Off-Grid Garage has released a video on this topic, his preferred solution is a voltage controlled relay: I prefer to control based on actual cell voltages (and I enjoy the challenge of implementing that, plus other features), but I can't argue with the simplicity of the relay...

I've hit dead-end with the CAN communication. I've completed all the code and hardware to implement it according to the protocol documentation, however the inverter simply won't accept communication over its CAN bus. I've validated the output against other examples of the Pylontech protocol and...

Care to elaborate?

There was a helpful block diagram posted over here: https://diysolarforum.com/threads/growatt-spf-very-high-idle-power.51322/post-689822 It helps to compare how high/low frequency systems operate from a functional perspective. If you want more detail, we have some very knowledgeable members on...

You're right to be concerned about the noise, the 3500 is about the same as an average computer when idling, but very noticeable when under load. I have mine inside my workshop where noise is somewhat acceptable, but I wouldn't be putting it adjacent to living spaces. When running hard it would...

Ok, I suppose it'd work, but don't underestimate the extra space you'll need for cabling, fuses, compression hardware (which is recommended, but not absolutely essential). Some people find peace of mind in using boxes that are somewhat fire resistant too.

I think the problem here is that we're making a false equivalency. The experiment in the video tests just one thing - whether connecting two panels with a very dissimilar Voc/Vmp together operate better in series or parallel when illumination is equal. The results speak for themselves: series is...

I've been reading through this thread, while it seems like you're having some (very dangerous) fun, I feel like I must be missing something. As far as I can tell you've built a system that: Switches specific loads between mains and solar-only supply based on availability of daylight. The...

I'm interested in doing the same thing. Like you, I've found lots of talk about this but no evidence that anyone has been able to get it working end-to-end yet. I've done plenty of work with RS485 before so all it should take is some experimentation and time. I have a SPF 3500 ES as my target...

The voltage regulator is only for running the ADUM1201, which uses a few milliamps of current. It would be powered through the UART port on the BMS which seems to have a variable output voltage between 10v and 14v. This obviously won't ever be higher than the voltage of your actual battery...

I like your angle with this, I'm tempted to build one myself to do a side-by-side test with a standard MPPT controller to record and compare efficiency. I've always wondered how much a panel's Vmp would actually change with clouds, shading and all the other real-world challenges. Looking at the...

Here you go: https://watts247.com/manuals/gw/GrowattModBusProtocol.pdf Looks like the maximum charge current can be set using register #34.

I'm using the 2.4" non-touch version. The STM32, the TFT display and the ADUM1201 combined consume 51mA or 0.26W while running. If you're going to buy the TFT display, make sure it's the 3.3v version. I've attached a photo of the back of mine to help you identify a similar model. Other than...

I'd be interested to hear how your experiments with peltiers go. My initial work with a small phase change refrigeration system didn't produce enough cooling to be worthwhile. I estimate that it was able to move about 100W of heat, however this wasn't enough to keep an insulated container much...