Anker Solix F3800 solar charging with Victron SCC

[RockPaperScissors]

This is a FYI post for anyone searching on this.

I've gotten my Anker Solix F3800 to charge using a Victron MPPT solar charge controller.

The issue is that the solar charge port on the Anker can only take a max of 60 volts (VOC rating on the solar panel). So it will not take 84 VOC from 3 series connected panels as a non specific example. However, the Victron can charge batteries from 12v, 24v and 48v. Using the preset lithium battery setting for 48v and limiting output to 20 amps (max input limit of the F3800) (update: the new max current input is 27amps??) converts the high solar panel voltage into acceptable voltage and current for the Anker. And best of all, the Victron sends a "wake up" voltage signal to the Anker port, activating it and allowing the Victron to monitor the power to regulate the charge.

There is also custom settings in the Victron which appear to allow increasing the charge voltage to the max 60 volts. I will update later on how/if this works.

 

[wwu123]

Friend of mine is using Victron SCC's the same way, nice thing about Victron's is that they can wake up directly from the PV solar input, and so can feed their output directly to the F3800.

I just got a Growatt SCC set up two weeks ago to do the same, but like most SCC's, needs to be powered up from the battery (maybe about 5W max). However, the Anker input does not present itself as a 48V battery per se, and cannot provide even 5W of power output to wake the Growatt. So I do have a 51.2V (16S) server rack battery in parallel, to wake the Growatt and also acts as a buffer to abosrb any solar production in excess of the max 1150W the Anker F3800 can take in.

My friend has also since added server rack batteries to his Victron setup, for more aux storage capacity, but initially used the Victron's with the F3800 without any other battery needed.

 

[Techman Solar]

You could try using a DC buck converter. Set the output to 60v + however many amps you can push through to the F3800. I have used a buck converter to charge power stations from a high voltage battery. Very little energy loss during conversion too.

 

[wwu123]

You could try using a DC buck converter. Set the output to 60v + however many amps you can push through to the F3800. I have used a buck converter to charge power stations from a high voltage battery. Very little energy loss during conversion too.

Interesting you mention this, as before I sprung for the Growatt SCC (which has dual MPPT trackers), currently running 4S, I was running in 4P directly to the F3800 DC inputs, but being clipped by the 27A limit of those inputs.

So before committing to getting a solar charge controller, I did try using a $40 buck converter first, esp since someone on another forum stated that he'd been able to drop a 96 Voc string down to about 58V and still get max Pmpp out of his strings. He was using smaller 200W panels though, and I had larger 41 Voc (31 Vmp) Qcells panels - so I ordered a different DC-DC buck converter rated up to 125V input, since I wanted to get at least 3S.

So the behavior I saw, was that the DC-DC converter did in fact drop the output voltage down to my setting of about 58V, and did allow solar power to the F3800 input without triggering the 60V max limit. That part was good, and worked in 2S and 3S. But strangely, what I noticed was that the actual solar power to the F3800 was not really much higher in 3S than 2S, nor much higher than in 2P in my control testing right before. Backing out the power reported by the F3800 and the 58V going in, I was not seeing the higher amps that should have been tracking at MPP in the V I curve.

Confused, I took the multimeter and tested voltages in the different configurations, and observed that the INPUT voltage to the DC-DC converter - basically on the solar panel side of the buck converter, was not at the expected 93 V (3 x 31 Vmp), but was being pulled down basically to the same voltage as the OUTPUT side of the buck converter, 58V. Similarly, in 2S, instead of seeing 62V (2 x 31 Vmp), I was seeing again 58V at the panels, same as the output side of the buck converter.

I searched high and low on the web for explanations, but finally I think found a few obscure threads probably on this site. That basically said that putting a DC DC buck converter into a MPPT basically renders the MPPT tracking useless and turns it into just a PWM controller, as the MPPT tracking gets confused, and basically is stuck at the low end of the V I curve, so in my case down at 19 V per panel, and correspondingly low current at that part of the curve. My crude understanding of PWM controllers, is basically that they do pull down the entire string to battery voltage, and that's where the panels run at that non-optimal part of the V I curve, so it made some sense.

If that is the case, might not have been bad if I just wanted to run in 2S, as the buck converter was only dropping slightly from ideal 62 Vmp, down to my DC converter setting of 58V, and only shifted slightly off from max Pmp. However, given I preferred to run 3S or 4S, and cheap Chinese controllers were as low as $80 vs $40 for the DC converter, I decided to commit to the MPPT SCC route (though opted for a $400 Growatt).

I still wonder though, did I just choose the wrong DC converter, and are there others that would not have confused the Anker F3800 internal MPPT (given there was one claim of success on another forum), and would have allow correct MPP tracking?

 

[Techman Solar]

Interesting you mention this, as before I sprung for the Growatt SCC (which has dual MPPT trackers), currently running 4S, I was running in 4P directly to the F3800 DC inputs, but being clipped by the 27A limit of those inputs.

So before committing to getting a solar charge controller, I did try using a $40 buck converter first, esp since someone on another forum stated that he'd been able to drop a 96 Voc string down to about 58V and still get max Pmpp out of his strings. He was using smaller 200W panels though, and I had larger 41 Voc (31 Vmp) Qcells panels - so I ordered a different DC-DC buck converter rated up to 125V input, since I wanted to get at least 3P.

So the behavior I saw, was that the DC-DC converter did in fact drop the output voltage down to my setting of about 58V, and did allow solar power to the F3800 input without triggering the 60V max limit. That part was good, and worked in 2S and 3S. But strangely, what I noticed was that the actual solar power to the F3800 was not really much higher in 3S than 2S, nor much higher than in 2P in my control testing right before. Backing out the power reported by the F3800 and the 58V going in, I was not seeing the higher amps that should have been tracking at MPP in the V I curve.

Confused, I took the multimeter and tested voltages in the different configurations, and observed that the INPUT voltage to the DC-DC converter - basically on the solar panel side of the buck converter, was not at the expected 93 V (3 x 31 Vmp), but was being pulled down basically to the same voltage as the OUTPUT side of the buck converter, 58V. Similarly, in 2S, instead of seeing 62V (2 x 31 Vmp), I was seeing again 58V at the panels, same as the output side of the buck converter.

I searched high and low on the web for explanations, but finally I think found a few obscure threads probably on this site. That basically said that putting a DC DC buck converter into a MPPT basically renders the MPPT tracking useless and turns it into just a PWM controller, as the MPPT tracking gets confused, and basically is stuck at the low end of the V I curve, so in my case down at 19 V per panel, and correspondingly low current at that part of the curve. My crude understanding of PWM controllers, is basically that they do pull down the entire string to battery voltage, and that's where the panels run at that non-optimal part of the V I curve, so it made some sense.

If that is the case, might not have been bad if I just wanted to run in 2S, as the buck converter was only dropping slightly from ideal 62 Vmp, down to my DC converter setting of 58V, and only shifted slightly off from max Pmp. However, given I preferred to run 3S or 4S, and cheap Chinese controllers were as low as $80 vs $40 for the DC converter, I decided to commit to the MPPT SCC route (though opted for a $400 Growatt).

I still wonder though, did I just choose the wrong DC converter, and are there others that would not have confused the Anker F3800 internal MPPT (given there was one claim of success on another forum), and would have allow correct MPP tracking?

Thank you for the detailed explanation. I must admit - my application was solely with a high voltage battery. In your case, there are a number of factors that may contribute to what you experienced, such as current fluctuations (due to weather changes), the MPPT algorithms of the F3800 and (as you rightly pointed out) the DC buck converter itself.

 

[stephenrx7]

Friend of mine is using Victron SCC's the same way, nice thing about Victron's is that they can wake up directly from the PV solar input, and so can feed their output directly to the F3800.

I just got a Growatt SCC set up two weeks ago to do the same, but like most SCC's, needs to be powered up from the battery (maybe about 5W max). However, the Anker input does not present itself as a 48V battery per se, and cannot provide even 5W of power output to wake the Growatt. So I do have a 51.2V (16S) server rack battery in parallel, to wake the Growatt and also acts as a buffer to abosrb any solar production in excess of the max 1150W the Anker F3800 can take in.

My friend has also since added server rack batteries to his Victron setup, for more aux storage capacity, but initially used the Victron's with the F3800 without any other battery needed.

I have an EG4 MPPT100-48V charge controller going into one of the XT60 ports of my Anker F3800. I think I am running into the same issue as you all where the MPPT is not charging the F3800 because there is no pulse signal. So, I actually had the same idea of putting a battery in between the EG4 charge controller and the F3800. My question is, what battery do I need and do I need some sort of charge controller between the battery and F3800.

Have you guys found another way of faking the pulse signal between the F3800 and an MPPT charge controller?

Thanks!
Stephen

 

[wwu123]

I have an EG4 MPPT100-48V charge controller going into one of the XT60 ports of my Anker F3800. I think I am running into the same issue as you all where the MPPT is not charging the F3800 because there is no pulse signal. So, I actually had the same idea of putting a battery in between the EG4 charge controller and the F3800. My question is, what battery do I need and do I need some sort of charge controller between the battery and F3800.

Have you guys found another way of faking the pulse signal between the F3800 and an MPPT charge controller?

Thanks!
Stephen

Does that EG4 unit power itself up each morning from just the PV power alone, without a battery connected? If not, then I think it needs a nominal battery voltage and probably a few watts from the battery it is charging (which the F3800 will not supply even a bit of power from the DC MPPT inputs).

So yeah, you would wire up a buffer battery, even a small one to the charge controller, and the F3800 input as well to the battery - so all in parallel. What I (with Growatt) and my friend (with Victron) have seen is when the charge controller is producing solar power, the first (up to) 1150W will be drawn into the F3800 if not full, in fact if PV is not enough it will draw from the buffer battery until it gets its 1150W. Only any excess above 1150W will go into the buffer, and if the F3800 AND buffer battery get full, then the charge controller will stop according to its bulk and float voltage settings as usual.

This behavior makes sense, as the F3800 input can take in anything from 12 to 60V, while also having an internal 51.2V 16S nominal cell structure, so it has a buck-boost converter to raise/lower voltage and MPPT tracking - so it should be able to set its input voltage to be lower than the buffer battery and thereby draw from PV and buffer battery til it reaches its 1150W max.

With the EG4 being a 48V charge controller, you'd want to use a nominal 48V (or ideally a 51.2V 16S LFP one) so the EG4 MPPT can still taper solar power when the buffer reaches full. But it can be a very small 48V buffer battery if you just want to use it to wake the EG4 to feed the F3800. However, many find you may as well get a larger server rack battery (5kwh), or even multiple, as it's cheaper way of expanding the effective capacity of the F3800 rather than buying the proprietary expansion batteries.

The other thing you'd want is a low voltage cutoff between the buffer and the F3800 input, as the F3800 will take suck the buffer down til empty, and if you trip the buffer's internal BMS protection, then the EG4 would not be able to wake up on its own the next morning. Whereas the separate LV cutoff allows the buffer to still keep the EG4 nominally powered all the time.

 

[RockPaperScissors]

An update

My experimentation time got cut off as the F3800 failed on me. It wouldn't power back on at all. Plugging the unit into 120v AC would bring the display back online, but the logs were showing failures from each of the relay controlled outputs. Of course it refused to charge. Took a week to get support to accept an RMA. Sent it back after which they sent out a new unit. Thankfully I kept the box and packing materials. Shipping back and forth took another 2 weeks. New unit works, but the lack of cloud free days is hampering my enthusiasm.

I think the BMS failed on the old unit since it would not accept a charge or allow discharge unless it had external power. Logs are terrible on the consumer side for figuring that out. There is an option to send more complete logs to support in the mobile app, but I didn't get access to the file.

 

[stephenrx7]

Does that EG4 unit power itself up each morning from just the PV power alone, without a battery connected? If not, then I think it needs a nominal battery voltage and probably a few watts from the battery it is charging (which the F3800 will not supply even a bit of power from the DC MPPT inputs).

So yeah, you would wire up a buffer battery, even a small one to the charge controller, and the F3800 input as well to the battery - so all in parallel. What I (with Growatt) and my friend (with Victron) have seen is when the charge controller is producing solar power, the first (up to) 1150W will be drawn into the F3800 if not full, in fact if PV is not enough it will draw from the buffer battery until it gets its 1150W. Only any excess above 1150W will go into the buffer, and if the F3800 AND buffer battery get full, then the charge controller will stop according to its bulk and float voltage settings as usual.

This behavior makes sense, as the F3800 input can take in anything from 12 to 60V, while also having an internal 51.2V 16S nominal cell structure, so it has a buck-boost converter to raise/lower voltage and MPPT tracking - so it should be able to set its input voltage to be lower than the buffer battery and thereby draw from PV and buffer battery til it reaches its 1150W max.

With the EG4 being a 48V charge controller, you'd want to use a nominal 48V (or ideally a 51.2V 16S LFP one) so the EG4 MPPT can still taper solar power when the buffer reaches full. But it can be a very small 48V buffer battery if you just want to use it to wake the EG4 to feed the F3800. However, many find you may as well get a larger server rack battery (5kwh), or even multiple, as it's cheaper way of expanding the effective capacity of the F3800 rather than buying the proprietary expansion batteries.

The other thing you'd want is a low voltage cutoff between the buffer and the F3800 input, as the F3800 will take suck the buffer down til empty, and if you trip the buffer's internal BMS protection, then the EG4 would not be able to wake up on its own the next morning. Whereas the separate LV cutoff allows the buffer to still keep the EG4 nominally powered all the time.

Thanks for this detailed writeup. This will be very helpful.

The EG4 powers up as soon as it received power from solar, but there is no output from the battery charge output- so I think you are right that it needs battery voltage to start charging.

Very good suggestion regarding the LV cutoff. Are you using one on your system that you can recommend?

Thanks again, this is all very helpful!!!
Stephen

 

[stephenrx7]

An update

My experimentation time got cut off as the F3800 failed on me. It wouldn't power back on at all. Plugging the unit into 120v AC would bring the display back online, but the logs were showing failures from each of the relay controlled outputs. Of course it refused to charge. Took a week to get support to accept an RMA. Sent it back after which they sent out a new unit. Thankfully I kept the box and packing materials. Shipping back and forth took another 2 weeks. New unit works, but the lack of cloud free days is hampering my enthusiasm.

I think the BMS failed on the old unit since it would not accept a charge or allow discharge unless it had external power. Logs are terrible on the consumer side for figuring that out. There is an option to send more complete logs to support in the mobile app, but I didn't get access to the file.

My first F3800 experienced the same issue. They sent me an advanced replacement that's been working fine since (about 4 months).

 

[wwu123]

Thanks for this detailed writeup. This will be very helpful.

The EG4 powers up as soon as it received power from solar, but there is no output from the battery charge output- so I think you are right that it needs battery voltage to start charging.

Very good suggestion regarding the LV cutoff. Are you using one on your system that you can recommend?

Thanks again, this is all very helpful!!!
Stephen

I've been using a $16 low voltage cutoff from Amazon that claims it's rated for over 60V and 30A output, it's mostly worked fine for two months now, but I won't post it here because I don't think it's really going to last. Some of the Amazon reviews point out how small the traces are on the circuit board, how they are very thin compared to 10AWG wire rated for 30A - indeed mine gets very hot, some of the board components exceed 130 deg F for hours each day while the circuit is passing 27A for 5-6 hours a day.

So I'm back searching on the forum for better quality LV cutoff options myself. The Victron BatteryProtect would be a quality, reliable option, but also 10X the price of the Amazon one. Not sure I'm finding anything in-between in price or quality, but will keep searching....

 

[roberty3k]

Do you guys think this will/should work:

Because the F3800 does not allow for parallel charging through the XT60 ports off a single source, I'm thinking of doing 2 individual solar arrays(roughly 1200w per), each with a PowMr 100A MPPT controller charging a 48v 100ah battery and each plugged to an XT60 port.

Would it be correct to assume, the XT60s will draw 27 amps of current, and that should theoretically leave 73amps charging each battery?

 

[BarryBenson]

This is a FYI post for anyone searching on this.

I've gotten my Anker Solix F3800 to charge using a Victron MPPT solar charge controller.

The issue is that the solar charge port on the Anker can only take a max of 60 volts (VOC rating on the solar panel). So it will not take 84 VOC from 3 series connected panels as a non specific example. However, the Victron can charge batteries from 12v, 24v and 48v. Using the preset lithium battery setting for 48v and limiting output to 20 amps (max input limit of the F3800) (update: the new max current input is 27amps??) converts the high solar panel voltage into acceptable voltage and current for the Anker. And best of all, the Victron sends a "wake up" voltage signal to the Anker port, activating it and allowing the Victron to monitor the power to regulate the charge.

There is also custom settings in the Victron which appear to allow increasing the charge voltage to the max 60 volts. I will update later on how/if this works.

Can I ask which particular Victron MPPT you are using for this?

 

[RockPaperScissors]

Can I ask which particular Victron MPPT you are using for this?

I have 2.
The 100/20 and the 150/30. (Or was it 150/35?)
The important part is to check the specs or the side of the case as it should mention that it will support 12, 24, 48v charging.

 

[schmenkee]

This is a FYI post for anyone searching on this.

I've gotten my Anker Solix F3800 to charge using a Victron MPPT solar charge controller.

The issue is that the solar charge port on the Anker can only take a max of 60 volts (VOC rating on the solar panel). So it will not take 84 VOC from 3 series connected panels as a non specific example. However, the Victron can charge batteries from 12v, 24v and 48v. Using the preset lithium battery setting for 48v and limiting output to 20 amps (max input limit of the F3800) (update: the new max current input is 27amps??) converts the high solar panel voltage into acceptable voltage and current for the Anker. And best of all, the Victron sends a "wake up" voltage signal to the Anker port, activating it and allowing the Victron to monitor the power to regulate the charge.

There is also custom settings in the Victron which appear to allow increasing the charge voltage to the max 60 volts. I will update later on how/if this works.

Would this also work for the earlier Anker Powerhouse 767 that had similar input restrictions?

 

[mjrkong]

I have 2.
The 100/20 and the 150/30. (Or was it 150/35?)
The important part is to check the specs or the side of the case as it should mention that it will support 12, 24, 48v charging.

Wondering if you have any more on your results with the setup ? And couple of specific questions:

Have you been able to confirm the 60v output option on the Victrons ?

Do you know if it would be possible to use a single SCC to feed 2 separate F3800s ?

 

[Andrewr05]

Would this also work for the earlier Anker Powerhouse 767 that had similar input restrictions?

Yes.

Power is power, no matter where it comes from.
A Power Station/Solar Generator doesn't know or care where to the Volts and Amps come from, as long as they're within the units limits.

 

[wwu123]

Wondering if you have any more on your results with the setup ? And couple of specific questions:

Have you been able to confirm the 60v output option on the Victrons ?

Do you know if it would be possible to use a single SCC to feed 2 separate F3800s ?

Not sure what you mean by 60V output option, but you would use the Victron with its 48V settings, as nominal "48V" 16S battery actually is 48-56V from empty to full, so already near the 60V limit of the F3800 input.

The challenge with feeding two F3800's directly from the SCC is the two MPPT's may start hunting or fighting with each other. When one is sweeping the voltage range, can pull down the SCC output voltage and confuse the other F3800 input. So can have inconsistent behavior.

But if you use a SCC with a 48V battery in parallel as a buffer, then likely to work fine, as the F3800 MPPT's can't really pull down the voltage of the buffer battery. Many are feeding two F3800's fine with a single SCC and 48V server rack battery.

 

[mjrkong]

Not sure what you mean by 60V output option, but you would use the Victron with its 48V settings, as nominal "48V" 16S battery actually is 48-56V from empty to full, so already near the 60V limit of the F3800 input.

The challenge with feeding two F3800's directly from the SCC is the two MPPT's may start hunting or fighting with each other. When one is sweeping the voltage range, can pull down the SCC output voltage and confuse the other F3800 input. So can have inconsistent behavior.

But if you use a SCC with a 48V battery in parallel as a buffer, then likely to work fine, as the F3800 MPPT's can't really pull down the voltage of the buffer battery. Many are feeding two F3800's fine with a single SCC and 48V server rack battery.

OP @RockPaperScissors mentioned that he was trying to get his Victron SCCs to output 60v via some kind of hidden setting. I searched around and could only find mentioned of additional 36v option for the SCCs. So I was wondering if OP had any luck with 60v output.

I am asking if anybody had any luck charging 2 F3800 via single SCC precisely because of the potential conflict between 2 F3800s. Would be nice to know if it worked for anyone.

I read in this thread that one reason to use Victron SCC vs cheaper MPPTs is that they are confirmed to work without additional buffer battery.

What I am looking to do is connect a 6x 400w-ish array to the F3800s with 150ft of wire. I thought it would be great to have them on in single string with high voltage, low current going into both F3800s via a single SCC. However, it seems like 2x 150-35 SCCs are cheaper than a single high voltage model, so it just comes down to single vs dual strings and wiring.

 

[dellmed]

This is a FYI post for anyone searching on this.

I've gotten my Anker Solix F3800 to charge using a Victron MPPT solar charge controller.

The issue is that the solar charge port on the Anker can only take a max of 60 volts (VOC rating on the solar panel). So it will not take 84 VOC from 3 series connected panels as a non specific example. However, the Victron can charge batteries from 12v, 24v and 48v. Using the preset lithium battery setting for 48v and limiting output to 20 amps (max input limit of the F3800) (update: the new max current input is 27amps??) converts the high solar panel voltage into acceptable voltage and current for the Anker. And best of all, the Victron sends a "wake up" voltage signal to the Anker port, activating it and allowing the Victron to monitor the power to regulate the charge.

There is also custom settings in the Victron which appear to allow increasing the charge voltage to the max 60 volts. I will update later on how/if this works.

Were you able to figure out increasing the charge voltage to 60V

 

[mjrkong]

Were you able to figure out increasing the charge voltage to 60V

Start with 48v, then create a custom profile. Under Charge Voltages, change Absorption to 60. It ends up being very close to regular lithium preset , just a little bit higher voltage.

 

[mjrkong]

The challenge with feeding two F3800's directly from the SCC is the two MPPT's may start hunting or fighting with each other. When one is sweeping the voltage range, can pull down the SCC output voltage and confuse the other F3800 input. So can have inconsistent behavior.

@wwu123 This does in-fact happen. 2 F3800 continually adjust the power between very low and full string output.