Epoch 460AH Marine V2

[guy with a boat]

What is there were a device at the alternator output that monitored the output voltage and switched off the alt if the voltage went outside adjustable voltage brackets? It could also trigger a light or alarm to signal a problem. Seems like a very simple device with minimal complications.
The main threat to the LFP batts overcharging (and BMS action for that fault) seems to be the alternator. I think it is pretty rare for a modern charger to go wonky enough to produce damaging charge voltage. We have surge suppressor devices to save the alternator in a BMS disconnect, but maybe a device to eliminate the risk of over-voltage from the alternator/regulator would get around the risk and concern of BMS disconnect from that cause.
It might be belt and suspenders similar to an exhaust temp sensor to warn you if you lose an impeller.
The other causes for BMS disconnect are less likely to take out the whole bank, so maybe alternator voltage risk management deserves a closer look. I agree the risk of all this is low, but I still run Sterling APDs on both alternators.

Sorry to veer so far off the Epoch track.

 

[Marco Flamingo]

I'll start right out with sorry to veer so far off the Epoch track.

I appreciate that in a well set up charging system the disconnect *should* never occur. However whenever possible systems should be made so that failure is graceful with little collateral damage. Old LA systems were better at this, and in that respect LFP are a step backwards.

Did you miss the lead acid "graceful failure with little collateral damage" shown in post #48? But wait, there's more. This is what comes to mind when I think about the collateral damage of "good old lead acid batteries." Doesn't even make the news anymore, but I don't want a LA in parallel with an LFP battery to "fix" charging problems.

Just a reminder to wear safety glasses when checking out a lead acid battery charging issue.

On a less gruesome note, I'm still waiting for somebody to step up and determine whether a bank of LFP batteries (i.e., a bank of BMSs) "solves" the alternator problem. I'm not even sure that using the term "problem" is accurate. I simply haven't seen the evidence yet. Some of us have probably taken measures to protect from a lightening strike, but how many have taken measures to protect from a meteorite strike? You can see the risk analysis issue. You have to know the likelihood of the occurrence before you can assess an appropriate mitigation. Does adding several more LFP batteries (i.e., several more BMSs) reduce the risk such that it is "solved?" And by solved, I only mean reducing it to the likelihood of a lightening strike, not a meteorite strike. I haven't seen anybody claim that a BMS shutdown is worse than a lightening strike, yet it is getting more attention at present despite the fact that a lightening strike will definitely take out your electronics (and maybe you and your crew).

Complicating matters when assessing BMS shutdowns is the quality difference in LFP batteries and their corresponding battery management systems. How will the ABYC ever decide which, if any, battery or battery manufacturer makes a product that shuts down without incident and additional fusing? The external electronic gadget "fixes" seem to have their own issues. I've seen a couple Victron failures on the web that make me question whether the various external gadget solutions are the best "belt" to wear with the BMS "suspenders."

Finally, now for some Epoch 460AH Marine V2 content. It looks like Epoch has done their homework for this battery for some applications. But they are servicing a large and divergent market. I'm not convinced that a single big battery is what is needed for some applications, especially those using an alternator. First, there is the weight. I just pulled out my 75# lead acid batteries. Putting in a couple of 92# LFP batteries doesn't sound appealing. For my use, I would have one in each bank, meaning that if using the 1 or 2 setting on my battery switch, I would be in the position of alternator issues if a BMS shut down because of only a single BMS in each bank.

A single Epoch 460AH batteries looks like a more compact (actually svelte) solution to having battery banks of smaller batteries, say using 5 100AH LFP. The Epoch cabling also looks simpler, assuming that the 5 BMSs of the smaller bank doesn't provide the requisite alternator protection. If an Epoch 460AH system requires DC/DC or other items to protect the alternator, then the cabling would be more complicated than at first glance, and the overall cost is probably more than a bank of 5 100AH LFP. On a boat, where a 5 battery bank isn't an impossible feat, simpler/cheaper may be the better choice. Granted, they don't look as cool.

 

[twistedtree]

Yeah, lead batteries. Full of:

- Hydrogen, one of the most explosive gasses on the planet

- Sulfuric acid, one of the most corrosive substances on the planet

- Lead, one of the most toxic substances on the planet.

I know, let's mix them together and use then everywhere. What could go wrong?

 

[Barking Sands]

Assuming that's how the battery is designed/built.

Pretty much all drop ins operate this way now. There might be a few stragglers or odd builds, but I havent seen one recently.

 

[Barking Sands]

And by solved, I only mean reducing it to the likelihood of a lightening strike, not a meteorite strike.

LOL...great way to put it. How do we define acceptable risk. I can say that Rod Collins has stated many times that a properly programmed external regulator for the battery envelope along with an APD is more than acceptable. I can't disagree.

 

[BDofMSP]

If an Epoch 460AH system requires DC/DC or other items to protect the alternator, then the cabling would be more complicated than at first glance, and the overall cost is probably more than a bank of 5 100AH LFP. On a boat, where a 5 battery bank isn't an impossible feat, simpler/cheaper may be the better choice. Granted, they don't look as cocool.But

But a 5 battery bank of 100 AH FLAs gets you 250 usable AH on a regular basis. So not even close to the 400+ from the Epoch. So it's more than just looking cool. Plus 5 Group 31s weighs well over 300 lbs vs 97 so...

 

[twistedtree]

Pretty much all drop ins operate this way now. There might be a few stragglers or odd builds, but I havent seen one recently.

I agree WRT drop-ins, but there are battery systems that are not self-contained cells+BMS. A number of companies make cell packs w/ monitoring & balancing circuitry, and external BMS that controls multiple cell packs, and sometimes even a separate contactor.

 

[Barking Sands]

Just about done testing the V2. But I have to say..the CAN control is MUCH more sophisticated than I expected. In a good way. Ill have a thorough write up soon. Some of what I have stated is not exact. So I'll have to revise those details too.

 

[Barking Sands]

This is my own personal experience so far with the V2 and it just my best observations to share with the group. Since it is my own observations, some things could be in error, but I have done my best to figure out the details. Keep in mind there are already some updates coming very soon to refine the various parameters as well as information reported on the app and to Victron. So some of these details may change slightly.

* Main harness and gauge connected to master battery into the blue port. The bluetooth is located in the gauge so the gauge must be plugged in to utilize the bluetooth app.
RJ45 cable out from master black port and into slave blue port. Additional slave's daisy chained out the black port and into the next slave blue port and so on.

* Grey Victron plug of harness plugged into BMS-CAN port on Cerbo. Or if you have the new Cerbo with no BMS CAN then you plug into either of the VE-CAN 1 or 2 plugs. You will have the change the port characteristics to BMS 500KBPS
You will only see 1 battery on the details in the Victron/touch 50/70 since all batteries are combined by the master and then reported to Cerbo.
DVCC on, select controlling BMS to Epoch, Select battery monitor to Epoch (do not use Victron shunt for battery monitor), Shared voltage sense, shared temp sense and shared current sense off.

* When the battery SOC is <90 the CCL is at maximum (Charge current limit as seen on the touch 50/70 Epoch Battery > parameters). Initiating a charge at this time will result in the inverter/charger outputting maximum current to charge the battery. The inverter/charger will be in BULK mode at that time. When the voltage gets to the absorption voltage as directed by the Epoch BMS-CAN the Inverter will show ABSORPTION and CCL will be reduced from maximum to a mid-range value. When absorption is done, the Inverter will again show bulk, but CCL will drop to 0 and battery voltage will fall to whatever float voltage is commanded by the Epoch BMS-CAN. Usually between 13.8 and slowly falling to 13.3/13.4 and an SOC between 98 or 98%. It may take some time for battery voltage to naturally fall. If you discharge from this point to an SOC less than 98% but more than 89.9% and re-engage the charger, CCL will remain at 0 but the charger will charge at a reduced current between 10-20 amps back to the 97 to 98% SOC range and stop.

* Re-bulk: <90% SOC. CCL is raised to full values when charging is engaged at <90% SOC)

* Commissioning the batteries, especially multiple batteries, will require some patience while the system balances and aligns the cells/packs. If you are unable to reach the CVL voltage without the red-light alarm occurring, enable and set the “Limit managed battery charge voltage” setting in the DVCC. Lower the voltage until the red-light warning stops. Cycle the system and leave it to absorb. Over time, as the batteries balance, Increase the voltage and cycle the battery to less than 90% SOC to initiate another bulk charge cycle. Repeat until you can turn off the “Limit managed battery charge voltage”. The larger the number of parallel batteries the longer it will take to align the cells/packs. When all cells/packs are aligned and the battery reaches full absorption as determined by the master BMS the battery will read 100% at that time and CCL will immediately drop to 0.

* Commissioning can be done on the bench with a charger such as an IP22 or power supply with the batteries isolated or set up as master/slave with a gauge. Raise the charge/absorption voltage as needed with cycles down to 95% in between each .1 volt increase. Yu can do this with batteries isolated but eventually it would be a good idea to do this with the batteries set and communicating as master/slave since the balancing is not only for each battery but tween batteries as well.

* If you are not using comms for Victron equipment, set absorption to 14.2 volts and float to 13.5 volts. If you get the red warning on initial charges, lower the absorption voltage until you can cycle without the red light. Increase absorption voltage over subsequent cycles until you reach 14.2 without the red light.

One additional observation is there does not appear to be an FCP on this. What I initially thought was FCP engaging around 14.3 was just cell balancing. I have seen as high as 14.5 momentarily with no shut down. The commissioning of these batteries will take a bit of patience because when communicating as one bank the system not only balances the cells in each battery but between packs from what I can see. A bank of 8 would take some time to align. Could be a few weeks. It appears the way these run is very similar to Pylontech bay making the CCL dynamic instead of CVL.

The SVS off may just be required for commissioning. Later after some additional cycling I will tun it back on and see how it affects the Victron world.

Attached is a picture of some of the details of the commissioning process towards running semi normally. I suspect over subsequent cycles the pattern will calm down to a more typical discharge > bulk > absorb > float pattern.

 

[rslifkin]

In normal use you should have Shared Voltage Sense turned on. That will cause the Multiplus and MPPTs to use the voltage from the DVCC selected source (the Epoch BMS in this case) rather than the voltage sensed at their terminals, so minor voltage drop won't affect charging. I'm curious why you found it necessary to turn SVS off for commissioning, as it shouldn't prevent you from doing any of the steps you described.

Shared Current Sense probably doesn't matter in this application as you're not using the built in charge algorithm of the MPPTs and such, so being able to use shunt amps to end absorb (instead of MPPT output amps) doesn't matter. And Shared Temp Sense doesn't matter with LFP as you're not doing temperature compensation.

 

[Barking Sands]

In normal use you should have Shared Voltage Sense turned on. That will cause the Multiplus and MPPTs to use the voltage from the DVCC selected source (the Epoch BMS in this case) rather than the voltage sensed at their terminals, so minor voltage drop won't affect charging. I'm curious why you found it necessary to turn SVS off for commissioning, as it shouldn't prevent you from doing any of the steps you described.

Shared Current Sense probably doesn't matter in this application as you're not using the built in charge algorithm of the MPPTs and such, so being able to use shunt amps to end absorb (instead of MPPT output amps) doesn't matter. And Shared Temp Sense doesn't matter with LFP as you're not doing temperature compensation.

For whatever reason as we were in a lengthy absorption for cell balance turning on svs was causing voltage rise which caused overvolt quicker this was on a freinds boat. We found leaving it off smoothed out the process. I had svs on during my commissioning and had no issues. I am going to cycle again in an hour and I'll turn on svs from here on out and see. I agree svs should be on long term. I'm pretty sure it will be ok.

 

[MV Cricket]

I am trying to commission my two bank batteries as well. I took the bank down to 30% and charged them back up. Got the red light flashing on the master after about 5 minutes.
In the morning, the red light was flashing on the slave battery as well.
The app won't tell me any history on the batteries. When pressing the history, I get the no devices connected message. That is kinda frustrating.
I have SVS on right now and during that 30% discharge. I might try to turn off the SVS and see what happens.
Should I also low the voltage limit during the next charge cycle after I discharge them?

Thanks for all the great info here! Really appreciate the time Barking Sands has been putting in and sharing with the forum.

 

[Barking Sands]

I am trying to commission my two bank batteries as well. I took the bank down to 30% and charged them back up. Got the red light flashing on the master after about 5 minutes.
In the morning, the red light was flashing on the slave battery as well.
The app won't tell me any history on the batteries. When pressing the history, I get the no devices connected message. That is kinda frustrating.
I have SVS on right now and during that 30% discharge. I might try to turn off the SVS and see what happens.
Should I also low the voltage limit during the next charge cycle after I discharge them?

Thanks for all the great info here! Really appreciate the time Barking Sands has been putting in and sharing with the forum.

There is an update maybe in a week for the battery to add the missing history as well as a few details reported to Cerbo. There are several other updates as well. Do you have them set as mater/slave?

 

[MV Cricket]

There is an update maybe in a week for the battery to add the missing history as well as a few details reported to Cerbo. There are several other updates as well. Do you have them set as mater/slave?

Good to know about the updates.
Yep, they are set as master and slave.

 

[MV Cricket]

Just a bit of an update.
I discharged the batteries down to 45% or so and then charged them up with the SVS off as Barking Sands suggested.
Batteries went to 99% and floated. No error lights.
Also, the batteries aren't doing what they have previously when floating, which was bouncing, slowly, between 90% and 99%. They are just staying at 99%. So that is new.
Here is a screenshot from this morning. About 36 hours after the discharge/charge cycle. I am a bit concerned about the VDIFF but I think that will settle once I get a few more cycles in. Have three cycles only so...
I am thinking I'll try to cycle the bank with SVS on next time. Thoughts?
Should I keep SVS off? Currently it is.
Thanks for all the help!

 

[Barking Sands]

Check the parameters on the device list in the battery tab. CCL should be at 0 and it should reduce voltage to around 13.4 and 98% SOC. Have you done the latest update? Looks like you still need the update. On your app it shows v0.0. The latest update will be v2.4. Be sure to delete and reload the app before doing the firmware update. Click OTA and follw the prompts. you only need to do the update on the top line after clicking OTA. When done it should say V2.4 in your app.

The latest update will not show the few missing items in the Cerbo info such as highest cell, lowest cell, high and low temps, number of modules online and total AH.

I still have some erratic cycling at the top of the charge, either hanging in absorption or cycling to CCL of 0 but with a red light on one battery which shuts off the charge mosfet on that particular battery. I spoke with the big brains at Epoch and they stated they will have another update to address this. My guess is they will need to relax a few parameters regarding the targets.

 

[MV Cricket]

Check the parameters on the device list in the battery tab. CCL should be at 0 and it should reduce voltage to around 13.4 and 98% SOC. Have you done the latest update? Looks like you still need the update. On your app it shows v0.0. The latest update will be v2.4. Be sure to delete and reload the app before doing the firmware update. Click OTA and follw the prompts. you only need to do the update on the top line after clicking OTA. When done it should say V2.4 in your app.

The latest update will not show the few missing items in the Cerbo info such as highest cell, lowest cell, high and low temps, number of modules online and total AH.

I still have some erratic cycling at the top of the charge, either hanging in absorption or cycling to CCL of 0 but with a red light on one battery which shuts off the charge mosfet on that particular battery. I spoke with the big brains at Epoch and they stated they will have another update to address this. My guess is they will need to relax a few parameters regarding the targets.

The Epoch firmware version in the Ekranos says it is 2.3 or something but sure enough, in that screenshot, it says V0.0
The attached photos are from my Android. When I look at the my Apple screenshots, they don't have a version number.
Here is the screenshot from my iPad.

I wonder if the version number is in a different place on the iPad.
I'll have to recheck all this when I get back to the boat.
Thanks for the info! I'll update this later today and let know you what I find.

 

[MV Cricket]

As it turns out, with the Android phone app anyways, when you click on the Dashboard in the Epoch app, it takes about 5-6 seconds for the Firmware version to show up. I was grabbing screenshots immediately hence the V0.0.
I went about updating the firmware to V2.4 and was unsure if I should just leave the batteries as one big virtual bank and update it that way (it seems like it would let me) or separate them. I decided to make them individual batteries again and changed the DIP switch reflect that. I then ran the update to each battery individually. No problems, they both updated fine.
I tried to use the iPad to do this update but the OTA password is not "test". No idea there.
I am sure that I have to do more cycles as suggested on page 2 of this thread but I am seeing the batteries at 99% but they never go into float. They are always hanging out at 14.2 volts. That concerns me.
The batteries in the app show they are full or nearly so but in the Ekranos (with the new update showing cells and stuff) I am at 820 amps with a total of 920. I'll add some screenshots.
I also woke up to both batteries having the error light flashing and the Charge MOS was switch off in the app.
I turned the SVS on and maybe that had something to do with it.
Hopefully I am not burning up too much goodwill here. I am very appreciative of the help here!

 

[MV Cricket]

I don't know how to edit my above post.
My CCL is at 220. I can't adjust it from there.

 

[SteveK]

I don't know how to edit my above post.
My CCL is at 220. I can't adjust it from there.

you should see EDIT bottom left corner after report

 

[Jeff F]

For whatever reason as we were in a lengthy absorption for cell balance turning on svs was causing voltage rise which caused overvolt quicker this was on a freinds boat. We found leaving it off smoothed out the process. I had svs on during my commissioning and had no issues. I am going to cycle again in an hour and I'll turn on svs from here on out and see. I agree svs should be on long term. I'm pretty sure it will be ok.

Thanks for sharing your learning and experience. I tuned out battery discussions while I was off living with mine and am just getting caught up.

I took interest in your commissioning steps. They make sense, and I can see how turning off svs would allow faster balancing.

I hope this all gets turned into end-user guidance and future design decisions. I understand your comments on different use cases, but having the BMS take on a routine charge termination role sounds a bit problematic to me.

Is Epoch looking for product testers for their 48v marine battery?

 

[mvweebles]

Knowledgeable tear-down video on Epoch 460 V2. Reviewer is pretty good in general. Not as flashy as Will Prowse but solid nonetheless

It's a bit nerdy but on point for this thread.

Peter

 

[Jeff F]

One more question, getting back to conditioning/balancing: do you have any insight from the battery into active balancing activity?

Put another way, is repeating charge cycles beneficial only because it allows the BMS to actively balance, or is the benefit entirely in just allowing the battery to spend a bit more time at the top and self-balance? Or some combination?

 

[Barking Sands]

Thanks for sharing your learning and experience. I tuned out battery discussions while I was off living with mine and am just getting caught up.

I took interest in your commissioning steps. They make sense, and I can see how turning off svs would allow faster balancing.

I hope this all gets turned into end-user guidance and future design decisions. I understand your comments on different use cases, but having the BMS take on a routine charge termination role sounds a bit problematic to me.

Is Epoch looking for product testers for their 48v marine battery?

Keep in mind all drop in batteries terminate the charge if you exceed the various criteria. But the V2 does need some tweaks on the CAN control side. Luckily The OTA updates allow this to happen. I have given my feedback and they are working on tweaks.

 

[Barking Sands]

As it turns out, with the Android phone app anyways, when you click on the Dashboard in the Epoch app, it takes about 5-6 seconds for the Firmware version to show up. I was grabbing screenshots immediately hence the V0.0.
I went about updating the firmware to V2.4 and was unsure if I should just leave the batteries as one big virtual bank and update it that way (it seems like it would let me) or separate them. I decided to make them individual batteries again and changed the DIP switch reflect that. I then ran the update to each battery individually. No problems, they both updated fine.
I tried to use the iPad to do this update but the OTA password is not "test". No idea there.
I am sure that I have to do more cycles as suggested on page 2 of this thread but I am seeing the batteries at 99% but they never go into float. They are always hanging out at 14.2 volts. That concerns me.
The batteries in the app show they are full or nearly so but in the Ekranos (with the new update showing cells and stuff) I am at 820 amps with a total of 920. I'll add some screenshots.
I also woke up to both batteries having the error light flashing and the Charge MOS was switch off in the app.
I turned the SVS on and maybe that had something to do with it.
Hopefully I am not burning up too much goodwill here. I am very appreciative of the help here!

The 920/820 is the same thing I got after my update. I did some testing yesterday and did a deep discharge to 2% SOC on the smart shunt and was at total discharge of 918 ah with 11.1 volts. So the 820 shown must be something they put in there. Maybe a 10% reserve? Either way I don't like it so I gave them some feedback to change it to reflect the actual capacity.

If you have V2.4 and hit OTA again it probably wont update until they have another one. It took me quite a few cycles on these to get to be able to hit 14.2 and even 14.3. These batteries have a pair of 230ah cells and then have 4 or these pairs in series for the arrangement. So balancing will probably always be a bit wider than a smaller 4 cell series arrangement. I have mine down to around .020-.025 Vdiff or so at 14.2 volts. Its passive balancing not active. The main thing is to ensure the cells come closer together at the top of the charge on each successive cycle and that they dont grow apart over time. So far 2 channels have capacity tested these, probably with cells not even well balanced and both are getting in the 475 to 476 ah range.

For now just keep cycling and absorbing until you can have 14.2 or 14.3 set in the Limit max voltage and enter absorption without the red lights. But it looks like there is still a bit of work to do in the programming of the CAN BMS comms to get them to cycle more consistently. As I learn things I will definitely share the updates.

 

[Barking Sands]

One more question, getting back to conditioning/balancing: do you have any insight from the battery into active balancing activity?

Put another way, is repeating charge cycles beneficial only because it allows the BMS to actively balance, or is the benefit entirely in just allowing the battery to spend a bit more time at the top and self-balance? Or some combination?

My experience so far....On any of the very large capacity batteries I have used so far a bit of commissioning has been needed before reaching higher charge levels. usually this is due to balancing. They assemble the batteries with matched cells but never cycle them. So typically, you will get cell overvolt of 3.65+ on initial cycles for the leading cell. This is true of many of the server rack batteries or larger drop ins such as the 48 volt CHINS in my Zero Turn mower.. Driving the pack voltage up to the upper knee where the cells will tend to split apart is really the opportunity for the passive balancing to take place. The passive balancer acts on the split and can do so because they are split. If you go through a charge cycle of bulk/absorption and then float there is limited time for this passive balancing to take place in that upper knee range unless you can manually increase absorption time. Many times with stubborn batteries i will set up a power supply to a higher absorption voltage such as 14.3 give or take and let them sit for 24 hours to allow additional time for the passive balancing to take place. This usually works pretty good. Sometimes you have to start at 14v and work your way up. Thats one of the drawbacks of very large batteries is a little extra work to balance them.

 

[twistedtree]

This seems like an awful lot of fussing and cajoling to commission the batteries and get them into service. Once fully commissioned, do they stay in Balance OK, or is this an ongoing process?

 

[Barking Sands]

This seems like an awful lot of fussing and cajoling to commission the batteries and get them into service. Once fully commissioned, do they stay in Balance OK, or is this an ongoing process?

No..once balanced they stay that way.

 

[Jeff F]

Many times with stubborn batteries i will set up a power supply to a higher absorption voltage such as 14.3 give or take and let them sit for 24 hours to allow additional time for the passive balancing to take place. This usually works pretty good. Sometimes you have to start at 14v and work your way up. Thats one of the drawbacks of very large batteries is a little extra work to balance them.

I'm not sure why a large battery would be any more effort than a smaller one to balance, unless the BMS balancing circuitry hasn't been upgraded to match the bigger cell capacity.

My understanding of passive balancing is that it works only as long as there is current flowing into the battery. If you hold at 14.3 for 24 hours does current continue to flow?

I'm still of the opinion that holding the battery in the shoulder range has benefits to battery balance and capacity that go beyond allowing more time for the BMS to fine tune cell voltages.

 

[Barking Sands]

I'm not sure why a large battery would be any more effort than a smaller one to balance, unless the BMS balancing circuitry hasn't been upgraded to match the bigger cell capacity.

That depends. For instance many 200ah, 300,ah and 460ah batteries all use the same BMS and have the same balance capacity. In those instances the larger batteries would be expected to take a bit longer to balance.

My understanding of passive balancing is that it works only as long as there is current flowing into the battery. If you hold at 14.3 for 24 hours does current continue to flow?

Again it depends on how the bms is set up. But generally if you set up a power supply at say..14.2..the voltage will settle to 14.2 but when balancing you will see a very small current on the power supply while balancing.

I'm still of the opinion that holding the battery in the shoulder range has benefits to battery balance and capacity that go beyond allowing more time for the BMS to fine tune cell voltages.

Could be, but I dont have any idea what that might be. Staying in an absorption voltage range in the upper knee only has 2 purposes that I can think of. Allowing maximum charge capacity at that particular voltage and balancing.

I tend to think you should stay in the upper knee voltages just long enough to get those 2 things done and then reduce voltage to a more natural resting voltage of a fully charged battery or slighly less as soon as practical.