Solid State Relay

[sailingharry]

I stumbled upon an interesting product today. A 100A DC SSR, with a form factor that mimics an ANL fuse, all for $40. Not sure where it is made, but sold by a US firm.
https://www.thornwave.com/products/t...tate-relay-ssr
No heat sinks, but the math indicates it dissipates 3W at full load, so not a lot of heat.
If they made a 150A version, it would be the solution to my charge bus.

[wholybee]

If it is for battery protection controlled by your BMS, I wouldn't trust it. I mean, I have a FET based BMS, and it seems 10x overbuilt compared to that. Maybe if it is meant to be like a fuse, where you have several spares?

Other than higher cost, why not the 220A Victron battery protect?

[UFO]

Originally Posted by wholybee

If it is for battery protection controlled by your BMS, I wouldn't trust it. I mean, I have a FET based BMS, and it seems 10x overbuilt compared to that. Maybe if it is meant to be like a fuse, where you have several spares?

Other than higher cost, why not the 220A Victron battery protect?

I used the 220A Battery protect, which died on me Mid dropping anchor (in 25kts with less than 10m of chain down).



Apparently anchor winches can sometimes give a small amount of reverse current, which totally fried it - Victron gave me a new one, but I never installed it after that.

[sailingharry]

Originally Posted by wholybee

If it is for battery protection controlled by your BMS, I wouldn't trust it. I mean, I have a FET based BMS, and it seems 10x overbuilt compared to that. Maybe if it is meant to be like a fuse, where you have several spares?

Other than higher cost, why not the 220A Victron battery protect?

I agree with the apparent issue with the physical size. But, the advantages over the Battery Protect are size and price. The fact that it is a US based company adds a touch of confidence in the function of the device (as compared to AliBaba sources).


I don't actually have a use for it -- it doesn't have enough capacity for my charge bus, and my battery disconnect will be a RBS. It just seemed an interesting option that may meet some needs. I hadn't seen anything like it. DC SSR's are typically hard to find and expensive.

[Emmalina]

I think this is very optimistic giving it a rating of 100A guessing that would be resistive not inductive but the heat dispersion from the FETs goes where ? I bought some chinese 100A ssr's for my domestic water pump and they lasted a few days !

[wholybee]

Originally Posted by UFO

I used the 220A Battery protect, which died on me Mid dropping anchor (in 25kts with less than 10m of chain down).



Apparently anchor winches can sometimes give a small amount of reverse current, which totally fried it - Victron gave me a new one, but I never installed it after that.

That is good to know. They do say that it will be damaged by reverse current, but I wouldn't have expected that. I wonder if a diode across the terminals would protect it, allowing the reverse current to flow through the diode.

And almost certainly, the SSR in this thread would be damaged by reverse current also.

[sailingharry]

Originally Posted by wholybee

That is good to know. They do say that it will be damaged by reverse current, but I wouldn't have expected that. I wonder if a diode across the terminals would protect it, allowing the reverse current to flow through the diode.

And almost certainly, the SSR in this thread would be damaged by reverse current also.

The manual for this SSR has a discussion on reverse current, which I found educational. Curiously, it seems that a MOSFET conducts equally well in both directions when ON, This one is rated at 100A in both directions (150A forward for 1 minute). However, when OFF, the math changes. When OFF, it blocks current in the forward direction (the basic purpose!). But, when OFF, it allows current in the reverse direction -- with a .7V drop (ie, a basic diode). This creates massive heat. This SSR is rated at only 4A in reverse, reflecting the much higher heat generation. Although, device temperature is a function of both heat generation and time -- I suspect a much higher current for a small fraction of a second wouldn't cause damage (but that's a suspicion, not a known fact!).

[wholybee]

I suspect you are right, but probably very short, and some of the limit may be the small traces and leads on the tiny semiconductors. I was once responsible for a protection device where the test was to put 30kV through 18" of 38 AWG wire, and the device was to measure the current flow, and turn off the supply before the wire vaporized. And if the device failed it would, with a flash and loud bang like thunder and lightning, some smoke, and not a bit of evidence there had been a wire in the test fixture. It usually passed the test.

I am a fan of FET BMSs, unlike most here. From the specs on mine, it is 120A continuous. In the software you can set the overcurrent cutoff, and the delay time, which can be a couple seconds. There is also a nonconfigurable hardware failsafe. 1500A (IIRC) it considers a short circuit, and will cutoff in a few milliseconds, and not damage anything. Given it's ability to withstand and protect itself from a direct short, I think it is much more durable than many give FETs credit for. Mechanical relays have a very high failure rate, especially if they don't cycle but stay closed. And they also can fail in the close position contrary to what many think.

I don't remember how many FETs it has, 10 or 12 total, I think. Each is individually rated for 100A (so I am told), which is probably BS, but with 5 for each direction provides a lot of overhead, and a large heatsink. It is pretty substantial, more comparable in bulk size and weight to the Victron than the ANL sized thing.

[sailingharry]

Originally Posted by wholybee

I am a fan of FET BMSs, unlike most here. From the specs on mine, it is 120A continuous.

Sliding off topic (this is CF, right?), I'm not terribly happy with the idea of a FET BMS -- but you are, so I'm listening.


First, I don't feel terribly confident in how they meet the ABYC standard (or, if they are even addressed). ABYC has an entire standard on battery switches, for instance. It's a fun read, if just to get insight into the thoughts that go on at their meetings. But, what standard or test does a FET BMS meet? Interrupt levels? Current carrying levels? Test standards? It's not that they can't do the job -- it's that the job hasn't been defined.


But, perhaps my bigger concern is the interface. ABYC alludes to, and I want, a few outside controls. I want an alarm before LVC disconnect (HVC is much more optional, since ideally an HVC is done in phases and never completely disconnects. I want to be able to directly turn off solar, alternator, charger, and inverter prior to any alarm or cutoff. I haven't looked closely at all the options (as I've been focusing on contactor based), but I have a feeling that they don't provide that level of control. If I directly disable all charge sources through a signal before reaching a HVC event....well, there won't be a HVC!. And if I turn off the inverter at about 15% SOC, the odds of a LVC go way down.


I do have a need for about 300A continuous (not really, but I have a 3kW inverter so it's a design input), which does limit (but not eliminate) FET based units.


Thoughts?

[wholybee]

Originally Posted by sailingharry

Sliding off topic (this is CF, right?), I'm not terribly happy with the idea of a FET BMS -- but you are, so I'm listening.


First, I don't feel terribly confident in how they meet the ABYC standard (or, if they are even addressed). ABYC has an entire standard on battery switches, for instance. It's a fun read, if just to get insight into the thoughts that go on at their meetings. But, what standard or test does a FET BMS meet? Interrupt levels? Current carrying levels? Test standards? It's not that they can't do the job -- it's that the job hasn't been defined.


But, perhaps my bigger concern is the interface. ABYC alludes to, and I want, a few outside controls. I want an alarm before LVC disconnect (HVC is much more optional, since ideally an HVC is done in phases and never completely disconnects. I want to be able to directly turn off solar, alternator, charger, and inverter prior to any alarm or cutoff. I haven't looked closely at all the options (as I've been focusing on contactor based), but I have a feeling that they don't provide that level of control. If I directly disable all charge sources through a signal before reaching a HVC event....well, there won't be a HVC!. And if I turn off the inverter at about 15% SOC, the odds of a LVC go way down.


I do have a need for about 300A continuous (not really, but I have a 3kW inverter so it's a design input), which does limit (but not eliminate) FET based units.


Thoughts?

There is nothing in E-13 that would suggest a FET based BMS would not comply. E-13 actually has very few hard requirements, and mostly says to follow the recommendations of the battery and BMS manufacturer. ABYC got a lot of criticism by not making more requirements. E-13 says that consideration must be given to how a disconnect might damage other components (I.E. the alternator) but doesn't require it be done in a specific way. DCDC is fine, and alternator protection device is fine, etc. It doesn't require, of even suggest, that the BMS actually control charging sources. It recommends that there be an alarm before a disconnect, and I have implemented that with my BMS, though it was with a Bluetooth meter/interface that I designed. But even that is only a recommendation, you would be compliant with out it. Arguably, using a Victron BMV would meet the recommendation. My solution does monitor at the cell level, however.

ABYC does require compliance with UL1642, UL1973, and UL2054. I think that applies to cells and/or drop-in battery assemblies, not the BMS. Not sure though. Maybe sometime there will be a more specific set of tests for a BMS. It would be simple enough for me to swap out, and it could just as well be that a REC BMS would not comply.

Standards aside though, the key thing to remember is that LFP is _safer_ than Lead-Acid, even with no BMS at all, excepting that a BMS will prevent you from overcharging the battery and leaving yourself without power. A lot of people get caught up in all the safety systems with LFP, or with the fact that other Lithium chemistries are much more dangerous, and worry too much.

A BMS should do NOTHING, unless something goes wrong. Your chargers should be set to not overcharge the battery, and you should pay attention and start the generator before it gets to 0%. I think more effort should be placed with making sure the system doesn't fail, instead of having a complicated BMS with all these features that get used only if it does.

A FET BMS is simple. It is simple to bypass if it fails. And it is simple to swap out, and a spare is cheap. Consider how expensive a REC BMS (or other comparable BMS) is, with everything needed to make it work, contactors, etc. and how much it would cost to carry spares of all that.

Consider what cells are being used, and how they were sourced. I put money into better cells. I nearly bought EVE cells from what was at the time a reputable seller. But ultimately, like all EVE cells, they were sourced from China as "Grade A" cells that really were not, or at least it couldn't be proven. At the last moment, I got cold feet on those cells, and spent 3 times as much on CALB cells, from a dealer in California that Calb recommended. They arrived in a wooden crate, from Calb(not the dealer I ordered them from), with all the Calb test data in it. Not long after I received them, reports came in that the seller of the EVE cells had some problem. All of the cells ordered from about the time I would have were defective, and he disappeared.

Anyway, I tell that story, because I think getting the best quality cells is better than getting the best quality BMS, so that is where I spent my budget. I read all the time about people adding active balancers, or that passive balancing isn't enough, or that you need a BMS with 2A to balance etc. All BS. That is what you need if you got grade B cells that are not matched. My 20mA of passive balancing has kept my 300Ah battery balanced within 15mV for 3 years. If my BMS failed and I had to bypass it, I would be comfortable doing that for several months and that my cells would stay in balance. I also read of people testing 280 Ah cells, and finding they test at 278-280Ah, and think that is fine. My 100Ah cells tested (and still test) at 115Ah.(specifically, the whole battery tests at 345Ah) Grade A cells have a substantial margin of having more than rated capacity.

The meter I made shows me at a glance the SOC, Voltage, Current in/out, just like a Victron would, but also gives me cell level voltages, and will alarm a configurable amount of time before the BMS cuts out(based on the fact the BMS has a configurable delay). It can also alarm at a voltage difference before cut off. Example, if cutoff is 3.65Vpc, it can be set to alarm at 3.63Vpc.

Anyway, long off-topic post, but that is why I like the FET BMS. I would probably use a contactor type if I used my LFP bank for starting, and I might do that at some point. LFP is so reliable, I lean toward not needing a separate (less reliable) battery for starting.