diy solar

diy solar

Blew up two Daly's, so just ordered a Batrium

vk2emp

New Member
Joined
Mar 2, 2020
Messages
101
I blew up a 100A Daly bms by over drawing it - we're talking a 12V circuit here, so bought a 250A unit. During operation from my CALB pack, the 100 did a fine job of what was expected of it, then I plugged in my fridge. I have a 1200VA Victron inverter that powers my experimental system before I go 48V and build a proper system. The peak of the fridge compressor blew the 100A daly, so I upgraded to a 250A model.

3 months later while I was at the pub hanging out with mates I get a call from my friend who is back as my place. "Hey mate, something just went bang and the power went off. The TV and the computer just turned off" (I have a server, a fairly efficient one).

When I got home I could smell that burned electronics smell. I bypassed the battery so my mate could continue to watch the show he was watching from my server.

I sat down beside him and got on my phone, and went to batrium.com and ordered these toys.

Now bear in min I plan on upgrading to a 48V system, however I have been experimenting with a 12V system, comprising 180Ah CALBs, Victron Smart Shunt, Victron 12/1200VA inverter and Victron 150/35 fed by 440W of solar on the roof.

All this gear has been kept in check by a raspi running venusos, and a Daly BMS.

Both BMS's died, blowing out their output fets, not the charging ones.

So. Heres what I purchased:
Batrium CORE watchmon
Batrium Expander Module for Watchmon
4 x LongMons
Batrium ISOmon

And a few other bits and bobs.

I got the expansion module so that if my batteries run low, the relays can turn stuff off like the inverter, and chuck on some fans during charging to keep the 150/35 cool. Once a month it can also kick on my "Flying Capacitors" brand balancer (really when I saw the name of this thing I thought, do I need to attach this to my battery via long leads and swing it around my head to make the capacitors fly and do what they need to do? haha).

When I Go to 48V I'll just need to buy 12 more long mons and reprogram the system.

Batrium aint cheap. but considering I have spen a fuck tonne on BMS's including cheap china, DALY and waiting ages for the one in the blue box whos name I forget only to find its the wrong type.

I plan on doing some you tube videos like Off Grid Garage and The Digital Mermaid (hi Andy and Madi!) when my Batrium arrives :)
 
Last edited:
In my opinion, Daly was good back when it was pretty much the only thing readily available and cheap enough to DIY. There are much better options nowadays, even on the cheap end. Daly hasn't kept up...
 
Yeah I'm getting this. Unmarked connectors, "you need to do this coz yours is that model" and the LOUD BANG when they die has put me off them.

The "Blue box" one I bought was a chargary, but it was just a monitor, and 7S plus, I didn't see that in its specs, nor its inability to balance. I'm not knocking it coz the dude is on here and helping peeps out, I just purchased wrong, for whos only blame is me. :p

I do NOT want to see a 48V Daly go bang. It will probably grenade. :(
I watched Madi (youtube channel The Digital Mermaid) short out a 280Ah EVE bank, and luckily she had a T class fuse to save her. Scary as, as I was watching her with the volume up then BANG! Poor Madi, made her jump!
 
Last edited:
I think that it all depends on the peak draw of your fridge. A typical 120 Volt "household fridge" in the USA (with a plate specifying perhaps "6A, 720 Watts" as the compressor-running power load) can momentarily pull 15-20A when starting the compressor. That momentary power draw (1800-2400 watts) does not exceed the Victron "peak" rating, but leads to a BMS draw of about 220A on the "12.8V" supply line -

which absolutely will not maintain 12.8V at that current load, unless it is BIGGER than AWG 4/0. (Your "household fridge" runs at higher voltage and lower current but has the same behavior and the same power draw through the Inverter.) High resistance in the supply wire causes the Voltage to drop, and the bus bars in your 12v battery packs will contribute high resistance as well. The "12v" supply voltage drops (probably a lot), and the demand for "more current" from the Victron goes even higher.

In the meantime, Daly "250" didn't have adequate time to recognize over-current and disconnect properly. Before Daly's microprocessor even saw the problem, the MOSFET output circuit fried.
- - -
If my preceding scenario was correct, then your configuration needs some adjustments. Your 180ah CALB batteries are rated for up to 2C discharge, but your bus bars are almost certainly unable to support that. When you go to 48V, you will need only 1/4 as much current to support that fridge - so you will possibly not need to build your new battery pack with significantly "better" bus bars (though you might want to do that anyway, because china-made bus bars supplied "for free" with battery cells are generally worth about what you paid - NOTHING.)

But, IMO, your instantaneous loads should never be exceeding a BMS "continuous max current rating" by a large factor. The transistors and solder don't care that the current "is only instantaneous and might go away soon", they have to handle it all. In the case of Daly, the headroom is definitely small. For the case of your 48V replacement: In your shoes, I would not exceed the continuous rating at all. For a "48v " battery pack built as 16s and actually running discharge at around 51 Volts (most of the time), I would use a 60A unit (and not a mere "50A").

I'm a Daly user, so my fear (for the case of "exceeding the continuous rating at all") falls in line with your unpleasant and costly experience. Would a different BMS be able to handle exceeding its "continuous rating" several times per day, for years and years of trouble-free operation? Maybe, but I wouldn't design a system to find out.
 
Oh you just brought to my attention something I DID NOT FACTOR, and its something I already know when firing up an inverter. Inrush current. This "pulse" of energy can be magnitudes greater than what the appliance wants. Its not just capacitors that do this, motor stall currents do it as well. And fridge compressors are notoriously bad for it - its why AC units have start capacitors to stop them blowing fuses or tripping breakers when they come on.

I cannot believe I overlooked this. Stall current of a compressor motor can be well over 8 times its normal run current, but only for a brief perios. Goes to show how resilient my Victron inverters FETS are!

Anywho no going back now, I'm a Batrium owner, just gotta await for it to arrive, and unlike JBD/JK/DALY its expandable.

Still, I'll keep the fridge off the circuit. My 1200VA inverter was quite pricy, and I don't want to blow it up too! :p

It would be interesting to get a scope or a peak reading wattmeter and see what the fridge actually draws when the compressor fires up. Mabye its time for a new fridge :)

Oh yeah, aussie here, so our circuits are 240V
 
Last edited:
Every once in awhile I think that maybe I should have just gone with a cheap all in one BMS, then I read stuff like this and it makes me glad I'm spending the time dealing with diyBMS. I can't afford battrium, but I get nearly all the same benefits with this system, and I have the full schematics, design, firmware, etc so there's just a lot more confidence in that for me.

I just wanted you to know I'm glad when people post failures like this here...
 
Thats how we learn. I watched Madi (The Digital Mermaid) on youtube short out a 48V 280aH EVE pack, and it made me jump. I'm very cautious around my batteries, and never had a problem. A mate was visiting while I was re-arranging my 260ah 48V LA pack and he came over (having had a few beers, me sober) and asked if he could help and I turned around and said "Andy, this is very dangerous. I need full concentration here, theres a lot of power here and if I fuck up, we both go to hospital". He promptly sat back down on the couch and watched netflix, until I pulled the plug, flicked on the "Reliable" brand inverter and plugged the entertainment systm into it.

I came back to the couch with a beer myself (I'd deserved one after all that re-wiring to make the pack smaller) with a fat jumper (I'd made more than I needed) and handed it to him. He said "(cuss word beginnign with f) thats a bigass wire)" and I said "yeah, don't be angry but this is why I didn't want you near the bank with a belly full of beer mate!" :)
 
Oh you just brought to my attention something I DID NOT FACTOR, and its something I already know when firing up an inverter. Inrush current. This "pulse" of energy can be magnitudes greater than what the appliance wants. Its not just capacitors that do this, motor stall currents do it as well. And fridge compressors are notoriously bad for it - its why AC units have start capacitors to stop them blowing fuses or tripping breakers when they come on.

I cannot believe I overlooked this. Stall current of a compressor motor can be well over 8 times its normal run current, but only for a brief perios. Goes to show how resilient my Victron inverters FETS are!

Anywho no going back now, I'm a Batrium owner, just gotta await for it to arrive, and unlike JBD/JK/DALY its expandable.

Still, I'll keep the fridge off the circuit. My 1200VA inverter was quite pricy, and I don't want to blow it up too! :p

It would be interesting to get a scope or a peak reading wattmeter and see what the fridge actually draws when the compressor fires up. Mabye its time for a new fridge :)

Oh yeah, aussie here, so our circuits are 240V
Your Victron handled it all for 3 months, it's impossible to guess whether the lifespan of Victron circuitry was reduced. But Victron is at least RATED to handle all of my SWAG for the inrush current. (I was unsure if you were UK or aussie, but I was pretty sure that you were one or the other.) Thanks for your reply! "Household Fridge" compressors don't have the expensive and big capacitor circuits which Air Conditioners do, and you're right abut the magniitude of the issue - some of them, especially in older models, can be ugly. What's the initial current capability of your Batrium, before expansion?
 
The batrium doesnt have one, unless you have the shunt but that is on the DC side. The shunt (500A or 1000A) will tell the batrium if the pack is being over drawn. The beauty of the batrium is, with the right setup, you can program it to say "No little inverter, no, turn off NOW" by using the expansion board to trip a relay. This relay can do numerous things, like turn on a fan, kick in a charger or indeed, tell an inverter to shut off. I wisely bought the expansion module.

By expand I meant go from 12V to 48V by just buying more Longmons, or indeed switching to a K-9 system where the cell monitors are contained on one board, and the data sent back over a 4 or 6 wire (I forget) to the watchmon.
 
hey Madi! Greeting from the other hemisphere! I haven't been on these forums for a little while - just saw the funny video where you popped one of your buck converters on your "big" battery. So looks like we're even so far - you've exploded a fuse and capacitor, I've popped a few fets in DAlLY's :)
 
hey Madi! Greeting from the other hemisphere! I haven't been on these forums for a little while - just saw the funny video where you popped one of your buck converters on your "big" battery. So looks like we're even so far - you've exploded a fuse and capacitor, I've popped a few fets in DAlLY's :)
Oof, how did you blow the FETs? I'd expect them to be shut down by the BMS before damage could occur. Was it a case of over-current and unbalanced disconnect of the FETs, like Andy ran into with the JK?
 
Fridge inrush current - over many cycles it eventually blew the output fets.
 
Yeesh, those must be high quality FETs!
I watched one of your recent videos where you did thermal imaging of all your batteries including BMS.
You observed that some of the BMSs were warmer than the others.
That might indicate with some blown fets.

I like your videos, especially the cats.
 
Possibly, but I think something else is at play. As I've been charging the packs, the outer two packs have been charging oddly. I'm going to finish with a proper top-balance, then do a capacity discharge test. When I finish that and start the recharge, knowing all cells are balanced, I'll see what the heat looks like again (with the cover off, so I can monitor cell temps as well as BMS temps).
 
Possibly, but I think something else is at play. As I've been charging the packs, the outer two packs have been charging oddly. I'm going to finish with a proper top-balance, then do a capacity discharge test. When I finish that and start the recharge, knowing all cells are balanced, I'll see what the heat looks like again (with the cover off, so I can monitor cell temps as well as BMS temps).
Did your verify the current mismatch with your clamp meter?
Another possible symptom of blown fets is inaccurate current accounting at the bms.
 
I was testing my Klein's amp reading versus the BMS's reading and all BMSes are ... off.

Yeah, and it's not so easy to calibrate correctly. For one, don't calibrate with a load through an inverter. The current drawn is not a constant (it's pulsed), which is also why you see the BMS current reading jumping. Ideally, you'd need a pure resistive load directly coupled to the battery (each individual) drawing something like 10 to 20A. Measure with a low amp clamp meter, and calibrate on that.
Now, you'd think that would be it. However, you have to make a decision: do you calibrate for low amps, or high amps. Because of the way the electronics work, you can either get accurate at low amps, or you can get accurate at high amps, but not both at the same time (for the same reason current clamps or lab meters are available for low or high current).
In the end, it doesn't really matter since you won't rely on the BMS for state of charge, but it's something to keep in mind.
 
Yeah, and it's not so easy to calibrate correctly. For one, don't calibrate with a load through an inverter. The current drawn is not a constant (it's pulsed), which is also why you see the BMS current reading jumping. Ideally, you'd need a pure resistive load directly coupled to the battery (each individual) drawing something like 10 to 20A. Measure with a low amp clamp meter, and calibrate on that.
Now, you'd think that would be it. However, you have to make a decision: do you calibrate for low amps, or high amps. Because of the way the electronics work, you can either get accurate at low amps, or you can get accurate at high amps, but not both at the same time (for the same reason current clamps or lab meters are available for low or high current).
In the end, it doesn't really matter since you won't rely on the BMS for state of charge, but it's something to keep in mind.
I wonder how much this would be a factor with the "true" sign wave (20khz, iirc) coming from the Quattro. Regardless, you point is well taken.
 
Back
Top