Why not to use Daly BMS with MPPT controllers

[Hedges]

SolarEpic 60amp controller (made by ipandee/ipanda) has 150v max PV input.

This particular array was 3x 310w 72cell panels in series for about 120VOC.

I use this right now for testing, and doing videos. The battery hooked up at the time was 4s 100ah LiFePO4

25% voltage headroom above data sheet VOC should be fine for any temperature.

If you got the transistors out (I didn't watch the entire video), maybe you can find what voltage rating they have.
I think the buck architecture has MOSFET withstanding voltage between PV Voc and ground, plus a diode drop. I wonder what voltage develops across diode before it starts conducting? Diodes are (usually) slow, MOSFETs are fast. There could also be a MOSFET for synchronous rectifier to reduce voltage drop and power loss.

Hmm, I'm not the first to think of that:

https://www.edn.com/diode-turn-on-time-induced-failures-in-switching-regulators/

Just one of many possible failure modes. Simple overheating is another. But you're at a fraction of the rated current.

https://www.amazon.com/SolarEpic-Controller-Temperature-Lead-Acid-Batteries/dp/B09Q5XV676

Your choices may be to do the engineering original designers didn't, or buy a unit from someone else who did.
Silk purse and sow's ear, you know.

 

[BradCagle]

25% voltage headroom above data sheet VOC should be fine for any temperature.

If you got the transistors out (I didn't watch the entire video), maybe you can find what voltage rating they have.
I think the buck architecture has MOSFET withstanding voltage between PV Voc and ground, plus a diode drop. I wonder what voltage develops across diode before it starts conducting? Diodes are (usually) slow, MOSFETs are fast. There could also be a MOSFET for synchronous rectifier to reduce voltage drop and power loss.

Hmm, I'm not the first to think of that:

https://www.edn.com/diode-turn-on-time-induced-failures-in-switching-regulators/

Just one of many possible failure modes. Simple overheating is another. But you're at a fraction of the rated current.

https://www.amazon.com/SolarEpic-Controller-Temperature-Lead-Acid-Batteries/dp/B09Q5XV676

Your choices may be to do the engineering original designers didn't, or buy a unit from someone else who did.
Silk purse and sow's ear, you know.

Yes I got the mosfets out, they are infineon 107n20n. 200v so should be good there. Only one was actually shorted, I have two on order for replcements.

Could just be bad engineering, or the other thing I was thinking may be fake infineon 107n20n mosfets LOL

I'll take a read on the "diode turn on time failure" link you provided, thanks for that. I might see if I can find faster/better diodes.

 

[Hedges]

Bad parts is a good thought.

Schottky is faster, but has high reverse leakage. Need to evaluate that it won't enter thermal runaway. I would start by assuming maximum ambient/case temperature under consideration and junction artificially heated to Tjmax, see if it would get hotter or cooler from there.

There are also fast silicon PN junction diodes with reverse recovery measured in nanoseconds, meant for switchers.

My plan for such protection is to use a PV rated breaker with remote trip or ganged to a small adjacent breaker to be tripped. It may not be fast enough to protect FETs but would prevent over-charge.

A "keep alive" design could be more failsafe than a "shutdown" design.

 

[BradCagle]

Bad parts is a good thought.

Schottky is faster, but has high reverse leakage. Need to evaluate that it won't enter thermal runaway. I would start by assuming maximum ambient/case temperature under consideration and junction artificially heated to Tjmax, see if it would get hotter or cooler from there.

There are also fast silicon PN junction diodes with reverse recovery measured in nanoseconds, meant for switchers.

My plan for such protection is to use a PV rated breaker with remote trip or ganged to a small adjacent breaker to be tripped. It may not be fast enough to protect FETs but would prevent over-charge.

A "keep alive" design could be more failsafe than a "shutdown" design.

Ok that diode read make sense, basically if the MOSFET switching frequency approaches/exceeds the turn on speed of the diode were going to have big trouble. Which may be the thing, considering I was running a 24v pack just fine for over a month, I reconfigured it for this 12v pack to test something, and bam! Did it go higher frequency, or lower duty cycle, or both? Not sure, but that's a possibility.

I just grabbed one of these https://www.amazon.com/dp/B01N6ATT8C?ref=ppx_yo2ov_dt_b_product_details&th=1
I'm not necessarily expecting it to be the solution, I pretty much expect to blow it out No where does it say (that I can find) what level of over volt protection it offers. I plan on hitting it with 120v directly from this array

 

[BradCagle]

Or maybe this https://www.amazon.com/Cole-Hersee-...tching+,automotive,111&sr=1-3#customerReviews

Use a 16v Zener to trigger it?

 

[Hedges]

Calculate power dissipation of zener, also voltage vs. current curve. It is pretty rounded, not a sharp knee.
Zener with resistor pullup can make a regulated voltage, compare to reference with an amplifier.
Maybe it works direct to coil, but I'd guess many volts on coil to operate plus zener volage. In other words, something other than 12V coil would be needed if no amplifier. I might use comparator plus transistor.

I used zener + darlington for shunt regulator on old motorcycle (had been designed to over-charge lead-acid battery as means to regulate.)
I used stack of zeners for low current high voltage reference in lab equipment (500V +/- to grids of a vacuum tube.)

If you engineer something and use quality parts, it could be reliable. But you have to think through the off-normal situations, like diodes not actually clamping instantly to 0.7V as an ideal one would.

I have a charge controller (made by MSTE, sold by SMA) that cautions about it shorting and over-charging battery. It has a control output for user-installed PV disconnect. So have the job is already done for me.

 

[BradCagle]

Calculate power dissipation of zener, also voltage vs. current curve. It is pretty rounded, not a sharp knee.
Zener with resistor pullup can make a regulated voltage, compare to reference with an amplifier.
Maybe it works direct to coil, but I'd guess many volts on coil to operate plus zener volage. In other words, something other than 12V coil would be needed if no amplifier. I might use comparator plus transistor.

I used zener + darlington for shunt regulator on old motorcycle (had been designed to over-charge lead-acid battery as means to regulate.)
I used stack of zeners for low current high voltage reference in lab equipment (500V +/- to grids of a vacuum tube.)

If you engineer something and use quality parts, it could be reliable. But you have to think through the off-normal situations, like diodes not actually clamping instantly to 0.7V as an ideal one would.

I have a charge controller (made by MSTE, sold by SMA) that cautions about it shorting and over-charging battery. It has a control output for user-installed PV disconnect. So have the job is already done for me.

Ok tried the victron battery protect. It blew up soon as It tried to disconnect the over voltage fault Exactly what I was expecting, just had to try.

 

[Hedges]

I thought about pointing out it couldn't withstand PV voltage, but figured you'd learn better this way.
You did say you bought it from Amazon, didn't you?

Rant: EG4 48v 100ah battery has faulty BMS. Signature solar customer service is TERRIBLE

You having Signature Solar update the Spec sheet from saying 250A for 1 second to what? How does this help the guy after he is out $700? If it did fail to meet spec sheet, that's a reason to refund customer 100% including shipping. Then update the spec sheet (and highlight the change on...

diysolarforum.com

 

[BradCagle]

I thought about pointing out it couldn't withstand PV voltage, but figured you'd learn better this way.
You did say you bought it from Amazon, didn't you?

Rant: EG4 48v 100ah battery has faulty BMS. Signature solar customer service is TERRIBLE

You having Signature Solar update the Spec sheet from saying 250A for 1 second to what? How does this help the guy after he is out $700? If it did fail to meet spec sheet, that's a reason to refund customer 100% including shipping. Then update the spec sheet (and highlight the change on...

diysolarforum.com

Yup, LOL. I seriously had my doubts, but figured it would be good for all to see. Also manufactures need to put the max protection voltage their products can withstand, nowhere in the documentation is this listed. Yeah came from Amazon.

EDIT: it is listed as pointed out by Hedges, totally my fault for missing that.

 

[Davismltc]

You might have a look at Crossy's solution to battery protection at post #191.

 

[Hedges]

"About this item​

• The Battery Protect disconnects the battery from non essential loads before it is completely discharged (which would damage the battery) or before it has insufficient power left to crank the engine"

Purpose was to disconnect for low voltage.

Can also be controlled by BMS.

https://www.victronenergy.com/upload/documents/Datasheet-Battery-Protect-65-A--100-A--220-A-EN.pdf

"Operating voltage range 6 –35V"

RTFM?

 

[BradCagle]

"About this item​

• The Battery Protect disconnects the battery from non essential loads before it is completely discharged (which would damage the battery) or before it has insufficient power left to crank the engine"

Purpose was to disconnect for low voltage.

Can also be controlled by BMS.

https://www.victronenergy.com/upload/documents/Datasheet-Battery-Protect-65-A--100-A--220-A-EN.pdf

"Operating voltage range 6 –35V"

RTFM?

We'll I am nearly blind

 

[BradCagle]

You might have a look at Crossy's solution to battery protection at post #191.

That looks interesting, basically using a buck convertor to drive a SSR.

 

[crossy]

That looks interesting, basically using a buck convertor to drive a SSR.

The important bit is the schmitt trigger, the buck converter is just a power supply and could easily be replaced by a zener/resitor arrangement, the actual power consumption is minimal (a few mA when the relay is on).

@Davismltc did a version for a 12V pack with a zener supply.

 

[BradCagle]

Ok guys, I came up with a solution.

 

[BiduleOhm]

The polarity on the contactor is probably because it has some kind of arc quenching (probably based on a magnetic field) that is sensitive to the current direction, like most DC breakers too for example. You connected it correctly

 

[BradCagle]

The polarity on the contactor is probably because it has some kind of arc quenching (probably based on a magnetic field) that is sensitive to the current direction, like most DC breakers too for example. You connected it correctly

Thanks for the info, I kind of suspected arc quenching.

 

[mrzed001]

Ok guys, I came up with a solution.

Grats !
Already shared it as the first cheap solution

 

[TorC]

Not from Battery Hookup, but I'm pretty sure there are 48V versions of that contactor. AliBaba has some, I believe, for a decent price.

 

[BiduleOhm]

Or you can use a zener in series to drop around 20 - 30 V, cheap and no need to source a different contactor. You'll need to check that the power dissipation is under the zener limit tho. At worst use multiple lower voltage zeners in series so you can divide the dissipated power across a few diodes.