EEVblog 1499 - EcoFlow Delta Pro 3.6kWh Portable Battery TEARDOWN!

[EEVblog]

I just realised that it's possible to overcome the main input switching problem by using a mains powered DC supply into the solar input. I have quite a few PSU's lying around...

[Billy Tree]

EEVblog Just like to clear up a couple things. The AC does not hiccup like that in a healthy unit.

It does, there is even a warning about it in the manual.

Where exactly does it say that in my Delta Pro manual?
I am quite active on the facebook Delta user group and an early adopter of the whole line, with two original Delta 1300's, a Delta Pro, and a new Delta 2.
It is NOT normal for any Ecoflow Delta to "hiccup" when outputting AC power - at all. None of mine do that.
Were you plugged in the wall when you were powering your fridges?
There are limitations on how much current the DP can pass through - but that's a operator issue to observe the device's parameters.
If the AC power was cut and the DP switched from the AC input to battery power, there is a 30ms switchover time. That is also a device parameter. My cable modems, routers and TV's do not even blink when that has happened, certainly my fridges, the portable and my Haier Quad have had no issues.

[EEVblog]

EEVblog Just like to clear up a couple things. The AC does not hiccup like that in a healthy unit.

It does, there is even a warning about it in the manual.

Where exactly does it say that in my Delta Pro manual?

3.10 Emergency Power Supply (EPS)
The product supports EPS. When you connect the grid power to the AC Input Port of the product
through an AC cable, you can power electrical devices through the AC Output Port (AC power will
come from the grid and not the power station in this situation). In case of a sudden blackout, the
product can automatically switch to the battery powered supply mode within 30ms. As a basic UPS
function, this function does not support 0ms switching. Please do not connect the product to any
device that requires 0ms UPS, such as data servers and workstations. Please test and confirm the
compatibility before using the product. We recommend that you only charge one device at a time
and avoid using multiple ones at the same time to avoid overload protection. EcoFlow takes no
responsibilities for any device failures or data losses caused by failures to follow instructions.

[/quote]If the AC power was cut and the DP switched from the AC input to battery power, there is a 30ms switchover time. That is also a device parameter. My cable modems, routers and TV's do not even blink when that has happened, certainly my fridges, the portable and my Haier Quad have had no issues.[/quote]

Yes, it's that 30ms. This is an issue for one or more of my fridges causes, I can physiucally hear the violent compressor restart. I'm not going to do this twice a day every day.

[electr_peter]

Other EcoFlow models also has ~30ms interruption in UPS mode (withing from mains power to inverter). It is mentioned in EcoFlow video EcoFlow Lab: Unveil R600 Smart Inverter [time stamp 06:36]

[Billy Tree]

Other EcoFlow models also has ~30ms interruption in UPS mode (withing from mains power to inverter). It is mentioned in EcoFlow video EcoFlow Lab: Unveil R600 Smart Inverter [time stamp 06:36]

It is disingenuous to characterize Ecoflow's 30ms switchover time between AC pass through mode and running off battery power as "a hiccup in AC power."
They are hardly unique in this - give me a list of other manufacturer's units that do this faster.
And you can avoid the switchover issue entirely, by powering the DP from DC power.
The dual direction inverter - charging board is what made Ecoflow the state of the art in So Gens all the other have been chasing.
That made an hour to 80% possible, previous legacy companies like Goal Zero and Jackery took 10 hours or more to recharge their So Gens.
That reversal and the relay switching takes time - but nevertheless
My Haier quad and my tenant's fridge certainly have never suddenly jumped into a "violent compressor restart" during the 30 ms switch over.
Even running off of Ecoflow's modest 600W output River Max unit, let alone an 1800W output Delta.
Really EEVblog's complaint - and configuration of 3 fridges is fairly unique, what kind are they exactly? Are they industrial/commercial units?
We can look at the specs.

[EEVblog]

Other EcoFlow models also has ~30ms interruption in UPS mode (withing from mains power to inverter). It is mentioned in EcoFlow video EcoFlow Lab: Unveil R600 Smart Inverter [time stamp 06:36]

It is disingenuous to characterize Ecoflow's 30ms switchover time between AC pass through mode and running off battery power as "a hiccup in AC power."

It literally is a hiccup, you can see it on a scope and I showed it in the video. Pointing out an aspect of a product that might cause problem (did so in my case) is not "disingenuous".

My Haier quad and my tenant's fridge certainly have never suddenly jumped into a "violent compressor restart" during the 30 ms switch over.

Good for you. Mine did.

Even running off of Ecoflow's modest 600W output River Max unit, let alone an 1800W output Delta.
Really EEVblog's complaint - and configuration of 3 fridges is fairly unique, what kind are they exactly? Are they industrial/commercial units?
We can look at the specs.

Does it matter? It caused a problem, end of story.
I'll use the DC input as I said.

I don't know why you have such a bee in your bonnet about this?

[Monkeh]

I don't know why you have such a bee in your bonnet about this?

You're daring to criticise his favourite product and he fancies an argument. Perhaps he gets paid to go review these things and feels he needs to defend them too..

[Ed.Kloonk]

I know you don't want to, but having looked at it, how hard would it be to manipulate the oscillator freq?

With old plain-jane inverters, I've wondered the viability of adding a circuit that 'leans' on the oscillator to sync it up with the mains, then switch over on the zero cross. (or 30ms before perhaps?)

Again, not asking can you. Asking could you? (with a modern model like this)

[EEVblog]

I know you don't want to, but having looked at it, how hard would it be to manipulate the oscillator freq?

With old plain-jane inverters, I've wondered the viability of adding a circuit that 'leans' on the oscillator to sync it up with the mains, then switch over on the zero cross. (or 30ms before perhaps?)

Again, not asking can you. Asking could you? (with a modern model like this)

Probably possible, but you totally wouldn't bother. The bigger dead period gap might cause you to come-a-gutsa anyway, so you are better off just getting a proper zero switch time UPS.

[Ed.Kloonk]

I know you don't want to, but having looked at it, how hard would it be to manipulate the oscillator freq?

With old plain-jane inverters, I've wondered the viability of adding a circuit that 'leans' on the oscillator to sync it up with the mains, then switch over on the zero cross. (or 30ms before perhaps?)

Again, not asking can you. Asking could you? (with a modern model like this)

Probably possible, but you totally wouldn't bother. The bigger dead period gap might cause you to come-a-gutsa anyway, so you are better off just getting a proper zero switch time UPS.

I wonder if the two freqs were indeed 180 degrees out of phase and that's why the fridge had a fit. Or a better question might be does the fridge still freak out with a 30ms interruption of steady power? 30ms seems too long anyway. Consumer ups are ~9ms aren't they?

Still be interested to know how the output frequency is derived. A discrete oscillator or from the micro.

The concern I have with the high impedance of the inverter, does one fridge starting muck up the other one running on the same line?

I wanna see some sacrificial fridges in the name of science (prior to entering summer lol).

[mikeselectricstuff]

As a comparison. the Bluetti AC500 switches to inverter in one half-cycle ( 10mS), and back with no obvious glitch at all - it re-syncs to the mains before switching back
https://youtu.be/RuEO17uvI8M?t=1954

[Billy Tree]

My objection is characterizing a specification stated up front as some kind of design flaw.
And your 3 refrigerators is not a typical use case, neither is your issue for any other user on the groups - and there have certainly been complaints.
Are they commercial refrigerators or just typical capacity household units?

[EEVblog]

My objection is characterizing a specification stated up front as some kind of design flaw.

It literally is a design flaw 
Which is why it's mentioned in the manual, so people are aware of the flaw, but I didn't see that before I did the video, which is why titles the video "RTFM Dummy"
You are just being pedantically annoying now.

And your 3 refrigerators is not a typical use case, neither is your issue for any other user on the groups - and there have certainly been complaints.
Are they commercial refrigerators or just typical capacity household units?

Just normal household fridges.
It's not the three fridges that is the issue, it's at least one of them that is disturbed by the brief interruption, but not always, it depends on what cycle it's on I suspect.
I will not discuss this further with you.

[EEVblog]

FYI, someone in the Youtube comments pointed me to fridge brownout protectors.
https://www.amazon.com/Exceline-Voltage-Electronic-Protector-Refrigerators/dp/B07Z5BH691
Silicon chip even did a project:
https://www.altronics.com.au/p/k6049-240v-mains-brownout-protector-kit/
And they don't seem to be a thing in Australia? I can't find one with an Australian plug

They detect the brownout and add a several minute delay for the fridges to lose pressure before connecting the mains.

I'm not aware of a DIN rail mount version though? A simple DIN timer may not dectect the brownout.
EDIT: Found one: https://d347awuzx0kdse.cloudfront.net/nawcontrols/product-download/17udt0.pdf

[electr_peter]

So theory that fridges are very fussy about brown-outs is confirmed. Probably because compressor has to overcome internal pressure "step", otherwise it stalls and loses momentum (and can't start again properly).

DC-DC converters, MCUs, FPGAs also may go crazy during brown-outs (via different mechanism, of course).

I see future video about this topic