Opening up my EcoFlow Delta Max to see what's inside and make a small change (which turned out to be pointless) Affiliate links: ecoflow.com?aff=138 (Works in USA only) www.amazon.com...
I’ll always respect the CZcamsr who shows their mistakes and does not try and blame others in the process. This video just reinforces this channel as one of my favourites.
I killed a JTAG debugger that way. USB debugger connected to board under test and desktop PC, accidentally touched the ground spring contact on the oscilloscope probe to the wrong side of a decoupling capacitor. Enough current flowed that the tiny spring contact was glowing briefly. All I can say is I'm glad the only thing that died was the $100 debug probe and not the $4k oscilloscope or the desktop PC.
Last thing I blew up was a lab made PSU, with some non standard features, lazy me did not isolate control board, and a DUT Made a condition where the loop was completed through that logic board.... Stupid as a fox!
By the way, and as Matthias mention Dave's eevblog; "how not to blow your oscilloscope" fits. The only thing that save this one was those old crt have a 1mm thick chasis, try that with a shiny new one... Lol!
The "epoxy" on that board is better known as conformal coat. One flavor of conformal coat can be dissolved with acetone, if you want to try the repair. Urethane CC is another variant and is much harder to repair.
I have a tip for you. If you want to change the transistors, just cut of the leads und desolder them afterwars. That makes it a lot easier. And for the transistor. You need not to you the same. Goggle the data sheet of the original ones and search ones with same specs
For what it's worth, some laptops actually do appear to be grounded between the USB ground and the charger neutral when plugged in. I discovered this while messing around with an Arduino Nano connected a Dell XPS laptop while examining output signal with a Tektronix 3052 digital oscilloscope. However, when the laptop is unplugged the USB ground then floats as you'd expect.
This is very likely just capacitive coupling. Switch mode power supplies (which all power supplies nowadays are) usually have an interference suppression capacitor between primary and secondary side, but are otherwise galvanically isolated. Connecting primary neutral to secondary ground wouldn't be possible in a lot of countries anyway, because they don't use polarized plugs. Which wire is live and which is neutral depends on which way round you plugged the power cable into the outlet. Most power supplies are designed for worldwide use and not one country, so they are built in a way that it doesn't matter which wire is live. (polarized plugs are a leftover from the beginnings of electrical wiring and haven't offered any benefit for decades as many countries in Europe, Asia and Africa will be able to tell)
Why do you say that? I feel like he's the most honest and forthright about the actual sponsored content and exactly what does and doesn't work with the product. Any revenue my preferred content creators can get is fine by me as long as the content remains the same and this unique situation they've actually provided since interesting content by being more free with the constraint on what he can and cannot show.
Pretty good sleuthing. One small correction: you need ~340V DC to generate 120V AC, not 160 as you said. If you say +/- 160V DC, that’s pretty close (slightly flat-topped, but it will run motors and rectify just fine).
It's like I once told my mother: you can't do anything to your computer that I can't fix. I, on the other hand, can break it much better. They design these things against normal people doing something stupid. It takes an engineer doing something engineer-level stupid! But at least at the end, he'll know whether it was his fault. 😁
So what you're saying is that after reverse engineering and tinkering you now have a couple of thousand dollars worth of boat anchor...that you can charge the trolling motor battery with. My electronics professor Ron Eli used to say, "Smoke test!". Great video Matthias, thumbs up.
@@Total_Egal The first amendment does not apply to private organizations, and certainly doesn't apply to anything in Canada (they have freedom of expression, which isn't quite the same thing). Had Matthias published this video without permission, he might have opened himself up to legal action.
@@F3Ibane i am not a USA Citizen.. almost the whole world ist not USA.. the Idea of fredeem of speach is what i was refering to. the point you start to argue about whats private contract, whats a nda, do you need permission to say something and whats the exakt meaning in the words in the first amendement, is there a legal action.. all that is the problem.. the solution is get your and everyones freedom of speach back dont sign such contracts and dont make such contracts and dont let them take legal actions against others.
Well that's a bummer that it's busted, but thanks for showing the innards! That is fascinating that it's using the same circuitry to charge and invert. I sometimes but don't often watch eevblog, too long winded and too many irrelevant opinions 😉 I do watch a lot of big Clive!
Based on what i've seen so far I'm pretty sure a few more hours of contemplation will allow you to identify suitable replacements for those transistors.
Yeah, I'm sure a suitable replacement shouldn't be too hard to find. It only needs to have similar specs (or better). Even if the specs are way different, he could overcome that by replacing all four, so they still match, as long as the new ones will work for the intended purpose.
I've been using the Delta Max for about a week now. Encountered no problem whatsoever. It is pretty well built from the ground up. It's nice to see the circuitry in all it's glory.
Good discussion! Yeah, the ac charger circuit is pretty sweet. Reverse inverter battery charging is so cool. Very unfortunate that you fried it! haha I hope you post a video when you get it running again.
Tough luck, but you are not the first victim of ground return damage. After all this effort, maybe Ecoflow could consider repairing the unit - that would be great PR for their company.
If he had only opened it up to inspect it sure, but once he took a sawblade to a pcb that's just lolno. The warranty is far beyond null and void at that point. We all know he knows what he's doing but such destructive modifications shouldn't be rewarded with a free repair
@@johnpaul9830 I'm not sure I can help much - either the charging circuit is faulty, or the batteries are dead. You need to disassemble it, isolate the batteries from the unit and use an external charger to check if they can receive a charge. If they don't, you need to source similar battery cells and reassemble a pack. Then, connect it to the unit and check if it charges. Not an easy task, that requires some skills and experience on these matters.
I would lose sleep. There is no way I would not get that thing up and running again. The transistors have to be available somewhere or at least cross-reference some that are compatible. Thanks for the vid!
You don't need exactly the same FET, check the maximum voltage and current rating and choose one with equal or smaller rdson (on resistance of the drain-source) and gate charge. It's a good idea to change the fet driver too if the gate is shorted.
That's still a terrible idea. Just buy a differential probe, and you will never have to worry about earth ground problems. There's plenty of affordable probes on Amazon.
If you hook up the oscilloscope to an isolation transformer and not the device-under-test, I really hope you know what you're doing. Don't suggest that to anyone who can't at the same time tell you why it's generally a bad idea.
That’s one heck of a pile of electronics. He’s right about how all of the power charging, converting and generating systems are integrated in one box including the batteries and connections. Managing all that with discrete components could be a pain.
An interesting thing I've learned is that most of these inverters, whether modified square or pure sine or grid tie, none of them use boost converters. Instead, they take the DC, convert it into a high frequency AC (10s of kHz, still battery voltage) with the first H bridge, then use a small step up transformer followed by a rectifier and capacitor to convert back to DC, and now it's the 170V or so we need for the second H bridge. Just like a standard switched mode power supply, except it steps up instead of down. Now that feeds into the second H bridge. So I wondered, why the hell have 2 H-bridges instead of just using one at battery voltage & the line frequency (60Hz) and stepping that up? Well, it turns out, if you want an equivalent transformer with the only difference being lower frequency, it has to be much bigger. I'm still not entirely clear on the details but it has something to do with the inductance/permeability of the core. Also, I believe for a pure sine inverter, the filter inductors would need to be absolutely massive, too, because they'd need to be pushing hundreds of amps. For anyone who wants to delve deeper into this, there's a guy named James Fotherby who has been building an inverter very similar to this one. He's got some incredibly good information, super valuable resource jamesfotherby.com/index-page/ EDIT: I also wanted to add, I have an old modified sine inverter that I've taken apart. One very interesting design feature is that the second (HV AC) H-Bridge is completely isolated from the rest of the circuit. It's got an opto-isolator going to the MCU, and aside from that its only connections to the outside world are the transformer (input) and the hot and neutral wires (output). The ground comes from the battery side of the board and is tied directly to battery negative. The only components that are in the isolated side of the board are the part that filters the output of the transformer, and the H-Bridge itself. Pretty elegant design really. I'm not sure why they didn't do that in this one, because it makes the neutral truly floating.
I think you would like the Fluke 43B power analyzer. It's a pretty old device so they can be found relatively cheaply on EBay. It measures and records pretty much everything you would care about in videos like these. Plus, it's a decent-ish scope meter which is fully battery operated and rated for these high voltages. I had to buy a new battery for mine since it was toast, but I have been absolutely loving the ability to watch and record transients...
That is why I always used an isolation transformer on my test equipment! but I can say I am from the days of Tube and transformers in my early days of electronics. hey I am still here and in Texas too! thanks for the video! I found you several years ago due to wood working like so many others. I like your approach to things.
This is the video explanation of what I have *always* wanted to watch! Really good, love to see more of these if it's possible with you. I do like this portable power station...so let's see what you found. As it's very late here in the UK, I'll be back later in the week.
They make cheap differential probes for scopes now. I alway use them for mains power stuff now having also thought I was too smart to ground short something…
Ok, this is the same failure I experienced on my EcoFlow DeltaMax2000 when trying to charge directly from my travel trailer breaker panel. So… exactly what would be the best work-a-round to prevent this failure? Not connecting the charging cable ground wire to the RV breaker panel neutral bar?
I love these deep dives, like you're solving a puzzle, it's great to see :D Can the batteries in that really self discharge below safe limits? That seems like a huge oversight on their part o.O
Ahhh, so this is why my ryobi batteries won't recharge. I got a few back to life by applying straight dc from a working battery to give it "drink" then put it on the charger. I know the charger needs to see some voltage before it starts charging. Didn't work for all of them.
@@richardpatterson4312 That is the same thing you can put a battery on the charger and it can show the normal voltage but soon after disconnecting the float drops back down
@@richardpatterson4312 If you have a good benchtop PSU, you can try to recover them. (But always assume the batteries are unsafe after, and keep them in a fireproof area when charging and storage). I've generally found that if the cell has more than 2 volts then it is fine to very slowly charge it up using a constant current source. Anything lower and you have to start worrying about dendrite formation.
@@richardpatterson4312 The "charger" is inside the battery itself. All the box does is provide a fixed DC voltage to the BMS circuitry inside the pack. If the logic has seen the pack go too low, it will not allow recovery. If you open it and attach an actual (CC/CV) charger to the cells directly, it might be salvageable, but any deep discharge damages the cell(s).
I got a RavPower station. I barely opened it and saw that there is such a substance all over the mainboard. How do I determine where does it come from? and what is that by the way? It is not charging well anymore and suddenly the fan turns on automaticallly for any reason and sometimes I can feel that is warm when I put it in the bag. I want to fix it.
I watched the whole video but I guess I missed it.....did it fail on its own or was it user induced? If user induced was it because of tinkering inside the unit with the hacksaw blade? Surely plugging the unit into a power strip or having a Kill a Watt plugged into it when powering a circular wasn't the caused for it to take a dump was it?
At about 10:14 when you're talking about how it switches over to direct line power it disables the inverter, you have some text saying 'wrong wrong wrong'. Is that not how it works when its plugged in? I have one of these plugged in and have my sump pump plugged into it to act like an EPS. Is this a bad solution for battery backup for the random storm and power outage?
So if you had not ruined the boards…. Would it be possible to swap out the battery either a 48v LFP drop on replacement or server rack battery? I know it won’t fit in the box but would be great to still be able to use the unit itself
There is an elegant topology for power electronics where the inherent large area PN junction in a power MOSFET is used as part of a rectifier in a H-bridge configuration. This looks like to be similar to that
At $1000 to $2000 for each of those power stations I'd seriously set about repairing the broken one. Transistors can be replaced easily enough. A Hakko FR301 desoldering pump will make short work of removing the solder from all the transistor pins. Use a Dremel wire-wheel to remove the conformal coating first, then pull the heatsink out as a unit. To search for the transistor try some of the electronics forums. The device is still in production so I have to assume the transistors are too. You just have to find out the correct designation as that's not always obvious by looking at the cases. If they really are not available then the forums can help you find equivalents.
I have a question. Do you know the new model which uses LFP battery. Everything looks same except that. So do you think we can replace old li ion battery with lfp battery? Thanks in advance. Delta max 2000 (2016wh) delta 2 max (2048wh)
I have a trashed Delta max. Error message 50hz overload, no ac output,. disconnected everything. did the reset. touched the power cord plug (not intentional!) and found it was live, gave me a good jolt. Turned off the delta max and plugged it into ac. Immediately big flash at the voltage protector and blew the breaker. What is going on with Ecoflow!
to look for short circuits in electronics I use a variable power supply AC or DC mostly DC, then I slowly turn up to about 100-500mA then you look with a thermal camera or feel forward with the hand, it works 90%. Another thing, never ground on an oscilloscope. You could use a transformer on the output of the ecoflow to go from 120v down to 15v AC with a 1Kohm resistor between the wires and measure there. I think it is the saw that makes these high frequencies but test it with an oil radiator or electric patio heater.
I always applaud people who fix their own stuff. However, Ecoflow has an excellent warranty. I had a Delta Pro that can take up to 150V solar input and a Delta 2 that only takes around 60V. Well I was messing around with the solar inputs one day and mixed them up like a moron. The Delta 2 wouldn't turn on and had the lovely burning phenolic smell. I checked the fuses and they were OK. After messing with it, I contacted Eco-Flow. They sent me a replacement free of charge even after I told them it was my mistake. (I did help them troubleshoot some SW bugs on their other models so maybe they took that into consideration. Who knows? Worth a try to warranty even if you screwed it up.
Sad that you can't source the transistors. With a unit that is nearly useless, I would certainly attempt a board repair to revive. Right to Repair hindered again by supply chain issues.
There's probably a real close alternative though. Look for the same package, the same pinout, the same voltage rating, and for Rds_on and input capacitances that are in the same order of magnitude. That'll probably work just fine.
Hi all, Thanks for the video. I have one of this DP. All working until it suddenly quit by takes no charge from any input😢 Now show blinking 0000. Please help.
For occasional emergency use, I like the idea of the power packs made by Ryobi, EGO and others that let you use your existing yard tool batteries. The downside is that they can be very expensive even without the batteries and extra batteries can be very expensive.
Why does everyone in videos disconnect the positive battery terminal first? it's much safer to disconnect the negative first if it's negative ground because that will make it so there's less places (metal chassis or ground planes) to short the terminals across.
Really well done all things considered. I think you're selling yourself short though about fixing it. I'm sure some nerds can find a compatible MOSFET/IGBT whatever that thing uses and it wouldn't be too bad to swap them out with a proper desoldering station. Desolder all of them and remove the heatsink and transistors in one shot so you're not trying to access those screws behind all the components.
Yeah, there's probably a sizable range of "close enough" compatible MOSFETS he could use. Even if he can't get two more with specs comparable to the two that still function, he could just go ahead and replace all four to keep them matching.
@@SuperDavidEF Agreed. Though finding parts like that can definitely be a needle in a haystack sometimes.
Před 2 lety+2
Ok... impressive design... Edit: It is fixable but you need a good hot air station, good soldering iron, solder pump and wick. You will need up-to-spec or better transistors not necessarily identical. Get their datasheets and get some similar or better (better rDS-on, current, lower or equal gate charge, etc) Also, when you bust a transistor or two in an H bridge, *replace them all and their gate drivers!* And, my free tip for you: put a set of S-D flyback diodes on all the H bridges. Have them as close to the transistors as physically possible. I put the tht ones on the other side of the pcb, on the transistor pins directly. Fast, but fast, 50ns fast, beefy like 2x mosfet/igbt rated drain current. The lack of flyback diodes killed your inverter. Inductive loads, huge short spikes, blew the fets and their gate drivers. Also, add more filtering. 0.5H, wound with litz wire for rated current on good ferrite torroid core. Make it identical to the existent one and use 4.7uF@250-350vac to null after the choke. A big NTC at the final output phase line will help extra high loads starting like fridges, compresdors and saws(10-5ohm max at 20°C, rated for your max current)
I know nothing about electronics, but I'm still fascinated. Is the epoxy coating on the board just for protection? When did they start doing that? I don't think I've ever seen a circuit board with epoxy before now, that I remember anyways.
It is fairly common. It is called conformal coating. There are other types besides epoxy. It may also help satisfy minimum spacing requirements between voltages. On low voltage boards that doesn't matter, but it can be an issue on higher voltages.
My guess for the reason you see the voltage spikes superimposed on the waveform is that the DC bus capacitors are specced for nominal load, whatever that would be. During startup, when a high current is drawn, the ripple generated by the full bridge boost converter appears on the output.
Taking the covers off a device and poking around inside is one way to learn about it. Of course it may destroy the device. This exercise may have accomplished your goal to reveal how this device functions and that has value. It would also be helpful to know what information the manufacturer would share with you or any other buyer of this product. Now you have a chance to find out what their technical service can offer to repair this power station. I imagine it’s well worth saving it.
Can't you replace the MOSFETs/transistor? I have a similar problem with 2 power MOSFETs in my inverter. I replaced them and they broke again. The solution. The MOSFET driver was also broken. I replaced the MOSFETs and drivers, and now it works again.
Matthias please just purchase a differential probe when probing AC systems. There are plenty of affordable probes on Amazon. They prevent all grounding issues you might encounter, since the probe does not have a true ground connection.
If the probe doesn't have a ground connctino, you need to connect the scope ground to the machine ground separately. if there's no common ground, you get garbage readings.
@@matthiasrandomstuff2221 That's the point of the differential probe? It isolates the ground of the scope with the measurement 'ground' (although there's no real 'ground' probe on the measurement side anymore - they're both floating so it's more of a positive - negative situation).
@@matthiasrandomstuff2221 This is exactly why a differential probe is useful. The probe takes the voltage across two input wires and uses it to generate a new voltage at ground level that is then fed into the scopes BNC connector. This means the red and black input wire can be connected to any voltage potential up to a kilovolt or two. The black wire on that differential probe is not connected to the scope ground in any way.
@@matthiasrandomstuff2221 No, that's not how a differential probe works. They have active circuitry inside, and do not require a common ground connection for the device under testing. They take a differential measurement between the positive and negative probes. It's effectively a two channel probe, each lead is referenced to a ground internally, but the measurement it presents to the scope is the difference of these two measurements. Since both "channels" in the probe are referenced to the same ground, this cancels out, and you're left with the true signal across the probe leads, without the need for common grounding.
I have seen CZcamsrs promoting this thing -- I feel like I have an application for this in that I have some land that is over a mile from the power lines. TBH, the cost coupled with what seems to be it being sensitive to abuse (those computer fans) makes me feel like a little gas generator is better. BUT, this would be tempting if the price per watt hour was half.
I want to work with them in the future. They are very reasonable, an engineering as opposed to marketing oriented company, so I'd rather not piss them off.
Nice the disclaimer at the beginning... I would verify if those are MOS or IGBTs, then find some that comply more or less with voltage and current requirements, and give it shot...
Such batteries like in the cars or in this box don't have high internal resistance. I have left my batteries for a 2 years and it only lost like 15% of charge. So if you charge it to 60% level which is safe and good for Lion batteries you still don't have to worry about it for like 4 or 5 years. LFP batteries are even better as they don't mind to sit at 90% charge level for long periods of time.
Your 'scope probe is connected to ground? Strange, usually the probe connection is floating, and there is a ground terminal on the scope for if you need to ground it.
@@sycc66 Strange- I have been using 'scopes for over 50 years- have even built them. Never met one with a grounded 0v line. Have seen front panel links so you can ground them, if desired. Isolated is essential for audio work - otherwise you have a ground loop.
@@Tensquaremetreworkshop I am guessing it has maybe been over 50 years since you bought a new scope then? As far as I can remember, all the scopes I have used were grounded. They have a pictorial indication on the panel that the outer shell of the BNC connector is grounded. Mostly compact tektronix models. I have been doing this for 20 years, so you have seniority on me.
@@mckenziekeith7434 You are correct that my 'scopes are of some vintage now. It may well be that recent models are grounded- my question would be - why? What advantage would permanent grounding convey? Optional, yes (front panel control) but always? Clearly, as this video shows, it has downsides, so why do it? Cheaper can be the only reason that I can see, and that translates to inferior. I guess they are usually used for logic circuits these days, but that is a poor excuse.
Just fix it dude, don't worry about the epoxy, you can buy conformal coating from any RC hobby store. You should get thermaltronics to sponsor you one of their high wattage soldering irons. Why no fuse.. that's just dumb design.
I’ll always respect the CZcamsr who shows their mistakes and does not try and blame others in the process. This video just reinforces this channel as one of my favourites.
Yes he needs lots of blame....let's not let him off the hook.....
Matthias "Big Clive" Wandel
Second this! I wonder what Big Clive would have to say about this.
Maybe, Matthias "Canada Clive" Wandel, more apt and any Big "Manx" Clive fans would certainly recognize the reference.
The old ground loop through an instrument.... happend to the best of us...
Lol!
I killed a JTAG debugger that way. USB debugger connected to board under test and desktop PC, accidentally touched the ground spring contact on the oscilloscope probe to the wrong side of a decoupling capacitor. Enough current flowed that the tiny spring contact was glowing briefly. All I can say is I'm glad the only thing that died was the $100 debug probe and not the $4k oscilloscope or the desktop PC.
Yep. BTDT
Last thing I blew up was a lab made PSU, with some non standard features, lazy me did not isolate control board, and a DUT Made a condition where the loop was completed through that logic board.... Stupid as a fox!
By the way, and as Matthias mention Dave's eevblog; "how not to blow your oscilloscope" fits. The only thing that save this one was those old crt have a 1mm thick chasis, try that with a shiny new one... Lol!
Flips circuit board over...."I think I'm getting to the bottom of it"
hmmm....yep lol 😉
Cackle, didn't notice that, nice one.
The "epoxy" on that board is better known as conformal coat. One flavor of conformal coat can be dissolved with acetone, if you want to try the repair. Urethane CC is another variant and is much harder to repair.
A wire brush on a dremel would clean enough off around the pads to desolder it without nasty fumes.
I understand little of what Matthias is saying, but I still find it interesting.
I have a tip for you. If you want to change the transistors, just cut of the leads und desolder them afterwars. That makes it a lot easier. And for the transistor. You need not to you the same. Goggle the data sheet of the original ones and search ones with same specs
I would desolder all of them and take them out still attached to the heatsink.
@@MattOGormanSmith that's a good idea.
For what it's worth, some laptops actually do appear to be grounded between the USB ground and the charger neutral when plugged in. I discovered this while messing around with an Arduino Nano connected a Dell XPS laptop while examining output signal with a Tektronix 3052 digital oscilloscope. However, when the laptop is unplugged the USB ground then floats as you'd expect.
This is very likely just capacitive coupling. Switch mode power supplies (which all power supplies nowadays are) usually have an interference suppression capacitor between primary and secondary side, but are otherwise galvanically isolated.
Connecting primary neutral to secondary ground wouldn't be possible in a lot of countries anyway, because they don't use polarized plugs. Which wire is live and which is neutral depends on which way round you plugged the power cable into the outlet.
Most power supplies are designed for worldwide use and not one country, so they are built in a way that it doesn't matter which wire is live. (polarized plugs are a leftover from the beginnings of electrical wiring and haven't offered any benefit for decades as many countries in Europe, Asia and Africa will be able to tell)
i love the fixing electronics content. Do more of it please.
to be fair he broke it twice and didn't fix it
@@natalieisagirlnow yea I was hoping he actually repaired it but the fact that he didn't probably mean his assumptions were innacurate.
I love watching bigclives teardown videos, it's super interesting to see something looked at from another perspective.
Matthias and sponsored videos is not a great marriage, great video tough.
I think bessey could be a sponsor where the two would fit perfectly.
Why do you say that? I feel like he's the most honest and forthright about the actual sponsored content and exactly what does and doesn't work with the product.
Any revenue my preferred content creators can get is fine by me as long as the content remains the same and this unique situation they've actually provided since interesting content by being more free with the constraint on what he can and cannot show.
That look at the end when he mentioned the offer 😅
Pretty good sleuthing. One small correction: you need ~340V DC to generate 120V AC, not 160 as you said. If you say +/- 160V DC, that’s pretty close (slightly flat-topped, but it will run motors and rectify just fine).
It's like I once told my mother: you can't do anything to your computer that I can't fix. I, on the other hand, can break it much better.
They design these things against normal people doing something stupid. It takes an engineer doing something engineer-level stupid! But at least at the end, he'll know whether it was his fault. 😁
"I, on the other hand, can break it much better" gave me the laugh I needed, thank you!
So: 'The more you know about something, the more you can break it.' Now that is something I have never thought of...
Nice one :D
It actually seemed like a good review overall, you were very pleased with the circuitry
So what you're saying is that after reverse engineering and tinkering you now have a couple of thousand dollars worth of boat anchor...that you can charge the trolling motor battery with. My electronics professor Ron Eli used to say, "Smoke test!". Great video Matthias, thumbs up.
The batteries, and integrated BMS (the board on the side he never shows) are likely just fine. That death-trap of an inverter should be thrown away.
"You must also watch EEVBlog videos" - why yes I do lol
great video. love this stuff. kudos to the manufacturer for letting you publish. actually makes me more willing to purchase.
its called freedom of speach
@@Total_Egal The first amendment does not apply to private organizations, and certainly doesn't apply to anything in Canada (they have freedom of expression, which isn't quite the same thing). Had Matthias published this video without permission, he might have opened himself up to legal action.
@@F3Ibane i am not a USA Citizen.. almost the whole world ist not USA.. the Idea of fredeem of speach is what i was refering to.
the point you start to argue about whats private contract, whats a nda, do you need permission to say something and whats the exakt meaning in the words in the first amendement, is there a legal action.. all that is the problem..
the solution is get your and everyones freedom of speach back
dont sign such contracts and dont make such contracts and dont let them take legal actions against others.
@@Total_Egal *Speech. For something you're so concerned about, you'd think you'd be able to spell it.
@@bittrippn6570 Or maybe he's secretly a strawman planted by the corporations to make us think that smart ideas are actually dumb.
Well that's a bummer that it's busted, but thanks for showing the innards! That is fascinating that it's using the same circuitry to charge and invert. I sometimes but don't often watch eevblog, too long winded and too many irrelevant opinions 😉 I do watch a lot of big Clive!
totally agree on eevblog
I rather like Dave :(
Thanks for posting an honest video. I’m sure Marco Reps would have added a laser to a repair of that.
Based on what i've seen so far I'm pretty sure a few more hours of contemplation will allow you to identify suitable replacements for those transistors.
Yeah, I'm sure a suitable replacement shouldn't be too hard to find. It only needs to have similar specs (or better). Even if the specs are way different, he could overcome that by replacing all four, so they still match, as long as the new ones will work for the intended purpose.
I've been using the Delta Max for about a week now. Encountered no problem whatsoever. It is pretty well built from the ground up. It's nice to see the circuitry in all it's glory.
Good discussion! Yeah, the ac charger circuit is pretty sweet. Reverse inverter battery charging is so cool. Very unfortunate that you fried it! haha I hope you post a video when you get it running again.
Hah, I don't know why I was surprised to find you here XD
Tough luck, but you are not the first victim of ground return damage. After all this effort, maybe Ecoflow could consider repairing the unit - that would be great PR for their company.
If he had only opened it up to inspect it sure, but once he took a sawblade to a pcb that's just lolno. The warranty is far beyond null and void at that point. We all know he knows what he's doing but such destructive modifications shouldn't be rewarded with a free repair
@@Koushakur They should at least give him a link to the transistors
My delta 2 is out of warranty, won’t charge anymore is there a way to change the battery cells? Or do I have a paper weight? Lol thanks for your time
@@johnpaul9830 I'm not sure I can help much - either the charging circuit is faulty, or the batteries are dead. You need to disassemble it, isolate the batteries from the unit and use an external charger to check if they can receive a charge. If they don't, you need to source similar battery cells and reassemble a pack. Then, connect it to the unit and check if it charges. Not an easy task, that requires some skills and experience on these matters.
I would lose sleep. There is no way I would not get that thing up and running again. The transistors have to be available somewhere or at least cross-reference some that are compatible. Thanks for the vid!
Good Video! , its good to see a more humble side of the wizard!!
Great to see what's inside these power generators. They are very well built.
You don't need exactly the same FET, check the maximum voltage and current rating and choose one with equal or smaller rdson (on resistance of the drain-source) and gate charge. It's a good idea to change the fet driver too if the gate is shorted.
I can't even find specs for the ones on the board.
@@matthiasrandomstuff2221 Want to share the part number? I work in electronics, might be able to find something here.
John Ward did a good explanation about using oscilloscopes with isolation transformers to avoid this.
How many do you think you will sell?
That's still a terrible idea. Just buy a differential probe, and you will never have to worry about earth ground problems. There's plenty of affordable probes on Amazon.
If you hook up the oscilloscope to an isolation transformer and not the device-under-test, I really hope you know what you're doing. Don't suggest that to anyone who can't at the same time tell you why it's generally a bad idea.
@@superdau hook the item under test to the isolation transformer
That’s one heck of a pile of electronics. He’s right about how all of the power charging, converting and generating systems are integrated in one box including the batteries and connections. Managing all that with discrete components could be a pain.
An interesting thing I've learned is that most of these inverters, whether modified square or pure sine or grid tie, none of them use boost converters. Instead, they take the DC, convert it into a high frequency AC (10s of kHz, still battery voltage) with the first H bridge, then use a small step up transformer followed by a rectifier and capacitor to convert back to DC, and now it's the 170V or so we need for the second H bridge. Just like a standard switched mode power supply, except it steps up instead of down. Now that feeds into the second H bridge.
So I wondered, why the hell have 2 H-bridges instead of just using one at battery voltage & the line frequency (60Hz) and stepping that up? Well, it turns out, if you want an equivalent transformer with the only difference being lower frequency, it has to be much bigger. I'm still not entirely clear on the details but it has something to do with the inductance/permeability of the core. Also, I believe for a pure sine inverter, the filter inductors would need to be absolutely massive, too, because they'd need to be pushing hundreds of amps.
For anyone who wants to delve deeper into this, there's a guy named James Fotherby who has been building an inverter very similar to this one. He's got some incredibly good information, super valuable resource jamesfotherby.com/index-page/
EDIT: I also wanted to add, I have an old modified sine inverter that I've taken apart. One very interesting design feature is that the second (HV AC) H-Bridge is completely isolated from the rest of the circuit. It's got an opto-isolator going to the MCU, and aside from that its only connections to the outside world are the transformer (input) and the hot and neutral wires (output). The ground comes from the battery side of the board and is tied directly to battery negative. The only components that are in the isolated side of the board are the part that filters the output of the transformer, and the H-Bridge itself. Pretty elegant design really. I'm not sure why they didn't do that in this one, because it makes the neutral truly floating.
I think you would like the Fluke 43B power analyzer. It's a pretty old device so they can be found relatively cheaply on EBay. It measures and records pretty much everything you would care about in videos like these. Plus, it's a decent-ish scope meter which is fully battery operated and rated for these high voltages.
I had to buy a new battery for mine since it was toast, but I have been absolutely loving the ability to watch and record transients...
I love watching your videos even when I have no idea what you’re talking about
That is why I always used an isolation transformer on my test equipment! but I can say I am from the days of Tube and transformers in my early days of electronics. hey I am still here and in Texas too! thanks for the video! I found you several years ago due to wood working like so many others. I like your approach to things.
This is the video explanation of what I have *always* wanted to watch! Really good, love to see more of these if it's possible with you. I do like this portable power station...so let's see what you found. As it's very late here in the UK, I'll be back later in the week.
Time for a collab with ElectroBoom :D
I used the same calculator 10 years ago in college. It's the best!
They make cheap differential probes for scopes now. I alway use them for mains power stuff now having also thought I was too smart to ground short something…
massive respect ecoflow
Pretty good technical review of the Delta and really see what's inside this heck of a powerful power station.
Very interested and helpful. Thnk You from Ukraine.
I understand very little of this but love watching it.
Ok, this is the same failure I experienced on my EcoFlow DeltaMax2000 when trying to charge directly from my travel trailer breaker panel. So… exactly what would be the best work-a-round to prevent this failure? Not connecting the charging cable ground wire to the RV breaker panel neutral bar?
I love these deep dives, like you're solving a puzzle, it's great to see :D
Can the batteries in that really self discharge below safe limits? That seems like a huge oversight on their part o.O
Yes, this is true for ALL lithium battery chemistry. External circuitry can only make it worse by taking electricity out of the battery.
Ahhh, so this is why my ryobi batteries won't recharge.
I got a few back to life by applying straight dc from a working battery to give it "drink" then put it on the charger. I know the charger needs to see some voltage before it starts charging.
Didn't work for all of them.
@@richardpatterson4312 That is the same thing you can put a battery on the charger and it can show the normal voltage but soon after disconnecting the float drops back down
@@richardpatterson4312 If you have a good benchtop PSU, you can try to recover them. (But always assume the batteries are unsafe after, and keep them in a fireproof area when charging and storage). I've generally found that if the cell has more than 2 volts then it is fine to very slowly charge it up using a constant current source. Anything lower and you have to start worrying about dendrite formation.
@@richardpatterson4312 The "charger" is inside the battery itself. All the box does is provide a fixed DC voltage to the BMS circuitry inside the pack. If the logic has seen the pack go too low, it will not allow recovery. If you open it and attach an actual (CC/CV) charger to the cells directly, it might be salvageable, but any deep discharge damages the cell(s).
That's why we love you, Matthias!
I got a RavPower station. I barely opened it and saw that there is such a substance all over the mainboard. How do I determine where does it come from? and what is that by the way? It is not charging well anymore and suddenly the fan turns on automaticallly for any reason and sometimes I can feel that is warm when I put it in the bag. I want to fix it.
I watched the whole video but I guess I missed it.....did it fail on its own or was it user induced? If user induced was it because of tinkering inside the unit with the hacksaw blade? Surely plugging the unit into a power strip or having a Kill a Watt plugged into it when powering a circular wasn't the caused for it to take a dump was it?
At about 10:14 when you're talking about how it switches over to direct line power it disables the inverter, you have some text saying 'wrong wrong wrong'. Is that not how it works when its plugged in? I have one of these plugged in and have my sump pump plugged into it to act like an EPS. Is this a bad solution for battery backup for the random storm and power outage?
What are the full implications of not being able to tie neutral to ground anywhere in your electrical system? It seems significant and limiting.
So if you had not ruined the boards….
Would it be possible to swap out the battery either a 48v LFP drop on replacement or server rack battery?
I know it won’t fit in the box but would be great to still be able to use the unit itself
There is an elegant topology for power electronics where the inherent large area PN junction in a power MOSFET is used as part of a rectifier in a H-bridge configuration. This looks like to be similar to that
At $1000 to $2000 for each of those power stations I'd seriously set about repairing the broken one. Transistors can be replaced easily enough. A Hakko FR301 desoldering pump will make short work of removing the solder from all the transistor pins. Use a Dremel wire-wheel to remove the conformal coating first, then pull the heatsink out as a unit. To search for the transistor try some of the electronics forums. The device is still in production so I have to assume the transistors are too. You just have to find out the correct designation as that's not always obvious by looking at the cases. If they really are not available then the forums can help you find equivalents.
I followed everything he said because i stayed at a holiday in last night lol 😂
I have a question. Do you know the new model which uses LFP battery. Everything looks same except that. So do you think we can replace old li ion battery with lfp battery? Thanks in advance. Delta max 2000 (2016wh) delta 2 max (2048wh)
I don't know if those chemistries are compatible. Voltages may be different.
I have a trashed Delta max. Error message 50hz overload, no ac output,. disconnected everything. did the reset. touched the power cord plug (not intentional!) and found it was live, gave me a good jolt. Turned off the delta max and plugged it into ac. Immediately big flash at the voltage protector and blew the breaker. What is going on with Ecoflow!
actually if you want one of these for emergencys, just cycle the battery every week or two to keep it fresh.
thx. i watched the whole thing.... understood basically zero, but one day I may learn.... keep up the videos!
That inverter circuit is so interesting. Thank you for breaking it (down) :)
I was thinking 'just like the eevblog video' then you said it. Haunting.
to look for short circuits in electronics I use a variable power supply AC or DC mostly DC, then I slowly turn up to about 100-500mA then you look with a thermal camera or feel forward with the hand, it works 90%. Another thing, never ground on an oscilloscope. You could use a transformer on the output of the ecoflow to go from 120v down to 15v AC with a 1Kohm resistor between the wires and measure there. I think it is the saw that makes these high frequencies but test it with an oil radiator or electric patio heater.
Of course, I was still there (11:00/01). 😅
I enjoy your channels!
I always applaud people who fix their own stuff. However, Ecoflow has an excellent warranty. I had a Delta Pro that can take up to 150V solar input and a Delta 2 that only takes around 60V. Well I was messing around with the solar inputs one day and mixed them up like a moron. The Delta 2 wouldn't turn on and had the lovely burning phenolic smell. I checked the fuses and they were OK. After messing with it, I contacted Eco-Flow. They sent me a replacement free of charge even after I told them it was my mistake. (I did help them troubleshoot some SW bugs on their other models so maybe they took that into consideration. Who knows? Worth a try to warranty even if you screwed it up.
We need a Matthias and electoboom collaboration video discussing this topic. Tons of confusion between neutral and ground to the uninitiated.
Sad that you can't source the transistors. With a unit that is nearly useless, I would certainly attempt a board repair to revive. Right to Repair hindered again by supply chain issues.
There's probably a real close alternative though. Look for the same package, the same pinout, the same voltage rating, and for Rds_on and input capacitances that are in the same order of magnitude. That'll probably work just fine.
I'm curious to know how it works backwards from AC source to charging the DC lithium batteries
Isolation transformer... Is what we should all learn from this video.
This is probably also a good video for lawyers. Why you need to include those lengthy legal disclaimers for your products.
yeah
Hi all, Thanks for the video. I have one of this DP. All working until it suddenly quit by takes no charge from any input😢 Now show blinking 0000. Please help.
2:49 are those four FULL BRIDGE RECTIFIERs used instead of puny diodes?
This unit uses MOSFETs instead of diodes for rectification. But they are in a full bridge configuration.
For occasional emergency use, I like the idea of the power packs made by Ryobi, EGO and others that let you use your existing yard tool batteries. The downside is that they can be very expensive even without the batteries and extra batteries can be very expensive.
Why does everyone in videos disconnect the positive battery terminal first? it's much safer to disconnect the negative first if it's negative ground because that will make it so there's less places (metal chassis or ground planes) to short the terminals across.
Really well done all things considered. I think you're selling yourself short though about fixing it. I'm sure some nerds can find a compatible MOSFET/IGBT whatever that thing uses and it wouldn't be too bad to swap them out with a proper desoldering station. Desolder all of them and remove the heatsink and transistors in one shot so you're not trying to access those screws behind all the components.
Yeah, there's probably a sizable range of "close enough" compatible MOSFETS he could use. Even if he can't get two more with specs comparable to the two that still function, he could just go ahead and replace all four to keep them matching.
@@SuperDavidEF Agreed. Though finding parts like that can definitely be a needle in a haystack sometimes.
Ok... impressive design...
Edit:
It is fixable but you need a good hot air station, good soldering iron, solder pump and wick.
You will need up-to-spec or better transistors not necessarily identical. Get their datasheets and get some similar or better (better rDS-on, current, lower or equal gate charge, etc)
Also, when you bust a transistor or two in an H bridge, *replace them all and their gate drivers!*
And, my free tip for you: put a set of S-D flyback diodes on all the H bridges. Have them as close to the transistors as physically possible. I put the tht ones on the other side of the pcb, on the transistor pins directly.
Fast, but fast, 50ns fast, beefy like 2x mosfet/igbt rated drain current.
The lack of flyback diodes killed your inverter. Inductive loads, huge short spikes, blew the fets and their gate drivers.
Also, add more filtering. 0.5H, wound with litz wire for rated current on good ferrite torroid core.
Make it identical to the existent one and use 4.7uF@250-350vac to null after the choke.
A big NTC at the final output phase line will help extra high loads starting like fridges, compresdors and saws(10-5ohm max at 20°C, rated for your max current)
You should use an isolating transformer!
I know nothing about electronics, but I'm still fascinated. Is the epoxy coating on the board just for protection? When did they start doing that? I don't think I've ever seen a circuit board with epoxy before now, that I remember anyways.
Probably for reliability and corrosion resistance
@@matthiasrandomstuff2221 Makes sense. Thank you for the response.
It is fairly common. It is called conformal coating. There are other types besides epoxy. It may also help satisfy minimum spacing requirements between voltages. On low voltage boards that doesn't matter, but it can be an issue on higher voltages.
@@mckenziekeith7434 Interesting. Thanks for the response.
My guess for the reason you see the voltage spikes superimposed on the waveform is that the DC bus capacitors are specced for nominal load, whatever that would be. During startup, when a high current is drawn, the ripple generated by the full bridge boost converter appears on the output.
Taking the covers off a device and poking around inside is one way to learn about it. Of course it may destroy the device. This exercise may have accomplished your goal to reveal how this device functions and that has value. It would also be helpful to know what information the manufacturer would share with you or any other buyer of this product. Now you have a chance to find out what their technical service can offer to repair this power station. I imagine it’s well worth saving it.
Wow! I'm the TX guy looking for a frozen winter power source. Thanks for the video.
for that, a generator would be better, like I said. Unless the power fails very often.
Can't you replace the MOSFETs/transistor? I have a similar problem with 2 power MOSFETs in my inverter. I replaced them and they broke again. The solution. The MOSFET driver was also broken. I replaced the MOSFETs and drivers, and now it works again.
I'm impressed by your electronic knowledge. Respect!
Next time, you can use a differential probe for best results.
Guess I'm confused - if I want to do a review or teardown of a product I paid for, why the hell do I need the manufacturers permission?
He hasn't paid for it
He didn't buy the box, they gave it to him as part of a previous sponsored video. This particular video, however, is not sponsored.
Talking about circiut boards like that....is like magic to me o.O
“By seeking and blundering we learn.” ― Johann Wolfgang von Goethe
Matthias please just purchase a differential probe when probing AC systems. There are plenty of affordable probes on Amazon. They prevent all grounding issues you might encounter, since the probe does not have a true ground connection.
If the probe doesn't have a ground connctino, you need to connect the scope ground to the machine ground separately. if there's no common ground, you get garbage readings.
@@matthiasrandomstuff2221 That's the point of the differential probe? It isolates the ground of the scope with the measurement 'ground' (although there's no real 'ground' probe on the measurement side anymore - they're both floating so it's more of a positive - negative situation).
@@matthiasrandomstuff2221 This is exactly why a differential probe is useful. The probe takes the voltage across two input wires and uses it to generate a new voltage at ground level that is then fed into the scopes BNC connector. This means the red and black input wire can be connected to any voltage potential up to a kilovolt or two. The black wire on that differential probe is not connected to the scope ground in any way.
@@matthiasrandomstuff2221 No, that's not how a differential probe works. They have active circuitry inside, and do not require a common ground connection for the device under testing. They take a differential measurement between the positive and negative probes. It's effectively a two channel probe, each lead is referenced to a ground internally, but the measurement it presents to the scope is the difference of these two measurements. Since both "channels" in the probe are referenced to the same ground, this cancels out, and you're left with the true signal across the probe leads, without the need for common grounding.
Any idea's my storage battery is not charging from my main Unit
OMG the look on your face at the outtro made me laugh so hard.
This video is a not-so-subtle reminder of how I nearly failed basic electrical in electrician A-school
I have seen CZcamsrs promoting this thing -- I feel like I have an application for this in that I have some land that is over a mile from the power lines. TBH, the cost coupled with what seems to be it being sensitive to abuse (those computer fans) makes me feel like a little gas generator is better. BUT, this would be tempting if the price per watt hour was half.
I want to upgrade the battery in mine, possible?
Did EcoFlow replace the device in exchange for the discount ad? If not, why worry about consent?
I want to work with them in the future. They are very reasonable, an engineering as opposed to marketing oriented company, so I'd rather not piss them off.
Nice the disclaimer at the beginning... I would verify if those are MOS or IGBTs, then find some that comply more or less with voltage and current requirements, and give it shot...
The closest to it I could find I think was a darlignton transistor.
What transistor model is it? I've had some good luck with finding replacements (or suitable substitutes).
Also, your link goes to a 404 page.
I guess you'll never hear from them again!
Such batteries like in the cars or in this box don't have high internal resistance. I have left my batteries for a 2 years and it only lost like 15% of charge. So if you charge it to 60% level which is safe and good for Lion batteries you still don't have to worry about it for like 4 or 5 years. LFP batteries are even better as they don't mind to sit at 90% charge level for long periods of time.
So the AC output is from full H-Bridge, does it mean you cannot have one pole of the AC bonded to the safety ground/circuit ground?
Who makes the cells?
the input circuit is a DAB (dual active bridge), quite a popular topology for smart grids applications
Does this unit use a 12v 200ah battery?
Your 'scope probe is connected to ground? Strange, usually the probe connection is floating, and there is a ground terminal on the scope for if you need to ground it.
Not strange at all, probes are almost always grounded. That's why you never measure differential voltages without a differential probe.
@@sycc66 Strange- I have been using 'scopes for over 50 years- have even built them. Never met one with a grounded 0v line. Have seen front panel links so you can ground them, if desired. Isolated is essential for audio work - otherwise you have a ground loop.
@@Tensquaremetreworkshop I am guessing it has maybe been over 50 years since you bought a new scope then? As far as I can remember, all the scopes I have used were grounded. They have a pictorial indication on the panel that the outer shell of the BNC connector is grounded. Mostly compact tektronix models. I have been doing this for 20 years, so you have seniority on me.
@@mckenziekeith7434 You are correct that my 'scopes are of some vintage now. It may well be that recent models are grounded- my question would be - why? What advantage would permanent grounding convey? Optional, yes (front panel control) but always? Clearly, as this video shows, it has downsides, so why do it? Cheaper can be the only reason that I can see, and that translates to inferior. I guess they are usually used for logic circuits these days, but that is a poor excuse.
Just fix it dude, don't worry about the epoxy, you can buy conformal coating from any RC hobby store. You should get thermaltronics to sponsor you one of their high wattage soldering irons. Why no fuse.. that's just dumb design.
the problem is not adding epoxy. The problem is that there is epoxy on it.
" NO "
Solar Generators are great !
Just Don't Let You Around Them ! lmao
Thanks for the info, guess it's time to get myself a Delta.