Bars on top are arc catch plates, you will have a big arc on breaking under current, and closing on current. Contacts are large to handle the inrush current, which can be 2kA, and the flexible copper braid has to be rated to handle this inrush current. Common parts in all AC contactors and all AC breakers, to quench the arc and connect the moving contact to the terminal. Ideally you want a separate delay relay in the feed to the incoming supply, so that power has to be applied for 30 seconds before it is reconnected, so that all the transients on the line can die down, and also the massive inrush load can subside from all the other loads being turned on cold.
I used to work for a company that manufactured very large backup power units (our largest at the time was 1 MVA). The contacts and bus bars used were massive as were the arc extinguishing plates. Any large switchgear manufacturers will have similar (Siemens, Square-D, Eaton, etc.) devices. I could not find any other test videos of contacts up close but this give the general idea of just how violent it can be for large gear and high currents: th-cam.com/video/DS06Q6oMRYY/w-d-xo.html
Agree with all you've said. Came here to comment those plates are 'arc chutes' to quench the opening arc. Now, when you get into entire switchgear being transferred, you even get into 'load-shedding'. Some rotary loads with high moments of inertia have to be stripped before allowing second power source to close in (otherwise the sudden reenergizing of those loads will create massive torque transients). 'Fast transfer', vs. 'slow transfer' versus 'dead transfer'. With BIG loads, it's a lot of careful study.
When you get into larger transfer switches they of course get a lot more sophisticated. Automatic delays to ensure the grid is actually stable before switching over, voltage/frequency checks, and feedback to modulate the generator speed to match phases and frequency before switching.
That would be a concern of mine. Voltage sag with a battery source could produce a rapid on off condition as would an overloaded generator. There should be a delay circuit on the line in if used in this configuration.
I was thinking that's really missing on here, when the power here comes back on it's often very dirty and often has to "try" more than once before it's reliable
None of the 2,000A switches I work on ever bother with generator synchronization. They just do a 1s break before make. If the load is critical, you slap a UPS on it.
Hi Dave, Imagine this scenario: Your freezers are powered by battery in the night and the battery is empty. Then the ATS switches to mains. However, since the battery is not under load anymore, the voltage increases back again. Then the ATS switches back to battery. Then because of not having enough power, ATS switches back to mains...then loop.... I am sure that your BMS is smart enough to cut entire battery output off (high Z) after the battery depleted and not enable it again until it is charged. But anyway keep this problem in mind ;) Cheers.
LOL as I sit and watch this video in South Africa 🇿🇦, we currently have whats called "load shedding" stage 4. Power off for 2 hours at a time, 3 times a day minimum. In stark contrast to experiencing only a couple of hours in a lifetime in Sydney.🙈 Great video as always! Thanks Dave!
I have a similar setup at home to use battery power when electricity is expensive. I have found that switching while a fridge is running and the voltages are too far out of phase between inverter and grid the fridge kompressors will stop abruptly. I am aware double positive (or negative) ac pulse essentially is DC and acts as a brake on induction motors when using vfd's, I suspect my fridge compressor have a very low rotating mass as well and out of phase power upsets it. However measuring voltages with a multimeter and transfering between sorces when both are near 0 volts betwheen line 1's keeps the fridge happy and it continoues to run. Inverter fridges dont care, but I recommend testing this in your setup.
Cute little baby transfer switch.😁 The data center at the bank I used to work for was more the size of a refrigerator, and had a huge lever, not a tiny knob. Those microswitches are effectively the negative feedback side of the circuit. Typical sort of a limit switch setup: apply power until you're in the right position and then stop.
18:30 Those metal fins are there to help quench any arc that's formed as the contacts separate, I'd imagine -- same as for the fins you find in protective circuit breakers.
My guess is that the switching time varies depending on where in the sine wave the switch happens. If it's near a zero crossing it will take longer because the solenoids may not pull the rod until closer to peak power. So you would get some ~10ms (half cycle) variation in the timing.
The quality of the first unit you tore down is much better than that of the second with the green selector switch. The one with the green switch is a "stacked" design used to add more switch sets. I have some of these I used for my generator transfer switch with 4 contacts (I only needed 3) and the outer most switch fails spectacularly when there is a surge. I already lost 2 during power outages. Always the outer most switch assembly. It seems since these are stacked and each one "plugs" into the adjacent switch plastic lever pivot, the strength holding the contacts closed gets weaker the more poles are added due to poor tolerances! The original one you tore down is not stacked and should not suffer from this defect.
I wonder if the plastic in those transfer switches meets the necessary flammability requirements? I have a feeling it doesn't. But I at least like how the slightly crooked "QC PASS 08" stamp in the manual is actually just an image in the pdf :) The metal sheets at 18:38 look like arc splitters to cool and extinguish any arcs from breaking loads (particularly inductive ones), not sure if I'd trust them at 400 VAC 63 A but probably good enough at 240 V 10 A. However one thing I don't understand is that both units have speed holes on the front plastic cover presumably either to ensure the impending explosion is vented forwards towards the user to cause maximum damage or maybe they had issues with the low quality contacts overheating and their solution was just to add ventilation slots, but seems like it's just begging for contamination issues. I also assume this change over switch and battery inverter will not be considered a 'fixed installation'?
I'm betting not, I do not see any real separation between Gen and Out, I'd expect that thing to be a fire hazard, and your screwed as insurance wont pay out. I do not see any Aussie Compliance "RCM" markings, just the fakish "C"hina "E"lectrics mark (note the non existence of the spacing between C and E which exists on the real European thing)
yeah, not that I care (because we do it at work all the time), but I'm certain this wouldn't pass AUS electrical code. I doubt DIN rail is certified for use anywhere inside the home. Sadly, as its really useful stuff.. wish my breaker panel was all DIN.
@@Ressy66 Agreed, except for the whole "China Electrics" thing, which isn't real (and wouldn't make sense anyhow). CE markings are self-issued so if it's on there, the manufacturer/importer claims it meets the required standards.
Good informative video. I bought similar from same manufacturer and now I learned how to wire it up thanks. $25 here in Thailand off Lazada. Same project, running guest house off solar and a Bluetti AC200MAX with mains grid as backup.
These are very simple, very crude, but also very effective for small single phase loads. As pointed out by many others, the strengthening bars are intended to break/split/dissipate the arc when switching loads. For small single phase it obviously works, I doubt they would come anywhere near the rating claimed (50kA) or the full load for the no. of operations claimed. As also pointed out by others, professional units will have multiple time delays built in to prevent hunting in the event of voltage fluctuations, and also run-on timers for generators. But they also cost accordingly :-) Great video, thanks.
congratz on 1500 videos(+ or - a few not in the main series) dave! i've found a few pieces of industrial kit like this before, some beefy switches and motor contactors from some pumps, amazingly simple and yet surprisingly complex for what they do
I like this switch, it's so simple, not over-engineered and should be reliable. Automatic transfer switches make sense 1) when you have two separate utility connections (primary and backup), or 2) you've got a generator that can be remotely started. I wonder if the relay uses its auxiliary contacts for controlling the generator startup; it would make sense since as the mains dies and the relay turns off, the aux contact gets toggled, starts the generator, and when the generator is up and running and supplying power, the solenoid in the ATS gets energized and the changeover happens. Fridges won't care, but for any servers, computers etc. get yourself a UPS that can keep them powered through the changeover process. You got a lot of distortion in your mains. All that SMPS rubbish - forget about clear radio transmissions... It's sad.
It indeed seems that the relay contacts on the primary (A) side are intended for a generator start signal - C and NO at the top of the device. The manual for these can be found if you look up Geya Electrical, model W2R
I've seen these used within the water industry along with other sine wave normalizers, delays, soft starts - soft stops, VFD switches and current shunts for very high voltage electric motors.
I have Tested with Timer, and it works fantastic. The Timer just have to be connected to the green poles AR and AN or BR and N. And it work with the Time you want. :)
These ATS are not for rotating devices as one cannot ensure that both power supplies are (were) in-phase. The worse case scenario is opposed phase which would surely stall an rotating device.
Small motors will survive. But really large equipment, you're absolutely right. Our system would check and if drifted out of phase, strip the motors before closing the alternate source feeder.
@@animarkzero it would trip any upstream breakers. The battery inverter would likely cop it too. The mechanical delay should be sufficient at 180° however the waveform is going to look like the McDonald's M. It will cause massive inrush currents on any connected transformers and the harmonic will take several cycles to dissapate.
Hi Dave, very interesting. I wanted to use this but was concerned about safety, so built a servo to a manual transfer switch (video on channel!). But this looks great - and now I can see inside I'm happy to spend the $50. Cheers!
@@AndyMcBlane These are chinesium devices and at the pricepoint I would not take the current rating seriously! When you pay up to 10x as much for brand ATS at the same current rating(63A) you know something is fishy.
Very good useful video I just ordered one of these (Jotta Brand same thing off Ebay) Thank you! cant beat the price of these $35.26. Of course I am sure they are not UL rated, but reviews are all good.
I'm actually doing the same thing but my continual load is a rack of servers so the changeover needs to be smooth and phase synced. Instead of a timer though and a commercial battery/inverter, I picked up an old 3000VA UPS and modified it so it can't charge the batteries attached to it. I gutted the batteries from it and connected a 14.4KWh LiFePO4 array to it's "extended runtime" input and built a little circuit with a STM32F103C8T6 at it's heart that measures current consumption, battery charge state, among other things. When the charge state is high and solar input is high it disconnects the UPS from grid power making use of it's internal switchover function. The batteries are just connected as per normal to a MPPT via a BMS to charge from solar. Because I am using a high grade UPS that is designed for 100% duty, it has the added benefit of locking it's AC output phase with the grid phase, and when it takes over the phase is in sync with the grid, and when it goes back to mains power again it takes about 3 seconds while it re-syncs it's output to match the grid phase before switching back over.
Just a thought Dave... Do you really want to switch your fridges (essentially single phase motors under load) from one source to another almost instantaneously??? Will the solar inverter power source be synchronised to mains frequency? If so, no wukkas But I'd think you might get into difficult territory if the two sources are not in synch When I was responsible for a data centre, we were always in synch (courtesy of a UPS on the output side of the transfer switch) for most things, but we still enforced a minute or two gap between power sources (so we didn't damage the HVAC)
Does that circuit have sufficient isolation between the grid and backup sides? The distance between the relay's pads is awfully small. With the right fault condition it seems like there would be a possibility of back-feeding power into the grid through the control circuitry (creating a hazard for linemen etc.).
Nice Video!!! Some risk could be when your battery inverter (connected to your A input) goes into a "brownout. ---> (i) Battery "empty" - turns off the power - (ii) Relay swaps to B (grid in your case) - Battery has no load at that time (iii) battery and inveter recovers - since there is no load - (iv) relay swaps back to source A ---> we are back at the start.
For your 3 fridges, I have seen someone make a 3 relay box, to prevent all 3 loads starting at the same time. It's a very low chance of this happening, but with thermostats opening & closing, it can happen.
@@slypig24 but the compressor will turn off after a time, so how do they detect when it would turn on again and prevent 2 from turning on at the same time?
@@Philip8888888 it is most likely just for the initial switch on, when all three compressors would kick in at the same instant... I. E. Turn them on in sequence with a 10 second delay or such, and then whatever happens happens. Yes, you could get fancy with power and even temperature monitoring, but more risk of something failing then.
Used to work on megawatt backup gen's - when the xfer switch on those kick over, you better be paying attention and plug your ears! Some sounded like a car crash!
How’s there people commented 8 hours ago while the video just been uploaded 20 seconds ago? 😅 Btw, this subject is literally important for any beginner electrical engineer, thanks in advance ❤️
@@EEVblog dont lie to us! we know you opened a time portal in your yard and have been allowing patreons to play with the course of history! if this wasnt true, WHY IS MY COPY OF BERNSTEIN BEARS CALLED BEARSTAIN BEARS?????
Time travelers of course. Now you really don't know what the answer is do you? Oh and I'm trying to find an IBM 5100 so I can fix the future and my time machine.
Thanks Dave! Just about to put one in my caravan to switch between mains (aka “shore”) power, and my inverter.. allows me to use my microwave and coffee machine (and any other 240vac) when travelling
Heat pumps generally can't immediately restart. If there's a power failure, it should stay off for a minute or two before applying power again. They have a soft-start accumulator tank which fills when first turned on. Applying power while that is full may stall the compressor. I suspect inverter-driven units take care of this electronically, but I'm not certain.
At about 18:45, the two perforated bars may be arc extinguishing means? You see such on higher voltage and power switches or any breakers capable for over 10 kA breaking. The idea is to let the arc extend by its own magnetic effekt and push it to multiple gaps that also cool the arc.
You might consider using one of the easun/powland cheap Chinese inverters with automatic bypass and built in charger when you have direct 24/48v battery connection. 5kW are around $400 and unlike typical ATS when battery is low it switches like ups without circuit disruption to grid power. Your ats is here really very fast! You may also configure safe buffer, to reserve a bit capacity should grid fail. those inverters are really good and configurable as for that price.
The first box internally it looks real cheap-arsed. The way the solenoids are positioned seem a tad wrong. The later switch box looks a bit better... 10/10 for the vid
Was very impressed with the portrait DaveCAD. Indeed rare and it looks like you have plenty of portrait cycles left based on the thickness of the DaveCAD Pad. DCP
Here in europe where power is expensive i do measure actual power into/from the grid and switch extra consumers according to solar surplus. Measuring is done with 3-phase shelly-3em over WiFi, switching load is done with sonoff-basic over WiFi.
Our CBD office has independent connections to two separate grids because it's in a bank building, which is great because it basically meant we didn't have to have a UPS for our data centre. It's practically impossible for both grids to be down at the same time, because that would mean the head offices for most of the major banks and financial institutions in the country would go down. Well, one day there was a major power outage in the CBD, it took out one of our grid feeds, but when our automatic transfer switch switched over, it decided half way through that it was too much work and exploded, taking out the second power feed and our entire office and data centre. Everyone was basically left standing around twiddling their thumbs for a couple of hours until they could get an electrician to bypass the transfer switch and reconnect the one good power feed. Unfortunately the disruption probably cost the business about half a million Aussie dollary doos. But that's not the end of the story. Do you think this convinced them to install a UPS? Nope! Instead they decided that in the event of losing one grid feed, they'd have an enormous generator on the back of a truck, which someone would drive into the CBD and park on the street outside the loading dock. Then they'd roll out a huge cable up the ramp, through the loading dock, and into a panel to feed another transfer switch that they could use to power up the missing feed. Yep, a flawless plan. Thankfully we haven't yet needed to implement it, since the grids have been stable since then. Unfortunately the UPS in our head office wasn't as reliable ... but that's a story for another time.
Thats a pretty nifty setup, my grid here in the rural US sucks. During the winter we routinely loose power for week's/month's. Ive been wanting to build a system to automatically switch my homelab server rack (with UPS) to a generator.
Hopefully you can tell your battery inverter to not re-energize until it reaches a certain SoC, otherwise this might chatter as the inverter energizes, then quickly browns out. Nicer ones have time delays to prevent that (wait 30+ seconds to make sure grid is actually back on), and also zero crossing switching. Funny timing on this video. My campus is doing ATS testing today and switching to “island mode” (local power generation) off the grid and making sure everything works.
Bridge rectifiers and associated circuitry are possibly only protected by the 100A main fuse. The control circuits need their own low amp protection device.
I wonder if you could use a UPS to do this exact same setup, depending if the UPS had certain features that you could utilise? I'm thinking something like a self test feature that you can set on a time schedule, and that will allow the battery to drain down almost completely, then switch back to AC power afterwards. So you could make it at a certain time every day, it disconnects AC power and uses it's batteries until their depleted, or until a certain timeout is reached perhaps.
Hi Dave ! futher to the SeanBZA comment below, there is the problem also fixed by the timer, the compressors having a power cut and reapplication with the compressor still under head pressure usually that time is around five or ten minutes otherwise the compressor just keeps cycling till that pressure reduces hard on the overload and the motor, Cheers!!! Rick Armstrong from Canada
I know you are running freezers through this but I would think the delay of it switching over wouldn't even cause your digital clocks to go to "12:00" and blink. (If you were using it as a backup for grid failure). Hey! Try that, Dave! "Plug" a digital clock into it and see if the toggle action is fast enough to keep it going. I would try that before seeing if it crashes a desktop computer (or not)! 😱
Be worth using an ON DELAY circuit on the output, domestic refrigeration equipment usually needs the pressure generated by the compressor to have equalised off cycle for the compressor to start equalised and therefore unloaded, short breaks in supply don’t allow this to happen and the compressor will generally trip on over current and it’s not a good thing to do frequently. What you described as strengthening bars on the first unit are most likely arc chutes to break the arc current during switching
Do your refrigerator/freezers have protection against short cycling? I don't know if the 20-50ms glitch is enough to get the compressor motors to lock up.
@@EEVblog If the battery pack starts to fail or underperform, You will observe shortcycling of the automatic changeover switch when the compressors are called in by refridgerator control system which will kill the refridgerators for sure (got to know about that after watching several videos from @HVACR VIDEOS channel). My suggestion is to use a delay line(delay setting of 5 minutes) between the battery pack and the change over switch for additional protection,
Your optimal cost timer should be set to a random time offset from exact hours, to avoid grid impact from other people running similar clocks. This smoothes out the load discontinuity for the grid backbone system, giving them time to startup power sources.
In theory, maybe. But those refrigerators are intermittent loads anyway. They cycle on/off somewhat randomly throughout the day. If you have a load big enough to seriously impact the grid, the grid operator already knows about you and your staff have discussed such issues with them.
Would love to see Dave explaine the PEN and MEN connections, when using a transfer switch. There's a good article from WorkSafe QLD on the dangers of switching Neutral
Element14 part 2787143 seems to switch pretty fast. I'm thinking of using it for an always-on inverter ready to replace mains 240v in a blackout for PC & display & external hard drive. I'm considering options to include a timer so that when the mains does come back up, it waits a few minutes before switching back over to it.
The problem with AC is if you want zero switching time, you need to phase lock the supply's. Or convert both to DC then invert back to AC the same way a lot of high power UPS do.
Hi Dave. Have you also considered using a power monitor GPO? These monitor the line voltage and automatically cut off power for 5 mins during a brown-out. I could potentially see a feedback loop, if the battery failed under demand (current) but was able to produce enough voltage to trigger the transfer switch. It would just switch back and forth. I do like the transfer switch for redundancy but make sure it has a failure mode lockout. You could also "possibly" use current transformers to siphon off excess solar, instead of a timer.
For fridges I would better look for some that would add delays so the power stays off for at least a minute before switching to the other source. The reason is, the 50ms could be way too long gap for the compressors to actually stall. Moreover when the phase is not correct. Without the delay, the motor starting PTC won't cool down, so it won't energize the phase shift winding, so the motor stay stalled. The fridges would then rely on the overload protection switches to switch off, wait for the PTC to cool down and then restart again, now in a proper way. Although the fridges are supposed to respond that way, you rely really only for the overload switches to operate properly. If these fail short circuit, it will cause the compressors to overheat and significantly increase the risk of fire. Overheated compressor may cause refrigerant leak and because the refrigerant in non CFC fridges uses to be an isobutane, so a pretty flammable gas, then a small spark and you get quite "nice" flame thrower. Only the most modern fridges use timers delaying the start after power restore (the actual delay being designed as random, in order to prevent excessive mains surge when multiple fridges are connected there), but I would not rely on that (some may not have that feature or by the short 50ms gap it may not get triggered reliably while the bad phasing may still stall the compressor)... By the way the "famous" Grenfell tower fire is suspected to be started exactly that way (officially the ignition was determined to be a fridge-freezer fire...)
Nice idea. I guess you got the Battery from the last video for free, so it could be put to good use here. There are combined Solar/Battery inverters out there, that have all this built in. If someone is thinking about implementing such a system, the purpose made inverter is probably the best solution.
@@EEVblog I've got one of those in my setup. I use the Schneider XW+ 6848 NA with Midnite Solar Classic 250 MPPT chargers. Works great. Transfer time is just long enough to reset my server though.
@@EEVblog Actually we have to distinguish 1) grid-tie/off-grid hybrids which run grid-tied during normal operation and can power the load with a mix of solar, battry and utility power at the same time (and even feed excess power into the grid), and 2) off-grid inverters with utility input and built-in transfer switch, which can power the load either from the inverter output or from utility, but not simultaneously.
@@ccoder4953 What maximum milliseconds would you recommend for keeping computers alive? I realise it depends on how much charge the PSU can hold vs its loads, but I'm trying to figure out what a typical PC can last for before going dead.
@@jimmybrad156 Yeah, it varies quite a bit, depending on load and PSU. Some of the very small form factor supplies achieve their size by switching fast and using small output caps. Others have huge output caps. Pretty much any supply should be able to ride though perhaps a few tens of milliseconds since that's about on the order of a line cycle, particularly if it isn't all that close to its maximum load rating. Beyond that, really hard to say. Could be a line cycle or two, could be seconds.
At 18:36 those are not straightening bars, they are electric arc dissipation plates, also the second model that you have is not 100% identical because the secondary contact's are made once the lever is moved with the big micro switch's on the other one with the 4 pols relay when the main power goes of they switch immediately and it doesn't matter the position of the handle. Other than that your video is great p.s. you should rename the video because I have searched all over the TH-cam and I couldn't find an internal look inside these units and you have dismounted it and I have seen what I need. Good luck
Clive, I think you know that all this can be done much easier with some available home automation thingies. They measure the consumption and control without steps the output of your batteries as well as the charging. But well, your oldschool solution might not be too bad.
can you recharge the battery directly from the solar and avoid the cutoff timer? Plug the battery into the grid if the solar doesn't keep up with demand.
I'm guessing the solenoids are so powerful that it can't run them continuously, so the microswitch turns off the solenoid as soon as it has done its job.
My house was prone to brownouts. If the power went out then the generator kicked on, and the switch transferred power: great. But if the powercame back in brownout mode, the generator would turn off, and leave the main solenoid on brownout power (i.e., not enough to switch back). Yup: melted solenoid!
Why not replace the mechanical timer you mention at about 09:50 with a means -- measuring solar panel output -- to activate the switchover to battery when the output of the solar panel rises above, say 10 to 15 percent of the solar panel's maximum output. In the summer time in Australia your solar panels will be supplying power for a loner period of time than during Australian winter.
in your setup, you were talking about using a timer.. why not use a DD (dusk to dawn) switch instead? save yourself some hassle and automate that part as well! (note: there are also dawn to dusk switches as well- and most of them can be made to operate in either direction... piece of cake!)
a side note about DD switches, I happen to come across one that has ZERO electronics whatsoever! it consists of a small magnet, a 10 Kilo-Ohm carbon comp resistor, a CDS (photo) cell, 2 bi-metal strips and 2 contacts- and thats it! and how it works is elegant- yet complete.. it even compensates for temperature differences! it came as a dusk to dawn (night) switch, but, by simply moving the resistor to the other strip, it became a dawn to dusk (day) switch! its a really nifty little gadget!
Seems like a design flaw to me. Shouldn't the tamper switches be arranged so that the solenoid circuit is broken at the end of the throw, not the beginning? Breaking the circuit at the beginning of the throw could result in oscillations if the initial force of the solenoid is not sufficient - perhaps due to increased friction after years of use and accumulation of dirt.
@tim g I watched some videos and I can see the risk if it was mounted outside of his switch board, but I would assume this will be mounted in his switch board. Couldnt both supplies tie to the house neutral bar at the switch board? that way they are both near the MEN / PEN link and keep both supplies close to ground potential.
You want to make sure you don't have neutral connected earth or you could put any fault current back onto the mains. If you just have an earth spike in the ground you will be ok, but for a neutral connected earth you need to switch over the earth to a local earth too.
is there a posibility that one of those switches gets stuck and you have 2 supply connected at the same time to the load... or is there something that prevents that...
With this complexity of your home power and the hardware you plan to install in the future... can you imagine what selling a house like this will be like in 20 years? When 1/3 of the gear has faults and most companies making this stuff aren't around anymore.
dont know if I got lost in the sauce, but I assume the battery its only being charged when solar is working? from the diagram it seems that battery is charged from mains (which it may be) but I assume there is some configuration to make work only on sunlight time, right?
You can use a raspberry pi running a program called sunwait to control a relay. That will turn off the mains based on where the sun is in the sky. Its adjusts due to timezone and season. You set it to turn on/off the relay based on the angle above the horizon the sun is at. I use one for my garden lights, sun is at -5 degrees so below horizon, biggest issue is that the Pi can't tell the time for toffey so needs an internet connection and if it loses it (and it will lose it) the time goes out of whack, have to schedule in a reboot of itself every day to solve it.
You may use a $5 sonoff basic flashed with tasmot for this. It will sync time per NTP and you can set it to switch the internal relay on sunset/sunrise. I use a sonoff POW2 which also measures the power generated by my grid-tie inverter. The micro inverters run on a shelly pm1 which does the same thing but costs $20. Forget the rpi, way too expensive and complicated.
So, this transfer switch is pretty neat but I'm wondering where it will transfer when the voltage is low but not completely lost, and at what voltage it will pull back in. In my experience with similar relays here in the states (meaning the 120 vAC version), an “ice cube” relay like the one in this transfer switch might not drop out until its voltage drops to about 60%’ish of rated and going the other way, the relay will pick up at about 75%’ish of rated voltage increasing. While I’m not sure how or at what voltage your inverters turn on (or off), but when used with Grid as Priority, input voltages can be anywhere and not just on or off as the grid degrades. I’m not saying this is a problem in your application, just pointing out to others to always consider how a device works under all conditions and if this is satisfactory in the intended application.
Instead of all this stuff. You can simply set your fridges to -25C and use a timer to turn them off during the night. That way they are only powered by solar, and you don't need the battery or a transfer switch. You are in essens using the frigedes as batteries.
What happens if the control wires aren't connected. Can it still be used manually as a mechanical switch? It's too late to get my head around the circuit.
Or, place a big can of water in each freezer. That will act like a thermal buffer to keep the temperature low for some hours. Then just turn off the freezers each night. Or, if you want to be really fancy: Add a larger tank of coolant gas to each freezer.
Is that timing OK for the fridge compressors? What would it look like with different sources, the sine wave wouldn't nicely match up from two different sources would it? Any issues cutting the power on/off/on to the fridges frequently? I hear compressors don't like that. But also read that fridges have a timing circuit to leave the compressor off so pressures return to normal before they restart. Any thoughts?
That thought crossed my mind as well. AC compressors are supposed to be left off for two minutes once they stop so the pressure in the lines can bleed off. But even if that's also the case for refrigerator compressors, I think this thing switches quickly enough that the compressor motor would barely slow down at all before the backup kicked in and it got synchronized.
Has anyone looked into what a typical blackout looks like over ~1 second? I'm wondering if there's normally behaviour that would make a 240v relay across it "chatter". Do UPS's normally have a minimum time to run off battery before re-connecting to the mains once it's back up?
Automate the battery disconnect? Why not have something connected to the solar panel input, so that it's controlled by if the solar panels are producing power. Or just get a photo sensor, so that it follows when there is light outside, so it disconnects at dusk and connects at dawn.
Bars on top are arc catch plates, you will have a big arc on breaking under current, and closing on current. Contacts are large to handle the inrush current, which can be 2kA, and the flexible copper braid has to be rated to handle this inrush current. Common parts in all AC contactors and all AC breakers, to quench the arc and connect the moving contact to the terminal.
Ideally you want a separate delay relay in the feed to the incoming supply, so that power has to be applied for 30 seconds before it is reconnected, so that all the transients on the line can die down, and also the massive inrush load can subside from all the other loads being turned on cold.
Yes, some of these transfer switches have 30sec delay built in.
I used to work for a company that manufactured very large backup power units (our largest at the time was 1 MVA). The contacts and bus bars used were massive as were the arc extinguishing plates. Any large switchgear manufacturers will have similar (Siemens, Square-D, Eaton, etc.) devices. I could not find any other test videos of contacts up close but this give the general idea of just how violent it can be for large gear and high currents: th-cam.com/video/DS06Q6oMRYY/w-d-xo.html
@Pharisee Spotter Even 500W load can cause arc over if switched between different potentials without the arc chambers, ask how I know this._
Agree with all you've said. Came here to comment those plates are 'arc chutes' to quench the opening arc.
Now, when you get into entire switchgear being transferred, you even get into 'load-shedding'. Some rotary loads with high moments of inertia have to be stripped before allowing second power source to close in (otherwise the sudden reenergizing of those loads will create massive torque transients). 'Fast transfer', vs. 'slow transfer' versus 'dead transfer'. With BIG loads, it's a lot of careful study.
I agree, you can see the gaps also in the plastic cover on the back right of the DaveCAD segment.
When you get into larger transfer switches they of course get a lot more sophisticated. Automatic delays to ensure the grid is actually stable before switching over, voltage/frequency checks, and feedback to modulate the generator speed to match phases and frequency before switching.
Right. Scada tech here. Grid size and it involves multiple switches, load switching, and feedback from the ECC before it is allowed to click back on.
That would be a concern of mine. Voltage sag with a battery source could produce a rapid on off condition as would an overloaded generator. There should be a delay circuit on the line in if used in this configuration.
Yes, more complex models have a 30sec delay and brown out detection etc.
I was thinking that's really missing on here, when the power here comes back on it's often very dirty and often has to "try" more than once before it's reliable
None of the 2,000A switches I work on ever bother with generator synchronization. They just do a 1s break before make. If the load is critical, you slap a UPS on it.
Let's take a moment to appreciate how strong those solenoids are. Incredible!
Hi Dave,
Imagine this scenario: Your freezers are powered by battery in the night and the battery is empty. Then the ATS switches to mains. However, since the battery is not under load anymore, the voltage increases back again. Then the ATS switches back to battery. Then because of not having enough power, ATS switches back to mains...then loop....
I am sure that your BMS is smart enough to cut entire battery output off (high Z) after the battery depleted and not enable it again until it is charged. But anyway keep this problem in mind ;) Cheers.
LOL as I sit and watch this video in South Africa 🇿🇦, we currently have whats called "load shedding" stage 4. Power off for 2 hours at a time, 3 times a day minimum. In stark contrast to experiencing only a couple of hours in a lifetime in Sydney.🙈 Great video as always! Thanks Dave!
I have a similar setup at home to use battery power when electricity is expensive. I have found that switching while a fridge is running and the voltages are too far out of phase between inverter and grid the fridge kompressors will stop abruptly. I am aware double positive (or negative) ac pulse essentially is DC and acts as a brake on induction motors when using vfd's, I suspect my fridge compressor have a very low rotating mass as well and out of phase power upsets it. However measuring voltages with a multimeter and transfering between sorces when both are near 0 volts betwheen line 1's keeps the fridge happy and it continoues to run. Inverter fridges dont care, but I recommend testing this in your setup.
Cute little baby transfer switch.😁
The data center at the bank I used to work for was more the size of a refrigerator, and had a huge lever, not a tiny knob.
Those microswitches are effectively the negative feedback side of the circuit. Typical sort of a limit switch setup: apply power until you're in the right position and then stop.
18:30 Those metal fins are there to help quench any arc that's formed as the contacts separate, I'd imagine -- same as for the fins you find in protective circuit breakers.
My guess is that the switching time varies depending on where in the sine wave the switch happens. If it's near a zero crossing it will take longer because the solenoids may not pull the rod until closer to peak power. So you would get some ~10ms (half cycle) variation in the timing.
Simplicity is the ultimate sophistication! Even with a 5,000-cycle rating, that thing will probably last forever.
Heads up with refrigeration !
Watch out for locked rotors on switchover .
A off delay of ten minutes is advised .
The quality of the first unit you tore down is much better than that of the second with the green selector switch. The one with the green switch is a "stacked" design used to add more switch sets. I have some of these I used for my generator transfer switch with 4 contacts (I only needed 3) and the outer most switch fails spectacularly when there is a surge. I already lost 2 during power outages. Always the outer most switch assembly. It seems since these are stacked and each one "plugs" into the adjacent switch plastic lever pivot, the strength holding the contacts closed gets weaker the more poles are added due to poor tolerances! The original one you tore down is not stacked and should not suffer from this defect.
I wonder if the plastic in those transfer switches meets the necessary flammability requirements? I have a feeling it doesn't. But I at least like how the slightly crooked "QC PASS 08" stamp in the manual is actually just an image in the pdf :)
The metal sheets at 18:38 look like arc splitters to cool and extinguish any arcs from breaking loads (particularly inductive ones), not sure if I'd trust them at 400 VAC 63 A but probably good enough at 240 V 10 A. However one thing I don't understand is that both units have speed holes on the front plastic cover presumably either to ensure the impending explosion is vented forwards towards the user to cause maximum damage or maybe they had issues with the low quality contacts overheating and their solution was just to add ventilation slots, but seems like it's just begging for contamination issues.
I also assume this change over switch and battery inverter will not be considered a 'fixed installation'?
It's not a "fixed installation", as it's not physically permanenty connected at all.
I'm betting not, I do not see any real separation between Gen and Out, I'd expect that thing to be a fire hazard, and your screwed as insurance wont pay out. I do not see any Aussie Compliance "RCM" markings, just the fakish "C"hina "E"lectrics mark (note the non existence of the spacing between C and E which exists on the real European thing)
yeah, not that I care (because we do it at work all the time), but I'm certain this wouldn't pass AUS electrical code. I doubt DIN rail is certified for use anywhere inside the home. Sadly, as its really useful stuff.. wish my breaker panel was all DIN.
@@Ressy66 Agreed, except for the whole "China Electrics" thing, which isn't real (and wouldn't make sense anyhow). CE markings are self-issued so if it's on there, the manufacturer/importer claims it meets the required standards.
Good informative video. I bought similar from same manufacturer and now I learned how to wire it up thanks. $25 here in Thailand off Lazada. Same project, running guest house off solar and a Bluetti AC200MAX with mains grid as backup.
These are very simple, very crude, but also very effective for small single phase loads. As pointed out by many others, the strengthening bars are intended to break/split/dissipate the arc when switching loads. For small single phase it obviously works, I doubt they would come anywhere near the rating claimed (50kA) or the full load for the no. of operations claimed. As also pointed out by others, professional units will have multiple time delays built in to prevent hunting in the event of voltage fluctuations, and also run-on timers for generators. But they also cost accordingly :-) Great video, thanks.
congratz on 1500 videos(+ or - a few not in the main series) dave! i've found a few pieces of industrial kit like this before, some beefy switches and motor contactors from some pumps, amazingly simple and yet surprisingly complex for what they do
I like this switch, it's so simple, not over-engineered and should be reliable.
Automatic transfer switches make sense 1) when you have two separate utility connections (primary and backup), or 2) you've got a generator that can be remotely started.
I wonder if the relay uses its auxiliary contacts for controlling the generator startup; it would make sense since as the mains dies and the relay turns off, the aux contact gets toggled, starts the generator, and when the generator is up and running and supplying power, the solenoid in the ATS gets energized and the changeover happens.
Fridges won't care, but for any servers, computers etc. get yourself a UPS that can keep them powered through the changeover process.
You got a lot of distortion in your mains. All that SMPS rubbish - forget about clear radio transmissions... It's sad.
U are on like every electronics video comment section LOOL
@@mysterious_czrs well, kinda. Some science stuff too.
It does not look safe or useable for your main household - not to mention code
It indeed seems that the relay contacts on the primary (A) side are intended for a generator start signal - C and NO at the top of the device. The manual for these can be found if you look up Geya Electrical, model W2R
I've seen these used within the water industry along with other sine wave normalizers, delays, soft starts - soft stops, VFD switches and current shunts for very high voltage electric motors.
great video. I love understanding the magic behind the lid.
This was a really great tear down and explanation. I really like the simplicity of the design for this device. Thank you for making this video.
I have Tested with Timer, and it works fantastic. The Timer just have to be connected to the green poles AR and AN or BR and N. And it work with the Time you want. :)
These ATS are not for rotating devices as one cannot ensure that both power supplies are (were) in-phase.
The worse case scenario is opposed phase which would surely stall an rotating device.
...it would destroy the Device😱 it would.....Explode
For some smaller appliance it may get away with it...
Small motors will survive. But really large equipment, you're absolutely right. Our system would check and if drifted out of phase, strip the motors before closing the alternate source feeder.
@@animarkzero it would trip any upstream breakers. The battery inverter would likely cop it too.
The mechanical delay should be sufficient at 180° however the waveform is going to look like the McDonald's M. It will cause massive inrush currents on any connected transformers and the harmonic will take several cycles to dissapate.
Great content. I have learned so much, Bob is now my uncle.
Transfer switches are an extremely important part of backup generators
Hi Dave, very interesting. I wanted to use this but was concerned about safety, so built a servo to a manual transfer switch (video on channel!). But this looks great - and now I can see inside I'm happy to spend the $50. Cheers!
Ah yes, I want to be able to switch grid off programmatically... maybe a smart plug will do.
There are also much more expensive versions that I'm sure are certified to meet various local standards etc.
@@EEVblog Yep. I am doing this to save money not spend it :D
@@AndyMcBlane These are chinesium devices and at the pricepoint I would not take the current rating seriously!
When you pay up to 10x as much for brand ATS at the same current rating(63A) you know something is fishy.
@@animarkzero only switching
Very good useful video I just ordered one of these (Jotta Brand same thing off Ebay) Thank you! cant beat the price of these $35.26. Of course I am sure they are not UL rated, but reviews are all good.
18:27 - these are for absorbing and quenching the arc produced when making and breaking the current.
I'm actually doing the same thing but my continual load is a rack of servers so the changeover needs to be smooth and phase synced. Instead of a timer though and a commercial battery/inverter, I picked up an old 3000VA UPS and modified it so it can't charge the batteries attached to it. I gutted the batteries from it and connected a 14.4KWh LiFePO4 array to it's "extended runtime" input and built a little circuit with a STM32F103C8T6 at it's heart that measures current consumption, battery charge state, among other things. When the charge state is high and solar input is high it disconnects the UPS from grid power making use of it's internal switchover function. The batteries are just connected as per normal to a MPPT via a BMS to charge from solar.
Because I am using a high grade UPS that is designed for 100% duty, it has the added benefit of locking it's AC output phase with the grid phase, and when it takes over the phase is in sync with the grid, and when it goes back to mains power again it takes about 3 seconds while it re-syncs it's output to match the grid phase before switching back over.
Just a thought Dave...
Do you really want to switch your fridges (essentially single phase motors under load) from one source to another almost instantaneously???
Will the solar inverter power source be synchronised to mains frequency? If so, no wukkas
But I'd think you might get into difficult territory if the two sources are not in synch
When I was responsible for a data centre, we were always in synch (courtesy of a UPS on the output side of the transfer switch) for most things, but we still enforced a minute or two gap between power sources (so we didn't damage the HVAC)
Does that circuit have sufficient isolation between the grid and backup sides? The distance between the relay's pads is awfully small. With the right fault condition it seems like there would be a possibility of back-feeding power into the grid through the control circuitry (creating a hazard for linemen etc.).
18:34, those "bars" are not for strenghtening, they are spark arrestors.
Nice Video!!! Some risk could be when your battery inverter (connected to your A input) goes into a "brownout. ---> (i) Battery "empty" - turns off the power - (ii) Relay swaps to B (grid in your case) - Battery has no load at that time (iii) battery and inveter recovers - since there is no load - (iv) relay swaps back to source A ---> we are back at the start.
For your 3 fridges, I have seen someone make a 3 relay box, to prevent all 3 loads starting at the same time. It's a very low chance of this happening, but with thermostats opening & closing, it can happen.
How does that work?
They used an arduino for sequential control of power outlets, to avoid heavy start up loads.
@@slypig24 but the compressor will turn off after a time, so how do they detect when it would turn on again and prevent 2 from turning on at the same time?
@@Philip8888888 it is most likely just for the initial switch on, when all three compressors would kick in at the same instant... I. E. Turn them on in sequence with a 10 second delay or such, and then whatever happens happens. Yes, you could get fancy with power and even temperature monitoring, but more risk of something failing then.
Used to work on megawatt backup gen's - when the xfer switch on those kick over, you better be paying attention and plug your ears! Some sounded like a car crash!
Yeah a lot bigger possibility for arcing, it's gotta be quick and fast.
Not to mention the size of the contacts. Was like a bunch of small frying pans all going BANG at the same time - at about 200mph. Hahaha
How’s there people commented 8 hours ago while the video just been uploaded 20 seconds ago? 😅
Btw, this subject is literally important for any beginner electrical engineer, thanks in advance ❤️
Patreon supporters got to see it last night.
@@EEVblog dont lie to us!
we know you opened a time portal in your yard and have been allowing patreons to play with the course of history!
if this wasnt true, WHY IS MY COPY OF BERNSTEIN BEARS CALLED BEARSTAIN BEARS?????
Time travelers of course.
Now you really don't know what the answer is do you?
Oh and I'm trying to find an IBM 5100 so I can fix the future and my time machine.
@@frogz LOL
@@imark7777777 🙈😂
Happy 1500!
Brilliant tear down and explanation. Thanks
12:12 Check section 7.4 of the instruction sheet at the lower right - it’s on about Amphetamines and other stimulants - what’s that all about???!!!
Don't worry, they're environment as normal working collision avoidance measures.
Thanks Dave! Just about to put one in my caravan to switch between mains (aka “shore”) power, and my inverter..
allows me to use my microwave and coffee machine (and any other 240vac) when travelling
Just make sure you're installing to AS3000 and AS3010. You might also want to use a PE cell instead of your manual switch.
Heat pumps generally can't immediately restart. If there's a power failure, it should stay off for a minute or two before applying power again. They have a soft-start accumulator tank which fills when first turned on. Applying power while that is full may stall the compressor.
I suspect inverter-driven units take care of this electronically, but I'm not certain.
At about 18:45, the two perforated bars may be arc extinguishing means? You see such on higher voltage and power switches or any breakers capable for over 10 kA breaking. The idea is to let the arc extend by its own magnetic effekt and push it to multiple gaps that also cool the arc.
You might consider using one of the easun/powland cheap Chinese inverters with automatic bypass and built in charger when you have direct 24/48v battery connection. 5kW are around $400 and unlike typical ATS when battery is low it switches like ups without circuit disruption to grid power. Your ats is here really very fast! You may also configure safe buffer, to reserve a bit capacity should grid fail. those inverters are really good and configurable as for that price.
The first box internally it looks real cheap-arsed. The way the solenoids are positioned seem a tad wrong. The later switch box looks a bit better... 10/10 for the vid
Was very impressed with the portrait DaveCAD. Indeed rare and it looks like you have plenty of portrait cycles left based on the thickness of the DaveCAD Pad. DCP
Here in europe where power is expensive i do measure actual power into/from the grid and switch extra consumers according to solar surplus. Measuring is done with 3-phase shelly-3em over WiFi, switching load is done with sonoff-basic over WiFi.
Our CBD office has independent connections to two separate grids because it's in a bank building, which is great because it basically meant we didn't have to have a UPS for our data centre. It's practically impossible for both grids to be down at the same time, because that would mean the head offices for most of the major banks and financial institutions in the country would go down.
Well, one day there was a major power outage in the CBD, it took out one of our grid feeds, but when our automatic transfer switch switched over, it decided half way through that it was too much work and exploded, taking out the second power feed and our entire office and data centre.
Everyone was basically left standing around twiddling their thumbs for a couple of hours until they could get an electrician to bypass the transfer switch and reconnect the one good power feed. Unfortunately the disruption probably cost the business about half a million Aussie dollary doos.
But that's not the end of the story. Do you think this convinced them to install a UPS? Nope! Instead they decided that in the event of losing one grid feed, they'd have an enormous generator on the back of a truck, which someone would drive into the CBD and park on the street outside the loading dock. Then they'd roll out a huge cable up the ramp, through the loading dock, and into a panel to feed another transfer switch that they could use to power up the missing feed.
Yep, a flawless plan. Thankfully we haven't yet needed to implement it, since the grids have been stable since then.
Unfortunately the UPS in our head office wasn't as reliable ... but that's a story for another time.
Thats a pretty nifty setup, my grid here in the rural US sucks. During the winter we routinely loose power for week's/month's. Ive been wanting to build a system to automatically switch my homelab server rack (with UPS) to a generator.
Hopefully you can tell your battery inverter to not re-energize until it reaches a certain SoC, otherwise this might chatter as the inverter energizes, then quickly browns out.
Nicer ones have time delays to prevent that (wait 30+ seconds to make sure grid is actually back on), and also zero crossing switching.
Funny timing on this video. My campus is doing ATS testing today and switching to “island mode” (local power generation) off the grid and making sure everything works.
Bridge rectifiers and associated circuitry are possibly only protected by the 100A main fuse. The control circuits need their own low amp protection device.
I wonder if you could use a UPS to do this exact same setup, depending if the UPS had certain features that you could utilise? I'm thinking something like a self test feature that you can set on a time schedule, and that will allow the battery to drain down almost completely, then switch back to AC power afterwards. So you could make it at a certain time every day, it disconnects AC power and uses it's batteries until their depleted, or until a certain timeout is reached perhaps.
Hi Dave ! futher to the SeanBZA comment below, there is the problem also fixed by the timer, the compressors having a power cut and reapplication with the compressor still under head pressure usually that time is around five or ten minutes otherwise the compressor just keeps cycling till that pressure reduces hard on the overload and the motor, Cheers!!! Rick Armstrong from Canada
Big Clive took a similar unit to bits a while back, but it used a motor, much smother but the transfer time is longer
I know you are running freezers through this but I would think the delay of it switching over wouldn't even cause your digital clocks to go to "12:00" and blink. (If you were using it as a backup for grid failure). Hey! Try that, Dave! "Plug" a digital clock into it and see if the toggle action is fast enough to keep it going. I would try that before seeing if it crashes a desktop computer (or not)! 😱
Our fridges don't have clocks.
Be worth using an ON DELAY circuit on the output, domestic refrigeration equipment usually needs the pressure generated by the compressor to have equalised off cycle for the compressor to start equalised and therefore unloaded, short breaks in supply don’t allow this to happen and the compressor will generally trip on over current and it’s not a good thing to do frequently.
What you described as strengthening bars on the first unit are most likely arc chutes to break the arc current during switching
Do your refrigerator/freezers have protection against short cycling? I don't know if the 20-50ms glitch is enough to get the compressor motors to lock up.
No idea, but it's rare this would actually switch.
@@EEVblog If the battery pack starts to fail or underperform, You will observe shortcycling of the automatic changeover switch when the compressors are called in by refridgerator control system which will kill the refridgerators for sure (got to know about that after watching several videos from @HVACR VIDEOS channel). My suggestion is to use a delay line(delay setting of 5 minutes) between the battery pack and the change over switch for additional protection,
Your going to need to think about RCD protection when running off the battery.
Came here for that comment - lots of people seem to ignore rcd protection of downstream devices after genny or battery system
Your optimal cost timer should be set to a random time offset from exact hours, to avoid grid impact from other people running similar clocks. This smoothes out the load discontinuity for the grid backbone system, giving them time to startup power sources.
In theory, maybe. But those refrigerators are intermittent loads anyway. They cycle on/off somewhat randomly throughout the day. If you have a load big enough to seriously impact the grid, the grid operator already knows about you and your staff have discussed such issues with them.
Would love to see Dave explaine the PEN and MEN connections, when using a transfer switch. There's a good article from WorkSafe QLD on the dangers of switching Neutral
The vents on the stationary contacts are for arc flash suppression.
Barring aside the manual switching, would it be feasible to use a DPDT contactor with the coil wired to your priority source to do a job like this?
Absolutely! You do lose the ability of controlling it manually, but there should not be any major harm in doing so.
I think that would work.
Disadvantage of around 20VA constant power consumption, as opposed to the small relay only being around 1VA coil current.
Element14 part 2787143 seems to switch pretty fast. I'm thinking of using it for an always-on inverter ready to replace mains 240v in a blackout for PC & display & external hard drive.
I'm considering options to include a timer so that when the mains does come back up, it waits a few minutes before switching back over to it.
The problem with AC is if you want zero switching time, you need to phase lock the supply's. Or convert both to DC then invert back to AC the same way a lot of high power UPS do.
Hi Dave. Have you also considered using a power monitor GPO? These monitor the line voltage and automatically cut off power for 5 mins during a brown-out. I could potentially see a feedback loop, if the battery failed under demand (current) but was able to produce enough voltage to trigger the transfer switch. It would just switch back and forth. I do like the transfer switch for redundancy but make sure it has a failure mode lockout. You could also "possibly" use current transformers to siphon off excess solar, instead of a timer.
The second switch is a GeChi Electric GCDQCN available on Taobao for $16.50 US including a free screwdriver.
Is the degeneration of the 3000$+ battery and the inverter running every night, or 24/7 actually, (caps?) actually worth the lower energy intake?
lithium titanites seem to've come down in price. claim to do something like 20k+ cycles.
For fridges I would better look for some that would add delays so the power stays off for at least a minute before switching to the other source. The reason is, the 50ms could be way too long gap for the compressors to actually stall. Moreover when the phase is not correct. Without the delay, the motor starting PTC won't cool down, so it won't energize the phase shift winding, so the motor stay stalled. The fridges would then rely on the overload protection switches to switch off, wait for the PTC to cool down and then restart again, now in a proper way. Although the fridges are supposed to respond that way, you rely really only for the overload switches to operate properly. If these fail short circuit, it will cause the compressors to overheat and significantly increase the risk of fire. Overheated compressor may cause refrigerant leak and because the refrigerant in non CFC fridges uses to be an isobutane, so a pretty flammable gas, then a small spark and you get quite "nice" flame thrower. Only the most modern fridges use timers delaying the start after power restore (the actual delay being designed as random, in order to prevent excessive mains surge when multiple fridges are connected there), but I would not rely on that (some may not have that feature or by the short 50ms gap it may not get triggered reliably while the bad phasing may still stall the compressor)...
By the way the "famous" Grenfell tower fire is suspected to be started exactly that way (officially the ignition was determined to be a fridge-freezer fire...)
In my experience anyway, those contacts on the Source B/Generator are used to trigger the generator's start sequence.
Nice idea. I guess you got the Battery from the last video for free, so it could be put to good use here.
There are combined Solar/Battery inverters out there, that have all this built in. If someone is thinking about implementing such a system, the purpose made inverter is probably the best solution.
That's the Hybrid inverter I mentioned. I'm thinking of replacing my sunnyboy with one of those and adding a battery.
@@EEVblog I've got one of those in my setup. I use the Schneider XW+ 6848 NA with Midnite Solar Classic 250 MPPT chargers. Works great. Transfer time is just long enough to reset my server though.
@@EEVblog Actually we have to distinguish
1) grid-tie/off-grid hybrids which run grid-tied during normal operation and can power the load with a mix of solar, battry and utility power at the same time (and even feed excess power into the grid),
and 2) off-grid inverters with utility input and built-in transfer switch, which can power the load either from the inverter output or from utility, but not simultaneously.
@@ccoder4953 What maximum milliseconds would you recommend for keeping computers alive? I realise it depends on how much charge the PSU can hold vs its loads, but I'm trying to figure out what a typical PC can last for before going dead.
@@jimmybrad156 Yeah, it varies quite a bit, depending on load and PSU. Some of the very small form factor supplies achieve their size by switching fast and using small output caps. Others have huge output caps. Pretty much any supply should be able to ride though perhaps a few tens of milliseconds since that's about on the order of a line cycle, particularly if it isn't all that close to its maximum load rating. Beyond that, really hard to say. Could be a line cycle or two, could be seconds.
The “strengthening bars” are probably arc extinguishers
Yes, that seems to be what they do.
At 18:36 those are not straightening bars, they are electric arc dissipation plates, also the second model that you have is not 100% identical because the secondary contact's are made once the lever is moved with the big micro switch's on the other one with the 4 pols relay when the main power goes of they switch immediately and it doesn't matter the position of the handle. Other than that your video is great p.s. you should rename the video because I have searched all over the TH-cam and I couldn't find an internal look inside these units and you have dismounted it and I have seen what I need. Good luck
Clive, I think you know that all this can be done much easier with some available home automation thingies. They measure the consumption and control without steps the output of your batteries as well as the charging. But well, your oldschool solution might not be too bad.
can you recharge the battery directly from the solar and avoid the cutoff timer? Plug the battery into the grid if the solar doesn't keep up with demand.
Yes, the one I have has solar input. But if the solar for that day sucks then you risk not having enough power and your fridges turning off.
@@EEVblog still have the automatic transfer to avoid the risk of your fridges turning off. Is recharging directly from the solar more efficient?
@@danielmcnerthney yep I was thinking the same thing and alternative would be a daylight sensor. Concern the timer might lose time.
I'm guessing the solenoids are so powerful that it can't run them continuously, so the microswitch turns off the solenoid as soon as it has done its job.
My house was prone to brownouts. If the power went out then the generator kicked on, and the switch transferred power: great. But if the powercame back in brownout mode, the generator would turn off, and leave the main solenoid on brownout power (i.e., not enough to switch back). Yup: melted solenoid!
Why not replace the mechanical timer you mention at about 09:50 with a means -- measuring solar panel output -- to activate the switchover to battery when the output of the solar panel rises above, say 10 to 15 percent of the solar panel's maximum output. In the summer time in Australia your solar panels will be supplying power for a loner period of time than during Australian winter.
in your setup, you were talking about using a timer.. why not use a DD (dusk to dawn) switch instead? save yourself some hassle and automate that part as well! (note: there are also dawn to dusk switches as well- and most of them can be made to operate in either direction... piece of cake!)
a side note about DD switches, I happen to come across one that has ZERO electronics whatsoever! it consists of a small magnet, a 10 Kilo-Ohm carbon comp resistor, a CDS (photo) cell, 2 bi-metal strips and 2 contacts- and thats it! and how it works is elegant- yet complete.. it even compensates for temperature differences! it came as a dusk to dawn (night) switch, but, by simply moving the resistor to the other strip, it became a dawn to dusk (day) switch! its a really nifty little gadget!
Seems like a design flaw to me. Shouldn't the tamper switches be arranged so that the solenoid circuit is broken at the end of the throw, not the beginning? Breaking the circuit at the beginning of the throw could result in oscillations if the initial force of the solenoid is not sufficient - perhaps due to increased friction after years of use and accumulation of dirt.
Why switch the neutral? I would have thought only switching hot would be the prefered solution.
@tim g I watched some videos and I can see the risk if it was mounted outside of his switch board, but I would assume this will be mounted in his switch board. Couldnt both supplies tie to the house neutral bar at the switch board? that way they are both near the MEN / PEN link and keep both supplies close to ground potential.
You want to make sure you don't have neutral connected earth or you could put any fault current back onto the mains. If you just have an earth spike in the ground you will be ok, but for a neutral connected earth you need to switch over the earth to a local earth too.
is there a posibility that one of those switches gets stuck and you have 2 supply connected at the same time to the load... or is there something that prevents that...
brilliantly simple ..
my concerned is the lifespan of the coil of relay that continously energized
Same that's always what I think of with permanently on relay coils and I haven't really found an answer.
They are designed for 240V AC power
Because the mains disconnect time switch always goes wrong time ,,can you switch it with a small -day-night solar panel 🤪
I like the idea of the timer but timers tend to lose time. Maybe a daylight sensor? Or maybe you can get DC from the Solar into the DC input?
Technically I can get an output relay board for my Sunnyboy that could do this.
"I'm not going to take it apart further "
One of the first bigclive videos i watched was about a transfer switch like this.
With this complexity of your home power and the hardware you plan to install in the future... can you imagine what selling a house like this will be like in 20 years? When 1/3 of the gear has faults and most companies making this stuff aren't around anymore.
I believe those big "reinforcing bars" are arc arrestors but i might be wrong
dont know if I got lost in the sauce, but I assume the battery its only being charged when solar is working? from the diagram it seems that battery is charged from mains (which it may be) but I assume there is some configuration to make work only on sunlight time, right?
nevermind, I continued washing the video and he immediately explained the timer switch, lol
You can use a raspberry pi running a program called sunwait to control a relay. That will turn off the mains based on where the sun is in the sky. Its adjusts due to timezone and season. You set it to turn on/off the relay based on the angle above the horizon the sun is at. I use one for my garden lights, sun is at -5 degrees so below horizon, biggest issue is that the Pi can't tell the time for toffey so needs an internet connection and if it loses it (and it will lose it) the time goes out of whack, have to schedule in a reboot of itself every day to solve it.
You may use a $5 sonoff basic flashed with tasmot for this. It will sync time per NTP and you can set it to switch the internal relay on sunset/sunrise. I use a sonoff POW2 which also measures the power generated by my grid-tie inverter.
The micro inverters run on a shelly pm1 which does the same thing but costs $20.
Forget the rpi, way too expensive and complicated.
A simple mechanical timer works just fine. Technically I could use an optional auxillary relay output on my Sunnyboy inverter.
just check your earth connection from utiluty and genset. some genset are centre tap depending on origen.
So, this transfer switch is pretty neat but I'm wondering where it will transfer when the voltage is low but not completely lost, and at what voltage it will pull back in. In my experience with similar relays here in the states (meaning the 120 vAC version), an “ice cube” relay like the one in this transfer switch might not drop out until its voltage drops to about 60%’ish of rated and going the other way, the relay will pick up at about 75%’ish of rated voltage increasing. While I’m not sure how or at what voltage your inverters turn on (or off), but when used with Grid as Priority, input voltages can be anywhere and not just on or off as the grid degrades. I’m not saying this is a problem in your application, just pointing out to others to always consider how a device works under all conditions and if this is satisfactory in the intended application.
Instead of all this stuff. You can simply set your fridges to -25C and use a timer to turn them off during the night. That way they are only powered by solar, and you don't need the battery or a transfer switch. You are in essens using the frigedes as batteries.
I dare say not all units have a set point that low?
Huh. The last Automatic Transfer Switches I saw somewhere (Bigclive, maybe?) were motorized jobbies - took several seconds to switch between sources.
What is the power consumption of this thing? I'm guessing no less than 5-8W?
What happens if the control wires aren't connected. Can it still be used manually as a mechanical switch? It's too late to get my head around the circuit.
Yes
Or, place a big can of water in each freezer. That will act like a thermal buffer to keep the temperature low for some hours. Then just turn off the freezers each night.
Or, if you want to be really fancy: Add a larger tank of coolant gas to each freezer.
Is that timing OK for the fridge compressors? What would it look like with different sources, the sine wave wouldn't nicely match up from two different sources would it? Any issues cutting the power on/off/on to the fridges frequently? I hear compressors don't like that. But also read that fridges have a timing circuit to leave the compressor off so pressures return to normal before they restart. Any thoughts?
Don't know, but this would very rarely activate, only during a power outage which is incredibly rare here.
That thought crossed my mind as well. AC compressors are supposed to be left off for two minutes once they stop so the pressure in the lines can bleed off. But even if that's also the case for refrigerator compressors, I think this thing switches quickly enough that the compressor motor would barely slow down at all before the backup kicked in and it got synchronized.
@@chitlitlah Yes but the inertia of the compressor might be against the new AC's inertia eg. too far out of phase.
Has anyone looked into what a typical blackout looks like over ~1 second? I'm wondering if there's normally behaviour that would make a 240v relay across it "chatter".
Do UPS's normally have a minimum time to run off battery before re-connecting to the mains once it's back up?
Automate the battery disconnect? Why not have something connected to the solar panel input, so that it's controlled by if the solar panels are producing power. Or just get a photo sensor, so that it follows when there is light outside, so it disconnects at dusk and connects at dawn.
SCRs and IGBTs are awesome