Current draw from a Fridge compressor circuit

[NorthernGuy]

Please forgive any newbie errors in this post, my electronics days are long ago! I just wanted to see if my assumptions are correct regarding a fridge current draw.

The fridge operates from 240v AC, 50Hz. The brushless compressor motor 'run' winding has a cold resistance of 18ohms and is wired in series with a 4uF run capacitor. At cold startup, a second 'start' compressor winding (23 ohms cold) is briefly energized in parallel with the run winding, via a PTC (thermistor) which has 13 ohms cold resistance. The PTC is designed to quickly heat up with current draw, until when hot it is a few K ohms, so effectively ceases powering the start winding. I've no idea what the reactance is of the two windings. The compressor is nominally about 125w from tech info.

Looking at the 4uF Capacitor, the reactance is 1/(2 Pi f C) = 796 ohms. So if the compressor motor side was, at a worst case scenario, completely shorted out, the current draw would be limited by the capacitor at from this formula 0.30amps or 72 watts. Any lesser compressor failure would I assume give more impedance so less current/power draw.

Are these assumptions correct? I'm trying to get my head around the (rare) fire risks associated with fridges/freezers. Are my sums correct, that a run capacitor in series with the motor, which not all machines have, greatly limits the power that can be drawn, or am I missing something?

 

[Diver300]

I am a bit sceptical that a capacitor as small as 4 µF is in series with the run winding. I would have expected that the run winding would be directly connected to the mains, and the start winding has the capacitor, but that may be left running anyhow.

Does the PTC get hot enough to make a difference in normal running? I suspect not.

Fridges need surprisingly little power and the motor doesn't need much start torque, as the gas pressures will be equalised when it starts, and so that is why they can use a small start capacitor and there's little point in turning off the start winding once the motor is running.

If the motor is stopped and then restarted, there can be problem. The gas pressures haven't equalised and the motor can fail to start.

That only has to be handled a few times in the life of the fridge because the thermostats have so much hysteresis that there is plenty of time for the pressures to equalise in each off cycle. Electronically controlled fridges will have timers.

However those few times need to be survived. Often the motor will overheat, some thermal protection will prevent damage, and by the time that the motor has cooled down and the thermal protection has cut back in, the pressures have equalised.

I think what you have there is a motor with as start winding which has close to 240 V across it when running, and the PTC is only to prevent damage if the motor fails to start due to back pressure from turning on a fridge that has just been stopped.

 

[rjenkinsgb]

The PTC replaces the start relay used in older systems. I believe that is correct, but the cap is in series with the start winding and the run winding is directly powered.

Reference:

 

[NorthernGuy]

Thanks for both your replies. rjenkinsgb - the diagrams you post show start capacitors, I have the less common run capacitor only system. Online parts for this model confirm it as a run capacitor.

Diver300 - I will post the diagram when I re-find it of the series capacitor I described, but which you were a bit sceptical of. They really exist. When measuring the resitance at the plug between Live & Neutral, with the fridge in a warm condition so thermostat calling for cooling, I get the ever climbing resistance of a capacitor charging up. If the mains was straight across the run winding as more normal, then regardless of any other wiring perhaps in parallel with it, I would see the low stable resistance of a run winding, which I don't.

The PTC disk is wired across the start & run terminals in a plastic housing (also including an overload) that clips straight onto the three compressor pins. I wasn't sure what you meant when you questioned the PTC getting hot enough to make a difference in normal running? If you mean safety wise, as I raised this concern, they can mildly char the plastic casing they are nowadays housed in, that design choice being baffling thing to me. A few mins of running left the plastic housing around the PTC warm to touch when I felt it.

I guess what you are saying is that if there is no capacitor in series with the windings, the stalled current can lead to overheating, and the overload clicking out, which I've heard on other fridges with faults. What would the heat draw be then? 240v across 13 ohms run winding would be a lot of heat by my calculations, but I don't know the reactance of that winding. Does a stalled winding have a different impedance to a spinning one?

 

[Diver300]

Does a stalled winding have a different impedance to a spinning one?

I don't know if the reactance is difference, but when the motor spins there is a back-voltage that opposes the supply voltage and massively reduces the current.

From what you've said, your fridge is connected in a very different way to ones that I have seen.

 

[schmitt trigger]

Fridges and A/C units have almost always an attached label with a schematic diagram. Could you locate it and take a clear focused picture?

Lastly, if you would like to know the starting current, your best bet is to borrow a clamp ammeter with a peak hold function. I believe that nowadays all of them have the function.
The starting current is a function of not only the series impedance, but in the case of an electrical motor, the airgap reluctance and the rotor’s resistance, both of which you cannot measure from the outside.
Those are carefully balanced to provide the required start/ run torque characteristics, and its resultant locked rotor and full load running amps.

 

[NorthernGuy]

I don't know if the reactance is difference, but when the motor spins there is a back-voltage that opposes the supply voltage and massively reduces the current.

Well that is an interesting explanation I can live with. I've been trying to understand how such low resistance windings can result in such low power consumption, but that fits, and so too the idea that at rest the rotor/windings suddenly heat up a lot, whilst their impedance likely doesn't change.

Fridges and A/C units have almost always an attached label with a schematic diagram. Could you locate it and take a clear focused picture?

I'm afraid there is no such diagram there. I feel certain that the capacitor must be in series with the winding(s) as stated before; I'd get a low, stable winding resistance between neutral & earth pins otherwise. If this is correct, does it represent a safer way to power a fridge motor? You could never get such a dangerously high stall current with a capacitor offering the 796 ohms impedance I calculated.

 

[rjenkinsgb]

This in principle is the only wiring variation I can find that uses only a run cap and thermistor:

The winding resistances would presumably depend on the motor power rating.

 

[NorthernGuy]

Sorry for slow reply, I tried but failed to find the alternative circuit diagram I thought I'd found. I accept what is drawn by rjenkinsgb is the very normal way to wire a run capacitor motor circuit; unfortunately, my ohmmeter did not show anything like the low resistance of a run winding, further lowered by parallel start winding, as illustrated in the diagram.

Putting that point aside for one moment, I recently repaired this fridge after it failed, with the PTC disk part crumbled away. Initially the remains of the old PTC in its housing gave 8 ohms, later 200-300 ohms, intermittent and slightly burnt up one side. It came in a plastic moulding (incorporating overload) which I eventually sourced the very image of from eBay/China; I'm sure from the same factory. They couldn't provide any tech info, and neither could the UK manufacturer of my fridge/freezer (too old). I don't like unknowns like that in a repair, but that's where I was.

The new PTC unit on measurement had a cold resistance of 13 ohms. Once fitted, the fridge started every time without hiccup, and ran until fridge & freezer hit correct temperatures. As helpfully suggested by schmitt trigger, I used a clamp-on ammeter to monitor both run and start peak current. It runs drawing around 0.46 amps; the peak of several starts is just under 6 amps. The run current fits nicely with the compressor's data sheet (109 W at -25 / +55 °). The compressor cuts in/out periodically, I assume on the thermostat. All seems normal.

But coming back to the PTC and that start-up peak of 6 amps, would any knowledgeable person have an opinion on this? If the new PTC is wrong/higher resistance (start/run windings are 23/18 ohms if helpful), I'm trying to work out if this is OK, or if the start peak current is too low, maybe towards a stall situation which I obviously don't want.

 

[Diver300]

6 A peak doesn't seem too much. It is about 13 times the running current, which is well in the normal range for a motor that has no electronic control.

I had a freezer that took something like 0.7 A when running, and I fitted a 3 A fuse in the plug, which lasted about 2 weeks. I assume that it took way more than 3 A when starting, and weakened the fuse each time. I changed to a 5 A fuse and that never blew.

In my experience, fridges start really quickly, probably in less than 1/4 of a second. There is no "winding up" like you get with high-inertial loads like fans, so there will be minimal heating during starting.

A mains powered fridge is designed to be run from domestic mains supplies, so plugged into the same socket that could run a fan heater at 10 A or more continually, so there is no advantage in reducing the starting current. Anything that makes the start current less is likely to reduce the starting torque, and on a fridge that risks the compressor stalling at startup if the refrigerant pressures haven't had a chance to equalise since the last time it ran.

The highish start current could be a problem if you ran the fridge from a low-power source of 240 Vac, such as a small generator or an inverter, but the manufacturers of the fridge aren't going to care about that.

It seems to be that the PTC heats up very quickly each time the fridge starts. There can be an issue with self-heating thermistors that the hotter parts have higher resistance and generate more heat than the colder parts, leading to a thermal runaway until the heat has distributed to the whole of the thermistor. That could be why the PTC wore out on one side first. A similar thing happens in incandescent bulbs, which is why slow starting, giving time for the whole filament to get to the same temperature, can increase the life a lot.

 

[NorthernGuy]

Diver300 – that’s very useful, thanks. I noticed it comes with a 13A fuse in the plug. I had wondered about dropping it to 5A, on the basis that the starting pulse is too fast to blow a wired fuse, but wasn’t quire brave enough.

Yes the fridge just starts suddenly, this one has always clunked when it does. I read elsewhere that the PTC heats in hundreds of milliseconds, so effectively gives that short sharp pulse to the starter circuit before heating to K ohms and effectively contributing nothing, it just sits there consuming about 2w, apparently.

Hadn’t thought about the PTC being inhomogenous in construction, but thinking about it, anything that heats up that fast will be under a lot of thermal shock, which as you say may not be evenly distributed as there isn’t time for the heat to distribute. I knew a slow start-up greatly extended a filament bulb’s life, but had never heard that explanation, but it does make sense.

 

[Nigel Goodwin]

Diver300 – that’s very useful, thanks. I noticed it comes with a 13A fuse in the plug. I had wondered about dropping it to 5A, on the basis that the starting pulse is too fast to blow a wired fuse, but wasn’t quire brave enough.

Yes the fridge just starts suddenly, this one has always clunked when it does. I read elsewhere that the PTC heats in hundreds of milliseconds, so effectively gives that short sharp pulse to the starter circuit before heating to K ohms and effectively contributing nothing, it just sits there consuming about 2w, apparently.

Hadn’t thought about the PTC being inhomogenous in construction, but thinking about it, anything that heats up that fast will be under a lot of thermal shock, which as you say may not be evenly distributed as there isn’t time for the heat to distribute. I knew a slow start-up greatly extended a filament bulb’s life, but had never heard that explanation, but it does make sense.

Plug fuses are fast blow, it needs to be 13A, a 5A fuse will intermittently blow (due to the high surges), disastrous for a fridge. For the same reason CRT TV's needed 13A fuses in the plug.

 

[Diver300]

I knew a slow start-up greatly extended a filament bulb’s life, but had never heard that explanation, but it does make sense.

A filament bulb will take much more current when it turns on if there is nothing to limit the current. A slow start means that the current will basically never exceeds the rated current by much, which extends the life, and is fine as long as the application doesn't need quick switching.

It's a bit like the fuse ratings we have been discussing. Keeping the peak current to not much more than the fuse / filament's rating means you don't get premature failures.

 

[NorthernGuy]

Plug fuses are fast blow, it needs to be 13A, a 5A fuse will intermittently blow (due to the high surges), disastrous for a fridge. For the same reason CRT TV's needed 13A fuses in the plug.

Well I must say that is the opposite of what someone once told me, i.e. that an overloaded plug fuse took an age to blow unless the overload was massive like a short; I thought fast blow fuses were quite special? Please have a look at this link from a PAT testing website and what it says below in particular:

BS 1362 Fuse Operation Characteristics

It may not be immediately apparent, but a 13A rated fuse is not designed to actually blow at 13A. In fact, a 13A fuse will allow a current of 20A to pass indefinitely without blowing. If we look at the graph in Fig 1, it shows the operating characteristics for both 3A and 13A BS 1362 fuses. The grey shaded area is the point where the fuse is designed to operate. So for example, a 13A fuse will blow within 0.01 - 0.3 seconds for a fault current of 100A; shown in red on the graph. For a current of 20A, shown in blue on the graph, a 13A fuse will not blow!

If this is applied to a 5A BS1362 fuse, it should never blow with a 6A pulse at start. What do you all think?

 

[Nigel Goodwin]

Well I must say that is the opposite of what someone once told me, i.e. that an overloaded plug fuse took an age to blow unless the overload was massive like a short; I thought fast blow fuses were quite special? Please have a look at this link from a PAT testing website and what it says below in particular:

BS 1362 Fuse Operation Characteristics

It may not be immediately apparent, but a 13A rated fuse is not designed to actually blow at 13A. In fact, a 13A fuse will allow a current of 20A to pass indefinitely without blowing. If we look at the graph in Fig 1, it shows the operating characteristics for both 3A and 13A BS 1362 fuses. The grey shaded area is the point where the fuse is designed to operate. So for example, a 13A fuse will blow within 0.01 - 0.3 seconds for a fault current of 100A; shown in red on the graph. For a current of 20A, shown in blue on the graph, a 13A fuse will not blow!

If this is applied to a 5A BS1362 fuse, it should never blow with a 6A pulse at start. What do you all think?

You seem rather confused about fuse specifications? - a fast blow fuse still requires a considerable overload to blow - where as a slow blow fuse also requires a longer time of the considerable overload.

Fridge manufacturers fit 13A fuses in the plug because that's what it needs - otherwise the large surges (and 6A is only a low estimate) will cause the fuse to blow intermittently.

Why would you want to change something the manufacturer does, and recommends?.

On one make and model of CRT TV the manufacturer stupidly fitted a 5A fuse in the plug (presumably designed by a new university graduate), needless to say fuses in them blew in huge numbers during the 12 month warranty, and had to be replaced with 13A ones.

Still your choice, but make sure to keep a good stock of replacement fuses, and keep a very close eye on the fridge to make it's not blown, I would suggest checking every night when you go to bed, and again in the morning when you get up. You 'may' be lucky, it may never blow - but why risk it for no reason?.

 

[NorthernGuy]

You seem rather confused about fuse specifications? - a fast blow fuse still requires a considerable overload to blow - where as a slow blow fuse also requires a longer time of the considerable overload.

Fridge manufacturers fit 13A fuses in the plug because that's what it needs - otherwise the large surges (and 6A is only a low estimate) will cause the fuse to blow intermittently.

Why would you want to change something the manufacturer does, and recommends?.

On one make and model of CRT TV the manufacturer stupidly fitted a 5A fuse in the plug (presumably designed by a new university graduate), needless to say fuses in them blew in huge numbers during the 12 month warranty, and had to be replaced with 13A ones.

Still your choice, but make sure to keep a good stock of replacement fuses, and keep a very close eye on the fridge to make it's not blown, I would suggest checking every night when you go to bed, and again in the morning when you get up. You 'may' be lucky, it may never blow - but why risk it for no reason?.

No I'm not confused at all. You did have me doubting myself for a second, saying plug fuses are fast blow. But the link I posted shows my memory served me well. They're hardly fast if BS1362 allows 20A to pass indefinitely through a 13A fuse. I have come across fast blows in scientific equipment I use, but they are a rarity.

The reason someone might deviate from a manufacturers spec and fit a 5A fuse, in this case are twofold:

Firstly, because they know that manufacturers have standardised on either 3A or 13A in the appliances they sell. Harder for the public to get wrong, but running with a 13A fuse 'if' a 5A was fine is doing no one any favours. They are safety devices, after all. In a simple short with huge current, probably any of the three ratings here would disconnect in a flash. But what about more gradual, insidious overloads? My sister used to run two fan heaters from one extension, maybe (gulp) 5 or 6kW total, and wondered why the plug fuse only blew after hours.

Secondly, because in the case of a fridge/freezer, the normal running power might compare with an incandescent lightbulb, 120w for me, but the stalled motor dissipation, with no back EMF, might rival a portable heater, kilowatts if mains is across a winding. In a motor sitting in oil. The bimetallic overload is the only device that 'should' disconnect this dangerous situation - it doesn't look like a 13A fuse will do much in a hurry.

So if the fridge can safely run on a 5A fuse, because (pro rata from 13A spec) its indefinite current is 7.7Amps, and the split second start-up is a pulse not over 6A, that gives a lot better chance that the plug fuse will pop before we have a fire. Motor stalls combined with Overload failure are probably rare, but I did mention safety in my original post on this subject. The Grenfell fire was caused by a fridge/freezer catching fire.

Perhaps Diver300 could add some factual experience by letting us know how many 5A fuses he's blown, and over what period, in his machine drawing 0.7A?

 

[Nigel Goodwin]

Like I said, your choice, but you obviously don't understand what are fast blow and slow blow fuses - plug fuses are fast blow, just a piece of wire.

 

[Diver300]

Perhaps Diver300 could add some factual experience by letting us know how many 5A fuses he's blown, and over what period, in his machine drawing 0.7A?

It was a long time ago and I can't remember how long the freezer was in use for, but I never had another fuse blow.

Of course, the fact that the 3A fuse only lasted a couple of weeks shows that there were considerable current surges in normal use.

There will be some piece to piece variation in the fuses, and in the installations. A poor contact will lead to more heating so the fuse will blow at a lower current, and other things will alter the temperature as well.

Just because I got away with a 5 A fuse in my freezer for the whole time I had it, doesn't mean that all will do the same.

Nigel's considerable experience with TVs shows that there is variation. He saw lots of CRT TVs blow 5 A fuses, but the fact that they blew in the 12 month warranty period implies that they lasted at least a few days, and probably more, and if even 10% of TV fuses failed, that would be a huge cost to repair.

I would be interested to know how big the surge current was with the CRT TVs and how long it lasted.

However, given that a 5 A fuse is not supposed to blow at 7 A for 20 minutes or so, if the 6 A surge has been measured reasonably accurately, and lasts for less than 1/4 of a second, a 5 A fuse seems ample.

Having a lower fuse rating can help reduce damage if there is a fault, and it can prevent upstream circuit protection from operating, but it also increases the chances of blowing when it shouldn't.

The increased current that appliances take at turn-on can have all sorts of different causes and therefore characteristics, so the considerations for each type are different.

Incandescent bulbs take much more current when warming up. The resistance of the supply doesn't make so much difference on 230 V circuits, but at lower voltages the more supply resistance there is, the less the peak current and the more time it takes for the lamp to illuminate.

Motors take much more current when starting, and the motor type and load inertia alter the maximum current and the time. The current waveform of some motors is far from being a sine wave. Induction motors often have a very poor power factor at start up.

Transformers often saturate at turn-on and there can be up to 1/4 of a mains cycle where they take a lot of current and the primary winding resistance is all that is limiting the current. The point in the cycle where the transformer is turned on makes a lot of difference to the peak current.

Any circuit that charges capacitors will take a lot of current at start-up, but the peak current and the duration depend on loads of different things.

 

[rjenkinsgb]

Just to conform what others are saying:

A normal plain-wire type fuse is an "F" or "GF" rated one; fast blow. "T" or GL / GG is time-delay aka slow-blow, which come in various types and speeds; there are also Motor rated fuses, which have extra high surge ratings.

The fuse rating is generally the maximum continuous current that will NOT cause it to fail. In quite a few types, it takes around 3x rated current to cause a failure in one second.

I have come across fast blows in scientific equipment I use, but they are a rarity.

I believe you are getting confused (ouch, bad pun..) between common "F" fast and semiconductor rated fuses; "FF", AR or UR types etc.

Those are frequently used in power semiconductor systems such as industrial motor drives, and some multimeters etc.

(Those ultra-fast types typically have elements that alternate between short, narrow fusing points and dissipating sections several times wider).

None of that is counting the variations in rated voltage, breaking current rating [supply source current capability], HRC or indicator types etc.

An example of variations - this one is only 40A. You can get 40A fuses with very small bodies.

However, this one is semiconductor rated ultra-rapid and can break a fault current up to 200,000A (eg. if something caused a dead short across the supply).

You can get a basic industrial 40A 14x51mm cartridge fuse for around £1.00

These are around £50 each.

 

[NorthernGuy]

It was a long time ago and I can't remember how long the freezer was in use for, but I never had another fuse blow.

Of course, the fact that the 3A fuse only lasted a couple of weeks shows that there were considerable current surges in normal use.

There will be some piece to piece variation in the fuses, and in the installations. A poor contact will lead to more heating so the fuse will blow at a lower current, and other things will alter the temperature as well.

Just because I got away with a 5 A fuse in my freezer for the whole time I had it, doesn't mean that all will do the same.

Nigel's considerable experience with TVs shows that there is variation. He saw lots of CRT TVs blow 5 A fuses, but the fact that they blew in the 12 month warranty period implies that they lasted at least a few days, and probably more, and if even 10% of TV fuses failed, that would be a huge cost to repair.

I would be interested to know how big the surge current was with the CRT TVs and how long it lasted.

However, given that a 5 A fuse is not supposed to blow at 7 A for 20 minutes or so, if the 6 A surge has been measured reasonably accurately, and lasts for less than 1/4 of a second, a 5 A fuse seems ample.

Having a lower fuse rating can help reduce damage if there is a fault, and it can prevent upstream circuit protection from operating, but it also increases the chances of blowing when it shouldn't.

The increased current that appliances take at turn-on can have all sorts of different causes and therefore characteristics, so the considerations for each type are different.

Incandescent bulbs take much more current when warming up. The resistance of the supply doesn't make so much difference on 230 V circuits, but at lower voltages the more supply resistance there is, the less the peak current and the more time it takes for the lamp to illuminate.

Motors take much more current when starting, and the motor type and load inertia alter the maximum current and the time. The current waveform of some motors is far from being a sine wave. Induction motors often have a very poor power factor at start up.

Transformers often saturate at turn-on and there can be up to 1/4 of a mains cycle where they take a lot of current and the primary winding resistance is all that is limiting the current. The point in the cycle where the transformer is turned on makes a lot of difference to the peak current.

Any circuit that charges capacitors will take a lot of current at start-up, but the peak current and the duration depend on loads of different things.

You raise some valid points. And we are mixing a lot of different technologies here too. Fridges from an era before electronic controls have a pretty ancient, motor with PTC/Relay startup system. CRT TVs were, correct me if wrong, mostly transformer driven, well the ones I knew were. But transformers have all but vanished it seems now, in favour of switch mode supplies, well known for heavy start-up transients.

My Ctek car battery charger probably can only deliver 75w output at full pelt, but is fused at 13A. Presumably a 3A fuse would lead to in-warranty fails, so financially ruinous to the manufacturer. I had the 12v 2A adaptor to an external hard drive fail once at switch-on. This didn't touch the 13A fuse in the socket adaptor, but it instantly took out the 32A ring main MCB (not a residual current device). The response of metal wire fuse clearly being very slow compared to even a much a higher rated MCB.

I suppose a fuse blows by melting metal, and as alarming as a transient may be on your peak hold clamp meter, you have to factor in its duration, less visible, to work out if the heating affect will melt metal. My guess is that it rarely does; melting metal isn't fast.

I do think the demise of the 5A fuse as a manufacturer fitted item was a bad turn for the consumer. A corded power drill (remember those?) drawing several hundred watts must now be fused at 3kW, or in reality far more as the BS1362 standard says a 1.9x overload (say 6kW) should blow it 'within 30 mins'. If something hasn't caught fire within than power/time window, it likely never will. These plug fuses seem little more than short-circuit protection in safety terms.

The reason for my original post was with safety in mind, fixing up an old fridge and pausing to ponder about that device in almost every house, buzzing away 24/7, using very little power but with that little risk we all overlook. All safety seems to hang on the bimetal overload, the only thing that should disconnect upon a high stall current if the plug fuse won't. And it must repeatedly disconnect every time it cools to return on again, perhaps unnoticed for days in a garage etc. Yes it is my risk to fit a 5A, but with nothing more than beer in this one, I think I'll see if I am as lucky as Diver300 has been.