I’m seeking for help to understand the behavior of this pump at certain compressor speeds because I can’t quite understand what’s going on. First I want to say I’m using FRC to regulate max. compressor power. Until know there were mild outside conditions and I was able to run at 50% and 60% FRC limit - this means at 50% the max. compressor frequency is 22Hz and at 60% 27Hz. All was fine and the pump run well.
But it got colder and I had to switch to 70% FRC which limits the compressor speed to 31Hz. And here goes the strange regulation issue. When compressor changes frequency to 31Hz you can see the EEV starts slowly decreasing and then it jumps down pretty much and then open back for a while and again down and all over again. Power input is not stable as you can see. I don’t understand why there are such big jumps in EEV at 31Hz and it’s not stable ? What the pump is doing ? I can’t get this… Yesterday it happened that after 2h of this erratic behavior the pump somehow locked at 30Hz and it didn’t react at FRC changes at all. I’ve commanded the pump to switch to 50% FRC (22Hz) and nothing, it was still locked at 30Hz.
@SarahH do you maybe have an idea what’s that ? Thx
Sorry @Michal_S I have not seen EEV behaviour like that. If compressor speed is still a constant 30Hz then it sounds like the controller software may have got itself into an unexpected place. Personally, beyond maybe switching the MIM off and on again, I would 1) pose the question to UK Helpdesk, then 2) call an engineer in. There could be a problem like liquid refrigerant breakthrough into the suction, or a problem with the seal oil circuit, or low refrigerant inventory, or a jammed vapour injection EEV or SOV etc. - afaiac these all need specialist diagnosis and fixing, and I wouldn’t want to make things worse by fiddling…
PS I couldn’t resolve the x-axis titles in your time trends even with a magnifying glass - perhaps you could enlarge?
Hi Sarah, thanks for reply. This 30Hz I saw only once yesterday. Somehow it looks like it is some safety mode but at 30Hz didn’t help at all. I’ve stopped the pump and start again and it fixed itself. Today it is running at 31Hz because it’s under -4C and it is limited by FRC at 70%.
But what is strange is this EEV behavior at this frequency. The pump is closing EEV smoothly but pretty aggressively as you can see the power input increase is quite steep and I guess superheat is decreasing dramatically so it must close the EEV radically , otherwise fluid could get into compressor. But I simply understand why this is doing it at 31Hz. At 22Hz and 27Hz no issue whatsoever, the only side effect is that -4C outside the water is max 35C@22Hz or 40C@27H but everything stable. Once I allow compressor go to 4Hz more to 31Hz it starts doing this craziness.
Oh, I hope there is no hardware issue at higher compressor speeds but it looks like indeed. But with DHW heating I haven’t noticed this I will forward this topic to Nick, maybe he will have a clue what could be the reason fort this…
I can see that EEV closure is preceding the power draw increase, presumably because compressor discharge pressure - thus pressure ratio - will increase as a result. Can you give us a plot showing compressor suction/discharge pressures along with EVI data (EVI EEV/SOV/bypass positions)? I’ve noticed that the EVI circuit is engaged much more frequently in cold ambient conditions, and, since you first noticed your problem when it got colder, I wonder whether the EVI circuit is the culprit (e.g. letting refrigerant liquid into the compressor interstage)?
Hi SaraH, sure I will provide it to you but I guess this is a regulation bug in Samsung firmware. Look at this… I’ve increased compressor limit to 36Hz and here we go - ALL fine, no issues with EVV… I guess there is wrong map and 31Hz band is unstable. What do you think about this ? You can see once it increased the speed to 36Hz EEV become stable and even opened a bit. In 31Hz EEV immediately started closing until it reached some threshold and pump panicked and closed it rapidly, then opened it and again drop. Seems like 22Hz, 27Hz, 36Hz and probably higher are stable but for some reason 31Hz is problematic.
Good work Michal. I wonder if 31Hz causes problems on non-FRC controlled HTQs? I’ll go back over my old logs but I don’t recall any. (Can’t check today - my HP packed up overnight and I’m waiting in the freezing cold for my installer to come round… ).
Oh, sorry to hear that, what happened ? Hope they will fix it fast for you, don’t have a good feeling from this Samsung.
This is my next test, but when I disable FRC it will ramp up compressor like crazy but I will observe the behavior for sure. The next few days are good freezing temps
Hmm I’m no control engineer but I’m thinking that if Samsung have coded two separate regressions into their Comp Speed/EEV opening algorithm - one for 31Hz and below, and another for 31Hz and above - and they are a bit discontinuous at 31Hz, the software may be leaping from one regression to the other .
I’ve just noticed something interesting…. This Bypass. Up to 27Hz is completely open… At 31Hz it’s open/closed/open/closed and 36Hz bypass is closed ? Like the pump can’t decide at 31Hz… Shall I open or not ? Very interesting. Probably this is the issue only with FRC limit at 31Hz… But I guess it’s a bug or ?
Our posts crossed but it looks like we read each other’s minds…
I can’t reset the trip in my MIM, so there must be a short-circuit somewhere. I’ve checked the circulating pump (easy to unlug to test), but can’t easily check the secondary pump or 3-way valve. Unfortunately my installer is very busy today, so to keep warm I guess I’ll just have to go to the pub…
I agree. It’s really a bypass causing those drops but question is why Samsung programmed this like this especially at 70% FRC. Seems I can’t use this frequency for my regulation between -4C and -8C outside because it simply unstable. I think it’s not good for the pump to constantly switch between two circuits. I guess this is why it went into “limp” mode after few hours yesterday at lock to 30Hz. But funny thing is even at 30Hz it’s not stable and bypass switch on and off
So lesson is to not use 70% FRC limit with this Samsung heat pump, otherwise it cause stability issues.
I guess I simply use FRC only up to -3C outside where I limit it to 60% FRC (27Hz) and where there is colder I will disable FRC and use weather curve and simply accept the fact the pump will hunt once target temp is reached. I have old firmware so the hysteresis can’t be set and I saw it starts to panic often 1.5C below target. Well, this is life with Samsung heat pump
Do you think it makes sense to report this to Samsung or it is waste of time ?
That sounds bad, hope you will have it fixed soon, I don’t know what’s the weather in UK right know but here it’s pretty cold this week -5C to -10C so without heating it would be bad. Hope you have at least a fireplace or something.
If reporting your observations to Helpdesk wouldn’t take much time, then I’d do so. They may be unaware of this particular quirk, and it may help them with other folks’ problems .
I don’t know why this is controlled so stupidly…. there is no PID regulation or what ??? Do you also have such graphs or it’s just me ? This happens always when water outlet temperature approaching target. Compressor starts to oscillate like this creating this ugly zigzag pattern… You can see at 15:40 I’ve disabled FRC and immediately it starts behaves like this. I’d rather use FRC to be honest and set target temp limit at 60C. It’s would be better than to see this but that regulation issue at 31Hz is putting me off. But when I look at this I can see it enables bypass anyway so maybe it does not matter ??? The only difference is that EEV is no dropping that dramatically.
(Ignore the defrost at 07:35) I can see that the controller is converging LWT asymptotically towards target (so there’s 2- or 3-term control in there somewhere), but look at that compressor inverter frequency (magenta line) - bouncing around quite a lot, just like yours whan not on FRC.
The only interesting thing I noticed from this is the effect of LWT on CoP over the period 07:00 - 07:30 when the compressor speed was constant:
Oh, and my HP is working again. My installer tried to reset the power supply trip in the MIM several times and eventually it stayed in. The good news is that I have heating again, the bad news is that there may be an intermittent fault (the worst kind) or there is a slight earth leakage somewhere and this may happen again. Oh the joys of being a Samsung owner…
Thanks for charts. Indeed. Looks like same behavior. In your NASA monitoring it seems it is more precise than when I’m getting this data via Modbus as there is obviously some delay. It’s really strange way how the pump is regulated. I saw some Daikin and Mitsubishi and no way such crazy behavior. Even my two Air-conditions never seen such erratic handling. This is very rough in Samsung.. I saw on my unit sometime the compressor change is in 5 - 6Hz steps up or down… crazy…. Driving it with FRC run much better, of course you can’t exactly hold the room temperature but the pump works much better. Well, until 70% FRC
BTW it would bounce also with FRC enabled, this has nothing to do with bouncing. It start bouncing as soon as LWT is approaching the target flow temperature. Mine didn’t bounce only because I virtually disabled temperature curve by setting it flat to 60C So the pump always run as high as possible but because I limited with FRC it never reached this 60C target.
I decided to use method that up to -3C outside I will use FRC - two steps 50% and 60%. Then from -3C I will disable FRC and set my temperature curve to 45C@-5C and 55C@-20C. Maybe I will need to set 60C@-20C I will see. Thankfully freezing days are much less in last years so maybe I can live with it. What I need is to set 50% FRC from midnight until 6a.m. I don’t want to shut down the pump because of the famous DHW heat stealing and other freeze protection mechanism. I think it simply better to keep it running like to stop it in freezing temps. Also the house will not get cold that fast. Maybe temp drop by 1-2C until morning.
I also have a graph where I monitor how much the compressor frequency changed from previous state and as you can see it’s crazy. Sometimes there is 8hz change and I’m asking why the hell this is working like this. Such big change you can here if you stand near the pump and wonder if such high frequency changes are healthy for the pump ?? On the left FRC enabled, right after FRC disabled. I thought that with colder temps it will be able to regulate it more smoothly but it’s horrid.
Hmm, this is strange becasue you should not have such low COP at 40C water even at 0C. At this parameters I have COP around 3, yours is 2 ? But then at 6C outside you have 3.4 at around 42C water temperature - this seems like OK.
Do you really need such high outlet temperature at outside temps above 0C ?? What is your target room temperature ? I know in UK people heat pretty low Here I have 22 - 23C and 22 is pretty cold already
Happy to hear your pump is working again and you will not get frozen ?
Indeed, probably the biggest single improvement you could make. Maybe start by adding extra K2 radiators in any downstairs living spaces.
With large emitters, we are able to run at 30-33C flow temps for most of the winter and achieve our target 20C room temp. During design we simply replaced all existing radiators with the largest K2 sizes that would reasonably fit the space available (plus added one additional large K2 to the main living area that was previously hard to heat), and then checked output vs heat loss to ensure we weren’t stupidly out of balance (e.g, twice the required heat output in one room with half in another), and we tried to end up over specified in our main living space and didn’t mind if we were under in rooms such as a spare bedroom. As there is no internal insulation between rooms and doors are left open, heat is free to flow between rooms so it doesn’t need to be an exact science (given how inaccurate the overall heat loss figures may be anyway).
On our coldest day to date this winter, we have just about been able to maintain a daily COP of 3, including our DHW. On milder days such as yesterday (OAT 10C), we are at a daily COP of 4.5 and my spreadsheet tells me we are averaging around 3.5 for January so far. If we can stay at or above 3.0 even on the coldest days, then we are no more expensive than the oil boiler we replaced on the coldest days, and significantly cheaper the rest of the time.
That said, I should note I think our 12 replacement Stelrad radiators cost around £2k, so not an insignificant outlay. However, we did go with a lot of non-standard sizes to maximise available space and output in each room, where more common sizes (e.g, 600x1600) are significantly cheaper.
).
.
.
