You will see that this implies that a defrost cycle starts when the R32 temperature leaving the EEV is at or below -6degC, and stops when various other parameters combine/apply. This starts the compressor, and flips the 4-way valve to defrost/cool position, which directs hot compressor discharge into the evaporator to melt any ice build-up on the airfin. It takes heat from the emitter circuit or the DHW circuit, depending upon which was in use at the time of the “start” request.
We have also been advised there is a frost-prevention algorithm:
“This is the anti-freeze logic, when ambient is below 3C and the system has been idle for 60 mins the primary pump will run for 5 mins with the DHW valve energised to ensure return temperature is above 6C. This will happen every hour these conditions are met. This logic is fixed and cannot be turned off.”
This algorithm just starts the circulation pump (not the compressor) on a regular (60 minute) basis for a few minutes to ensure that the pipework is comfortably above freezing point. And it takes heat from the DHW tank!
I infer from the above - and several posts on the forum - that some Samsung users may be seeing a frost prevention cycle but think that they are seeing a defrost cycle. It might be useful to be aware of this distinction when observing heat pump operating data.
Are you talking about recognising a defrost in tech readouts? Because a real defrost is pretty obvious if you look at the machine - all fins hidden beneath a lovely white frosting?
Nice post Sarah. Quiet useful information describing real defrost cycles (exactly what I’ve asked Nick). But I believe this special anti-freez cycles every hour are unique to monoblok heat pumps. I would correct information that it’s not 3C but 5C unfortunately so it’s pretty often if the heat is satisfied and pump is idle. Like today it was nice sun though under 5C and pump was off for like 3 hours. Of course I’ve witnessed 2 anti freeze cycles and heat from DHW was gone.
I would be perfectly fine if I have a choice if I want to circulate water over CH or DHW. Stealing heat from DHW is not very great solution if we taking into consideration that without DHW it worked perfectly fine just with CH circuit installed. Why Samsung don’t incorporate this into the controller - by default take heat from DHW I don’t care but give us the option to change it just like it is without DHW installed.
I’m really debating to put a relay which disconnects phase opening DHW for my 3-way valve in case there is below 5C and idle time approaching one hour and water temperature is above start threshold. I really don’t have nerves from it an watch how the water is getting colder and colder over a course of colder night. That way pump gives the voltage for opening circuit but nothing happen and water will be cycled through CH circuit and everyone will be happy. Here are cycles when I did not have a DHW tank, absolutely fine… So I don’t understand why with installed DHW Samsung needs to take heat from the DHW. Also causing unnecessary load on 3-way valve…
Pump running from 22:30 - 4:13 (5h41 mins) without defrost. HE temperature dropped below -6 at 3:17. As I understand to trigger a defrost it needs to be below -6 for 120 mins. OK this is not fulfilled but at 3:40 temperature dropped under -8C so then if it stays like that for 3 minutes defrost is triggered. Defrost was triggered at 4:13. So strane. Or is that L2 curve not linear ? It’s hard to say from that graph.
Could be. I only posed the question to UK Helpdesk in the context of the AE080BXYDEG series. I looked at the EHS Data Book as a result of @Ringi’s question:
and noted that the Twin versions do have a PHE (whereas Monoblocks do not) so there could well be a difference in the controller algorithms for each.
It’s the sensor labelled “Cond Out (T_CO)” on the PFD (between the main EEV and the evaporator):
which appears on your table as “Heat Exchanger Outlet Temperature” (as you subsequently guessed, it appears from your next post). My reference to LWT>3degC was an error - I had misread the symbol on the Samsung defrost strategy as > rather than < under F, and have edited my top post accordingly. I did ask Helpdesk what “EEP_Option(min)” under D was (months ago) but never got a reply.
I asked Helpdesk about the L1/L2 lines too (I quote my Q and their A):
Q. How do I know what the T_L1_deice and T_L2_deice settings in EEPROM are? Without these I don’t know the slopes of the L2 and L1 lines.
A. These have not been shared with us by HQ so I cannot clarify sorry. I have attached an example below.
No idea. I assume that the numbers are the “Communication Status” figures in the table in your previous post, which your MIM-B19 gives you (but the MIM-C02N doesn’t give me). I looked at the MIM-B19 manual and saw a reference to 7 meaning “Ready” - might this mean something?
Hi. I’m new to this forum so apologies if this is going over old ground.
I’ve been wrestling with the frost prevention cycle ‘stealing’ hot water since my Samsung heat pump was installed 3 years ago.
My installers argued they had installed the system as specified by the manufacturer, and couldn’t help, so I contacted Samsung tech support. After an exchange of emails, this is the key part of their response:
“This is the way the anti-freeze logic works.
The system needs to pump heat from a part of the system during this logic and its fixed to use the DHW cylinder, the reason for this is that there would generally always be available heat in the cylinder to ensure the cycle completes. If the system took heat from the heating portion of the circuit is some situations, with poorly applied systems and low water volumes the antifreeze logic would use the residual heat available. If the circulating water temp drops below 20C the heat pump will not have a safe condition for prolonged operation, which would lead to a lag in response the next time the heat pump was called for.
The logic does call for a backup heater to energise when this logic is carried out which, if you had one fitted would, I imagine reduce the temp drop on the cylinder. However in the UK most installers opt to not use a backup heater, also if you did have on it could lead to slightly reduced efficiency as an inefficient back up heater would be turning on periodically during the cycle.”
For most of the year, the ‘anti-freeze’ logic isn’t a problem for me, but with the recent cold weather it’s been an absolute pain. I heat the water overnight on a cheap Octopus tariff, and that gives enough hot water for the full day. The antifreeze logic quickly knocks much of the tank from around 50c to below 40c (what’s needed for a shower in my view), so leaves me with insufficient hot water unless I run the system to heat it back up.
I’m currently wondering if anyone has found a way to fix the issue? A firmware update from Samsung would be my first choice (there has got to be a better way of doing this), but they don’t seem interested - although I’m quietly hoping they wake up to this as increasing number of customer work out what’s going on and complain! My best idea at the moment is to fit a 3 port valve so the flow only goes through the hot water tank when it’s set to heat up. Any other ideas?
Hi @Rob_P, a good first post, and welcome to the pit.
Frost prevention only cuts in if 1) ambient is below 3degC and 2) the circulating pump hasn’t run for an hour.
So the “low and slow” brigade (looking at you @Ringi) don’t have a problem – when it’s cold the circ’n pump will invariably have run at least once per hour, when the HP has been operating.
But the “big setback” club have to either 1) fool the outdoor temperature sensor into thinking that it’s >3degC (the “heated egg-cosy” approach) – possibly unwise if you value your water piping integrity when it’s really cold – or 2) prevent the lapse of an hour without the use of the circulation pump after heating but before consumption (not easy with modern TOU tariff timings or unpredictable consumption timings).
The option of somehow disabling the 3-way valve so it doesn’t enable the DHW tank heat routing after a heating run, when put to Samsung, elicited the rather brusque response “We do not offer any guidance on work arounds for protection logics as they are designed to protect the equipment.”
So we have been warned , and the search for a satisfactory solution goes on…
@SarahH Rather than the bitterly cold days when the system is running hard, it’s actually the days when the outside temperature is hovering around zero that triggers the frost prevention cycle most.
I have a setup where I can send commands via modbus to the unit, so was wondering if the pump could be triggered to run without the DHW valve open. If that’s possible I might be able to run my own frost prevention cycle with the CH valve open and keep the system happy that way. I’ve not spotted anything in the modbus documentation yet, but thought I’d mention it here in case anyone can think of a way.
I have tried adding a relay into the DHW zone valve circuit, so it didn’t open when the frost prevention kicked in, but that threw a system error. I backed that out sharpish!
So I think I’m still heading towards an extra 3 port valve and plumbing solution (so the flow is maintained but the heat not extracted from the tank), but still quietly hoping a software/logic fix is possible.
Why can’t the setback be on the flow temperature with the pump continuing to run at a low speed? (If needbe create a fake zone with a low required flow temperature and a call for heat that last a few minutes a few times an hour.)
I control my Gen 6 over Modbus. Something I discovered during development may be of use to you. I enable/disable the heatpump with the external thermostat dry contact input driven from a modbus relay. I found that with this contact disabling the heatpump the pump would still be cycled every 5 minutes or so, presumably sampling temperatures in case it was enabled at some point. In order to stop this behaviour it was necessary to turn the indoor unit off (Modbus register 52). This stopped the pump cycling.
Sooo …
Over Modbus set the desired flow temperature to 15degC, disable the heat pump via the dry contact input, turn the indoor unit off. The pump should not run in this state.
At some period less than the freeze protection cycle time (1 hr) turn the indoor unit on. The pump should now run, maybe after a short delay. Use the Modbus Flow Rate register to detect the end of the pump run, then turn the indoor unit off.
Repeat
IF (big IF) the internal logic of the controller only does a freeze protection cycle if the pump hasn’t run in the last hour, then this should defeat it. It remains to ensure that any 3 port valve is pointing at the heating circuit (I don’t use the heatpump for DHW, so can’t comment on this aspect).
If you run the heat pump overnight, it won’t need to run any frost protection cycles. You could set a lower setback temperature overnight to avoid the house overheating. If the house does overheat, the flow temperature is probably set too high. Running for longer on a lower flow temperature should result in higher scop and less energy use overall.
Electric is cheapest at night if you’re on dynamic tariff and gird carbon intensity is lowest, so it’s the perfect time to run the heatpump.
Either way the amount of energy lost during an anti frost cycle is very small, about 100-150Wh per cycle for my 5kW unit, so I wouldn’t worry too much about it
But if the DHW coil continues high into the tank, an anti frost cycle will take more heat then that from the middle/top of tank, due to bottom of tank being cold.
The measure above is the total amount of heat being extracted from the DHW tank, it’s very small e.g three anti frost cycles extracts 300Wh of heat from the DHW tank:
Assuming a very pessimistic COP of 2.5 to re-heat the DHW, this is only 0.04kWh of electric per anti frost cycle, about £0.01. @Michal_S screenshot above shows 6 anti frost cycles during an overnight period, this will be using about 0.24kWh or £0.06 of electric (assuming price cap fixed tariff).
From my observation I can say what I’ve already stated in some of my previous post. That amount of “stolen” heat is significant. Basically I had my water temp set to 50C. On a colder night below 0C there were 4 or 5 cycles with the circulation pump running every hour for 6 or 7 minutes. Each hour it took around 1.5C from my tank and over course of the night it was around 8C maybe taken from my tank. I don’t think this is low. I need maybe 1.5KW of electricity to re-heat it back to 50C. Definitely not 300W.
COP2.5 is maybe truth in + temperature or really low DHW temperatures but this night there was -3.5C and COP for DHW was 1.96 to get the water to 50C (middle part so top maybe 56C) not talking about defrost cycle.
Frost prevention is running when outside temp is below 5C. Sarah got info from technician it’s 3C but my observations says 5C. And yes, I’ve “fixed” it with a relay. Basically i connected relay between DHW terminal on the controller and the 3-way valve. I have the relay (shelly 1) ON all the time when space heating is running. When the pump wants to really heat the water I switch this off by automation in home assistant and the valve change its position since is without power. But in case the pump is only opening DHW circuit to steal the heat for the frost protection cycle the relay is still on and won’t allow 3-way valve to switch to different position. I know, nasty, but it works So basically similar what you did so I wonder what error you’ve got ?
@Michal_S That’s really interesting, I’m glad people are exploring solutions, and it’s got me thinking too. I’m wondering if tricking the outdoor temperature sensor may be a simpler approach?
Maybe a switched resistor in place of the temperature sensor, corresponding to a value of 10C for example, may work. One could then switch between the genuine ambient temp sensor and a fixed value (that is above 5C) to temporarily ‘disable’ the anti-freeze protection.
).
) don’t have a problem – when it’s cold the circ’n pump will invariably have run at least once per hour, when the HP has been operating.
, and the search for a satisfactory solution goes on…
So basically similar what you did so I wonder what error you’ve got ?