3 way valve installation protip

Dear all,

I wanted to share my experimentation with heat pump installation.

In my setup, there is a 3 way valve, to connect the underfloor circuit to the heatpump aside the regular radiator circuit.
My samsung gen 7 5kW handles the 3 way valve to regulate the temperature inside the underfloor circuit. However, it does only support proportional valves, which perform a complete open->close or close->open movement within a range of time between 60 and 120 seconds.

From past experience, I know that those valves tend to get clogged quite easily thanks to underfloor oxydation :frowning: Therefore, spending hundreds in a valve that I know will fail in the same amount of time that a low cost valve was worth the try to reduce the expense.

On amazon, I found a 240VAC operated 3/4" 3 way valve, however, it opens/closes completely with a 10/12 seconds period. Naively, I mounted it, and observed that each 2 minutes (a FSV controls this time) the valve was operated, closing or opening to “regulate” the temperature, but very poorly. But this was expected, as the heatpump expected to for example open by 50%, therefore powering the valve for 30 seconds (when FSV configuring the total time is set to 60 seconds). In 30 seconds, the valve with a 10 seconds full swing was obviously open completely!

I tried regulating the voltage given to the valve by adding a thyristor (gradator circuit), but the valve, was only operating with a minimum voltage which implied running full speed. Wrong path!

Eventually, I designed a NE555 and triac based circuit, to power the valve 0.5s each 3 seconds (1/6 ratio) to convert the 60 seconds full swing into a 10 seconds of effective powering. And this works very smoothly. I went from 2 minutes spikes of temperature curve to a very smooth curve. The only circuit design trick was that I wasn’t willing to have a separate power for the NE555, therefore extracting power from the valve command by the heatpump. The only problem was that the heatpump has 3 wires, 1 neutral, 1 clockwise live, and 1 counterclockwise live. The circuit therefore had to command only a single line at a time.

If anyone is interested, let me know, I’ll post up the circuit.
Cheers

Also, I forgot to mention that nowhere in the documentation is to be found the mapping of clockwise and counterclockwise on the flow of water (from the circuit or from the heatpump).

I posted the full details there Mixing valve installation - Samsung EHS Wiki. But here is the essential digest:

  • CW corresponds to water flow from the UFH return path. Wire the B2 connector on the control unit to the phase of the valve to point toward the return water flow of the UFH to circulate again in the UFH (100% mix).
  • CCW corresponds to heated water flow from the AWHP path. Wire the B3 connector on the control unit to the phase of the valve to point toward the AWHP water flow to circulate into the UFH (0% mix).

Cheers,

I’m not familiar with Samsung units, but if the output is for PID control of a mixing valve for UFH etc, then why not just use a mixing valve with a 3-point controlled actuator (with two switched lives for CW & CCW), like a ESBE ARA661 etc and set the total runtime to suit, it will then incrementally open/close the valve to achieve the desired temperature, as governed by the flow temperature sensor. I guess it may cost a little more money upfront, but it’s what it’s designed for and should be pretty reliable and saves reinventing the wheel! Or if you are not using weather compensation etc for this circuit and want to save a pound, then you could just use a simple thermostatic mixing valve to blend it down as required.

Not sure why you’re having problems with oxidation/clogging on a clean & sealed system with barrier pipe etc, but the ESBEs can be easily stripped down and cleaned/serviced if need be.

Thanks for your message,
Actually it’s a controlled actuator I used, but a cheap one, 30€ is not much for an actuator, it just required a bit of delaying on its command to be usable :slight_smile:

I’m having oxydation perhaps mainly because of the pipes (PER) and the fact that the previous heating piece was a gas boiler. So probably boiling the water or almost was more subject to oxydation and clogging of the underfloor system. I’ll know for sure after one or two years and checking the water state and if the valves gets clogged.

Cheers

Have you tried flushing/treating the system, especially some biocide for low temperature heat pump systems.