community, your help is required on this one. I need to control opening and closing (up to 50% opening) of a 3 way valve (see pictures). I use the following hardware:

15W 230 - 24 v transformer; Arduino DUE (the software side works correct); LTC1665 (octal 8 bit digital to analog converter); Regulated switching power supply on LM2596(see photos);

The idea is to control valve with 0-5V signal from DAC, the software part of the project works perfectly, I am able to get the 0-5V DC signal. The problem is that I need to power the valve directly from the transformer 24VAC connectors as indicated in the picture below:

enter image description here

When I do so, I am able to open and close the valve as required, but the transformer fails shortly after that. When I measured the current flow at the output of the transformer I measured 1,65A , surely the transformer fails due to overheating.

The (BAD) solution to the problem was to rearrange the circuit in the following way:

enter image description here

The valve now works according to the 0-5V signal and the current measured at the output of the transformer is 0.14A.

The problem is that I really need to power the valve directly from transformer, while in the bad solution I have to connect 24VAC (-) pin of the valve to GND pin of the Arduino and 24VAC (+) pin to transformer, that might cause some troubles in the future.

Any suggestions to my problem ?

Thank You.

P.S.: As I understand, when the valve is powered directly from the pins of the transformer, the 0-5V signal I'm giving from the DAC just got no reference on the valve side. The power supply thing is connected properly and it provides stable 12V DC power to the system , also the in (-) and out (-) pins are connected (checked using multi-meter).

  • \$\begingroup\$ The 3way valve is controlled using 0-10 DC signal for 0-100% , but I only need 0-50% here. \$\endgroup\$ Aug 31, 2016 at 11:48
  • \$\begingroup\$ Why do you "really need to power the valve directly from transformer"? \$\endgroup\$
    – Andy aka
    Aug 31, 2016 at 11:50
  • \$\begingroup\$ Well, I've got myself into hobbyist air ventilation controller thing recently. Professionally I'm designing software for such systems (industrial air vent systems) but well I was stupid enough to try to design an Arduino based controller myself, just as a hobby. And generally I've seen a lot of professional air vent system designs and for some reason it's a rule of thumb there to power the valves from transformers directly. Just trying to fiure out how they do it. \$\endgroup\$ Aug 31, 2016 at 11:59
  • \$\begingroup\$ This whole project depends on the internal circuit of the valve and you haven't provided any information on that. Please add a link to the datasheet. You have no ground return for your 0 - 5 V control signal. \$\endgroup\$
    – Transistor
    Aug 31, 2016 at 12:43
  • \$\begingroup\$ If "surely the transformer fails due to overheating" is the problem, is not the solution a bigger transformer that is rated to handle the rated power of the valve? \$\endgroup\$
    – scorpdaddy
    Aug 31, 2016 at 13:24

2 Answers 2


Generally 0-10Vdc controlled valves are two-way not three-way. I mean 2 way electrically, but they say 3 way mechanically, meaning 3 ports. (SPDT action)

They are usually optoisolated drivers to switch the solenoid powered by a floating transformer both isolated from grid AC and requires no SMPS DC supply (for this purpose) and no ground connection to the supply output, except a DC ground between Valve and logic driver of Arduino. (Verify with a diode test and DMM.)

Thus because the user has a choice of using 0 to 10Vdc voltage control (isolated from 24Vac) or 50 Ohm current control using 5V logic direct drive. (which is what I think you mean by 50% ... of 10Vdc.)

Most CMOS now is 50 Ohms in 5Vmax logic and 25 Ohms in 3.3V max logic. This drives about 10-20 mA in current mode thru the IR optoisolated valve control input often with a small current limiting R.

  • (5-1.2V)/20mA = 190 Ohms ... subtract 50 gives Rs = 170 Ohms



http://www.axiomatic.com/10vdc-din-controller.pdf (linear)

The SMPS is NOT needed at all for the valve and may be floating output for other circuits isolated both from the 24Vac and the 5Vdc logic. If powering the SMPS from the same 24Vac transformer used by the valve, then the VA rating must be capable of supplying VA output requirements for both.

Now getting back to your "three-way" title. Thus implies tertiary logic levels for forward, reverse and center OFF using 0mA 5mA and 10mA.

Please confirm 3 way valve with specsheet. But may be terminology for 3 port connection with a 2way solenoid. i.e. SPDT


Both circuits are really really questionable!

1st) I does not work properly because you supply the valve with AC, but your 0-5V signal is referenced to GND. I now assume that the LTC GND pin is connected to the Arduiuno GND. Thus your 0-5V signal becomes a -5V to 5V signal. Just connect you valve to the 24V after the rectifier, so there is DC! This should be possible, you even namend the pin AC+ and AC-.

2nd) This is really bad! You produce an unsymmetrical load over your DC/DC-Konverter. But this helps to fix your GND problem.

Use solution #1 ;-)


Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.