Weekend Challenge... Heater Controller

[Zerax]

Challenge
I need to come up with a way to control an electric convector heater from a simple external on/off switch.

I have been given several convector heaters that match the ones I already have, but I've been told that "the thermostats" have broken. I'd like to take the opportunity to re-use these heaters but have my own external control via an external wifi type thermostat such as the Honeywell Y87RF2024.

As the heaters have their own programmer/thermostat module, simply switching the supply does not work as the programmer needs buttons pressing to initiate any output. Looking inside the units, they have a heater control board and a separate programmer/thermostat module connected via a ribbon cable.

So... me thinks that if I can identify the output from the programmer that switches the heater board, I could probably re-configure this to be switched by the external unit. However, my limited electronics knowledge and the fact that the heater is currently not working is hampering the plan.

Pics are attached...

I've powered up the heater, and checked for voltage at each combination of pins on the ribbon cable. I only got mains voltage across Pins 1 & 8 (pin 8 being the one on it's own)

I know that the gate of the Triac needs to go high... but that's as far as I get...

 

[Lucien Nunes]

I'll need to look at on a full size screen later. The low voltage DC that powers the control board probably comes from a capacitive dropper, rectified by one of the MELF diodes, smoothed by the 470uF electrolytic and possibly shunt regulated by a zener. The negative is probably common with one side of the mains. See if both sides of the capacitor go to the power pins of the IC. If so, that's two cores of the ribbon identified. You can then see if DC is present and with luck, using 0V as a reference, see which of the other lines does or does not change state when you operate the controls. If the DC is lacking, a likely cause is the cap going low or a dropper resistor going high.

FWIW what actually happens... does it respond to buttons but not correctly control the temp, or is the control completely dead?

 

[Zerax]

Thanks for that LN... back at it now after a few busy days working... so from what you can tell... the board with the triac on should be providing a DC output via the ribbon cable ? Because, when I powered up the board... there was no DC on any of the pins, only 240v across the two end pins.

The electronic timer/programmer at the end of the ribbon cable lights up fine, but just doesn't actually switch the heater on.

I'll examine the cap later on... to see where it goes.

I'm intrigued about how the gate on the triac is switched... does it need DC or AC ? and at what sort of voltage ?

 

[Lucien Nunes]

There should be a sensible (e.g. 5V) DC voltage across the 470μF electrolytic if the controller seems to be alive. If it is suspect, check the resistor standing on end next to the triac for being high (the other possibility is that the cap is low) Keep in mind that all low voltages here are dropped directly from the mains so treat '5V' as mains safety-wise. Triacs typically fail S/C so if it is permanently off, it is less likely to be a triac failure which would have been a simple fix.

Triacs are triggered by a current (not voltage) impulse every half-cycle. The impulse causes the 4-layer construction to avalanche and it stays on until the end of the half-cycle. As the main circuit current waveform (between MT1 & MT2) falls below the holding current it turns off, ready to be triggered again at the next half-cycle. The sensitivity of the triac is defined by the gate threshold current, a few mA is typical, but it also depends which of the four quadrants it is working in (i.e. the relative polarities of MT2 and G to MT1), which in turn depends on the design of the firing circuit and whether the nature of the load requires four-quadrant operation.

A resistive load requires only two quadrants; if the firing circuit is symmetrical e.g. in a simple 2-wire twist dimmer, it will be Q1 & Q3, i.e. the gate drive polarity will reverse every half cycle, in sync with the main circuit voltage present on MT2. In this case, because it is being fired from an electronic control without an opto-isolator, it is probably Q2 & Q3, i.e. the gate is always driven negative w.r.t. MT1 on both half-cycles of the mains. If this is the case, you will find the DC supply +ve connected to the mains and MT1. I cannot see the circuit around the triac clearly enough to be sure whether it provides symmetrical firing.

 

[marconi]

Zerax: As you are using the Honeywell wifi t'stat which has a mains powered receiver with relay - you could use the volt free contacts to either turn an all new mains coil contactor on and off to control the element or instead a solid state relay with heatsink (albeit you would have to faff about arranging a low voltage 4-32 V low current dc supply to power the control SSR input - capacitor voltage dropper, some resistors, a zener and capacitor - or DIN PSU see link below).

https://cdn.plumbnation.co.uk/site/...round-thermostat/y87rf-installation-guide.pdf

solid state relays | Rapid Online - https://www.rapidonline.com/Catalogue/Search?Query=solid%20state%20relays

Turn-on voltage min.	4V DC
Turn-off voltage min.	1V DC
Input current max.	25mA

Mean Well MDR-10-12 12V / 10.08W Mini Din Rail PSU - https://www.rapidonline.com/mean-well-mdr-10-12-12v-10-08w-mini-din-rail-psu-85-5651

 

[Simon47]

An obvious observation is that if you want to use an external switch, then just discard the controller board and wire the switch in it's place - or just link across the switched terminals on the board.
It's too late in the day for my mind to start looking into how the circuit is arranged. The fault may be as simple as the triac having failed - but I'd need to have the unit in front of me so I could poke and prod it ...