The motor is a central heating circulating pump which has a
single-phase motor with capacitor driving a quadrature-phase winding. The
unit is rated at up to 82W, 0.36A @ 230V AC
The flow switch is rated at 15VA, max 1A AC, into resistive loads.
I've picked up a couple of BTA16 600B triacs and some 250V AC MOV
suppressors which seemed like a good starting point (OK a better starting
point might have been to design the circuit first ... :-).
I'm now wondering about how to connect it all up. From the point of view
of triggering the triac I'm thinking of something like this:
-------------- LIVE 240V AC
| MOTOR |
FLOW \ | MT2 |
SW o | |
| | TRIAC |
------| G |
| MT1 |
However if the switch closes at peak mains voltage then there's going to
be a surge of current into the gate before the triac gets fully turned on,
which could damage the triac and/or reed switch. Therefore I think I
should put a resistor in line with the switch. The peak gate current of
the triac is 4A for 20uS. (This is more than the steady-state rating of
the reed switch but I'm guessing/hoping that a mechanical switch can take
orders of magnitude bigger transient peaks than a semiconductor can.) I
don't know how long the triac will take to turn on and therefore the
voltage across it to drop but 20uS sounds about right, so allowing for 4A
max gives a resistance value of (340V peak / 4A = 85 ohm) about 100 ohm.
The triac gate current to turn on is 100mA worst case (quadrant IV) so the
triac should turn on when the supply voltage has risen to (0.1A * 100 ohm
My other concern is switch-off transients. Since the load is inductive
then, even though the triac should turn off at a mains zero crossing,
there will still be current flowing in the motor winding which will
generate a big voltage spike. I'm hoping that the MOV will catch
this. I'm thinking of connecting it across the triac, since that's what
I'm trying to protect.
I now have this circuit:
-------------- LIVE 240V AC
| MOTOR |
| ------------ |
FLOW \ | MT2 | |
SW o | | -----
| | TRIAC | | MOV |
------| G | -----
| MT1 | |
OK what do folks reckon? Will it fly, or crash and burn?
Any better suggestions/improvements etc.
In particular I wonder if I should be using a snubber network as well as
an MOV? (And if so, why?) If so, what components? I've heard 0.1uF and 100
ohm mentioned. And what sort of capacitor? 400V polyester OK?
A bit of background: I've been out of the electronics game
for more years than I care to remember and never did much with triacs and
suchlike even when I was involved. However I do plenty of mains work (so
please feel free to skip the health warnings :-))
Use a zero crossing optoisolator like a MOC3031.
>I'm trying to protect.
>I now have this circuit:
> -------------- LIVE 240V AC
> | MOTOR |
> 100R |
> | ------------ |
>FLOW \ | MT2 | |
> SW o | | -----
> | | TRIAC | | MOV |
> ------| G | -----
> | MT1 | |
> ------------ |
> -------------- NEUTRAL
>OK what do folks reckon? Will it fly, or crash and burn?
It will possibly work. Best to put a snubber across the triac as well.
Eg. 0.1uF/600V+ in series with 100 ohms. This is a classic triac
'static switch'. Make sure your flow switch is rated for the voltage.
Make sure your heatsink is more than adquate.
>Any better suggestions/improvements etc.
>In particular I wonder if I should be using a snubber network as well as
>an MOV? (And if so, why?) If so, what components? I've heard 0.1uF and 100
>ohm mentioned. And what sort of capacitor? 400V polyester OK?
Ah, now that I read down to here.. no 400VDC is NOT okay for across
the 240V line. 600VDC is about minimum, better to use a cap rated for
X-line AC service. Eg. Panasonic ECQ-E2A104MW (about $1 one-off). Use
a 1W MOF power resistor for both the 100R resistors.
>A bit of background: I've been out of the electronics game
>for more years than I care to remember and never did much with triacs and
>suchlike even when I was involved. However I do plenty of mains work (so
>please feel free to skip the health warnings :-))
Okay, but don't get cocky.
"it's the network..." "The Journey is the reward"
sp...@interlog.com Info for manufacturers: http://www.trexon.com
Embedded software/hardware/analog Info for designers: http://www.speff.com
Hi, John. A couple of things:
Apart from the "health warnings", it's considered normal practice these
days to use low voltage for control circuits. In the intervening years
since you've been out of the electronics game, using line voltage for
control circuitry has become very bad form. Not only that, but it
could get you in serious legal trouble if something goes wrong and
someone gets hurt.
Second, your reed switch is rated for 15VA, 1A max. Assuming your
switch makes contact at the peak of the AC line cycle, you're switching
400V, which is a mistake. I'll bet if you look in the fine print,
you'll find the maximum switching voltage of the reed switch is either
100V or 150V. Also, the 100 ohm series resistor means the peak gate
current will be in excess of 1 amp, also a problem. Your circuit will
spot weld closed the contacts of the reed switch in short order.
It might be best to forget the triac. Get yourself a small DC wall
wart, a proper control relay with the same coil voltage, and a
transistor and couple of resistors sized properly to drive the relay.
You can then use the reed switch to apply voltage to the base of the
transistor to turn on the relay. This will give you much better
reliability, and is really what small control reed switches are meant
This circuit with a 12VDC unregulated wall wart and the components
chosen will drive just about any reasonably sized 12VDC coil control
relay (view in fixed font or MS Notepad):
| +12V _
| + / \ L
| | .---------( M )-----o
| .---o--------o---. | \_/
| | | |RY1 |
| o 1N4002| | |
| '\ SW1 - C| |
| \ ^ C| / |
| o \ | C| - - - - --- ---
| | | | -------
| | | | / |CRY1
| | '---o |
| | | | N
| | ___ |/ '---------------------o
| ---o--|___|- -| TIP100
| 10K | |>
| .-. |
| 10K| | |
| | | ===
| '-' GND
(created by AACircuit v1.28.6 beta 04/19/05 www.tech-chat.de)
Questions of this type usually get a good reception at
Many good points. Around here (USA) it's also very bad form to put a
contact between the load and neutral.