why not a Diode before a relay ?

Thread Starter

Pastilhas

Joined Jun 1, 2020
10
Hi all, and sorry for my poor English.
I have a water controller (24v) having many zones. The same controller always start the master valve and one programmed zone. The fertilizer pump start because I have a relay connected to the master (I use a relay 24v / 220v to start the pump) Fig 1.
My problem is every zone got fertilizer and some time they don't need it.
So I connected the relay to all zones using a switch to turn on or off where needed and used a diode on every positive line zone, for no return current (Fig 2). But the relay just don't stay on, he goes crazy intermittent and pump don't start. If I remove the Diode it works correctly and pump start.
What is wrong ? Can't we connect a diode before a relay ?

Tanks in advance and hope people got what I'm trying to say....:mad:

Fig 1:
Fig 1.jpg

Fig 2:
Fig 2.jpg
 

Thread Starter

Pastilhas

Joined Jun 1, 2020
10
What are the details on the 24v controller??
Max.
Hi MaxHeadRoom.
Tank for your reply.
It's a esp lx modular controller from Rain Bird.
• 230VAC ± 10%, 50Hz.
• Output: 26.5VAC 1.9A.
• Multi-valve station capacity: Up to two 24VAC, 7VA solenoid valves per station plus a master valve.
Master valve and the others are:
• 24 VAC 50/60 Hz (cycles per second)
• solenoid power requirement: 0.450A inrush current; 0.250A holding current
• Solenoid coil resistance: 38 Ohms

What did I try so far ?
Master + another valve it works
Master + relay for pump it works
2 Valves in one station (Zone) it works
1 valve + relay in one station it work
Master + Diode 1N4007 then relay (24v / 220v) dint work, relay go crazy.
1 valve + Diode 1N4007 then relay (24v / 220v) connected dint work, relay go crazy.

I think relay don't like to have a diode, just don't know why...
 

Delta prime

Joined Nov 15, 2019
329
When an electromechanical relay is de-energized rapidly by a mecanical
switch or semiconductor, the collapsing magnetic field produces a
substantial voltage transient in its effort to disperse the stored energy
and oppose the sudden change of current flow. A 12VDC relay, for example,
may generate a voltage of 1,000 to 1,500 volts during turn-off. With the
advent of modern electronic systems, this relatively large voltage transient
has created EMI, semiconductor breakdown, and switch wear problems
for the design engineer. It has thus become common practice to suppress
relay coils with other components which limit the peak voltage to a much
smaller level.
IUn5v.png
 

Thread Starter

Pastilhas

Joined Jun 1, 2020
10
When an electromechanical relay is de-energized rapidly by a mecanical
switch or semiconductor, the collapsing magnetic field produces a
substantial voltage transient in its effort to disperse the stored energy
and oppose the sudden change of current flow. A 12VDC relay, for example,
may generate a voltage of 1,000 to 1,500 volts during turn-off. With the
advent of modern electronic systems, this relatively large voltage transient
has created EMI, semiconductor breakdown, and switch wear problems
for the design engineer. It has thus become common practice to suppress
relay coils with other components which limit the peak voltage to a much
smaller level.
View attachment 208705
Tank for your input.
But I'm not "de-energized" the relay, I'm energizing it.
What you're saying is about when the current through a coil is switched off, we use a diode in parallel to avoid damaging some nearby components. In the case you point the purpose of the diode is to allow the current flowing through the coil to continue circulating when the relay is deactivated

My case is different as I say.
When I energize the coil from the relay whit a diode, the coil just go crazy. Whiteout the diode it works as supposed
Diode  and relay.jpg

This is the circuit I did:
How it work ?
The controller can run a master valve and one and only one station valve same time.
The controller turn on the master valve + station 1 then station 2 etc..
So for fertilizing station 1 and station 2 I need to turn switch 1 and 2 on, but I need to put diodes or when station 1 get energized it going to send 24v back to station 2.

My problem is when I energize station 1 and having the diode, the relay just go crazy.
ferlilizer.jpg
 

Thread Starter

Pastilhas

Joined Jun 1, 2020
10
Those are AC relays, which require the current to go in both directions, so you can't put a diode in series (and there's no reason to).
That why he goes crazy intermittent ? Only get current in one direction? So what are the way to archive my purpose ?
I put the circuit on Delta prime reply.
Thx in advance
 

crutschow

Joined Mar 14, 2008
24,938
So what are the way to archive my purpose ?
If you use a diode for isolation, you need to use a 12Vdc relay.

Edit: That will likely cause chatter in the relay due to the half-wave signal.
Likely better to use a 36Vdc relay with a large capacitor (e.g. 100μF) across the coil to smooth the signal.

Alternately you could use a separate 24Vac relay from each output and parallel the relay outputs to the fertilizer pump.
 
Last edited:

MaxHeadRoom

Joined Jul 18, 2013
20,661
Tank for your input.
But I'm not "de-energized" the relay, I'm energizing it.
What you're saying is about when the current through a coil is switched off, we use a diode in parallel to avoid damaging some nearby components. In the case you point the purpose of the diode is to allow the current flowing through the coil to continue circulating when the relay is deactivated
My case is different as I say.
When I energize the coil from the relay whit a diode, the coil just go crazy. Whiteout the diode it works as supposed
View attachment 208708



My problem is when I energize station 1 and having the diode, the relay just go crazy.
You cannot use an AC coil or relay on the same value DC, the resistance is way too low.
Max.
 

Irving

Joined Jan 30, 2016
308
You cannot use a diode on an AC relay like that. The resulting magnetic field from the 1/2 wave rectified current is insufficient to maintain pull-in, hence the chatter. Changing the relay for a different coil voltage and adding a smoothing capacitor is overkill.

As Max suggested, the simplest & correct solution is to use a separate 24vAC relay on each valve circuit, and connect the relay contacts in parallel for the fertilizer pump. That means each relay needs to be able to switch the pump current. It may be cheaper to use a lower specification 24v AC relay on each valve circuit to drive the bigger pump relay off the master circuit.
 

Irving

Joined Jan 30, 2016
308
As Max suggested, the simplest & correct solution is to use a separate 24vAC relay on each valve circuit, and connect the relay contacts in parallel for the fertilizer pump. That means each relay needs to be able to switch the pump current. It may be cheaper to use a lower specification 24v AC relay on each valve circuit to drive the bigger pump relay off the master circuit.
Circuit for above...
relays.JPG
 
Last edited:

Irving

Joined Jan 30, 2016
308
Alternate using lower spec relays...

The other advantage of this arrangement is that it keeps the 230v supply away from the rest of the low voltage circuit and is therefore prefered from a safety aspect.

relay2.JPG
 

Alec_t

Joined Sep 17, 2013
11,274
When I energize the coil from the relay whit a diode, the coil just go crazy.
If your relay coil is rated for an AC input and you have a series diode then you are driving the coil with half-wave rectified AC. It's not surprising it chatters. An alternative to the suggestions above would be to rectify each controller output to DC and then use a diode-OR arrangement to drive a DC-coil type relay.
 

crutschow

Joined Mar 14, 2008
24,938
Here's the schematic for the approach using diodes, a capacitor, and a DC relay.
100μF should be sufficient for the typical 36Vdc relay coil resistance of about 3500Ω.

1591106252444.png
 

Thread Starter

Pastilhas

Joined Jun 1, 2020
10
Ok I got it now... Tank all for your tips and good will to help :p
I'm filling a little smarter today after reading all.;)

I think I'm going to do crutschow approach using a diodes, a capacitor, and a DC relay .

The esp lx modular controller have 32 sectors (valves) so using 32 relays is to overkill for that project. o_O
 

Irving

Joined Jan 30, 2016
308
I was just looking at same, it's 3600Ohm, so your circuit looks good. Will give 32v average output into 3600ohm, and relay needs 75% = 27v to pull in. Drop out is 3.6v min so plenty of leeway. drop out time constant ~ 0.75sec.
 
Top