Relay driver with holding current

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hello

I am looking for a clever/simple/few parts technique of activating a relay and then dropping off the current to a "holding" level to minimise the energy spent on the coil.

I am working with a 12V relay that activates on 7.5V and deactivatres on 3V.

I want to start it off at 12V, hold it there say for 0.5 seconds and then gradually drop it to 5V-8V which will be the "holding" state. If I switch it off it should clear itself immediately and forget it was in a holding state before.

I have experimented with adding an RC controlled bypass transistor but am not getting anywhere. Simple schematic attached. The idea is that you press the switch, the capacitor is short circuit and Q2 shorts the series resistance allowing full current to pass through. Once the cap is charged Q2 switches off and then the current though the coil has to go through an extra resistor. But when you try to switch off the relay and back on this circuit fails.


Thanks
 

Attachments

  • relay-driver-holding.JPG
    relay-driver-holding.JPG
    33.5 KB · Views: 174
A capacitor appears as a short at power-on, so you could try a completely passive approach by connecting the resistor and capacitor in parallel and putting them in series with the relay coil. When switched, the cap will pass current until the resistive DC voltage divider takes over to set the holding current.
When switched off, the cap discharges through the resistor. If you need a faster reset, use a DPDT switch the quickly discharge the cap. You'll probably need something bigger than 10uF for an acceptable delay.
 
Last edited:
I tried exactly that approach to start with and needed a very big cap (coupled with a 300ohm resistor , or even less for some relays), so then I thought that I should "multiply" the capacitor by using the transistor, so that I can set an RC independent of the coil resistance. This is great to start with, and you can adjust the RC to suit, but I have not yet found a way to discharge the capacitor quickly. The schematic is very simplified, the control signal will be the output of an op-amp, which will be "high" for "relay on" and "low" for "relay off".
 
Administrator
Joined 2007
Paid Member
How quick does the repetition rate have to be ?
You could use a series resistor calculated to give say 4 volts across the relay coil. The coil is switched at the "earthy" end. An electroylitic cap is connected from the top of the coil to ground. When off the cap charges to 12 volts and this voltage is available to "kick" the relay smartly closed. Its a technique that works well.
 
Hi,
Attached it is a schematic of a circuit that allow you to do what are you looking for. You can doing it using a voltage regulator as the LM317 with 2 resistors. What you do it is using two resistor to adjust the voltage output one for 5 volt and the second one for the 12 volt. To energize the relay you set the voltage to 12 volt and to drop to 5 volt you short the R3 to drop the voltage to 5 volt. You can use an LM555 to set the delay or use a resistor and cap for the delay. Sorry but I am not technical writer. Hoped you understand the principle.
 

Attachments

  • Switching voltage output.jpg
    Switching voltage output.jpg
    131.4 KB · Views: 153

Attachments

  • PWMcircuitwithmotor.jpg
    PWMcircuitwithmotor.jpg
    53.8 KB · Views: 129
You haven't given any performance reason for the long delay, and "very big cap" is relative, especially considering a half dozen other parts. If I were using an op amp control signal, I'd probably look to pair it with a solid-state relay. Or perhaps you could use a "logic-activated" FET, similar to that used in the power management section of the O2 HPA.
 
This is a good idea I had not thought of.

I have a (battery) circuit that switches on a relay after a delay, using a PIC micro. I have a spare pin on the PIC. It's a 5V relay with a low enough current to run off a PIC pin.

I can turn on the relay with one pin, then hold it on with a second pin, dropping some voltage using a diode or two, maybe a LED, then turn off the first pin. This will save a bit of current.

A latching relay is no good to me because the relay must drop out when the PIC is powered off.

Thanks. I mean, thanks a lot.
 
It's worth knowing that if you have a freewheeling diode across the relay (protecting a transistor or other driver) then it delays the relay dropout by some milliseconds. If you want the relay to drop out as quickly as possible, then include a resistor in series with the diode, of about the same value as the DCR of the relay coil. This will dissipate the energy in the coil quicker than the diode alone.
 
"...thought that I should "multiply" the capacitor by using the transistor,..."

"...but I have not yet found a way to discharge the capacitor quickly..."

That would be the best idea...and put a reverse diode on the cap amplifying transistor's base-emitter junction. In the off state the cap will discharge through the diode and R3.

just to clarify, see jpg
 

Attachments

  • hold-relay.JPG
    hold-relay.JPG
    16.4 KB · Views: 101
Last edited:
I have two projects where relays are used: a "slow-start" to power a big toroidal, which currently uses two relays at full power all the time (but it has worked fine for years), and a battery charger where the relay must disconnect the battery when a condition is met.

Latching relay requires a signal to switch it off, and if the power goes down I'd like the relay to disconnect too (battery charger).

I cannot use a solid state relay (but I would prefer it) because I will passing 12V/1A/2A, and have not found anything over 200-500mA (within reason).

Here's what I have come up with so far. It does use a few extra components than the basic resistor transistor relay driver: an extra two transistors, four resistors and one capacitor.

When the signal is "on" then:
Q1/R3 sets the holding current.
Q2 provides the initial energising current.
Q3 and the RC combination C1/R4/R6 provide a small delay after which Q3 shorts the base of Q2 and thus stops it from conducting. R4 provides the necessary base current to Q3.

When the signal goes "low" then R4 serves to discharge C1, so the system gets "reset" when the cap is more or less empty.

On the scope with 100ms divisions, you can see the initial energising voltage, and after about 50ms the steady holding voltage. When the signal is off, it takes about 40ms to drop to 0.

This is all done on the simulator, I have not built it yet.

I think the next step might be to try using a 555 or find the PWM relay driver chips that someone mentioned before.
 

Attachments

  • relay-driver-holding.JPG
    relay-driver-holding.JPG
    106.1 KB · Views: 110
emf drives a series combination of dropping resistor + relay coil + C to E of transistor switch.

If you have a 12V coil of 300r and adopt a 300r dropping resistor then in the long term with CE closed the relay coil sees half the emf.

Now add a capacitor from E to top of coil.

when emf is first applied and CE is open the capacitor charges up via r and reaches ~2/3rds of final voltage in 1 RC time period.

Close the CE switch and the capacitor partially discharges to half emf through the relay coil. This fires (triggers) the relay. and the remaining half emf holds the relay active.

I usually use 15V from a 7815 to drive a 12V relay. R= coil the relay fires on ~14V and holds on ~7.5V. Cool running, fast ON and fast OFF.

All it "costs" is one 33uF 25V electrolytic.
 
"...thought that I should "multiply" the capacitor by using the transistor,..."

"...but I have not yet found a way to discharge the capacitor quickly..."

That would be the best idea...and put a reverse diode on the cap amplifying transistor's base-emitter junction. In the off state the cap will discharge through the diode and R3.

just to clarify, see jpg

Ahhh, that seems to work nicely. There is a problem reaching full energising voltage, and the drop out is very gradual rather than stepped, but it should work. Here it is.
 

Attachments

  • relay-driver-holding-2.JPG
    relay-driver-holding-2.JPG
    91.6 KB · Views: 84
emf drives a series combination of dropping resistor + relay coil + C to E of transistor switch.

If you have a 12V coil of 300r and adopt a 300r dropping resistor then in the long term with CE closed the relay coil sees half the emf.

Now add a capacitor from E to top of coil.

when emf is first applied and CE is open the capacitor charges up via r and reaches ~2/3rds of final voltage in 1 RC time period.

Close the CE switch and the capacitor partially discharges to half emf through the relay coil. This fires (triggers) the relay. and the remaining half emf holds the relay active.

I usually use 15V from a 7815 to drive a 12V relay. R= coil the relay fires on ~14V and holds on ~7.5V. Cool running, fast ON and fast OFF.

All it "costs" is one 33uF 25V electrolytic.

Hi,

Here is how I understand your suggestion with realistic values - looks nice, and will definitely go and try it because it is the simplest. Only thng I am not sure about is the energising "spike" which I'd much rather it were a "energising plateau". But I will build it and test it in a minute.
 

Attachments

  • relay-driver-holding-3.JPG
    relay-driver-holding-3.JPG
    89.7 KB · Views: 83
The spike you are seeing is the charge held on the capacitor.
It cannot be higher than the emf you decide to adopt.
15V if your use a 3pin 7815 or 12V or rectified Ac.

It's that current spike that pulls in the relay a little faster if emf is >coil voltage. But don't worry about it. It works.
It also works with emf = coil voltage. The cap helps meet current draw when the transistor first switches ON.

BTW,
I adopted this on the channel selector and the mute relay of the Mesmerize. The cap charges fast enough between the break of the break before make of the channel selector switch.
RC = 300ohms * 33uF = 10ms.
5*RC for >90% charging is just 50ms.

post16, r1 is too biug.
A transistor switch works by ensuring the transistor is saturated. That requires the base current to be ~ 10% of the collector current.

If you know the relay coil current then the base current target should be ~ 10% of this.
a coupe of k should be about right. You can go to as much as base current ~=20% of collector current, but not less than 5%.
Don't use hFE for this !!!!!
 
Last edited:
Hello

I am looking for a clever/simple/few parts technique of activating a relay and then dropping off the current to a "holding" level to minimise the energy spent on the coil.
Wow..all these fantastic, re-inventions of the square wheel when the "round" version has been around for decades.
Have you ever heard of an "Economy resistor" invariably used with larger contactors?
Economy Resistor in DC Coils | Electrotechnik - The Website on Electrical Engineering

Frank
 
The spike you are seeing is the charge held on the capacitor.
It cannot be higher than the emf you decide to adopt.
15V if your use a 3pin 7815 or 12V or rectified Ac.

It's that current spike that pulls in the relay a little faster if emf is >coil voltage. But don't worry about it. It works.
It also works with emf = coil voltage. The cap helps meet current draw when the transistor first switches ON.

BTW,
I adopted this on the channel selector and the mute relay of the Mesmerize. The cap charges fast enough between the break of the break before make of the channel selector switch.
RC = 300ohms * 33uF = 10ms.
5*RC for >90% charging is just 50ms.

post16, r1 is too biug.
A transistor switch works by ensuring the transistor is saturated. That requires the base current to be ~ 10% of the collector current.

If you know the relay coil current then the base current target should be ~ 10% of this.
a coupe of k should be about right. You can go to as much as base current ~=20% of collector current, but not less than 5%.
Don't use hFE for this !!!!!

Hi

I have used a coil with R=300Ohm (approximately) and inserted a series resistor of 560R, and tried a few caps. Here are pics with 470uF and 100uF - oscilloscope probe on the coil (working off battery so not earthed). Resistor divider now feeds 1.6mA to base.

Edit: on initial power up, the relay remains in the off position because there is no time to charge up the cap and create the spike that is needed. That is a problem that needs to be solved.
 

Attachments

  • IMG_20120920_111244s.jpg
    IMG_20120920_111244s.jpg
    291.5 KB · Views: 37
  • IMG_20120920_111356s.jpg
    IMG_20120920_111356s.jpg
    316.2 KB · Views: 46
Last edited:
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.