April 1967 Electronics World
Wax nostalgic about and learn from the history of early electronics. See articles
Electronics World, published May 1959
- December 1971. All copyrights hereby acknowledged.
Relays are another topic that never goes out of date. Even with the advent of fully solid state relays that use semiconductors in the conduction path, there are still many applications that only mechanical contacts can satisfy. Very high power, whether it be high voltage, high current, or both, cannot yet be handled by semiconductors - at least not economically. Yes, there are switching diode arrays that can handle very high powers, but they are typically very expensive. This article is one of a group of articles about relays in the April 1967 Electronics World. All will eventually be re-posted here on RF Cafe.
Here are links to the other relay articles: Operate and Release Times of Relays, Reed Relays, Time-Delay Relays, Finding Relay Operate and Release Times, Arc, Surge, and Noise Suppression
By Jerry E. Elpers
Solid-State Products, Potter & Brumfield (Div. American Machine & Foundry)
The author is presently employed as Solid State Products Sales Manager. Prior to this, he did circuit design in the Solid Stale Switching Group. He holds a BSEE from Evansville College, Evansville, Ind. and has done graduate work at Purdue. He is a member of IEEE and Sigma Pi Sigma.
Factors to consider in selecting a relay that produces a predetermined delay. Included is data on thermal, motor-driven, pneumatic, RC, slugged, hydraulic, escapement, and solid-state types.
The fantastic growth of the field of automatic industrial control has increased the demand for new and more versatile devices to perform the basic electrical switching functions required. The use of time-delay relays has grown rapidly to keep pace with the demand for the basic function which they can perform: that of obtaining a predetermined delay from one switch operation to another.
Time-delay relays perform in a manner quite similar to a standard relay in that they have contacts that open and close when power is applied and removed from the input terminals. The basic difference is that a delay is incorporated into the contact opening or dosing. Time-delay relays are used in a wide range of applications: from determining how full your coffee cup will be when you put a dime in a vending machine, to shutting off the cutting oil on a milling machine.
The most popular time-delay relay is the delay on operate, or de-energization, in which the normally open load switching contacts transfer at a predetermined time after power is applied to the input. The contacts drop out immediately upon removal of the input power (Fig. 1A).
Often a time delay on release, or de-energization, is required. In this case the normally open load switching contacts operate immediately when the input power is applied and remain in this position as long as the input power remains "on". Upon removal of this power the timing begins, and after a predetermined delay, the contacts drop out (Fig. 1B).
Several variations on these two basic timing modes are used, such as interval "on", automatic recycle, combined "on" and "off" timers, and sequence timers. Many of these can be made by simple connections of the two basic types.
A selection of typical solid-state time-delay relays, many of which provide an adjustment for the amount of delay provided.
Factors to Consider
There are many types of time-delay relays available which will provide the timing action desired, including thermal. motor-driven, pneumatic, RC circuit, solid-state, slugged, hydraulic, and escapement.
There are many factors which must be considered when choosing one of these time-delay relay types. Consideration should be given to how each fulfills the following criteria: accuracy, reset time, repeatability, load-switching capabilities, price, life, mounting configurations, size, length of delay, temperature effects, and adjustable or fixed time delay.
Also, various time-delay relays have certain peculiarities in their operation which should be determined in order to select the type that will do the job reliably and economically, Some of these peculiarities are covered below.
Thermal Time Delays
The basic operation of this timer takes advantage of the difference in the thermal expansion of two metals. A bi-metallic element is placed in close proximity to a heating element, and when power is applied the bi-metal deforms and closes or opens a contact. The time required for the contact to operate is generally determined by the physical characteristics of the bi-metallic strip and the amount of power applied to the heating element.
Thermal time-delay relays are usually used where a time delay on energization is required and the accuracy of the time-delay period is not critical. One manufacturer states an over-all accuracy of ± 30% for a miniature or octal plug-in timer, with delays available from 2 to 180 seconds. Another manufacturer gives accuracies of ± 20% from 0.75 to 1 second, ± 15% from 1 to 4 seconds, and ± 10% for delays up to 360 seconds. This is also an octal plug-in type time-delay relay.
Contact forms are usually limited to s.p.s.t., N.0. or N.C. with ratings that normally do not exceed 5 amperes, 115 volts a.c, resistive (100,000 operations). Since the thermal time delay uses the I2R heating effect, the device is somewhat sensitive to input voltage variations. Voltage variations of ± 10% will change the delay period approximately ± 5%.
The biggest drawback to the thermal time-delay relays is their long reset time (the time required for the contacts to open and to achieve an appreciable percentage of the nominal delay time on the succeeding cycle of operation). This reset can be as long as 50 to 200 percent of the operate delay in order to achieve 80% of the nominal delay period on the next cycle.
One thermal-delay relay manufacturer recommends the use of an auxiliary relay to overcome this reset time problem. This unit uses two sets of contacts, one that closes at the end of the heating interval which pulls in the auxiliary relay and interrupts the input power. The cooling interval then begins, after which the second set of contacts drop out completing the load-switching circuit. Using this method, approximately 85% of the nominal operate time is achieved on the succeeding cycle.
Most of the better known thermal time-delay relays are of the plug-in type and range in price from $2 to $20.
Fig. 1. (A) Delay on energization. (B) Delay on de-energization.
Pneumatic or Air-Operated
The term "pneumatic" (meaning air-operated) immediately indicates the basic principle of operation of this type of time delay. A mechanism is used in which a controlled amount of air under pressure is displaced from one place to another. (This can be a unit isolated from the surrounding air, in which the air is displaced from one chamber to another, or where the air is drawn from or dispelled into the atmosphere.) A typical unit uses a diaphragm, a coil, a plunger, and an orifice. When power is applied to the coil. the plunger (which is mechanically tied to the diaphragm) is drawn into the coil. The rate of plunger movement is controlled by the rate of air escape from the diaphragm, which is controlled by the orifice adjustment. When a predetermined position is reached, a set of contacts operate as a result of the plunger movement.
The pneumatic time-delay relay has been in use for 25 years and has a well-established reputation in heavy industrial applications. Several manufacturers have units available which lend themselves to these types of application. They are available for control-panel mounting, have screw terminals, and are designed for use in severe industrial environments. This type of time delay is available as an adjustable unit: some with time-calibrated dials and some with screwdriver-slot adjustments. The repeat accuracy is usually ± 10% and the reset time is approximately 25 milliseconds (this reset also applies in case of power interruption).
Units are available in both time delay on energization, delay on de-energization, and also with both of these functions packaged in the same enclosure. Some can be converted in the field from delay on energization to de-energization by a simple mechanical change. Contact ratings are available up to 20 amperes, 120 V a.c., 60 Hz, resistive (100,000 operations life).
Time-delay periods are available from 0.2 second up to 30 minutes from one manufacturer and 0.050 second to 3 minutes from another. The temperature range of operation is in the area from -50° C to +65° C. Supply voltages of 6 V a.c. to 550 V a.c., 60 Hz and 6 V d.c. to 250 V d.c. are available. Input power requirements range from 5 to 8 watts. Prices range from $18 to $100.
Many of the pneumatic units are fairly large due to the space required for the mechanical mechanism, although some smaller versions are also available with lower contact ratings (10 amperes) and shorter delays (180 seconds) at a higher cost. Pneumatic delays can cause a problem in application where a dirty atmosphere exists. Any clogging of the orifice will cause changes in the delay period.
A synchronous motor is normally used in motor-driven timers to drive a gear train which controls the load-switching contacts. When power is applied, the movement functions until a predetermined time has elapsed, at which time the output contacts are switched. This timing method depends upon the input line frequency for its basic accuracy, in a manner similar to a standard 12-hour, 120-V a.c., 60-Hz clock. The majority of these timers use a magnetic clutch in conjunction with the clock movement which serves the function of engaging the movement when power is applied, and allowing it to reset when the power is removed. The unit is reset by a return spring when the clutch is released.
The time-delay period is set on these units by a pointer on the front or top of the relay (this may be continuously adjustable or in increments). A second pointer is usually used to indicate the elapsed time. The setting accuracy of the continuously adjustable timers is generally ± 0.5% of full-scale, and the repeat accuracy is ± 1% of full-scale, or better. The reset time is proportional to the time required to reset the spring-loaded mechanism and will be less than 500 milliseconds, depending on time setting relative to full-scale.
Motor-driven time-delay relays are available in both delay on energization and delay on de-energization. The load contacts of the delay on de-energization operate immediately when power is applied to the clutch and removal of clutch power starts the timing interval. When this interval is completed, the contacts drop out. The delay on de-energization will reset when a momentary power loss occurs.
Delay lengths of 5 seconds to 60 hours are available from one manufacturer and 5 minutes to 5 hours from another. Life ratings of these timers range from 500,000 operations for one manufacturer to a contact life rating of 5 million operations for another. Life is usually limited by clutch failure rather than contact life. Input power ranges from 5 to 15 VA, including clutch coil and motor. Contact switching ratings up to 15 amperes continuous are available. Most of the units available are for 120/240-V a.c., 60-Hz operation. D.c. units are not generally available. The temperature range of operation is -20°C to +50°C. Prices range from $5 to $50.
The majority of the motor-drive time-delay relays have some kind of adjustment feature, some of which are for front-panel mounting, with a knob to set the time-delay period and are intended for industrial control-panel use. Others are available with pointers that can be set to adjust the timing in increments and are not intended for applications requiring front-panel mounting.
Delay Slug Relays
A time delay can be produced on a telephone-type d.c. relay by placing one or more shorted turns around the magnetic circuit (usually the core) so as to produce a counter-m.m.f. which retards the build-up of the operating flux, and upon de-energization provides an m.m.f. to retard the collapse of the flux. This shorted turn. or turns. is called a slug. Usually it consists of a copper collar on the core of the relay. In some designs, a copper sleeve is used over the full length of the core, and the coil is wound on the sleeve.
The principle of operation of the slug is as follows: When the relay coil is energized, the flux build-up passes through the slug and by self-inductance produces an m.m.f that opposes the coil m.m.f. This opposing m.m.f. delays the build-up of the magnetic field in the air gap to a strength that will cause the armature to close. The time delay on drop-out occurs in the opposite manner. When the relay coil is de-energized, the field starts to collapse, thus inducing a current in the slug that provides an m.m.f. oriented so as to sustain the magnetic field and thus delay the drop-out.
Pickup delays up to 120 milliseconds and drop-out delays up to 500 milliseconds can be achieved by the use of slugs. The delay time will vary due to mechanical wear over life and ambient temperature and this type is not intended for high-accuracy applications. Slugged relays are not generally an off-the-shelf item and are available only on a special-order basis from most manufacturers of telephone-type relays.
Various methods have been used to devise time delays using combinations of resistors, capacitors, and relays. All of these circuits use the basic principle of charge and discharge of a capacitor and one of the simplest circuits using this principle is shown in Fig. 2A.
When the switch is closed, the capacitor charges toward the applied voltage and, after a period of time determined by the resistance and capacitance (assuming the relay coil resistance to be very large compared to the resistor), the relay pulls in. When the power is interrupted the capacitor discharges through the relay at a rate controlled by the size of the capacitor, the inductance, and the resistance of the relay coil. This circuit does not produce an accurate time delay since the time is dependent upon many factors. Variations in the resistance, the capacitance, the input voltage, and relay pull-in voltage will cause changes in tile time-delay period. Delay-period accuracies of ± 30% are to be expected. The reset time is also long, due to the high-resistance, capacitor-discharge path. In practice these timers can be purchased with delays up to 30 seconds and in both a.c. and d.c. versions.
Variations of this basic circuit are available which use additional components to increase the accuracy and speed up the reset time. One popular circuit uses a neon bulb to sense the voltage level on the capacitor and a photo-resistive cell to drive the relay. The photocell offers a low impedance in series with the relay when the neon bulb fires and causes the relay to pull in (Fig. 2B).
This circuit has several advantages over the previous circuit in that the neon bulb senses an accurate voltage level and does not rely on the relay pull-in voltage for the accuracy of the delay. Also, the relay drops out immediately when the power is removed, and the timing can be reset with another set of contacts across the capacitor if desired.
This circuit is used in several commercially available time-delay relays. The units are generally adjustable (a potentiometer is used in series with the resistor) , and the enclosure is all octal plug-in type. The accuracy of this unit is usually stated as ± 10% over a limited temperature range. Relay outputs on these timers are usually d.p.d.t., 5-ampere, 120-V a.c, resistive.
The latest entries into the time-delay relay field are the solid-state types. There are presently two basic types available using entirely different principles of operation. One manufacturer is marketing a relay that uses an accurate oscillator and a counter to perform the timing function. An output from this oscillator is initiated when power is applied and is fed into an amplifier circuit. This amplifier shapes the pulses and reeds them to a magnetic-core counter. When a predetermined count is reached. the output load switching contacts are operated through a logic circuit. The time delay period is determined by the logic and by how many counters are used after the oscillator.
This time-delay relay is intended for applications where a high degree of accuracy is required (± 2 %) . The standard operating voltage is from 18 to 82 V d.c, It has either a relay or a solid-state output and requires 0.4 sec to reset the delay to the stated accuracy. Enclosures are available with the popular plug-in feature or with hook solder terminals and a side mounting bracket. This unit is expensive as time-delay relays go and sells for $100 or more.
The most popular circuit used in solid-state time-delay relays employs the RC charge principle, mentioned previously. The reason for the popularity of this method stems from the use of the unijunction transistor.
The unijunction transistor has the inherent ability to offer a high input impedance to the capacitor voltage until a predetermined voltage is reached. At this point the device fires and discharges the timing capacitor.
The circuit of Fig. 2C can be used to illustrate this operation. When the switch is closed, the capacitor charges at a rate controlled by the RC product. At a voltage level controlled by the unijunction transistor, the capacitor discharges through the relay coil and causes the relay to operate. This pulse is only momentary and a set of relay contacts serves to latch in the relay.
The unijunction circuit has several advantages over the neon-bulb circuit in that the firing level is proportional to the input voltage. Therefore, any input voltage variations are compensated for by a proportionate change in firing voltage. Other devices call be added to this circuit to overcome some of the deficiencies of the previous circuits. Zener diodes are added for better compensation of time changes due to input voltage variations. An SCR can be added to give the relay more pull-in and hold-in power.
Fig. 2. (A) RC circuit may be used to provide time delay. (B) Neon bulb and photocell or (C) unijunction transistor may be used.
The circuit shown in Fig. 2C also has the drawback that the capacitor will discharge through the relay coil immediately if the power is interrupted and this causes the relay to pull-in momentarily. The addition of an SCR eliminates this problem.
One problem peculiar to the solid-state time delays has been that of line transients, particularly on time-delay relays used on 120-V a.c. line voltage. These transients, or momentary overvoltage conditions. are common on a 120-V a.c. line. They can be produced from a wide variety of conditions including lightning striking the line, switching of inductive loads on the line, or making and breaking of the transformer input supplying the time-delay relay. These transients are usually not a severe problem because they do not contain much energy.
Solid-state time-delay relays used on a 120-V a.c. line normally employ a silicon rectifier at the input in order to produce the required d.c, voltage for time-delay operation. This rectifier is susceptible to the energy contained in these transients if the voltage is allowed to exceed its reverse rating. Two methods are commonly used in better solid-state units to suppress these transients. One method is the use of a controlled-avalanche rectifier for this input rectifier. When the reverse input voltage exceeds the rectifier reverse voltage, the device avalanches and dissipates the transient energy. Special selenium breakdown devices are also used which can withstand even more energy. Transient problems are better understood now than in the past and adequate protection can be provided.
Since solid-state time delay relays use transistors in the timing portion, and transistors have long life, a very reliable timer can be built using these devices. :Many of the units available have an integral relay for load switching; typically, d.p.d.t., 10-ampere, 120-V a.c. resistive. In this case the life is limited to the life of the relay, usually10 million mechanical operations.
Several manufacturers market an all-solid-state timing module to drive an external load-switching relay. In this case, the timer life is nearly infinite and is very useful for applications where high cycle rates are experienced.
Solid-state units are available in fixed delays, internal potentiometer adjustable delays, and external resistor adjustable delays. The external resistor can be a potentiometer mounted remotely and wires run to the delay unit.
A wide range of mounting configurations is available, including screw-terminal dust covers, plug-in types, hermetically sealed military types, and panel-mounted types. Units are available in d.c. voltages from 12 to 100 V d.c. and 24 to 240 V a.c., 60 Hz and require approximately 3 watts of power. The timing range is usually limited to delays from 0.10 second to 5 minutes on commercially available units. Delay accuracy varies from one manufacturer to another and can be as good as ± 5 % over the temperature range from -40°C to +55°C, and a voltage range of ±10%.
Several manufacturers are marketing knob-adjustable solid-state time-delay relays with time-calibrated dials. The reset time varies, but is usually fast-from 40 milliseconds to 100 milliseconds - depending upon the circuit configuration. Time-delay relays are available in both delay on energization and on de-energization. (The delay on de-energization requires an auxiliary source of power during the delay period to hold in the load switching relay and cannot be used for a momentary power failure delay.)
The solid-state time-delay relays are available in small enclosures and range in price from about $10 to $60.
There are two types of driving mechanisms used in stepping relays (often called stepping switches): the indirect and the direct.
When the armature-pawl combination acts directly on the ratchet under the power of the electromagnet, the device is said to be directly driven, as shown in Fig. A. When the pawl acts on the ratchet wheel from force stored in a drive spring, the mechanism is said to be indirectly driven. An example of this method is shown in Fig. B. The indirectly driven system is the most commonly used. The spring-driven system is more consistent in performance, more efficient, and capable of faster stepping than the directly driven type, besides having a longer operational life.
In an indirectly driven unit, when the proper voltage and power is applied to the motor magnet coil, the armature is attracted and holds the drive spring in the "cocked" position. When the coil is de-energized, the energy stored in the drive spring pushes the pawl against a ratchet wheel tooth, causing the wiper assembly to take a step. Repetitive pulses will cause the switch to take as many steps as the number of discrete pulses received. The length of time the circuit is closed (and opened) in a series of fast pulses is critical.
Self-interrupted operation, such as shown in Fig. B, is used to step the switch rapidly from one point to another without the use of discrete pulses from outside sources. In this method, a circuit is closed to the coil through a set of interrupter contact springs that are opened by an arm of the armature before it is fully seated. Breaking the coil circuit causes the armature to fall away, driving the wiper assembly one step and simultaneously reclosing the interrupter contacts. The armature is again attracted, re-cocking the switch and thus causing a re-opening of the interrupter contacts. The switch runs self-interruptedly until the circuit is again opened.