ECN Forum Forums General Topics of the Trade General Discussion Area 12 volt motor delay

If you add a 50 cent 555 timer to this you get a lot more predictable time constant. It isolates the gate trigger level from the equasion and makes the turn on a digital event.

With the increase of posts of this nature (TDC's), maybe it would be of some benefit to draw a few simple + safe Time Delay control circuits, along with some other topics (suggestions???), and post these in the Technical Reference area.

I have a few more weeks of "Free Time", so now is the best time to do so!

p.s.: "JoeTestingEngr"

What "flavor" of CAD Application do you use?
I use AutoCAD (R14 and 2000).

If you have anything you would like to contribute to the ECN Technical Reference section, let me know.

Same goes for anyone else!

Contact me via E-mail, or by replying to this thread.
If you do contact me via E-mail, please include your ECN user name, along with "From the ECN Forums", in the "Subject" heading of your E-mail message. Otherwise, the message may be discarded without being viewed (message filters and such).

Send messages to me at this address:

setelectric at pacbell dot net

(just change the phonetic stuff for the "@" _at_ sign and a "." _dot_)

Scott35

It should probably be pointed out that nothing we make up ourselves will be a U/L listed product. If this is a product or service you sell it will be at your own peril.
Hopefully we all understand the separation requirements of 725 well enough to keep line voltage away from low voltage and can do this safely. Personally I like identified optoisolated solid state relays for this isolation, along with appropriate barriers.

Greg, I'm not sure I could go along on the 555. I use them alot but I can't figure an easy way to use one for BigB's application. I have never used one for timing at the same time power was applied with a trigger event. It's the same with the LS7210 that I like to use for switched timing. It too, has a negative trigger and I wouldn't trust it at power up. By the time we add all the goodies to make the timer fit the app, I'll bet the original SCR schemes we used are as good or better.

I thought of other means to do what he wants, using bipolars and TMOS (BS-170) transistors. Then I started wondering how stiff his supply might be and what would happen when the relay picked. Now we have to add hysteresis to account for that. I'm betting that BigB just needs a little time for things to settle down and wouldn't mind a little extra. I keep coming back to the SCR method being cheap, simple, reliable, and probably the best.

Scott, promise not to laugh at me. I use Generic Cadd 6.0 It was the last version before AutoDesk gobbled up Generic Software. I bought 3 packages for work and a co-worker and I bought it for home at $65.00 each. It was our department's first use of CADD. Since then, everyone else has switched to the latest version of AutoCAD. They try to get me to switch over but I resist. The AutoCAD 2004 in my computer isn't even working now because it's confused over a license file that must have been moved. I just export files as DXFs if we need them for AutoCADD. Fills are the main problem. I'll send you a sample so you can see what I mean.

Greg, you make a fine point. I'll freely admit guilt. The NEC or U/L don't even cross my mind when I'm thinking about 12 volt relay TDO circuits. That could certainly come into play, depending on the application.
Joe

I am doing this with a 4060 counter but my delay time is a big number, up to all day. It is a lot more parts tho.
A .001uf cap and a 100k resistor settles things down prety well on POR.
I'm working on a solar pool/spa controller. Folks are saying PIC but I still like CMOS, particularly when it is a fairly simple truth table. Still looking for a clock but I am seriously leaning toward a garden variety alarm clock with 2 events. The more I work on this problem the simpler the logic is actually becoming.

The application is automotive.

"It needs to be rated at least 30 amps 12 volts DC and be able to delay starting for one to two seconds."

If you can tolerate a delay before the relay opens, the brute force method is to hang a large capacitor across the coil terminals with a resistor in series with the wire driving the relay coil.

R(ohms) x C(farads) = Time Constant(seconds)

My guess is that you will need about 1 to 2 time constants to reach the pull in voltage of the relay. So R times C should equal about 1. Try 20 ohms at 2 Watts and 47,000 uF at 63 VDC.

LarryC

http://www.alarmcontrols.com/

Click on products then Timer Modules and look for the UT-1 Timer. Operates 6-18 VDC and has an on board 3 amp relay. Trigger a 30 amp with that and your good to go. I've got some squirreled away in my basement goodies if you cant find a distributor.

LarryC, I agree with your estimate, except that the capacitor will probably want to be closer to 100,000 uF. For the benefit of the curious, I offer a few more details.

A typical automotive relay, Omron G8W, has an 88-ohm coil and an 8-volt must-operate spec, so with the 20-ohm series resistor it will operate reliably down to 9.8 battery volts.

We have to use the parallel value of the coil resistance and the series resistor to obtain "R" (88 || 20) = 16.3 ohms. The actual RC time constant is then 16.3 ohms x 100000 uF = 1.63 seconds. But, the operating delay time will depend on the actual relay pull-in voltage, the battery voltage, and component value tolerances.

With a 13.7-volt battery we get a steady-state coil voltage ("Vcoil") of (13.7 x 88/(20+88)) = 11.2 volts. Assuming nominal R and C values and a typical pull-in value ("Voperate") of 6 volts, the delay time ("T") is calculated as:

T = -RC x ln ((Vcoil-Voperate) / Vcoil)

T = -1.63 x ln ((11.2-6) / 11.2)

= 1.25 seconds

(The ln () function is "natural logarithm." It's on your scientific pocket calculator.)

Here's the bad news:

1) The operating time will vary with battery voltage. At 11 volts, the same circuit will take 1.8 seconds.

2) The capacitor must be rated at least 16 volts. A 100,000uF 16V capacitor is big (about 1.5" diameter and 2.5" long) and expensive (about $10-12 new).

3) As LarryC mentioned: Assuming the relay is released by simply opening the circuit, the R-C product becomes (88 ohms * 100000 uF) = 8.8 seconds. Assuming the relay drops out at about 2 volts, the release time is:

T = -RC x ln (Vrelease / Vcoil)

T = -8.8 x ln (2 / 11.2)

= a whopping 15.2 seconds!

You can see why the solid-state alternatives, such as the one suggested by TwinCitySparky, are so desirable.

JoeTestingEngr, the key to using the 555 in an application like this is to forget that it's a timer and just use it as a really good inverting Schmitt trigger with a nice, high-current output. Tie the trigger and threshold pins together as an input, and leave the discharge pin unconnected. The output goes high when the input is below 1/3 VCC and low when the input is above 2/3 VCC.

Oh, and I used GCADD for years before finally biting the bullet and buying AutoCAD LT Didn't they have a "Miss Manners"-like character answering FAQs in their newsletter?

Edit: Didn't get those italics right the first time... Or the second...


[This message has been edited by John Crighton (edited 06-29-2006).]

John, this is where I run into the problem with the 555 here. If we do it the usual way, 2 & 6 to ground via a cap and + supply via a resistor, skipping pin 7, we power up with a trigger. We won't be at 1/3 Vcc, and might not be at theoretical 0 on the cap either. If there is a cap on the control voltage pin 5, it is charging. If not, noise could be a factor. So we effectively start out high on the OP and are holding off for a low. This is intuitively backwards for our SCR trigger.

So if I want the output to be better for my SCR timing, I would probably hook the cap to the supply and the resistor to ground. The cap will now be charging down to trigger while my control voltage stabilizes up. The SCR will get a clean firing pulse but when? Do we really know the charge on the cap at power on? Won't the supply voltage effect the time. It's not that I'm saying that this won't work, just that I doubt there will be more precision than a simpler circuit.

I started out using Generic Level 3 CADD in the mid-80s. I think I used 5.0 for a little while too. I still love it. Stuff happens if I forget to assign a printer to LPT1 before a dotplot but I'm lucky that ancient printer drivers still work. I just made 4 circuit boards this week, using it and the DynaArt system. But I had totally forgotten that news letter. I'm going to bounce it off of my EE brother, who got me started on Generic Cadd, to see what he remembers.

Joe

[This message has been edited by JoeTestingEngr (edited 06-29-2006).]