Designing your own delay circuit

[Vestax]

Designing a turn on/off delay circuit is quite simple and will cost you no more than $5 to make. All parts can be picked up from Radio Shack or Fry's if you have them as well as Parts Express. This will cover the mechanical aspects of it - theoretical topics can come later. If you suffer from pops on your amps or any other components, this will help you eliminate it, but it does not work in all cases.

You will need the following parts..

1 12 volt SPDT 30 amp relay
1 NPN Transistor
1 Diode (same one installer uses in alarms)
1 capacitor with either different values of 100uf to 1000uf, polarized electrolytic, 16 volts or higher
1 few small fuse holders (for safety issues)

Start off with your turn on lead or ignition switch from your head unit. Wire in series a diode with the stripe facing towards the amp. (you can put a 1 amp fuse here in case you short out the turn on lead.)

Next, wire the capacitor after the diode and ground it. The higher the impedance on the capacitor, the longer it'll delay.

After the capacitor, the transistor is straight forward, wire the turn on lead to B, 12 volts to C, then E out to the relay.

For the relay, use 86 as the input from your turn lead lead, and ground 85. Wire 12 volts to 87 (make sure you fuse this) and 30 out to your components. When the coils are energized, 87 and 30 will be connected and will send current to your devices.

I forgot where I found this saved imaged file from, but it's the best one I could find...

http://memimage.cardomain.net/member_images/7/web/151000-151999/151662_121_full.gif

[npdang]

Wow, thank you for the tutorial!

[MarkZ]

You can get away with a smaller relay, saving both physical space and current draw through the transistor (due to the coil resistance). Radio Shack carries a 1A and 5A relay and I think a 3A one too. Also, you can probably save space and parts by using a single fuse off the 12vdc source instead of two, as it's completely unnecessary to have to provide 30A to a remote turn-on lead. Generally, you'll be under 1A unless you have more than 10 devices to turn on.

[Vestax]

You can get away with a smaller relay, saving both physical space and current draw through the transistor (due to the coil resistance). Radio Shack carries a 1A and 5A relay and I think a 3A one too. Also, you can probably save space and parts by using a single fuse off the 12vdc source instead of two, as it's completely unnecessary to have to provide 30A to a remote turn-on lead. Generally, you'll be under 1A unless you have more than 10 devices to turn on.

Good point, a smaller relay would do just fine.

[chad]

Using a TIP dual darlington device you could eliminate the relay too. I think a TIP102, could be wrong.

[honfatboy]

D'oh!

I thought (hoped) this would be an explanantion of the very mysterious and fabled mechanical time delay circuit...

[fearlessfly28]

thanks for the tip...

u can do alot of shits with these li'l relays.

[lightspeed]

i need some help with this.

I am using my stock GM truck H/U (not Bose). I will have two amps to turn on, and I have been told that the remote wire will turn on both just fine without a relay, AND I have been told to use a relay. What's the real deal? i am fine with doing all of my install, but this starts to get into confusing electronics for me.

Any and all help is much appreciated.

[03blueSI]

^^^^ This circuit would not be needed for you. This circuit is for the situation when you have a turn on or off pop. For you all you would need is the relay. Just wire 12v constant to 87, ground 85, remote out of HU to 86 and 30 would go to your amps. With a stock head unit I would probably do it to be safe. With aftermarket you can usually do 2 amps safely, 3 amps is hit or miss depending on the HU and draw from the amps and any more than 3 amps I would use a relay.

[chad]

Amp triggers are voltage sensing designs with a very high input impedance. If it puts out 12V and the wire's not broke then I'm sure the requirements for a relay would be nil. Now if the OP were trying to turn on fans/bling/etc it would be a different story.

Chad

[lightspeed]

Cool. I guess what I wanted to avaoid, was having the turn on wire work for some time and then burn out the transistor or whatever inside the HU at a later date.

I was thinking about using Navone's solid state switch:
http://www.davidnavone.com/detail.asp?PRODUCT_ID=N-TRIG-LO

[Whiterabbit]

why has noone mentioned wiring a capacitor across the relay coil?

on an unrelated note, does anyone know the ballpark minimum voltage is to hold the coil closed on a standard bosch relay?

[ca90ss]

I built one of these to fix a turn on noise from my amp and I'm having a little problem with it. It works fine as a turn off delay but it doesn't delay the turn on at all which is the exact opposite of what I need it to do. Does anyone know how to make one of these that will delay the turn on?

[drake78]

good thread

[ca90ss]

I built one of these to fix a turn on noise from my amp and I'm having a little problem with it. It works fine as a turn off delay but it doesn't delay the turn on at all which is the exact opposite of what I need it to do. Does anyone know how to make one of these that will delay the turn on?

:confused:

[thechris]

not the best circuit, but it does remove a resistor.

the more common way to do this is with a common emitter amplifier, or a common source amplifier.

the more common delay on/off will be based on the attached pic.

in this circuit, at turn on, C1 is charge via R2. if optional diode D2 is installed, then C1 is charged by R4||R2. This allows inrush limiting if C1 is made large. the D2+R3 parts will cause the relay to turn on fast.

as C1 charges, it will reach the Vth of the MOSFET M1, shortly after, it will be able to turn M1 fully on, and K1 will engage.

When ACC is removed, C1 will discharge through R2+R1, eventually reaching Vth and turning off K1. Again, addition of the D2 diode will allow a faster turn off action.

Turn on and turn off times will not be equal in this circuit because the gate of M1 will turn on around 4V, but will charge to 12V. as such it must discharge 8V to turn off, but only charge 4V to turn on.

Diode D1 is the standard flyback diode designed to allow current flowing in K1 to natually stop flowing, instead of suddenly trying to break the magnetic field and causing high voltage spikes across M1.

this circuit can be simplified, or made more complex to acheive desired results.

[thechris]

for the case of a turn on delay with no turn off delay, i suggest the attached.

when Acc is engaged, the gate of M1 will charge to 12V, and M1 will engage*, but M2's gate is delayed by the time required to charge C1 to Vth through R1. Once M2 starts to conduct, M1 will also turn on, and K1 will engage.

when Acc is disengaged, the gate of M2 will stay high, discharging through R1+R2. M1's gate will quickly drop below Vth if R1 is much larger then R2. this will quickly turn the circuit off.

because this circuit uses mosfets, and is a common source design, the transistors will dissipate very little heat.


suggest making Vcc*R2/(R1+R2) < Vth, which means
Vcc/Vth < R1/R2 + 1
or:
R1 > R2*(Vcc/Vth - 1)

suggest making R2 10kohm. if a 2n7000 is used, Vth will be around 2V. meaning giving R1 > 10k(12/2 - 1), R1 >50k.

time delay will be based on the size of R1 and C1:
Vgs = Vcc - Vcc e^-t/R1C1
for Vgs = Vth
Vth/Vcc = 1 - e^-t/R1C1
1 - Vth/Vcc = e^-t/R1C1
ln(1-Vth/Vcc) = -t/R1C1
t = -R1*C1*ln(1-Vth/Vcc)

since we know R1 > 50k:
C1 > -t/(R1*ln(1-Vth/Vcc))

for a sample 1 second delay:
C1 > -1/(50k*ln(1-2/12))
C1 > 109uF

Hmm, 109uF is attainable, but if I wanted say, 60 seconds of delay, this would not bode well, and I'd be forced to raise R1, which i can do without much penelty.


* rather will be engageable.


edit -- and if you wanted a turn off delay with minimal turn on delay, you could place M1, M2 in parallel instead of series.

Keep in mind, that this circuit should not be turned on/off at a high rate, as C1 will need to discharge. for such apps, a speedup diode as shown above would be needed to clear C1 at a faster rate. if a short turn off delay is acceptable, M1 might be omitted and the speedup diode used.

[radioflyer07]

Forgive my noob-ness.... but what is the purpose of this circuit?

[Whiterabbit]

to eliminate turn off pops or turn on pops that are caused by an entire series of equipment that powers on in a "less than optimal" order.

not uncommon in complex systems.
http://www.selectronic.fr/article.asp?article_ref_entier=70.2750-5US

of course, why should anyone pay $200 euro when a creative installer can buy $5 worth of parts above (and probably has 90% of the parts laying around for free anyways) for each piece of gear they have and start swapping out caps for timing till they have the order right anyways?

And even farther, chances are good that there are only gonna be two pieces of gear in the car that "fight" over turn on and turn off speed anyways (one signal processor, and all amps lumped together), requiring only one circuit. But again, multiples are five bucks of parts. max.

[thatvan]

Here is a Time Constant calculator for figuring out resistors and caps on a timed turn on delay.

http://www.daycounter.com/Calculators/Capacitor-Energy-Time-Constant-Calculator.phtml


A TIP102 would work for use without a Relay. Just remember to use a heatsink.



Turn off delay is a little more complex. The best I ever found is to use an SCR, cap, Darlington and some resistors.

Remember to always use a Diode across all relays to prevent the contacts from burning.

I'll post the circuits I use when I find them.

[Whiterabbit]

question:

why should I make these more and more complex turn off delay circuits if the one in post #1 works?

[thechris]

the first post doesn't have a well defined turn on delay. also, a smaller transistor can be used in the common emitter configurations I have listed. The common collector setup will discharge the cap through the base of the transistor and the reflected load. this provides relatively poor tolerance for the delay as the delay is based on the current gain of the part (which changes over temp and production by a large amount). further, the turn off switching of the transistor will fall slowly from 12V to 0V. in the common emitter configs, there will be a delay after which a fast transition from 12V to 0V will occur. This reduces the power dissipation of the part as well. the common emitter also has better low-voltage characteristics. it provides within 0.25V of the supply voltage to the coil. the circuit in post 1 provides within 1.5V of the supply voltage to the coil.

In industry, you will see the common emitter config used -- better control, better switched edge, lower losses, lower drop out.

[Whiterabbit]

and will the regular audio hobbyist notice any of these improvements?

[thechris]

if they want a turn on delay or no turn off delay, yes.

[KeepItSimpleStupid]

I'm a newbie and I'm going to try to resurrect this thread, first by asking a few questions?

1. How important is the thumpy thing?
2. Is it true that no one seems to know how to fix it?
3. Is there a device that works?
4. No one seems to understand the problem?
5. No one seems to understand the automotive environment.

I'll attack (4) for a bit:

Problem: Slow turn-on and fast turn-off required for anti-thump

Remote Lead: Some remote leads are actually regulated outputs for antennae preamps. Very tough to use as a remote line because it tends to turn off slowly. Most are relay based. Some may not have the drive (Fan out) to drive all the other devices.

Accessory Power: This one is a big gotcha. It goes through an ON/(OFF)/ON sequrnce when you start the car. The (OFF) is the off that occurs when the starter engages. When the starter engages the accessories turn off. Delay circuit needs an extremely fast reset time.

Timing ckt: Must be at t=0 when entering vehicle. Must reset quickly to t=0 when starter is turning. Once delay has elapsed, the timing circuit must again be ready for the next t=0 event.

Protection: There are +50 and -200 V spikes from the alternator. You need to take steps to protect against that.

Delay time: 0-4 seconds sounds reasonable

Status: Device should have a status LED

Turn-off Thump: Is effectively eliminated with speaker relays. Stubbon cases require muting the input much faster than the speaker relays.

Turn on thump: Fast reset of the timing circuitry is essential otherwise you'll sometimes get thumping because of the pesky ignition switch. Generally delaying the remote lead works. Sometimes a load resistor must be placed on the remote in, so it will decay rapidly.

Thumping is a fault of some component in the loop The ignition sequence being a troublesome input.

Minimum Design: Accept Remote in, Gnd, and Remote Out. Solid state protected output to 1 Amp, short circuit protected. LED indicator, Adjustable delay.

Design Elements Missing: Independent comparator based trigger, Remote Out comes from Battery Terminal, Relay vs Solid state output. Low voltage drop out.

What would an installer/Doityourselfer be willing to pay for such a device (Delay of the remote signal), if one existed? Speaker relays not included, Fan out relays not included, a 3 terminal device without low voltage drop-out.

Comments anyone?

[Ge0]

Protection: There are +50 and -200 V spikes from the alternator. You need to take steps to protect against that.


Actually automotive conducted transients can get much higher than that. Newer standards state fast transistion rate positive transients can climb as high as +300V and negative transients up to -400V. Mind you, these have a weak energy content, but repeat them a few thousand times a second and this can add up. Slower moving, high energy content, transients can exceed +100V.

This might seem overwhelming at first, but a few precautions on the front end power conditioning circuit can migitate their effects.

Good stuff by the way, keep it up.

The circuit can be as simple as an RC network. Use the time constant of a low pass filter network to set turn on time. Place a diode across the resistor to speed up turn off time. Mind you, timing of this circuit will vary but would be good enough for 90% of the "thump" problems one may encounter.

Ge0

[KeepItSimpleStupid]

Back in the early 90's I designed the delay system with the specifications in the .pdf document. I just typed in this info from page 3. This stuff was done before the Internet and with software packages that are not common today. The board was designed using an Amiga computer.

I had about 200-250 boards made with solder mask but no silk screen. They are not stuffed. I can't even find my test to test the reset circuitry.

The board and components were EXPENSIVE. I haven't figured out the costs today, but there is a BIG problem. The IC I used is discontinued and I only have about 10 of them. The IC's are available, I just don't know how expensive they are and I have to buy a bunch like $200 to $300 worth.

I don't know if it's possible to replicate the functionality with multiple IC's or not, but space constraints would be a problem.

It might be possible to design the board with a new technology. I did solve the reset problem and I think I could do it with a new technology, but again I'm afraid the unit would be expensive.

What should I do?

[Whiterabbit]

1. How important is the thumpy thing?
2. Is it true that no one seems to know how to fix it?
3. Is there a device that works?
4. No one seems to understand the problem?
5. No one seems to understand the automotive environment.

Comments anyone?

Plenty. Thank you for the refreshing effort into small questions. I personally think it is important to devote TONS of energy into details.

1: the thumpy thing is so, so, SO important. There are two kinds of importance placed on turn on and turn off thumps: those who have blown a tweeter because of thumps, and those who have not. Puts thumps into perspective. I've never heard an OEM system with a thump. They did the job right. Our job is to emulate their design.

2: Some of us know how to fix it, then do. Some do not, and do not have the knowledge or understanding of how to fix it. So the answer to the question depends on whether the user knows how t wire a relay and know what a transistor IS.

3: The device that works costs less than a dollar in parts and is in a linked photo in post #1. It is used by many folks.

4: The problem is understood by some, used by a few more, and a few more experience the problem without knowledge. A standard statistical distribution.

5: The only person I personally know who understands the automotive environment finds the problems we work on so mundane as to be above working on. And thank god I know that one person, because I may as well have a bone as my only electrical tool and pick fleas out of my co-hobbyists for consumption without his advice and input.

It's nice you've got an expensive complex solution. We have one that uses one non-passive component (a transistor) and would find it ridiculous to exert more energy into the problem than that. The irregularities caused by the one transistor solution will not be seen by 999 of 1000 cases, and the one that does have a problem will be smart enough to research how to use a 555 timer or other cheap solution. No $10 chip required.

Even thechris' solution is more complex than it needs to be for our uses. And that too uses no DIP components.

Ultimately, our goal takes after your screen name. We want to keep it as simple as possible.

[KeepItSimpleStupid]

3. A dollar in parts, huh? Let's see the breakdown. Besides there are elements not included that are necessary.

1. Case - about $5.00 in my case, metal lid, 2 extra holes required. Lid can be purchased separately, but for a product would could be purchased drilled and silk-screened. 1 hole in plasitc case for grommet.
2. Transistor mounting hardware. Thermal pad, step collar, screw, nut
3. 3 pieces of wire
4. Grommet for wire
5. 3 DIN Barrier terminals (I left them out too)
6. PC board would be part of this equation too, but $2.50 sounds about right.

Packaging: A box with instructions.

What's this cost of the case/heatsink/wires and the labor to assemble it? In post #1 you have virtually the same costs and clearly it's over $1.00.

[KeepItSimpleStupid]

Protection - transient/reverse polarity Fuse that must be soldered in place, 2 special diodes; Call it $2.50. Since the output transistor is part of the integral protection, it's like $3.50. So that's about $6.00 invested in reliability.

Visual indication - LED + Resistor; about a $1.00

The circuit in Fig #1 could use them too. Callbacks, reputation?

[Whiterabbit]

can you explain the "callbacks, reputation" thing? I don't understand what you mean.

as for the rest, the TO-220 won't require sinking. When off, the transistor will have no current flowing so power consumption will be low. When on, the transistor will have a low voltage drop and thus very low power consumption. In the real world, it doesnt get hot at all.

My cases cost me 25 cents from surplus stores. However, in the realistic DIYMA style install, I suspect they (the folks this thread is supposed to benefit) wouldn't even bother to put it in a case.

Perf board is cheap, the price breakdown for a piece 1/2" by 1/2" is pennies. Same with a 1" piece of heatshrink to coat the whole thing as a DIYMA style ghetto case.

No need for the heatsink, LED indicator, or protection diode. the TO-220 is robust enough to handle the automotive environment for this simple use. I don't count the fuse because the installer who would have a concern for that would put a fuse inline anyways, regardless of this circuit. No need for a box and instructions, These are DIYers, these aren't folks looking to make these on a mass scale and sell them. Just make one for personal use.

No need to make it overly complex! And if the person who needs the thing has some of this stuff laying around already, it may be even cheaper. Maybe they have some scrounged perf board and capacitors and diodes, just need to purchase a 25 cent (we'll pretend they are overpaying) transistor. Throw the whole thing together, get it installed, and move on to the next problem.

In the real world, that literally is reliable enough to last for years! :)

[Whiterabbit]

honestly, I never answered your question about what DIYer's would be willing to pay for a designed device. Given they have access to the schematic on post #1, they will be willing to pay about a dollar. That of course is an estimate subject to error. Some of the folks here will be inclined to may much less, some much more.

[envisionelec]

can you explain the "callbacks, reputation" thing? I don't understand what you mean.

as for the rest, the TO-220 won't require sinking. When off, the transistor will have no current flowing so power consumption will be low. When on, the transistor will have a low voltage drop and thus very low power consumption. In the real world, it doesnt get hot at all.

My cases cost me 25 cents from surplus stores. However, in the realistic DIYMA style install, I suspect they (the folks this thread is supposed to benefit) wouldn't even bother to put it in a case.

Perf board is cheap, the price breakdown for a piece 1/2" by 1/2" is pennies. Same with a 1" piece of heatshrink to coat the whole thing as a DIYMA style ghetto case.

No need for the heatsink, LED indicator, or protection diode. the TO-220 is robust enough to handle the automotive environment for this simple use. I don't count the fuse because the installer who would have a concern for that would put a fuse inline anyways, regardless of this circuit. No need for a box and instructions, These are DIYers, these aren't folks looking to make these on a mass scale and sell them. Just make one for personal use.

No need to make it overly complex! And if the person who needs the thing has some of this stuff laying around already, it may be even cheaper. Maybe they have some scrounged perf board and capacitors and diodes, just need to purchase a 25 cent (we'll pretend they are overpaying) transistor. Throw the whole thing together, get it installed, and move on to the next problem.

In the real world, that literally is reliable enough to last for years! :)

Hmm, I did it for 10 cents on my DSPower. Suppose I could make a variable one (variable on-time and variable off-time) with polyswitch protection (auto-reset), 5mA standby current and no relay for $10? Wrapped in a heatshrink tube and it won't get hot. Who wants it?

[envisionelec]

Hmm, I did it for 10 cents on my DSPower. Suppose I could make a variable one (variable on-time and variable off-time) with polyswitch protection (auto-reset), 5mA standby current and no relay for $10? Wrapped in a heatshrink tube and it won't get hot. Who wants it?

Basic circuit is below. No values or part numbers shows (to protect the innocent). It may look complex, but this will get the job done for variable delay on and/or delay off with super fast reset. idle circuit power draw is in the single milliamp-range. :o Not shown is transient suppression, overcurrent protection.

And yes, I drew this up in under an hour. But that's what I do... ;)

http://i21.photobucket.com/albums/b292/envisionelec/envisionelec-timer.jpg

[Ge0]

on an unrelated note, does anyone know the ballpark minimum voltage is to hold the coil closed on a standard bosch relay?

Pulling this out of my ass, but...

Lookie here:
http://www.components.omron.com/components/web/pdflib.nsf/0/E6C4C0E4A0A85DDF85257201007DD648/$file/D20G8JR1005.pdf

In the data sheet the parameters you are looking for are pull in voltage and dropout voltage.


IMHO you'll want a strong pull-in voltage to preserve contact life (keep'em from bouncing and pitting). The minimum pull-in voltage for this style relay is typically 8V. However, I would bump that up to at least 10V. However, once pulled in, voltage can be reduced down to about 6V to hold the contacts on. You'll notice the datasheet I linked mentioned approx 1V dropout. I don't believe that. I can see contacts bouncing like wild.

How do you initiate a turn on at 10V and then drop to 6V (or whatever) once on? Fuel injectors do this to preserve their life (reduces power dissipation and thus heat). A number of IC manufactures sell a chip that does this for you. Just go to ST, Infineon, NXP, etc... and search uner the term fuel injector.

Just out of curiosity, why were you asking?

Ge0

[chad]

IMHO you'll want a strong pull-in voltage to preserve contact life (keep'em from bouncing and pitting).

Ge0

Jesus dude we are triggering a remote turn on not turning over an Allison :p

[envisionelec]

Jesus dude we are triggering a remote turn on not turning over an Allison :p

Heh. It's common practice to pop a high voltage to these things then drop the current once engaged. Usually on valves, though, not relays....

[chad]

Heh. It's common practice to pop a high voltage to these things then drop the current once engaged. Usually on valves, though, not relays....

Yeah, I'm working on a trans-brake for drag cars control using a decent solenoid, we are pulling it in pretty hard, and the hash that comes off of that Bad Oscar! WOW! It will nail you!

[Ge0]

Remember to always use a Diode across all relays to prevent the contacts from burning.


Prevent the contacts from burning? Please explain.

Lifes experience has taught me placing a diode across the coil clamped the huge inductive voltage spike that occurs to coil field collapse once current to the coil is abruptly removed. Prevents you from frying the solid state device that switches the coil on and off.

Running a relay without a diode is actually better for the contacts than running with one. The field collapes faster and the contacts pull away from each other cleaner. Adding a diode prolongs field collapse and allows the contacts to chatter a little more prior to disengaging.

However, if you're only pulling a few hundred millamps through the contacts (remote turn on loads are very low) contact wear may be a non-issue. If anything I'd be concerned more about if you are pulling enough current through the contacts to keep them oxide free.

Just my $0.02

Ge0

[Ge0]

Jesus dude we are triggering a remote turn on not turning over an Allison :p

Sorry, force of habit. The products I design pull 45A through a relay. We need to attain a minimum of 250,000 activations over the course of the products life. EVERY little trick you can do to preserve contact life is welcome. To my defense though. Wabbit did not necessarily say he was talking about using the relay to switch remotre power feeds. He did say "side note".

Question though. Remote turn-on inputs draw minimal current. Everyone is all gung ho on using relays to switch power to their devices remote turn on feeds. Has anyone thought that you may not be making the minimum wetting current requirement for the relay? If you do not wet the contacts properly they will oxidize. This can cause hidden problems that are not easy to track down.

Ge0

[envisionelec]

Sorry, force of habit. The products I design pull 45A through a relay. We need to attain a minimum of 250,000 activations over the course of the products life. EVERY little trick you can do to preserve contact life is welcome. To my defense though. Wabbit did not necessarily say he was talking about using the relay to switch remotre power feeds. He did say "side note".

Question though. Remote turn-on inputs draw minimal current. Everyone is all gung ho on using relays to switch power to their devices remote turn on feeds. Has anyone thought that you may not be making the minimum wetting current requirement for the relay? If you do not wet the contacts properly they will oxidize. This can cause hidden problems that are not easy to track down.

Ge0

Not everyone does. I think it's a crock to use a relay at 300mA when you could turn on six or ten amplifiers without that stupid relay and draw less current. Some old myths from the 1980s never die.

[Whiterabbit]

Has anyone thought that you may not be making the minimum wetting current requirement for the relay? If you do not wet the contacts properly they will oxidize. This can cause hidden problems that are not easy to track down.

Ge0

Yeah. like your aux. battery isn't charging and the stereo lasts for 5 minutes with the car off! :p

[chad]

Question though. Remote turn-on inputs draw minimal current. Everyone is all gung ho on using relays to switch power to their devices remote turn on feeds. Has anyone thought that you may not be making the minimum wetting current requirement for the relay? If you do not wet the contacts properly they will oxidize. This can cause hidden problems that are not easy to track down.

Ge0


Oh the dreaded "Mackie Ribbon Cable" plague :D