Knocker coil burnt

[dreamkey56]

Hey guys,

i have bad luck with my electronics again.

Today my freeplay knocker was running hot.

I enjoyed the knocker very much, but today, there was a lot of smoke, because the coil of my knocker burnt

It is connected to a relay,

but after exiting a table,

the relay still kepps the knocker switched, and after a while, the coil burnt.

I think, there was a Software Problem, so the Relais didnt Switch off again.

What can I do, to prevent this in future?

The game script tells the LEDWiz Output to Switch the Relais, correct?

So, is there a Kind of Relais available, that Switches only a second on, and then off by itself?

Lets say a "time-controlled" Relais, that is triggered by the Ledwiz ouptut?

To prevent the Relais beeing switched on for a longer time?

Please help me!

Thank you,

HP ("the smoker")

[mjrhp]

Hey guys,

i have bad luck with my electronics again.

Today my freeplay knocker was running hot.

I enjoyed the knocker very much, but today, there was a lot of smoke, because the coil of my knocker burnt

I think, there was a Software Problem, so the Relais didnt Switch off again.

What can I do, to prevent this in future?

So, is there a Kind of Relais available, that Switches only a second on, and then off by itself?

Lets say a "time-controlled" Relais, that is triggered by the Ledwiz ouptut?

To prevent the Relais beeing switched on for a longer time?

I've seen a couple of other people mention this happening, so I've been thinking along the same lines - adding some hardware protection against this in my build.

The easiest thing to do would be to add a slow-blow fuse at right around the peak current for the normal knocker operation. Add the fuse on the knocker side of the relay circuit, so that it protects the coil from staying energized continuously. That would let the knocker fire briefly, as it's meant to do, but the fuse would blow if the coil stayed energized for more than a few moments. The only snag is that it's tough to find exactly the right fuse size that would provide adequate protection without constantly blowing in normal operation.

You might be able to find a pulsing relay, but I haven't seen anything that does exactly this, and it would probably be really expensive if you could find one. The special-purpose relays I've seen are really pricey.

My plan is to add a little 555 timer circuit on the Ledwiz side of the relay that converts the signal from the Ledwiz into a single short pulse. That would guarantee that the coil would never stay energized continuously - no matter how long the Ledwiz "on" signal is, the coil would just see the same constant-length pulse, and wouldn't get another until the Ledwiz turned off and on again.

Here, just sketched out the circuit. I've built a couple just like this, but you should breadboard it first to make sure I got everything right. The 1K resistor and 100uF capacitor combo on pins 6 and 7 control the pulse time - 1K x 100uF = about 0.1 seconds. You might want to experiment with that with your coil to see if you need a little shorter or longer; adjust the resistor up or down proportionally to the time change (e.g., if you want to double the time, double the resistor value - a 2K resistor will give you a 0.2 second pulse).

It's a little more complicated than a fuse, but it's only a few dollars in parts, so it's cheaper than blowing a fuse repeatedly, and certainly cheaper than blowing a coil!

One thing to note - if you go this route, you'll need a relay with a 3V coil. The 555 output on pin 3 puts out about 3V when Vcc is 5V. The 555 can handle 200 mA on the output, which is plenty for most small relays.

My setup is a little different from yours, in that I'm using a Zeb's Boards booster instead of a relay to drive the coil. If you use that setup you'll need to replace the relay on pin 3 with an inverter going to the booster board input. You can use a NAND gate as an inverter - connect one NAND input to the 555 pin 3 output, and the other to +5V, and connect the NAND output to the booster input.

[russdx]

All real pinballs have hardware fail safes on the solenoid matrix and even the lamp matrix (if it stops scanning and sticks on a row those lamps are getting fried!)

Usually hardware watchdog timers checking each output and if it sees it on for to long it just disables it. Prevents burnt coils from bad programming / crashed code.

I would not hook a real coil upto a ledWiz / pinMame without a hardware watchdog timer as you have NOOO idea what it can set that output to hehe.

[zebulon]

Newer pinballs have the watchdog timers yeah. The older SS ones would just blow the output transistor, caps, diodes, resistors and sometimes the decoder chip on the driver board.... Usually because of a shorted diode on the coil itself.

Darkfall wrote a bit of code awhile back that reset the ledwiz upon exit every time. I'm sure it's still around somewhere.

For those using my boards, they'll accept a 3v input.

[dreamkey56]

Thank you, mjrhp.

I also was thinking about using a Monoflop, but I am not very good in electronics. And I dont know, if the time will be triggered again,

after the monfoflop Switches off, and there is still the Signal of the ledwiz on.

I think so, but I am not sure.

So I will try to adapt your diagram.

Thank you also, Zeb, for helping AGAIN.

I will search for this code...

Hopefully, I can find.

I ordered a new coil, so I have a few days to find the thread for it...

Edit: I think, the 555 will really be re-triggered, if the Signal is still at the Input, isnt it? Because I found this: http://electronicdesign.com/site-files/electronicdesign.com/files/archive/electronicdesign.com/files/29/3167/figure_01.gif

Best,

HP

[mjrhp]

Thank you, mjrhp.

I also was thinking about using a Monoflop, but I am not very good in electronics. And I dont know, if the time will be triggered again,

after the monfoflop Switches off, and there is still the Signal of the ledwiz on.

Yeah, the textbook 555 monostable circuit leaves its output turned on as long as the trigger is asserted, which would defeat the whole purpose in this case - we need it to turn off even when the trigger stays on. The trick in this version is the little RC network on the trigger pin (the two 100K's and the 0.1uF leading into pin 2) - that makes the trigger edge-sensitive, so it'll ignore a constant sustained "on" signal from the LEDWiz. The 555 won't fire the coil again until the LEDWiz turns its output off (high/open) and then on (low/ground) again.

So this circuit should handle what seems to be the known failure mode, where the software gets confused/crashes/terminates and just forgets to ever send the "off" command to the coil. The coil will fire once and turn off; the LEDWiz will stay on, the coil will stay off, and everyone will be happy.

The software probably *could* still blow the coil by hammering it with a constant 5 Hz on-off-on-off loop, but that doesn't seem like a likely bug scenario - I think someone would have to intentionally write malicious code for that to happen. And the coil could probably handle rapid cycling for quite a while anyway; probably much longer than it can handle being energized continuously. Plus, as a knocker, it would make such a racket that you'd have to pull the plug to avoid waking the neighbors before the coil was in danger of frying.

[dreamkey56]

Aahh, perfect, it is edged triggered, fine,

so all I need are the parts you putted in your drawing, and a 3V Relay, that triggers the knocker.

But the 3V Relay seems to be a very rare one.

I need one, that can handle the high power knocker, with a 3V Relay coil.

only found VERY expensive solid state Relays with this data

What is the best way to solve this?

A step-up- converter, to bring the 3V Output to a 12V Level for the relay?

Or a small 3V Relay (small 3V Print Relays are easy to get), that triggers again the 12V Relay?

Sounds a Little bit stupid, I know,

but do you have also a solution at this point?

Would be really nice.

Thanks a lot,

mjrhp.

HP

[mjrhp]

Aahh, perfect, it is edged triggered, fine,

so all I need are the parts you putted in your drawing, and a 3V Relay, that triggers the knocker.

But the 3V Relay seems to be a very rare one.

I need one, that can handle the high power knocker, with a 3V Relay coil.

only found VERY expensive solid state Relays with this data

What is the best way to solve this?

A step-up- converter, to bring the 3V Output to a 12V Level for the relay?

Or a small 3V Relay (small 3V Print Relays are easy to get), that triggers again the 12V Relay?

HP

Right, I think the parts in the schematic are the complete circuit.

Mouser.com has a couple of fairly inexpensive 3V relays: Omron G5LE-1-ASI DC3 ($1.17) switches 35VDC at 16A. Omron G6RL-1A4-ASI-DC3 ($3.55) switches 300VDC at 8A. Either of those should be plenty for a knocker coil - those need about 4A at 30VDC, right?

But you might be able to use your existing 12V relay just by using a higher Vcc on the 555 circuit. I drew up the schematic at 5V, but it'll work equally well with anything up to 15V - the 555 can handle 3-15 for Vcc, and the rest of the parts don't care (within reason ). The 555 output pin gets voltage equal to about Vcc minus 1.7V, so if you use 12V for Vcc, you can use a 10V relay. Check your existing relay to see if it'll switch on 10V - it might; the "must operate" voltage is often a couple of volts lower than the nominal coil voltage. (Just checked a couple of the Omron 12V relays - they say they'll switch on at 9V, which would work fine here.)

If your relay doesn't switch reliably on 10V, you could step up Vcc to 14V. Or you could add a driver transistor between the 555 and the relay to step up the voltage. Either of those solutions would require more parts, though, so it'd probably be cheaper and easier to add a 3V or 10V relay.

[mjrhp]

One more detail I forgot - you should put the usual flyback diode across the relay coil, to protect the 555 from back-current when the relay turns off. Wire it across the relay coil backwards (anode to ground) as always.

[Luckydogg420]

I'm really interested in this topic. My knocker is in a place that can't be easily replaced, so I want to do all that I can to insure that it won't burn out. I'm using zeb's boards to connect everything together.

What's the best way to prevent this from happening? Should I build the 555 circuit that's above? Will a fuse be enough (what value fuse) to prevent a coil from locking on and burning out?

I would literally have to build a new backbox if this happened, so I don't want to be worried about it happening.

[mjrhp]

I'm really interested in this topic. My knocker is in a place that can't be easily replaced, so I want to do all that I can to insure that it won't burn out. I'm using zeb's boards to connect everything together.

What's the best way to prevent this from happening? Should I build the 555 circuit that's above? Will a fuse be enough (what value fuse) to prevent a coil from locking on and burning out?

I would literally have to build a new backbox if this happened, so I don't want to be worried about it happening.

Wow, you definitely want some protection there.

I'm planning to put in the fuse *and* the timer circuit myself. Some people have mentioned scripting patches to guard against errant table scripts, but there are too many layers of software for that to give me a warm feeling. If VP itself crashes, or Windows reboots at an inopportune moment, a script patch won't help. The timer circuit should completely insulate you from software glitches on the PC or LEDWiz, since it's a separate hardware layer standing between the coil and everything else in the system.

As for fuse sizing, I don't think there's an easy answer. But we can figure it out with a little work. What you want is a fuse that allows the normal coil current briefly, but blows if the same level is sustained. To actually choose a fuse, we'll need to be more precise than "briefly". For the timer circuit, I made a wild guess that 100ms is long enough to get the knock - does anyone the actual pulse time the tables put out? That normal pulse time is our lower bound - we don't want the fuse to blow before that interval, otherwise it would blow constantly in normal operation. The upper bound is the amount of time of continuous operation that will melt the coil. Dreamkey56, do you have any idea how long your coil was stuck on before smoke started coming out? I'm guessing these coils can take at least a few seconds at full current - if that's true, then we have a pretty comfortable margin to work with: we want the fuse to blow somewhere between 100ms and 4s, say. Let's split the different and say we want the fuse to blow after 2s.

The other parameter we need is the coil current. Let's say you're using an AE-26-1200 coil at 30V; that's 10.9 ohms -> 2.75A. If you're using an AE-23-800 at 30V, that's 7A (I guess you shouldn't use this one with Zeb's booster board, as its outputs are rated for 5A). Let's assume the 2.75A for now.

Okay, so we want a fuse that blows after 2s at 2.75A, if all of our assumptions so far are correct.

Unfortunately, they don't rate fuses this way, at least not in the searchable ratings. What they do provide is a data sheet with a plot of the blow time vs current. For example, take a look at http://www.littelfuse.com/data/en/data_sheets/218p.pdf. There's a graph on page 2 of "average time current curves". This shows curves for the whole family, so it's actually pretty easy to search. What you do is figure out the coil current and blow time we want, find that point on the graph, and find the line comes closest to intersecting that point. Trace that line up to the top of the graph to see which fuse it's for.

So for 2s at 2.75A, it looks like the closest option in a Littlefuse 218 series is the 1A - its curve crosses 2.75A at ~1s. That's a little faster than we were looking for, but it should leave plenty of margin for normal operation. Energizing a knocker for a full second would definitely be an error, so I think it's okay to blow the fuse if it stays on that long. And we want to err on the side of more protection, meaning shorter blow time.

This is just the first fuse family I looked at - there are lots of these, so there's probably a better fit. A little time researching this on one of the electronics distributor sites would be worthwhile. They all link data sheets. Look for "slow blow fuse" with the form factor you're looking for, and check a few families to see if there's a good fit. I'll have to do this research myself soon; I'll post what I finally decide on when I get to that point (which might or might be useful to you, depending on whether you're using the same coil and voltage).

I'm also going to put in a separate fast-blow fuse with the usual 20-25% current margin, say a 3.5A, to protect against shorts or other hardware problems in the coil circuit itself. Put that in series (electrically) with the slow-blow (physically separated enough that one won't heat the other, since that could make both blow too fast). This is more for protecting the booster board from the coil than vice versa, but it would protect the coil too if the circuit got shorted to 120V or something like that.

[dreamkey56]

Hi.

Didnt get the Elkos yesterday, so I have to order 3 Elkos, with 1000times higher shipping costs than the price of the Eloks *grrr*. Also bought a assembled Monoflop board with 2 circuits, but a switch as starting input is mounted, so I have to check, how it is working, before I can connect it to the Ledwiz output. Or maybe I put another relay on, instead of the switch. Trigfered by the ledwiz. Lets see. But first, I want to gry it with the circuit mjrhp has attached.

Mjrhp, I cant say exactly, how long it took, until the smoke appeared, but I think, it was a lot of time. I think, I got a problem, when exiting a table, because when playing this table, the knocker worked.

After exiting, I did some software work. I guess, after 1 minute I smelled the smoke, but was thinking, my wife was ironing in theother room

And after further half a minute, I heared a silent *ping*. The noise, of the plunger, that was falliing down, because of the broken coil. And then I recognized.

When I looked inside the cab, there was a lot smoke.

I guess, the coil will survive many seconds, without any damage.

About the current I am confused. I switch the coil by a 16A relay, to be sure, because I have read in another forum, that "the knocker coil will eat small relays up to 8A for breakfast". I know, there are many different knockers on the market, so I think, the current consumption of the triggered coil should be measured directly? I will do, and attach also a very slow fuse near this value, if I can find. Together with this monoflop switch, there should be no problem, I think?

The software solution is difficult for me, as I am not very good in these scripting. And I couldnt find mamemans script here at the forum..

Wondering, that there are not more guys with this problem. BecauseI also have sometimes a problem with thesolenoids, because sometimes I also hear this pock* later, after exiting the table. The sound, when the solenoid releases the plunger.

No problem here, because of low power consumption. But the same symptom.

HP

[zebulon]

There is a simpler solution to this. For those with concerns about soldering up a circuit...

[zebulon]

and if a delay is needed....

The relays are simple 12v Bosch type automotive relays. Diodes are 1n4007.

How it works...

The power for the knocker coil is passed through the normally closed contacts of the relay. When the ledwiz fires, the coil for the knocker is energized and after a slight mechanical delay, the coil for the relay is energized cutting the power to the knocker. The capacitor on the coil is there to stiffen the voltage to the coil to ensure the knocker goes off before the power to the coil is cut. If the output of the ledwiz is locked on, the power to the coil remains off until the output goes inactive.

The second design pretty much ensures that the knocker will fire properly before the power is cut to the coil by the added delay created by adding the second relay.

[russdx]

You can get even simpler

Put a normally closed micro or leaf switch on the knocker that when the knocker is activated breaks the switch.

This way it will fire but will cut its own power once fired.

I guess a stuck signal would cause it to just keep firing over and over, so it would not fry, but by it probably sounding like a machine gun it would get your attention pretty quickly to reset the game lol

[zebulon]

Yeah I considered an end of stroke switch too....my thinking on it was that a knocker would smash up a switch pretty quickly. I did think that a magnetic reed switch near the coil might work though,,,

[MTPPC]

Isn't there a good chance of sinking more than 500ma into the ledwiz in that second diagram? WHat is the current required by a knocker?

[dreamkey56]

Was thinking about the same, as Russ wrote. Funny. But in case of failure, the machine gun action will damage the cab, if you have strong knockers

But the hardware would be quite simple and easy to assemble.

I am still waiting for the knocker. After, i will check current .consumption at 24V

But what is wrong with the solution mjrhp did? The monoflop has an "adjustable" timer, so it should work.

Any doubts?

The 2-relay solution Zeb showed us looks also fine to me. More "mechanical", so easier for me to understand

Seems, there is a solution for the problem I had.

My new knocker is on the way...

But next problem is waiting. Optocoupler board for the flippers

I´ll post a new thread..

HP

[mjrhp]

I cant say exactly, how long it took, until the smoke appeared, but I think, it was a lot of time. I think, I got a problem, when exiting a table, because when playing this table, the knocker worked.

After exiting, I did some software work. I guess, after 1 minute I smelled the smoke, but was thinking, my wife was ironing in the other room And after further half a minute, I heard a silent *ping*. The noise, of the plunger, that was falling down, because of the broken coil.

That's very useful information. That gives us a bit more margin than I was originally thinking in picking a fuse, which is good because there's going to be some variability between individual fuses with a given rating. I think we can pick a fuse that'll last for ~20 seconds of continuous coil power. That way the fuse should blow long before the coil is damaged, but also shouldn't blow in normal operation.

About the current I am confused. I switch the coil by a 16A relay, to be sure, because I have read in another forum, that "the knocker coil will eat small relays up to 8A for breakfast". I know, there are many different knockers on the market, so I think, the current consumption of the triggered coil should be measured directly?

A direct measurement is the best way to know. 8A or more for the coil is entirely believable. If your coil does take 8A, it's good that you're using a big beefy relay (and it'll do no harm to use a 16A-rated relay if the actual current demands are lower). It's always good to have some headroom in the specs.

You can measure the resistance of the coil and figure the current draw from Ohm's law, assuming you're using DC power to the coil. The current is the voltage you're supplying divided by the resistance of the coil in ohms.

[zebulon]

Isn't there a good chance of sinking more than 500ma into the ledwiz in that second diagram? WHat is the current required by a knocker?

I'm just indicating the signal, not the actual source. You would still need a relay to run between the ledwiz and the knocker or yes...you would greatly exceed the 500ma.

[zebulon]

That's very useful information. That gives us a bit more margin than I was originally thinking in picking a fuse, which is good because there's going to be some variability between individual fuses with a given rating. I think we can pick a fuse that'll last for ~20 seconds of continuous coil power. That way the fuse should blow long before the coil is damaged, but also shouldn't blow in normal operation.

A direct measurement is the best way to know. 8A or more for the coil is entirely believable. If your coil does take 8A, it's good that you're using a big beefy relay (and it'll do no harm to use a 16A-rated relay if the actual current demands are lower). It's always good to have some headroom in the specs.

You can measure the resistance of the coil and figure the current draw from Ohm's law, assuming you're using DC power to the coil. The current is the voltage you're supplying divided by the resistance of the coil in ohms.

Takes about 30 seconds or less at 50V to smoke a coil in a real machine....AMHIK...

[mjrhp]

Takes about 30 seconds or less at 50V to smoke a coil in a real machine....AMHIK...

Ah, another good data point. So I guess a 10-second slow-blow timeframe might be better after all. I was hoping that we'd have a little wider margin to work with, since it makes it easier to find a fuse that will be sufficiently protective but won't randomly blow out when nothing's wrong.

Joking, but maybe a thermocouple attached to the coil would be a good idea...

[dreamkey56]

Still a condensator missing...

But I got a set to build a monoflop board.

Time is adjustable at screws, works very well.

It is a board with two Monoflop circuits, so I have one left, I have another toy that has to be securely switched over a short time.

One funny Thing: At starting Computer, the knocker knocks once.

I will add a slow-blow fuse also, to be 100% sure, coil will never be damaged anymore.

[mjrhp]

Still a condensator missing...

But I got a set to build a monoflop board.

Time is adjustable at screws, works very well.

It is a board with two Monoflop circuits, so I have one left, I have another toy that has to be securely switched over a short time.

One funny Thing: At starting Computer, the knocker knocks once.

I will add a slow-blow fuse also, to be 100% sure, coil will never be damaged anymore.

[ATTACH=CONFIG]38432[/ATTACH]

Sounds like you have a good safe solution!

I have a guess about what's causing the power-on knock. If the timer you're using is 555-based, it might be an artifact of the 555 design. The trigger line on 555 is active low - putting 0V on the trigger line is what starts a pulse. When you first turn on power, the various capacitances throughout the circuit might be causing main power on the 555 to reach full voltage a few milliseconds before the trigger line itself comes up to 5V, so the chip might be interpreting that momentary low voltage on the trigger pin as a trigger signal, firing off an initial pulse on the output.

The way to fix this with a custom 555 circuit is to put a capacitor to ground and a resistor to Vcc on the RESET line (pin 4) of the 555. The reset line is also active-low; the idea is to hold the reset line near 0V for a few hundred milliseconds at power-on to let the voltage in the rest of the circuit reach equilibrium before enabling any timer pulses. The 555 ignores trigger signals as long as RESET is asserted, so this prevents a false trigger while powering up the chip. It probably isn't easy to add this to a pre-built board, though.

[dreamkey56]

Thanks for this, mjrhp.

But at the moment, i'll keep it, like it is. This extra 'BAMM' from the knocker at the startup sounds nice to me

I think, I will also wire a cree flasher LED parallel, to get a strobe effect also.

Without the monoflop board, could be also a simple indicator, if the coli got stock. If you have a very bright flasher connected, and it will stay on, you will know, you have to hurry up

I will check the current and a matching slow blow fuse later. I have checked, what temperature the coil gets after a few seconds beeing triggered. But after a few seconds, it isnt getting really mich warmer, because of its mass, I think. So, I guess, a very slow fuse will work well, If I can find one.

The next step is now, to get my opto circuits for the flippers working

But, here I ask for help in another thread.....

HP