PDA

View Full Version : Relay stutter



Madcow
02-19-2007, 03:07 PM
It may be that I've finally solved the mystery!

Recap: For those of us who have tried to wire a MR sound board with a relay (to allow full current to a K2 LED), it seems that any MR board other than the Mace, produces an annoying stutter effect that prevents the saber from turning on.

I recently wired up a mace board, Royal Blue K2, and a 4.8v rechargeable cell... and it stuttered!!!
So I continued experimenting and finally found that when I run 2 seperate power wires (one to the red and orange wires on the board, and one to the LED/relay) it helped. Then I made the power wire to the LED/relay much longer (looped it back and forth) it solved the problem entirely! I don't know why it works but it does!
http://rpetkau.photosite.com/Album7/IMG_1139.html

I have not tried it with my Luke board yet.

Post your results!

MC

[/img]

Strydur
02-19-2007, 05:52 PM
By making the wire longer you are adding resistance. Try putting a really low rated resistor in line and see what that does.

xwingband
02-19-2007, 05:58 PM
I tried a diode (at Do-Clo's suggestion) with my Ani board so I'm not sure resistance is what up is up with it there.

I think I'm lost on the two power leads thing there. Perhaps a diagram would help there?

Firebird21
02-19-2007, 06:06 PM
If the resistor doesn't do the trick then it may be a case of a magnetic field being created by you looping the wire. It may also have something to do with the radio emotions the resistor puts off.

I don't fully understand this phenomenon, but I do know that I have a wireless mouse that when I coiled the wire from the receiver to the PC it didn't work, when I uncoiled it, it worked fine.


Just a thought...

Madcow
02-19-2007, 11:21 PM
I thought it was resistance too, but I tried both a 1 ohm and a 3.3 ohm resistor on the LED and it didn't seem to affect it at all. Somehow the pos. wire to the LED being 10 inches longer than the one to the sound board makes a difference. I'm using 26 guage stranded wire - 10 inches shouldnt affect the resistance that much... right?

Looks like this:
http://rpetkau.photosite.com/~photos/tn/8630992_1024.ts1171952348874.jpg

MC

erv
02-20-2007, 07:00 AM
It could be a side effect of the fact that you are using a relay.
To diagnose this, please allow me to ask a few questions :
- does the problem occurs when the Luxeon is not there ? (the relay just clicks once ?)
- is your really a model with roll-off diode ? An inductive load is quite bad on a transistor commutation schematic like the MR board, you need a roll-off diode in // with the coil of the relay.
Erv'

xwingband
02-20-2007, 07:31 AM
The same one I used on your old set of boards was what I used when I tried it (I'd have to look up the number/name). I thought it wasn't a bad idea since when the batteries were low it didn't do it as much if at all.

erv
02-20-2007, 07:54 AM
if you want to check if your relay has a roll-off diode or not, simply use a ohmmeter and diode checker.
In ohm-meter, the coil is seen as a resistor for both polarities, about 100's of ohm. Go into diode check mode : you should read around .700 v in one polarity, and infinite on the other polarity (and this last point means that you have a diode).
If the relay doesn't have one embedded, wire an 1N4001 this way
http://www.kpsec.freeuk.com/images/diopro.gif

and test to see if you still have the problem.

Erv'

Madcow
02-20-2007, 11:22 AM
Thanks Erv,
I'm pretty sure the relay doesn't have a diode.

To answer your questions...
When the LED is not on the circut the relay works fine - in fact - this mace board works fine with a 3 watt lux on the relay. the Luke board I had the same problem with worked fine with a 1 watt lux. It only stutters when the heavy drain of the K2 is added. This led me to imagine that...
The instant the relay switches and the K2 draws power - the power drain disrupts the MR boards startup sequence... shutting off/trying to start again etc. in a cycle.
(What do you think of this theory Erv?)

I don't understand the need for a transistor in your schematic.

X-wing, did a diode parallel to the relay coil work in your application with a K2 and MR board?

MC

erv
02-20-2007, 11:45 AM
Hi MadCow,
you don't need a transistor here, the transistor is already there on the MR board (6 transistors in common ground actually). You're probably right MC, you get a little shortage of voltage when the power comes to the K2. It's so sudden that it causes the board to reset or go off (this is called a brown out reset on certain microcontroller chips). And then you get this oscillation phenomeon. Using a diode would not help it so much, but it should be tested. The role of the roll-off diode is when you cut the power on the relay : the self inductance of the relay coil stores current. When you turn off the relay, there is a big gap ofvoltage that generate some current... that goes backward in the circuit thru the positive wire. The roll-off diode absorbs this current at that moment and protects the circuit. But if there was a problem about this, it would occurs even without the luxeon plugged. However, it can be a combination of both phenomenoms.
Now, why is it working with some wire lenghts ? Just because you created another self inductance (a coil) by attaching the wire with many turns. A coil smoothen the current during the power on of the luxeon, and avoids the reset of the board. The main problem here is that the relay turn on and off the luxeon in a very sudden way, while the proper way would be to have soft start. The other problem is that you're taking the power of the lux directly on the batteries, not on the board. The solution would be to regulate the voltage of the board, with a little voltage regulator to make sure it does not reboot, cause an inductive load on the relay is not the best thing (well, okay, I'm proposing damn complicated things while your solution works :roll: :wink: )

would someone like a schematic to sum it up ?

Erv'

xwingband
02-20-2007, 12:02 PM
Yup I used the diode Erv stated. I looked back and it's the same one he recommended me before. It didn't change anything.

I'll have to try the extra wire. Honestly I'm getting lost with what erv is saying.

Firebird21
02-20-2007, 12:09 PM
I'll have to try the extra wire. Honestly I'm getting lost with what erv is saying.



lol

That's the technical and educated version of what I was saying. :mrgreen:

The difference is he knows what he's talking about and I only know that crap like this can happen, don't ask me why and how... :roll:

Something about a magnetic field created by coiling the wire (an electro magnet), after that I'm lost.

erv
02-20-2007, 12:54 PM
lol x 2 !!
well, ok, my appologies... It's true that it's clear for me, cause it's part of my daily work and background :roll: mmm, fasten your seatbelts, crazy Erv' is back...
Basically the board is rebooting cause the spike of current of the LED turning on. The voltage on the board is not regulated meaning that every variation of voltage on the batteries is spread all over the board, making it rebooting. Why the board doesn't do this when the lux is wired to the board ? cause it's turned on progressively by the 6 differents transitors. Each transitor has an equivalent resistor in serie with the led, and when the 6 transitors are on, all 6 equivalent resistors are in //, and the overall resistance is smaller compared to when a single transitor is on. It creates a kind of time controlled resistor give a little effect of brightness increasing in a progressive way (it's easier to see with a red/red orange/amber LED).
But with the relay, current is pumped instantly, and directly on the batteries, creating a "hole" of voltage, board rebooting etc.
I though that the board could be regulated in a simple way, but no. It's supposed to work at 4.5v. A classic regulator requires to have a powering voltage > regultated voltage + 1.5 volt, impossible here. Solution would be a low drop regulator, but we are (already ?) out of a "simple" solution with that.
What you need is a bit of stored energy when you turn on the relay. I would advise to test with a 1000 µF chemical capacitor (polarized) on the batteries, or better, close to the LED, but on batterie side : this way the capacitor is charged, waiting for the LED to be turned on. On commutation, the capacitor will help powering the LED while the batteries quickly adapt themselves to the current being taken.
Alternate solution would be to replace all the 6 transistors driving the led strip(they are NPN ones) by a bigger model, like a 2N1711 or equivalent. Each one can drive .5 amp, so you would need just 4 of them. No more relay, lux K2 would be powered by the MR board again !
I'll try this for you, if you want ! In the meanwhile, try the big capacitor for filtering the power spike, and let us know !
Erv'

Strydur
02-20-2007, 02:00 PM
We dont want to overpower the mr board though. So probably best to stick with the relay atleast with 5w setups for sure. Erv how would we go about wiring in a capacitor? Can you do a diagram for that?

Madcow
02-20-2007, 02:07 PM
Sweet!
What would a simple Jedi do without you guys!
Erv, you Rock! Let me know what happens when you replace the transistors - AND how many mA the MR board puts out afterwards.

I was all excited when I found a low tech solution - I had to share it with you guys... my wife just gave me a blank stare as I tried to explain to her how brilliant I am.

MC

vortextwist
02-20-2007, 02:15 PM
I'm still lost, My mace k2 started doing that when I went and put new batteries in, I never even thought to put the old ones back in, so I just took out the relay and led and put in an 1watt red. I can't wait to see what you guys come up with.

erv
02-20-2007, 02:24 PM
I'll try to draw a schematic by tomorrow morning in the train, cause for now I'm struggling against a bad cold and coughts...
Bed time stories here now :-)
(once unpon a time, there was this little capacitor in the castle of Manor Republic.... and it was feeling sooooo alone....)
Erv'

Madcow
02-20-2007, 02:24 PM
The Capacitor makes sense... but...
1. Will that drain the batteries when the saber is not in use?
2. Will that lower the power available to the LED?

MC

xwingband
02-20-2007, 02:59 PM
I understood why it was browning out. It was the solution that had me confused.

The capacitors make sense though. If you can show exactly which ones to take out and replace I'm not afraid of doing it. :D I'd actually spacewise prefer that to the space of a relay.

Ryma Mara
02-20-2007, 03:29 PM
I'm still lost, My mace k2 started doing that when I went and put new batteries in, I never even thought to put the old ones back in, so I just took out the relay and led and put in an 1watt red. I can't wait to see what you guys come up with.

Yeah but I dont think that relay was working right cause it wasent very bright at all. It was losing twards the tip like an already filtered led looks like.

vortextwist
02-20-2007, 08:35 PM
maybe i got a weird k2? heck if i know

erv
02-20-2007, 10:23 PM
The Capacitor makes sense... but...
1. Will that drain the batteries when the saber is not in use?
2. Will that lower the power available to the LED?
MC

no, and no :D
schematic soon !

here is the annotated schematic. I took a bit of freedom to reverse a few things for a better understanding. I've also transformed the relay into a normally open one (ie the switch of the relay that drives the luxeon is open when the coil is not powered).

Erv's Annotations & Mods on Mad Cow original schematic
http://www.plecterlabs.com/Media/RelayMadcowErvAnnotations.jpg

I'll soon come back to you with a detailed explanation about how a MR board works, and I have a nice idea to transform it into a progressive power on/off board for the luxeon, K2 capable, this for a 1.5$ mod (I hope so).



Erv'

Madcow
02-26-2007, 01:21 PM
Thanks Erv,
You da mannn!

First person to try this setup post your results!
I have to buy another MR board before I try it.

MC

xwingband
02-26-2007, 04:12 PM
Dang... I wish I had known about the capacitor. I would have tried it too because I had the diode part.

The mod you describe seem much more interesting though.

Psiberian
05-19-2007, 01:23 PM
LOL this actually answers some of the problem I am running into trying to wire up a K2 with an MR Vader.

IF I can get this up and running this weekend I will post some pics, and or explain what I did. (you know if it works :P)

Novastar
05-19-2007, 08:09 PM
Interesting! Maybe someday, when I care to bother re-wiring some or any of the MR conversions we've made for the BOP sabers... I'll see about making this wiring a reality for them.

In the meantime, I'm cool and the gang with buttered toast for a "pimped" saber... :)