DC-DC voltage converter for flashlight

[Dan Oelke]

We got a flashlight to use as a photon canon. And are having issues driving it.

The flashlight we are using is "LED Lenser P7"
http://www.amazon.com/gp/product/B001AT1F6S

It uses 4 AAA batteries so it needs 6V into it. We got a DC-DC converter specifically this one: http://www.amazon.com/gp/product/B008BHAOQO

And wired it up on the bench and everything seemed fine when we plugged it into a battery.

Then we wired it to a relay on the bot - so that the relay provided power to the DC-DC converter which was then wired to the flashlight. After that - it wouldn't work. Before bagging we found that the DC-DC converter output was only 1.6V instead of the 6V it had been adjusted to. Playing with the adjustement screw did not seem to give any different output.

So - any ideas on what we might have done wrong? I think the polarity was always right for the input to the DC-DC converter. Would have reversing it on accident blown something you think?

Any other ideas? Is there a better/more robust DC-DC converter we should be using?

Our regional starts in a few days and we would like to be ready to swap out the offending part for a quick recovery.

Thanks!
Dan

[apalrd]

How much current do you need?

You can source 6v from the digital sidecar. If you aren't using all 10 PWMs then you just have to put in the servo power jumper and then use 2 of the 3 pins of the PWM connector for 6v and ground. If you are using all 10 PWMs then you could wire up a 2-pin header to the jumper of a PWM port that isn't used for a servo to get 6v and another header somewhere on the DIO or Relay section to get a ground.

The spec says you can pull up to 3a from the 6v supply, I doubt you need anywhere that much.

[Al Skierkiewicz]

Dan,
It is entirely possible to have reversed the voltage. Remember that the Spike relay can supply positive on either or both pins depending on the state of the input.

[Brandon Holley]

I agree with Al on this one. It sounds like you may have flipped a set of wires somewhere, or the Spike is outputting in "reverse".

-Brando

[Dan Oelke]

Since this is powered by 4 AAA batteries in series - it shouldn't draw more current than what a single AAA could provide. I am having a hard time finding good numbers on that, but everything I have found has been way less than 1A so getting it from the digital sidecar is possible.

The more I think about it - the more I like it. It means one less component.

I assume that would then always have 6V present on that channel once I set the jumper? i.e. we couldn't turn the light on/off then???

[Dan Oelke]

I agree - with the spike being able to reverse the polarity AND knowing how many different hands were helping with getting the last things wired - I am ALL to aware the voltage could have been reversed at some point.

I am not sure if reversing the input polarity to that particular converter would "let out the magic smoke" - but considering the cheap price I wouldn't be surprised that there is no protection against that.

Any suggestions for a more robust DC-DC converter? I like the idea of using the 6V from the side car, but not if we can't turn on/off the light.

[apalrd]

True, you cannot switch the 6v power.

[Bruceb]

using power from the digital sidecar is exactly how we are powering our LED flashlight. As far as switching goes, how about using a switch? Nothing says you have to switch it with a spike. Our shooter goes up and down and when it is up it closes a switch that powers the flashlight.
Done

[EricS-Team180]

Kind of an aside, but we get our dc-dc converters from here:

http://www.current-logic.com/dcdc_converter_car.php

Unfortunately, I don't see one that fits your immediate needs.

We typically pick up a Mag light from Lowes or Home Depot and mod it to use a 12:5v (or 12:3.3v) stepdown converter. Then, use a SPIKE as a switch.

(I like Bruceb's and apalrd's approaches as well )

Thanks,
Eric

[EricVanWyk]

You can source 6v from the digital sidecar.

The spec says you can pull up to 3a from the 6v supply, I doubt you need anywhere that much.

My read of section 4.1.8 allows it, as does my read of the Inspection Checklist. The key is that your flashlight is not a motor or actuator. The rules focus on motor/actuator power paths, and are relatively loose on custom circuits.

As for whether this is a good idea - sure. That supply can more than handle any reasonable flashlight load, and is reasonably well protected against shorts/etc.

[Dan Oelke]

So are the Spikes "just" a straight relay with the coil powered completely from the digital sidecar? Or, are they using the voltage from the input side along with a signal from the sidecar to drive the coil.

The reason I ask - is *if* the coil is powered completely by the side car, then I would think we could wire the 6V from the sidecar's servo control thru a spike that is controlled from the relay control of the sidecar and use the output from the spike to the flashlight.

Everything I have found for the spike relay's shows them wired to 12V, but a straight up relay wouldn't care what that voltage is. I haven't found a circuit diagram of the guts of a spike.

[Joe Ross]

Everything I have found for the spike relay's shows them wired to 12V, but a straight up relay wouldn't care what that voltage is. I haven't found a circuit diagram of the guts of a spike.

The technical specs list an operating voltage of 6v-16v. http://www.vexrobotics.com/vexpro/motor-controllers/217-0220.html

[EricVanWyk]

I think you will encounter rule issues if you take this approach, as the Spike is called out specifically in several places as to how you can legally wire it.

I'm reasonably sure the spike has optoisolators on its input, but its been a few years since I looked too closely.

Personally, I'd just put a N-fet in line: 6V+, light, FET, ground. Attach the gate to a GPIO.

[Al Skierkiewicz]

Dan, the Spike is actually two relays powered from the input voltage. There are opto isolators on the input as Eric mentioned. You won't find a schematic, IFI doesn't publish that. If you look at the manual you will see the truth table for relay operations. This version can make either output at common or either output at + volts. Old versions (from about 2002 and earlier) did not have the full functionality.
Edit..
BTW, relays usually can stay energized when the voltage falls well below the operating voltage. The problem is pulling in at low voltage. Often the relay is speced to always pull in at a specific voltage. Often that is below the operating voltage specified.

In thinking about this, a LED ring is a custom circuit. There is nothing preventing one from using a transistor switch triggered from a digital output to provide either power or common to the LED string to turn it on. This would be considered part of the custom circuit as well. You can then feed it from the PD through an appropriate breaker to protect the wiring and custom circuit.

[DonRotolo]

Um, maybe I am showing my age, but what about a simple LM7806 regulator? Super cheap, an Amp or two, and rock stable.

[Travis Hoffman]

Perhaps overkill, but is there anything keeping someone from wiring the flashlight to a Victor 888 or Talon and commanding the required output voltage?

[Alan Anderson]

Perhaps overkill, but is there anything keeping someone from wiring the flashlight to a Victor 888 or Talon and commanding the required output voltage?

The output voltage from a speed controller is full battery voltage, just not all the time. That could be a problem for some LED flashlights, either by exceeding a maximum tolerable voltage even for a moment, or by messing with an internal DC-to-DC converter or a constant current source circuit by pulsing its input.

[billbo911]

Um, maybe I am showing my age, but what about a simple LM7806 regulator? Super cheap, an Amp or two, and rock stable.

C"mon Don, that would be way to logical!! It would also be EASY!!!

LM7806 (http://www.fairchildsemi.com/pf/LM/LM7806.html)

[gixxy]

If you want control and to source it via sidecar, use a Digital output, and some transistors to get the 6v from the servo pins.

If you want to continue using the Spike Relay and just stop it from reversing the polarity. You simply need to modify your PWM wire.

1. Cut the Red/Middle wire, leaving ~2 inches of the wire on the Relay side.
2. Cut the Black/Ground wire at about the same place.
3. Strip the red wire on the Relay side, and both of the Black ends.
5. Put heatshrink on the long black wire, and solder the three ends together.
6. Slide heatshrink over solder joint and apply heat.

The M- pin will now always be grounded, and by setting the relay forward/off in the code you can change M+ to +12v or GND, however if you set the relay to reverse, you will get GND on both pins because the middle pin is still pulled to (-).

Here is the spike User Guide that give info on how the input to the Spike works: http://content.vexrobotics.com/docs/spike-blue-guide-sep05.pdf

[DonRotolo]

The M- pin will now always be grounded...or, of course, you can just *ground* the wire instead of connecting it to the spike. That is, just use one side (M+) for power and connect the other wire directly to the PDB.

Naaaah...too simple.:p

[gixxy]

...or, of course, you can just *ground* the wire instead of connecting it to the spike. That is, just use one side (M+) for power and connect the other wire directly to the PDB.

Naaaah...too simple.:p

Eh, I am controls Captain, I think of fixes on the control side. I also like my wires going into the same place. Although I don't think grounding to the PDB will work unless you put it on a socket that already has a load on it, which I don't think is allowed (one wire per wago).

[Al Skierkiewicz]

Although I don't think grounding to the PDB will work unless you put it on a socket that already has a load on it, which I don't think is allowed (one wire per wago).

The black terminals on the PD are always connected to the negative terminal of the battery even if there is no breaker in place.

[Doc Wu]

No need to cut up a Pwm cable, just wire your load like this.

http://warlocks1507.com/wiki/spike.JPG

[EricVanWyk]

C"mon Don, that would be way to logical!! It would also be EASY!!!

LM7806 (http://www.fairchildsemi.com/pf/LM/LM7806.html)

Or you could use a linear regulator with an enable pin. That would eliminate the need for a spike or separate FET.