PDA

View Full Version : Flashing Light



H3kirven
07-30-2006, 05:03 AM
I am using the Basic stamp to flash an incandescent light source. THe light requires a 10 VDC power supply and I am using a mosfet for the switching. I need a faster turn-on time. I want to initially supply about 60v to the light source and after a few milli-seconds switch the light to the 10v supply. Attached is a PDF showing two floating npn-mosfets and optoisolators. Is this the correct approach? Note that it will take two·TTL signals. Does anyone have a better approach? Another approach would be to use a single 60v supply and a single mosfet and use a voltage divider on one of the two TTL signals. The second approach may create excess heat and could also be temperature dependent. Thank you in advance.

······





·

Bruce Bates
07-30-2006, 08:25 AM
H3Kirven -

You have a couple of potential problems here, whether you realize it or not. An immediate over-voltage with an incandescent lamp will, over the long term, provide you with one thing, and one thing ONLY - SHORT LAMP LIFE! This can be overcome.

I also suspect you have had to use some fairly heavy MOSFETs to overcome the cold inrush current provided by any incandescent lamp. This too can be overcome.

Here is what I would suggest. I can't give you an exact voltage, but try 10% as a starter. In other words, supply the lamp with a constant 1.0 volt supply. This "warming voltage" will NEVER be turned off, unless the entire apparatus is shut down for a period of time. You may need to protect the "warming supply" with a diode, based on how you impliment the flashing voltage.

If when you want to FLASH, you then switch to 10 volts and drop the "warming supply" out of the circuit temporarily, that's fine; no protection is needed (SPDT kind of operation). If, however, you slap the 10 volt flashing voltage into/onto the existing "warming supply" you may need to prevent a backfeed of the higher voltage through the "warming supply", by using an appropriate diode.

The simple theory behind this relies on the definition of incandescense:

DEF: incandescense - emission by a hot body of radiation that makes it visible

Here is the key - "Emission (of photons) by a hot body of radiation ..."

Thus, PRE-HEATING is the key to faster incandescence. Make sense? The ideal pre-heating would be just shy of incandescense, unless current consumption is an issue.

Regards,

Bruce Bates

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
<!--StartFragment -->

metron9
07-31-2006, 09:47 AM
Can the light be replaced by a Luxeon LED? instant on / off and 100,000 hours of life

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Think outside the BOX!

H3kirven
07-31-2006, 07:45 PM
Bruce,
Thank you. I like the idea of providing an initial warming current. Also the 60v was just a number. I would probably use 20v to give four times the power at start up. I have considered using PWM without a capacitor, but the timing can be complicated. I could PWM at about 25 kHz which is faster than the warm up /cool times of an incandescent light source. I could also use a single 20v power supply. If I look at the output with a photo diode and a good tektronics o-scope and keep the peak intensity below the steady "on" state I think the filament will survive. What do you think? I am considering a dual processor system. One to give on/off times and another to give the PWM.

Or.. I could keep things simple and PWM the warm up current and switch to a steady on state using 10v.

Metron9,
I plan to visit an LED manufacturer this week. They make blue and white LEDs. Hopefully, this work will lead to LED headlamps one day.
Thank you.

Doug

metron9
07-31-2006, 07:59 PM
Your 3 years late.... on the headlight thing....
http://www.theautochannel.com/news/2003/10/12/170520.html

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
Think outside the BOX!

Bruce Bates
07-31-2006, 08:37 PM
Doug -

You're going to "owe me" when I get done imparting all this incandescent lamp knowledge on you :-)

All incandescent lamps have a filament within. Feel free to think of this filament as a SPRING, as that's exactly how it's constructed. The next concept I'm about to introduce you to is mechanical, electical and electro-mechanical, so please pay close attention.

If you vibrate any spring, by whatever means, the "stronger" the force imparted upon it, the more, and usually the longer it will vibrate. If you have a problem with this concept, we'll pull it to the side and discuss it separately. Somewhat needless to say, vibration is deadly to any filament lanp!

Some filament lamps have very firm and structured support systems within. If you'd like to see this first hand, purchase a CLEAR 120 volt rough service bulb and you can see the care taken in filament support. So too vibration service lamps, except the support is in the opposite plane.

I suspect you're speaking of a miniature lamp in your application, so let me merely say this. As best they can,·small lamp manufacturers use multi-plane filament supports for miniature lamps intended for rougher services, such as automotive applications. Said differently, IF you can find a lamp intended or used in automotive service, you would probably be doing yourself a favor even if you had to compromise some other characteristic (base, lumens, gas vs. vacuum filled, etc).

When ANY voltage/current is sent through ANY filament lamp a vibration results. Part of this is thermodynamic, part of this is frequency related (if AC or _PWM_!) and part of this is merely mechanical due to an outside stressor applied to the "spring".

The thought of a warming current/voltage is really just part of what one must consider. IF you choose to use PWM to drive it (and I wouldn't if it were me)· you have now introduced another dynamic force upon this "spring". As they always say YMMV (your mileage may vary).

Lastly, any filament lamp only has a finite lamp life, since everytime it's turned·ON a bit of the (usually) tungsten filament is vaporized. This is why you see "black spots" within the envelope (glass) of a filament bulb which is reaching end of life! INVARIABLY, you will see the "black spots" appear opposite the lamp base location OR 90 degeees in opposition to· the base, if side (horizontally) mounted. This is the vaporization of which I just spoke.

Please feel free to "pick my brain" if you choose to. It needs some weeding :-)

I hope you find the "warnings" helpful and may your incandescent lamps "live long and prosper" :-)

Regards,

Bruce Bates

H3kirven
08-02-2006, 04:56 AM
Bruce and others,

Any response as to whether the floating MOSFETs will work?

Bruce Bates
08-02-2006, 05:30 AM
H3kirven -

I've no idea what you mean by "floating MOSFETs".

Regards,

Bruce Bates

▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔▔
<!--StartFragment -->

DumB
08-02-2006, 06:47 AM
Hat's off Bruce Bates...

I wonder..... why you are not in Parallax Support Team. ( May be u got... some brains!)

Unknown@hell.com

H3kirven
08-02-2006, 08:28 PM
At the beginning of this thread, I posted a diagram. In the diagram, there are two n-type MOSFETs. The MOSFETs are not tied directly to ground as they are found in normal use. The load is "below" the MOSFETs. Hence the MOSFETs and opto isolators are floating. Has anyone any experience with this type of circuit? If so, is there anything wrong with the circuit? Thank you.
Doug

Bruce Bates
08-02-2006, 10:04 PM
H3kirven -

Although I'm basically a software guy, I do know something about "floating grounds" and differential voltages, so let me offer the following, whether it applies to your situation or not. My apologies if it's not applicable.

Any circuit with "floating" components may do fine by itself. So long as the circuit or circuits which are a direct part of it are "tempered" to this·"floating component"·methodology. Other circuits which may interface with this circuity is where the problems _may_ lie.

Underlying this whole situation is the actual definition of "voltage". Without going into great detail, "voltage" is defined as the DIFFERENCE in two electrical potentials, as expressed in VOLTS.

Ex. 1

Considerations -

A circuit with "floating components" (somewhat nebulous definition) is interfaced directly with·some external circuit, with no consideration given to the "floating components" or the voltages output from the circuit in question. The voltages it passes on are +15 VDC and -3 VDC only.

Coments-

The "floating circuit" components recognize (by whatever means) that (say) the +15 VDC and the -3 VDC onboard is representitive of 12 VDC. If or when the +15 VDC and - 3 VDC (12 volt differential) are connected to an external·circuit which has no similar "floating components" of the same type and caliber, and a circuit which has (at least) a "solid" ground, and which makes no consideration for this "floating" basis, the - 3 VDC will be "drawn" to ground (shorted so to speak). The positive voltage may be altered, distorted or misunderstood as well. Problems are afoot!

Ex. 2

Considerations -

A different circuit with "floating components" is interfaced indirectly with an external circuit,·with consideration given to the "floating components" and the differential voltages output from the circuit in question.

Comments -

The "floating circuit" section recognizes (by whatever means) that +15 VDC and the -3 VDC is representitive of 12 VDC, and so can this external circuit, if it is designed to do so. There is nothing inherently "wrong" with differential voltages, and many times they can serve a very important purpose. If or when· the +15 VDC and - 3 VDC (12 volt differential) are connected to a this pre-prepared external circuit (through an interface) which is prepared to accept a differential, rather than an actual voltage, it can easily accomodate this actual voltage vs. differential voltage situation.

I hope that was somewhat helpful. More than that is up to the numerous analog gurus!

Regards,

Bruce Bates

pjv
08-03-2006, 12:26 AM
Hi Bruce;

Not to pick a fight, but would +15 VDC and -3VDC represent 18 Volts? I'm sure that's what you meant.

Cheers,

Peter (pjv)