why do electrons hate me today?
32 views
Skip to first unread message
Mykle James Hansen
Apr 22, 2024, 6:30:39 PMApr 22
to dorkbotpd...@googlegroups.com
Hi all,
There is something I don’t get about power regulators, grrrr …
I have this 3v xenon flasher that I want to power from a USB brick.
The flasher works great on my bench power supply at 3.3v. It seems to
draw close to an amp in the very brief moments when it flashes.
I previously had the flasher running straight from 5v,
so I know that the USB battery pack has plenty of current for the job.
However, 5 volts was damaging the bulb. So now, after swapping
the bulb & re-testing that it’s a good bulb, I have installed
a NTE1904 chip that down-regulates 5v to 3.3v & can in theory deliver
1 amp constantly. The regulator output reads 3.3v on my meter,
but when I connect the bulb to it the meter drops down flickering
around 1.5 -- and I also hear crazy amounts of powerline noise backflowing into
a nearby Teensy+Prop Shield powered by the same USB brick. Ugh.
I thought it might be some issue with the flasher taking too-short
sips of current, so I tried an electrolytic capacitor in
parallel with the xenon bulb. But when I do that, the entire
power system goes bleh; the upstream Teensy won’t boot.
Maybe the USB PS reads this as a short & shuts down?
I tried it with a 2200uf cap first, then again with a 70uf cap,
same thing.
Is there something strange about the load that xenon bulb flash
circuits put on a power supply? Or something I don’t understand
about how to use a voltage regulator? Am i just DUM?
Your opinions are solicited!
(FWIW this is all so I can pretend to be a submarine at FATHOM next month.)
-mykle-
There is something I don’t get about power regulators, grrrr …
I have this 3v xenon flasher that I want to power from a USB brick.
The flasher works great on my bench power supply at 3.3v. It seems to
draw close to an amp in the very brief moments when it flashes.
I previously had the flasher running straight from 5v,
so I know that the USB battery pack has plenty of current for the job.
However, 5 volts was damaging the bulb. So now, after swapping
the bulb & re-testing that it’s a good bulb, I have installed
a NTE1904 chip that down-regulates 5v to 3.3v & can in theory deliver
1 amp constantly. The regulator output reads 3.3v on my meter,
but when I connect the bulb to it the meter drops down flickering
around 1.5 -- and I also hear crazy amounts of powerline noise backflowing into
a nearby Teensy+Prop Shield powered by the same USB brick. Ugh.
I thought it might be some issue with the flasher taking too-short
sips of current, so I tried an electrolytic capacitor in
parallel with the xenon bulb. But when I do that, the entire
power system goes bleh; the upstream Teensy won’t boot.
Maybe the USB PS reads this as a short & shuts down?
I tried it with a 2200uf cap first, then again with a 70uf cap,
same thing.
Is there something strange about the load that xenon bulb flash
circuits put on a power supply? Or something I don’t understand
about how to use a voltage regulator? Am i just DUM?
Your opinions are solicited!
(FWIW this is all so I can pretend to be a submarine at FATHOM next month.)
-mykle-
Nathan McCorkle
Apr 23, 2024, 12:35:12 AMApr 23
to dorkbotpd...@googlegroups.com
My first feeling is you need some more, smaller, caps. Probably closer to the regulator than the load. I found this example circuit with 100nF on the input: https://www.bamlog.com/remotepotbox.htm
And they say it's equivalent other than pinout to lm1117, and that datasheet has 10uF on input and 100uF on output for most of the examples, it Even has one with a protection diode from output to input pin... and a xenon flash module sounds like it could have some high voltage spike with ringing activity associated... So it'd be an easy addition. You could also try ferrite beads in between the flasher module and the output of the nte1904.
Assuming the flasher module has some sort of architecture like the MAX8622, it shows the load in parallel with 150uF cap.
Hope you get it working! You may need to simulate in SPICE somehow, to really fine tune. But I think you'd need to collect some data first with an oscilloscope or maybe spectrum analyzer, when the load is hooked up to a working-configuration power supply.
-Nathan
--
You received this message because you are subscribed to the Google Groups "dorkbotpdx-blabber" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dorkbotpdx-blab...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dorkbotpdx-blabber/AF6EF3C1-CF7A-4134-AFBE-FBDB03DCAABE%40mykle.com.
jason
Apr 23, 2024, 1:15:05 AMApr 23
to dorkbotpd...@googlegroups.com
On 4/22/24 3:30 PM, Mykle James Hansen wrote:
> I have this 3v xenon flasher [snip]
I would stop right there and ask if a "3v xenon flasher" is a
common-enough thing that people know about....or is it some
random/specialized board/assembly you bought to solve a thing.
Assuming the latter, can you link to any information about what you
have? This seems like a good starting point...any specs or design
information beyond 3.3V?
How are you determining that it's drawing an amp when it flashes? That
flash is probably pretty quick. Quick enough for your meter not the show
a peak current draw I suspect. It's probably more.
USB power bricks are notoriously awful shit, which is maybe also
demonstrated by the fact that your device works well on a bench supply.
What are the specs of the brick, and how do those differ from how you've
dialed in the bench supply?
-jason
Dan Stahlke
Apr 23, 2024, 12:38:23 PMApr 23
to dorkbotpd...@googlegroups.com, Nathan McCorkle
Some regulators have both a minimum and
a maximum capacitance specification on the output side. The
NTE1904 lists only a minimum of 2uF and recommended of 10uF. So
apparently there is no maximum.
However, USB specifies a maximum 10uF capacitance. This is combined capacitance for the capacitors on both sides of the regulator plus whatever's in the strobe circuit. Otherwise, as you suspected, the USB supply sees it as a short. Now, it's not like 10uF is okay and 11uF is a guaranteed fail. It's more like it gets sketchier and sketchier as you go beyond 10uF and some supplies will work and others won't.
You might try a 1 ohm or 2 ohm resistor in series on the input power wire (between the regulator and the USB) to limit the current. It'll get hot and it'll not supply enough voltage to the strobe, but at least this will help to prove that inrush current is the problem. And maybe you can short the resistor once it gets going. Not as a permanent solution, but just as a first step towards debugging.
I had a similar problem recently with an incandescent light bulb. It's rated 12 volts and just under 3 amps, and my USB supply is rated for that (USB PD can give you 12 volts). If the bulb is warm, it works. If the bulb is cold then it doesn't. The cold bulb has low resistance and so draws a ton of current. Only for a split second until it gets warm, but my USB charger shuts it down immediately.
- Dan
However, USB specifies a maximum 10uF capacitance. This is combined capacitance for the capacitors on both sides of the regulator plus whatever's in the strobe circuit. Otherwise, as you suspected, the USB supply sees it as a short. Now, it's not like 10uF is okay and 11uF is a guaranteed fail. It's more like it gets sketchier and sketchier as you go beyond 10uF and some supplies will work and others won't.
You might try a 1 ohm or 2 ohm resistor in series on the input power wire (between the regulator and the USB) to limit the current. It'll get hot and it'll not supply enough voltage to the strobe, but at least this will help to prove that inrush current is the problem. And maybe you can short the resistor once it gets going. Not as a permanent solution, but just as a first step towards debugging.
I had a similar problem recently with an incandescent light bulb. It's rated 12 volts and just under 3 amps, and my USB supply is rated for that (USB PD can give you 12 volts). If the bulb is warm, it works. If the bulb is cold then it doesn't. The cold bulb has low resistance and so draws a ton of current. Only for a split second until it gets warm, but my USB charger shuts it down immediately.
- Dan
To view this discussion on the web visit https://groups.google.com/d/msgid/dorkbotpdx-blabber/CA%2B82U9KCXpwdWF-CSOPUVQzW%3DOK-McdVvqhOzawzD7rRikMz0g%40mail.gmail.com.
Jerry Biehler
Apr 23, 2024, 8:07:15 PMApr 23
to dorkbotpd...@googlegroups.com, Nathan McCorkle
Unless you are measuring with a scope you may have an inrush way higher than what you think you have. Best bet is try a resistor to limit the inrush. You could also use a PTC inrush limiter or just use a lipo to drive the flash directly.
-Jerry
To view this discussion on the web visit https://groups.google.com/d/msgid/dorkbotpdx-blabber/d89c4888-f641-4e9b-96ac-c9afb16fb0a1%40stahlke.org.
Mykle James Hansen
Apr 25, 2024, 3:14:30 PMApr 25
to dorkbotpd...@googlegroups.com
Thanks thanks thanks!
> On Apr 23, 2024, at 1:21 PM, dorkbotpd...@googlegroups.com wrote:
> jason <ja...@noisybox.net>: Apr 22 10:14PM -0700
to make before they switched everything to LEDs.
https://picclick.com/New-Visi-Flare-WP-Flashing-Emergency-Strobe-Safety-224350367780.html
I haven’t found a schematic though, and the bulb unit is sealed.
Normally it’s powered by two D-cell batteries in series.
I’m using it because the flash shines through a bunch of colored
gels, and the xenon bulb has a really nice wide flat color spectrum,
unlike an LED.
Every circuit variation I’ve looked at begins with charging a large capacitor.
That would cause a major inrush current def. But I’ve tried various
resistors in series (5 ohm, 8 ohm, 100 ohm, 10k ohm) to
limit that, and still got no blinks.
(BTW I find this whole site awfully charming:
"Here is the second, 220V xenon strobe light circuit.
It looks similar to before circuit.
I do not want to waste time.
You too, right?”
https://www.eleccircuit.com/xenon-strobe-light-110v-by-scr/ )
> USB power bricks are notoriously awful shit, which is maybe also
> demonstrated by the fact that your device works well on a bench supply.
> What are the specs of the brick, and how do those differ from how you've
> dialed in the bench supply?
To clarify: the bulb does flash when I hook it directly to the
brick, it’s just over-voltaged at that point & I damaged
one bulb already doing that. But the brick can provide the
full current at 5v to the bulb when the regulator is not involved,
and it provides constant 5v to the regulator despite the
bulb not flashing. So I doubt the brick is the problem.
> Nathan McCorkle <nmz...@gmail.com>: Apr 22 09:34PM -0700
So I have added that protection diode, and also those
input & output caps. I have re-tested that the
regulator works for other loads, and that the flash
bulb is still not damaged despite all my testing.
But no blinky, still.
> Jerry Biehler <jerry....@gmail.com>: Apr 23 05:07PM -0700
of the electronic junk in my scrap pile? Alas I think
I’ve run out of time for ordering a part for this.
(Now I’m wishing I’d just bought one of these 5v xenon flash
modules for drones that you can get from China if you
wait three weeks.)
-mykle-
> On Apr 23, 2024, at 1:21 PM, dorkbotpd...@googlegroups.com wrote:
> jason <ja...@noisybox.net>: Apr 22 10:14PM -0700
>
> I would stop right there and ask if a "3v xenon flasher" is a
> common-enough thing that people know about....or is it some
> random/specialized board/assembly you bought to solve a thing.
>
> Assuming the latter, can you link to any information about what you
> have? This seems like a good starting point...any specs or design
> information beyond 3.3V?
It’s a VisiFlare WP, a flashlight-sized unit that Tektite used
> I would stop right there and ask if a "3v xenon flasher" is a
> common-enough thing that people know about....or is it some
> random/specialized board/assembly you bought to solve a thing.
>
> Assuming the latter, can you link to any information about what you
> have? This seems like a good starting point...any specs or design
> information beyond 3.3V?
to make before they switched everything to LEDs.
https://picclick.com/New-Visi-Flare-WP-Flashing-Emergency-Strobe-Safety-224350367780.html
I haven’t found a schematic though, and the bulb unit is sealed.
Normally it’s powered by two D-cell batteries in series.
I’m using it because the flash shines through a bunch of colored
gels, and the xenon bulb has a really nice wide flat color spectrum,
unlike an LED.
Every circuit variation I’ve looked at begins with charging a large capacitor.
That would cause a major inrush current def. But I’ve tried various
resistors in series (5 ohm, 8 ohm, 100 ohm, 10k ohm) to
limit that, and still got no blinks.
(BTW I find this whole site awfully charming:
"Here is the second, 220V xenon strobe light circuit.
It looks similar to before circuit.
I do not want to waste time.
You too, right?”
https://www.eleccircuit.com/xenon-strobe-light-110v-by-scr/ )
> USB power bricks are notoriously awful shit, which is maybe also
> demonstrated by the fact that your device works well on a bench supply.
> What are the specs of the brick, and how do those differ from how you've
> dialed in the bench supply?
brick, it’s just over-voltaged at that point & I damaged
one bulb already doing that. But the brick can provide the
full current at 5v to the bulb when the regulator is not involved,
and it provides constant 5v to the regulator despite the
bulb not flashing. So I doubt the brick is the problem.
> Nathan McCorkle <nmz...@gmail.com>: Apr 22 09:34PM -0700
>
> My first feeling is you need some more, smaller, caps. Probably closer to
> the regulator than the load. I found this example circuit with 100nF on the
> input: https://www.bamlog.com/remotepotbox.htm
>
> And they say it's equivalent other than pinout to lm1117, and that
> datasheet has 10uF on input and 100uF on output for most of the examples,
> it Even has one with a protection diode from output to input pin... and a
> xenon flash module sounds like it could have some high voltage spike with
> ringing activity associated... So it'd be an easy addition. You could also
> try ferrite beads in between the flasher module and the output of the
> nte1904.
Ooof, ringing, yeah, I see the problem there.
> My first feeling is you need some more, smaller, caps. Probably closer to
> the regulator than the load. I found this example circuit with 100nF on the
> input: https://www.bamlog.com/remotepotbox.htm
>
> And they say it's equivalent other than pinout to lm1117, and that
> datasheet has 10uF on input and 100uF on output for most of the examples,
> it Even has one with a protection diode from output to input pin... and a
> xenon flash module sounds like it could have some high voltage spike with
> ringing activity associated... So it'd be an easy addition. You could also
> try ferrite beads in between the flasher module and the output of the
> nte1904.
So I have added that protection diode, and also those
input & output caps. I have re-tested that the
regulator works for other loads, and that the flash
bulb is still not damaged despite all my testing.
But no blinky, still.
> Jerry Biehler <jerry....@gmail.com>: Apr 23 05:07PM -0700
>
> Unless you are measuring with a scope you may have an inrush way higher than what you think you have. Best bet is try a resistor to limit the inrush. You could also use a PTC inrush limiter or just use a lipo to drive the flash directly.
Do you think there might be a PTC inrush limiter in any
> Unless you are measuring with a scope you may have an inrush way higher than what you think you have. Best bet is try a resistor to limit the inrush. You could also use a PTC inrush limiter or just use a lipo to drive the flash directly.
of the electronic junk in my scrap pile? Alas I think
I’ve run out of time for ordering a part for this.
(Now I’m wishing I’d just bought one of these 5v xenon flash
modules for drones that you can get from China if you
wait three weeks.)
-mykle-
Dan Stahlke
Apr 25, 2024, 5:35:04 PMApr 25
to dorkbotpd...@googlegroups.com, Mykle James Hansen
On 4/25/24 12:14 PM, Mykle James Hansen
wrote:
I’m using it because the flash shines through a bunch of colored gels, and the xenon bulb has a really nice wide flat color spectrum, unlike an LED.
White LEDs can have decent spectrum. Some are better than others.
They tend to have continuous spectrum between red and green, but
with a gap just before blue. Blue is kind of its own peak. At
least that's how it is for a few I tested.
Mykle James Hansen
Apr 26, 2024, 5:26:41 PMApr 26
to dorkbotpd...@googlegroups.com
> On Apr 25, 2024, at 2:21 PM, Neal <nse...@gmail.com> wrote:
>
> Anyway, since you're just replacing two D cells from what I get looking at the strobe company website, you're overthinking it. Just use two 4001 type diodes in series and call it good. You'll get quite close to the voltage from two alkaline cells after dropping across the two diodes.
>
> No regulation this way but who cares, batteries don't provide that either.
WINNER WINNER CHICKEN DINNER! Thanks Neal, you’re absolutely right.
This is way simpler than what I was trying to do, and it works.
Here’s some video of the flasher in action!
https://youtu.be/3512QfEk1Ho
I know I always say this but I so, so, so appreciate having a community
of people I can ask these sorts of questions. Not that I couldn’t maybe
find a helluva lot of help on the internet, but Dorkbot just feels
a lot more friendly to me. Thank you all!
-m-
Greg Peek
Apr 26, 2024, 5:38:25 PMApr 26
to dorkbotpd...@googlegroups.com
Loved the video!
Quite the contraption.
--
You received this message because you are subscribed to the Google Groups "dorkbotpdx-blabber" group.
To unsubscribe from this group and stop receiving emails from it, send an email to dorkbotpdx-blab...@googlegroups.com.
To view this discussion on the web visit https://groups.google.com/d/msgid/dorkbotpdx-blabber/76F798D6-035F-4C9C-82BA-E819E032163F%40mykle.com.
Russell Senior
Apr 26, 2024, 6:24:11 PMApr 26
to dorkbotpd...@googlegroups.com
Love the sonar pings. I am pretty sure you could spot a Jaguar Shark in that thing!
--
Russell
To view this discussion on the web visit https://groups.google.com/d/msgid/dorkbotpdx-blabber/CAE%2BDGWoGqUmxfon22SqQS7hRxA2m%3DJdiBYTnGNE0oBAgS3jTnw%40mail.gmail.com.
Erik Lane
May 2, 2024, 9:15:48 PMMay 2
to dorkbotpd...@googlegroups.com
I have a couple of these. I got them from someone local years ago, and it might have even been from someone on this list?
I put in some time, and got one of them working. The second one I started working on, but then got a bit discouraged when I realized two things: 1) I don't actually have a good spot to put it when I want to use it, and 2) I basically never need one. These days I could end up getting a small USB unit with even more capabilities, though I don't know if these old analog machines might have cleaner output? Maybe not - that's the only real advantage I can think of.
I took a quick look for specs, and noticed that on eBay they're listed for over $1000 for most of the listings. I don't believe that for a minute.
These both currently have their top and bottom covers off, so they could use vacuuming. I stored them in the back of the bedroom with the front panel facing up, so not too much would have gotten in there. I have the covers, and then the screws are in a little baggie, so they should at least mostly be there. I was comparing voltages and waveforms at different points to narrow down the issue when I stopped cold-turkey on these.
At this point, I'm also not sure which one of the two is the working one. It had a bunch of bad tantalum capacitors, and if I remember right that was the only issue. So I would like these to go together, and since I'm not sure which one works, that's best for both parties, I think.
Free to a good home. :) They're taking up space here and just going to waste, since they only sit around and never get used.
I'm in Vancouver, just barely over the I-205 bridge, but I'm sure we could also work out someplace to meet or something. I'm motivated, and also I *really* do want them to get used if at all possible.
Thanks,
Erik
Reply all
Reply to author
Forward
0 new messages