low cost challenge

The cockney /accent/ isn't strong. The problems arise because it is an /argot/. "I went up the apple to get my whistle" is about neither fruit nor music.

OTOH just before Billy Connoley hit the big time, there were serious discussions about whether people would understand him. I have no idea whether he modified his Glaswegian accent.

Reply to
Tom Gardner
Loading thread data ...

U2 is a PNP BJT. Marking = P23 Diodes Inc DDTA115TCA PNP 200mw 50v 0.1A

100K in series with base.
--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

You just power the uP from say 2 AAs -> boost -> 3.3 volts -> high side current sense -> Vcc input - > LEDs hung directly off the uP port pins which have hardware PWM. You know the uP supply voltage, total supply current draw (the bulk of which is being pushed right into the LEDs) the LED equivalent average DC voltages, and possibly even the die temperature. You can do a lot with those pieces of information continually updated in software and you only need 1 external resistor to get it

It's the equivalent of an analog feedback loop control scheme but everything is done with PWM/in software, there are no DC control voltages anywhere.

The e.g. TPS6120 mentioned before can do 300 mA at 3.3V output with Vin at 2.4 volts @ around 93% efficiency, and the 8 pin AVRs 40 mA per pin, limited by probably about 1/2 watt dissipation for a DIP package.

It can provide that current when it's up and running, but cold and dark LEDs are like a dead short and it'll fall down into say three of them as a load or something. Either the converter or LEDs have to be ramped up but that can also be done via PWMing the shutdown pin. The shutdown pin's response is pretty fast, 10s of uS, maybe.

Ya but the OP said he only had AA batts to use!

Optics not my forte there another guy here who does dat

Reply to
bitrex

Perhaps the next generation of Google Assistant or Google Duplex will have an idiom decoder. "New Google AI Can Have Real Life Conversations With Strangers" Each language and dialect have their own unique idioms, which often change in both composition and meaning over the years. In science and engineering, the problem is aggravated by the overuse of acronyms and abbrevs.

In the early 1980's, I worked for Granger Associates, purchased by DSC (Digital Switch Corp) in 1984. Granger was originally a UK company so much of the engineering staff and management were from the UK or AU. I was from neither and therefore had great difficult understanding the lingo. I could not reliably distinguish between the UK and AU contingents. Such errors were considered a capital offense. Instead of the UK and AU contingents adopting a US accent, I was beginning to show signs of adopting their accent, idioms, slang, and profanity. Since I had no intention of following anyones instructions, the lack of communications had little effect on my work. It was also helpful for avoiding meetings, where my inability to understand about half of those attending was sufficient to grant me a special dispensation to not attend certain meetings. Hmmm... maybe unintelligible accents and incomprehensible idioms do have some benefit.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

Microchip makes PICs that are designed to be SMPS controllers. The program sets up the connections between the comparators, DACs, ADCs, and PWMs, then the hardware controller works independently of the CPU. The CPU is out of the loop. Neat widgets.

How does it manage the battery life if it doesn't know how long the "ride" is?

It's also hell on efficiency.

Huh? There are 5A, and above, synchronous switching supply chips on the market.

But series LiIons complicate charging. It's a lot easier to parallel them, and deal with the boost, if necessary.

Reply to
krw

I constantly work on projects that are well scoped and budgeted -- but it's probably important that, I'm only seeing the engineering side of those projects, through a very well organized design company. Our projects are just the engineering work. The projects, as the customer sees them, who knows...

Tim

--
Seven Transistor Labs, LLC 
Electrical Engineering Consultation and Contract Design 
 Click to see the full signature
Reply to
Tim Williams

I want more narrow-scoped projects I run into the problem where the customer says "use your discretion" and then they don't like your discretion.

Reply to
bitrex

you're almost funny.

NT

Reply to
tabbypurr

Nice. That would work for the flashlight. The problem is that there are 145 switching power supply regulators listed: It's going to take me a while find something appropriate. I also see a problem. Most flashlights use one or more 7135 linear regulator to provide a current source for driving one or more LED's. The big benefit here is that a linear regulator doesn't need an inductor. The cost of the 7135 ($0.12) and a PIC or Arduino controller (about $0.50) is probably much less than using one of the MicroChip PIC's (about $1.00) plus whatever an inductor costs.

Incidentally, some flashlight driver boards do have inductors and are used in boost mode for driving a series string on LED's from a single LiIon (or alkaline) cell or in buck mode for running a single LED from a 12V source:

I didn't say "manage". There are too few features that could be turned off by the BMS (battery management system) to significantly affect the battery life. It's not like a smartphone that can turn things off, kill background apps, reduce brightness, decrease sampling rates, etc to save power. Some of the better bicycle lights and a few LED flashlights have SoC indicators. Something like this: Most are just bar graphs which simply display the LiIon battery voltage along with some kind of number to represent percent of full charge remaining.

More complexity. I'm all for it, and it sure sounds like a good way to do it, but I'm not sure I could cram it onto the COB (chip on board) with heat sink "star" found in most flashlights.

Nope. There are plenty of LiIon battery packs sold for flashlight power that don't do any balance charging. Usually, they're the external battery packs used by the really high power bicycle lights. They might have a BMS board inside, like on laptop power supplies, which do not equalize the voltage across each of the cells, but merely protect the cells against over/under voltage, over current load, over current charging, and short circuit protection as used in LiIon power tools and laptop batteries.. Something like this for 2 cells:

Almost 1AM. Time to get off the computah.

--
Jeff Liebermann     jeffl@cruzio.com 
150 Felker St #D    http://www.LearnByDestroying.com 
 Click to see the full signature
Reply to
Jeff Liebermann

One of mine - the oldest one - seems similar to your picture in that it has about the same number of same sized components, but the newest has only 1 SOP8, 2 chip Cs, 2 chip Rs and a surface mount red LED, so I guess the older one is a PIC or similar and the newer is a custom part.

Cheers

--
Clive
Reply to
Clive Arthur

Yeah - don't need no fancy schmancy boost converter chip with current regulation to within 1% just to light a frigg'n LED

Get the cheapest 6pin sot23 microcontroller AVR Tiny or PIC10F or similar add a high current switch transistor, inductor and if can afford it a supply bypass capacitor too. 4 parts (excluding LED and button)

Transistor charges inductor during ON time, LEDs connected across inductor, light during the OFF time flyback. Regulate peak current by timing the ON time. Most LEDs can tolerate 5V reverse bias so the 3V reverse bias seen during inductor charging on time is not a problem.

If micro can measure its supply voltage then it can vary on time to compensate for ageing cells but that is gilding. Many users would accept light dropping as cells fade. Cheap and cheerful.

piglet

Reply to
piglet

They're not listed as SMPS controllers, rather as PICs. If you're interested, I'll get the numbers tomorrow.

At the cost of some serious waste power (wasted capacity and heat). Also, the minimum useful battery voltage has to be above the forward drop of the LED(s). ...some pretty big design compromises, IMO. Inductors aren't expensive.

You're paying *WAY* too much for PICs. What's your production run? One?

Why a 12V source? Seems like a waste to boost voltage then buck current. Do it all at once.

Brightness. Better somewhat dimmer for an entire ride than dark for part.

Gas gauge. Integrate current. The PIC should be able to handle that, too.

formatting link

Not a good idea without some serious management.

Reply to
krw

Chinese CMOS ASIC, COB (chip+blob) construction on single-sided board. The chip in die form is probably in the couple cent range, FAR cheaper than a micro.

Eg.

formatting link

Packaged cost is still lower than any micro.

--sp

Reply to
speff

Any suggestions for a good glob-top epoxy?

--
John Larkin         Highland Technology, Inc 

lunatic fringe electronics
 Click to see the full signature
Reply to
John Larkin

formatting link

One component so no mixing required.

--sp

Reply to
speff

I'll try some. Thanks.

--
John Larkin         Highland Technology, Inc 
picosecond timing   precision measurement  
 Click to see the full signature
Reply to
John Larkin

Do you have die bonding capability now? Cool.

--sp

Reply to
speff

ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.