EDIT: Thanks to @Sensacell for spotting an error. The current requirements for the LED have been revised. Just to be clear, there isn’t intended to be anything tricky about these circuits, they are intentionally pedestrian but just complicated enough to allow for various solutions that use very different approaches. Don’t fear and gotchas or hidden problems, and if like @Sensacell you spot any, please let me know!

Here’s a challenge intended to tease out some information about the state of design philosophy among the peers here on AAC.

This challenge is open to anyone, no matter your level of knowledge or experience. In fact, it would be extremely helpful to hear from people who consider themselves neophytes or less experienced. Your understanding of the best choices in the cases below is, for purposes of this investigation, equally important to the most experienced and knowledgable among us.

The goal is to design circuits that meet the requirements of the three scenarios below. These solutions need only be in principle. Schematics, simulations, and other rigor are certainly welcome but they aren’t required to answer the questions. However, the answer should include part numbers for actives and values for any passives that are critical to operation where possible. If for any reason you can’t calculate those values, answer anyway so the rest of the information is available.

Each of the circuits starts with certain baseline parts, as required:

• A 200mAH CR2032 Lithium Cell for Power [datasheet]
• 1 or 2 Blue LEDs for Output [datasheet]
• 0 or 1 10kΩ Linear Taper Potentiometer for Input [datasheet]

The rest of the components are part of the design. There are no unstated constraints. The baseline components are given, and can‘t be changed, and the stated goal for each circuit is the only requirement for this challenge except for the unfortunately vague requirement the circuit be practical.

The goal for these designs is to come up with a solution you would actually build given the conditions. So, somewhat baked in are the optimizations you would make so the circuit could be useful. This isn’t a perfect specification, more of a spirit I ask you to employ.

There are three circuits and each has three versions for a total of nine. There is no requirement to provide all nine circuits to participate but for reasons of the data I would appreciate if you could do either three versions of one circuit, or one version of all three, Here are the version parameters:

• Lowest Part Count—the fewest parts aside from the required ones
• Lowest Cost—cheapest to build in round numbers, don’t worry about too much rigor here, good faith estimates are fine
• Longest Battery Life—this may or may not be the same thing as lowest power consumption, but that’s probably good enough.

So, here are the circuits, each involves blinking an LED according to parameters given:

1. Simple Blinker
Design a circuit that blinks an LED with a current of ~10mA ±2mA at a fixed rate between 5Hz and 10Hz. It does not require adjustability just a stable output in the given range. This should be accomplished with the fewest parts in addition to the given cell and one LED, so it starts at a count of 2. What would you choose, and why?

2. Variable Blinker
Design a circuit that blinks an LED from 1Hz to 100Hz at ~10mA ±2mA using the suppled cell, LED, and potentiometer. The accuracy of the interval just has to be basic, the adjustment range should be the full rotation of the pot, so very close to 1Hz when it is fully left and 100Hz when fully right.

3. Alternating Blinker
Design a circuit that alternately blinks two LEDs between 2Hz and 4Hz with a current of ~10mA ±2mA using the supplied cell and two of the LEDs. Like the first circuit, it must be a fixed rate in the range, no adjustment required.

So, what are your solutions to these challenges? I look forward to all answers.

 

The trick is the Vf of the blue LED's - higher than the battery voltage.

 

The trick is the Vf of the blue LED's - higher than the battery voltage.

Oops. I made an error in the specification. The current requirements should be 10mA, not 20. I had that in notes and forgot about it. There aren’t supposed to be any tricks here! I’ll revise it.

Thanks for spotting that!

 

Oops. I made an error in the specification. The current requirements should be 10mA, not 20. I had that in notes and forgot about it. There aren’t supposed to be any tricks here! I’ll revise it.

Thanks for spotting that!

No, don't change that, it's actually what makes this a sexy challenge.
Getting 10 mA to flow in a blue LED from ~ 2.75 V. is going to require a boost circuit of some kind.

 

No, don't change that, it's actually what makes this a sexy challenge.
Getting 10 mA to flow in a blue LED from ~ 2.75 V. is going to require a boost circuit of some kind.

Well, it changes the purpose of the challenge which is to get an idea about approaches to real world component selection but it can be a supplementary one that I will add as an addendum!

 

EDIT: Thanks to @Sensacell for spotting an error. The current requirements for the LED have been revised.

No, don't change that, it's actually what makes this a sexy challenge.
Getting 10 mA to flow in a blue LED from ~ 2.75 V. is going to require a boost circuit of some kind.

So here is the @Sensacell expert challenge (which I hope won’t cannibalize the main one):

Complete any one of the three circuits with the three versions but with a current to the LED of ~20mA ±4mA.

 

You first.

 

No spec for the number of LED pulses? The weak little coin battery can charge a capacitor to blink an LED. But how many times before the battery is dead?

 

No spec for the number of LED pulses? The weak little coin battery can charge a capacitor to blink an LED. But how many times before the battery is dead?

No, I purposely didn’t have a minimum battery life but that battery in particular is rated at 200mAH so it would be long enough to know it was doing something.

 

You first.

I have ideas but I am going to wait until people have a chance to propose some things to avoid skewing the results one way or another. I really want to see that the consensus is concerning the overall approach people choose—or avoid.

 

The little coin battery is rated for 200mAh with a 3mA load and a finishing voltage of 2.0V. A graph in its datasheet shows its voltage and capacity dropping quickly to 2.0V with a 10mA load.

 

The little coin battery is rated for 200mAh with a 3mA load and a finishing voltage of 2.0V. A graph in its datasheet shows its voltage and capacity dropping quickly to 2.0V with a 10mA load.

Yes, but it looks like that results in a runtime of at least hours. So it’s fine for this exercise.

 

Here are the version parameters:


Lowest Part Count—the fewest parts aside from the required ones
Lowest Cost—cheapest to build in round numbers, don’t worry about too much rigor here, good faith estimates are fine
Longest Battery Life—this may or may not be the same thing as lowest power consumption, but that’s probably good enough

For what it's worth, designing around such constraints might not be the best approach. Aiming for the lowest part count and it may not work as intended. Go the "lowest cost" route and reliability issues could arise.

First, get the darn thing working. Then refine the design until all of the cruft is removed, assuming energy consumption requirements have been met. Only then should one even consider looking into ways to lower costs.

Doing it the other way around imo is just going to lead to the same kind of quality one might expect from the kind of mass-produced Asian products that have been pawned off onto consumers for decades now. (Which is to say, junk.)

 

For what it's worth, designing around such constraints might not be the best approach.

Well, this depends on the goal, doesn’t it?

The goal here is to work out what the approach of the designer is in each case, not to create a product to sell.

 

well I just really want to be notified of the results, but to contribute. For the longest battery life my design choice would probably start around an led driver with a pwm output that I can fix with a resistor at like a 1 - 10% duty cycle. Then work my blinking circuit to interact with the enable/shutdown pin for the led driver chip. The blinking part of it, still working on that. But thats how I would start the design process if battery life was the main design criterion.

 

Hmmm. not a hot topic?
I love design challenges, but these blinkers are not getting my engine running.
Maybe another line of design is in order?

 

Hmmm. not a hot topic?
I love design challenges, but these blinkers are not getting my engine running.
Maybe another line of design is in order?

Well, it's unfortunate that things are not working out as I'd hoped. But I had a goal in mind for the results. On the general idea of challenges perhaps this post will provoke someone else to produce one that is more interesting.

 

Cant see any rationale in such design, but why not to meditate about 555?

 

Hello there,

I am not sure I understand the actual constraints, if there are any.

For example, I think I understand that the LED has to blink. As to lowest parts count, is a tiny microcontroller chip allowed or not? I am asking because that seems like the lowest parts count solution unless of course it is allowed to use a self-blinking LED with a resistor. The self-blinking variety come in the form of a random blink duration too so that it looks like a burning candle.

 

Cant see any rationale in such design, but why not to meditate about 555?

Design challenges do not always follow the usual rationale completely, they are for other purposes sometimes. They sometimes stand in their own category of design philosophies.