EEVblog #1284 - How Bad Product Design Kills The Environment

[thm_w]

"UL 217 – Single and Multiple Station Smoke Alarms, requires an A-weighted sound pressure level of at least 85 decibels (dBA) when measured at a distance of 10 feet from the horn" 10ft = 3m

This one is claiming 85dB 2mA at 5Vpp:
https://www.alibaba.com/product-detail/SS3038T3P-Pin-Type-Piezo-Piezoelectric-Buzzer_62050104559.html

But that would likely be at 1m, so you'd need something louder.. 90dB or so?

[nctnico]

What about a low-current dropper (say 1mA or less) + a couple of Farad supercap which can store enough energy for 15min sounding?

You would need a 15V supercap. That gets pricy.

Right. Supercap is "no go", dropper as well. NCP785A  is excellent, only it has max 10.5mA specs As a solution I see either external power transistor (?) or electrolytic cap + sound "chirps" with let's say 20% duty cycle resulting peak ~50mA sound driver current. - Obviously if such sound chirps allowed by regulations.

Better yet... run the beeper from rectified mains directly through a dropper capacitor. The circuit itself only needs 50uA.

[langwadt]

You just demonstrated typical trap for young players of "Design By Inspection". Engineers who made 80mA dropper instead of 1mA, made it for for some quite obvious reason - ALARM. You definitely can't produce audible alarm using 15v @50uA supply, unless you are building "smoke alarm earbud". Only actually making whole device and measuring SPL, you will see how much power is actually needed. Could be so that 10.5mA @ 15V of NCP785A is not even enough.

What about a low-current dropper (say 1mA or less) + a couple of Farad supercap which can store enough energy for 15min sounding?

use two capacitive dropper supplies. a low power one for normal operation, a high power one driving the sounder
only enabled hen needed

[thm_w]

use two capacitive dropper supplies. a low power one for normal operation, a high power one driving the sounder
only enabled hen needed

Wonder if a straight up dropper resistor would be any cheaper or not, probably not much difference. 50uA at 240V = ~12mW.
They have 220V rated piezos, but the one I found is not that cheap (~$5), active circuitry inside: https://www.cuidevices.com/product/resource/cpe-422ac.pdf

[nctnico]

use two capacitive dropper supplies. a low power one for normal operation, a high power one driving the sounder
only enabled hen needed

Wonder if a straight up dropper resistor would be any cheaper or not, probably not much difference. 50uA at 240V = ~12mW.
They have 220V rated piezos, but the one I found is not that cheap (~$5), active circuitry inside: https://www.cuidevices.com/product/resource/cpe-422ac.pdf

That one has internal electronics. The beeper in the smoke alarm is probably just a piezo disk. Connected in series with a capacitor so it doesn't get full mains and all you need to drive it is a mains rated MOSFET (and some other components).

[gf]

"UL 217 – Single and Multiple Station Smoke Alarms, requires an A-weighted sound pressure level of at least 85 decibels (dBA) when measured at a distance of 10 feet from the horn" 10ft = 3m

This one is claiming 85dB 2mA at 5Vpp:
https://www.alibaba.com/product-detail/SS3038T3P-Pin-Type-Piezo-Piezoelectric-Buzzer_62050104559.html

But that would likely be at 1m, so you'd need something louder.. 90dB or so?

The referenced page sais 85dB at 10cm, so it would need to be even more louder...
[ In Europe, EN 14603 also requires 85dB @3m. ]

Better yet... run the beeper from rectified mains directly through a dropper capacitor. The circuit itself only needs 50uA.

In case of mains power failure it still must run from the battery, though.

[kcbrown]

Have you talked to TI about buying a million parts? If not, you have no idea whatsoever what the volume price really is. However, you know by inspection that its not as high as $1, as there would be no market for it at that price.

No.  But neither have I talked to anyone about buying the parts for a million low-power SMPS sections, either.  The comparison is about as apples-to-apples as can be made.  Cheap SMPS controllers seem to be in the mid teens to low tens of cents each in quantities of 1000, from sites like Mouser.  The rest is jellybean parts.

Perhaps TI's thousands-quantity price is a substantially higher multiple of their millions-quantity price than the multiplier that applies to low-power SMPS components.  On that, I just can't say.  The only thing I have to go by is the thousands-quantity prices.  Why do you believe that those prices aren't comparable, or aren't indicative of the millions-quantity relative price?

As for the market for the TI part, well, if board space is at a premium and low standby dissipation is also a requirement, the TI part looks like just the thing, so it does fill a niche.

[nctnico]

Better yet... run the beeper from rectified mains directly through a dropper capacitor. The circuit itself only needs 50uA.

In case of mains power failure it still must run from the battery, though.

No problem. Feed the beeper circuit from the battery through a diode or so. With a couple of days to work on this I'm sure someone can come up with a clever circuit which is both cheap and efficient.

[coppice]

Perhaps TI's thousands-quantity price is a substantially higher multiple of their millions-quantity price than the multiplier that applies to low-power SMPS components.  On that, I just can't say.  The only thing I have to go by is the thousands-quantity prices.  Why do you believe that those prices aren't comparable, or aren't indicative of the millions-quantity relative price?

This is how the semiconductor industry works. If you have a part with some novelty there are small volume users who will pay a good price for it, so you set the small volume price high. However, high volume users are ALWAYS driven by the BOM above all other factors. You need to offer attractive volume pricing to win any high volume design. The markets for this device are high volume things like smoke detectors, where you can lose a design over a fraction of a cent on the BOM, and utility meters, which will be attracted to the lack of any magnetic components (they are sensitive to tampering issues), but still need to meet aggressive BOM goals.

[splin]

use two capacitive dropper supplies. a low power one for normal operation, a high power one driving the sounder
only enabled hen needed

This is a cost effective solution requiring the addition of a triac or back to back mosfet switch (approx $.05), along with a dirt cheap and (likely much more reliable) 30uA dropper to power the sensor.  All more than paid for by reducing the main dropper capacitance by 50% or more as it no longer has to have degradation factored in.  The terrible power factor will be the least of your concerns when it triggers for real...

[Alti]

I'd say that to find "the best" solution for powering smoke detector we need to rigorously specify requirements first and "the best" is going to be the one with the lowest total cost of ownership (design+investment+runningcost+recycling) that meets all requirements.

Otherwise we are going to have next discussion about preferences and not about solution of real life problem.

-I bet for some "the best" is a networked solution tied to fire department, that costs 1k$ in investments but hey, this smoke detector is for fire brigade to get here asap.
-Others prefer resistive dropper (ok, it is 1W but do not want to have a safety critical device that heavily reles on capacitors or SMPS. Besides, 1W heater just adds up to heating cost, if you happen to live in Norway).
-Then others will "the best" with Viper (bridge + buck) because it saves electricity at the cost of BOM.
Etc.

Can anyone specify a starting point of requirements that a smoke detector has to meet to switch discussion about personal preferences to the one about engineering?

[nctnico]

The primary function of a smoke detector is to wake people up so they don't suffocate from smoke while their home is burning. See it as an early warning system so you can get out of your home safely after a fire has started.

[TerminalJack505]

I'd say that to find "the best" solution for powering smoke detector we need to rigorously specify requirements first and "the best" is going to be the one with the lowest total cost of ownership (design+investment+runningcost+recycling) that meets all requirements.

Otherwise we are going to have next discussion about preferences and not about solution of real life problem.

-I bet for some "the best" is a networked solution tied to fire department, that costs 1k$ in investments but hey, this smoke detector is for fire brigade to get here asap.
-Others prefer resistive dropper (ok, it is 1W but do not want to have a safety critical device that heavily reles on capacitors or SMPS. Besides, 1W heater just adds up to heating cost, if you happen to live in Norway).
-Then others will "the best" with Viper (bridge + buck) because it saves electricity at the cost of BOM.
Etc.

Can anyone specify a starting point of requirements that a smoke detector has to meet to switch discussion about personal preferences to the one about engineering?

BBM.

Well, to start with, the idea that only 50uA is required at all times other than when powering the sounder, is incorrect.  There are other high-current components in the circuit.

There is the IR LED used in the detection chamber and (I'm guessing) two indicator LEDs--likely one green LED and one red LED.

If the unit is mains powered then the green LED will be always on.  (This is the case for the US/Canada market.)  This will be about 5mA.  I don't think Dave has factored this in since I think he did his measurements while the unit was under battery power--in which case the green LED will not be illuminated.

The red LED will come on and stay on when the alarm sounds.  This might account for about 5mA of the 15mA drawn as reported by Dave.

The IR LED (and red LED) will come on when sampling for smoke every 10 seconds.  The IR LED likely draws about 5mA.  The IR LED and red LED don't necessarily have to be on simultaneously but the red LED needs to flash to indicate that a sample was done.

Sampling will be done even while the alarm is sounding so that it can turn the alarm off automatically once the smoke clears.  The sample can be done during a quiet period of the sounder.  If, for example, the temporal-3 tone pattern is being used then it can be done during one of the 1/2 second quiet periods or the 1.5 second quiet period.

Finally, there is the MCU.  Most of the time it will be in an energy efficient sleep mode.  It will obviously have to come out of sleep mode for sampling, low battery check, etc.  If the alarm goes off then the MCU will have to drive the Piezo sounder.  The 50uA is likely the current drawn when the MCU is sleeping.  The current drawn by the MCU when it is awake has likely already been accounted for in the 15mA drawn when in alarm.

So, the circuit could be designed so that it uses no more than ~20mA at once--powering the sounder, the green LED, and the red LED.  Or 15mA when under battery power.  (No green LED.)

My guess is that this detector wasn't designed so that it could guarantee that all of the high current components aren't all powered at once.  They basically have to assume that all of them are going to be active at once.  They then added a safety margin of what seems to be about 100%.

There is one other high-current event that I can think of.  When a low battery check is done the MCU will engage a circuit that will draw about 50mA from the battery.  This current will not be supplied by the mains, however.

All of this is based on my experience with a smoke detector project.

The project I worked on was a battery-only (not mains powered) wireless 'smart' detector.  In addition to the smoke detection circuit, it had a wireless transceiver that would talk to a security/fire panel.  It required much more power than the one Dave has.  It was powered by a single 3V CR123 lithium battery (not the best choice in my opinion) and the battery would need to be replaced yearly.

The high power requirements were mostly due to the fact that the device would have to wake up frequently and check for commands coming over the air.  (If any of the units went into alarm then ALL of them would start sounding--in sync. with each other.)  It would also have to transmit a check-in message to the security/fire panel every three minutes.  If the unit ever stopped talking to the panel then it would constantly try to reconnect with no regard battery usage so this could kill a battery fast if the panel is down for a long period of time.

[dcac]

Found this article from 2011: "99% Waste: The Unexpected Energy Consumption of Smoke Alarms:"
https://reductionrevolution.com.au/blogs/news-reviews/5842566-99-waste-the-unexpected-energy-consumption-of-smoke-alarms

Here's a working link to the " comprehensive report":
https://prod-energyrating.energy.slicedtech.com.au/sites/new.energyrating/files/documents/sb200405-smokealarms_0.pdf


I think someone will recognize this:

[ogden]

Can anyone specify a starting point of requirements that a smoke detector has to meet to switch discussion about personal preferences to the one about engineering?

At least 10000x more energy-efficient smoke detector having same functions as one shown in the original video. Detector that actually works - not only during standby, but producing nominal alarm sound as well.

[TerminalJack505]

Allegro has what looks like a very efficient smoke detector IC.  (Digikey link provided since I couldn't find a link on Allegro's site.)  The datasheet outlines some of the requirements for such a product.

[DBecker]

Allegro has what looks like a very efficient smoke detector IC.  (Digikey link provided since I couldn't find a link on Allegro's site.)  The datasheet outlines some of the requirements for such a product.

That's the detector only.  It has logic level outputs to the alarm, which requires an external driver.  It is intended to be run from a 3V battery and thus doesn't address the issue here, which is a requirement for a line-powered alarm with battery backup and external interconnect wire.

You might argue that a line powered alarm is obsolete when 10 year battery life is achievable, and the sensor should be replaced after 10 years anyway.  But a technical argument doesn't change that there are widespread requirements for a hardwired alarm, and that's not likely to change for decades.

As a side note, the interconnect signal wire works with activated alarm applying a 5V-12V DC level (typically 9V) with the input sensing threshold is 4-5V for other alarm to echo the alert.  An alarm needs to generate that voltage at the same time it is operating the alert.

[TerminalJack505]

Allegro has what looks like a very efficient smoke detector IC.  (Digikey link provided since I couldn't find a link on Allegro's site.)  The datasheet outlines some of the requirements for such a product.

That's the detector only.  It has logic level outputs to the alarm, which requires an external driver.  It is intended to be run from a 3V battery and thus doesn't address the issue here, which is a requirement for a line-powered alarm with battery backup and external interconnect wire.

You might argue that a line powered alarm is obsolete when 10 year battery life is achievable, and the sensor should be replaced after 10 years anyway.  But a technical argument doesn't change that there are widespread requirements for a hardwired alarm, and that's not likely to change for decades.

As a side note, the interconnect signal wire works with activated alarm applying a 5V-12V DC level (typically 9V) with the input sensing threshold is 4-5V for other alarm to echo the alert.  An alarm needs to generate that voltage at the same time it is operating the alert.

I'm not sure what you think my intentions were of making that post.  Others were asking what the requirements are for a smoke detector product.  That datasheet gives a general idea about the requirements.  Also, see my earlier post.

[EEVblog]

Nice for a video down the track if you could build up those other two solutions (TI and OnSemi iirc?)  and then measure the three different options. Four if you count the full bridge.

Just for you!

You just demonstrated typical trap for young players of "Design By Inspection". Engineers who made 80mA dropper instead of 1mA, made it for for some quite obvious reason - ALARM. You definitely can't produce audible alarm using 15v @50uA supply, unless you are building "smoke alarm earbud". Only actually making whole device and measuring SPL, you will see how much power is actually needed. Could be so that 10.5mA @ 15V of NCP785A is not even enough.

It was in my first video, I measured 15mA peak buzzer current.
The entire point of the video (both of them) was not selecting a suitable replacement regulator for this exact design, it was about showing the design topology differences. And the 2nd video was about showing why I didn't have to build up and test the regulator circuit in order to know it was 1/1000th the consumption of the zener circuit.

[ogden]

It was in my first video, I measured 15mA peak buzzer current.

It means that NCP785A with it's max 10.5mA is not suitable for the product.

The entire point of the video (both of them) was not selecting a suitable replacement regulator for this exact design, it was about showing the design topology differences.

Right. It's like facepalming about high idle consumption of the car, then proving that you can "solve problem" using motor from the moped. For me such "engineering" do not make any sense.

[Kleinstein]

Nice for a video down the track if you could build up those other two solutions (TI and OnSemi iirc?)  and then measure the three different options. Four if you count the full bridge.

Just for you!

You just demonstrated typical trap for young players of "Design By Inspection". Engineers who made 80mA dropper instead of 1mA, made it for for some quite obvious reason - ALARM. You definitely can't produce audible alarm using 15v @50uA supply, unless you are building "smoke alarm earbud". Only actually making whole device and measuring SPL, you will see how much power is actually needed. Could be so that 10.5mA @ 15V of NCP785A is not even enough.

It was in my first video, I measured 15mA peak buzzer current.
The entire point of the video (both of them) was not selecting a suitable replacement regulator for this exact design, it was about showing the design topology differences. And the 2nd video was about showing why I didn't have to build up and test the regulator circuit in order to know it was 1/1000th the consumption of the zener circuit.

The linear regulator circuit is not 1/1000 the power: the linear regulator version takes some 70 µA from 300 V and thus some 21 mW. The capacitive dropper was ca. 1 W real power and maybe 15 VA.
The the current spikes from the rectifier gives an AC current that is considerably higher than the 70 µA DC, more like 200 µA RMS and also a relatively poor power factor, though for a different reason.

[EEVblog]

Right. It's like facepalming about high idle consumption of the car, then proving that you can "solve problem" using motor from the moped. For me such "engineering" do not make any sense.

Then go make your own videos.
I could make half a dozen videos on all the engineering factors that go into selecting, testing, and proving a simple regulator solution like this if you want to get into real practical engineering of it. I chose not to  in this case and had a different intention. You don't have to like that, and that's ok, I can't please everyone.
But if you can't at least understand my intention, even if you don't like it, then 

[EEVblog]

It was in my first video, I measured 15mA peak buzzer current.

It means that NCP785A with it's max 10.5mA is not suitable for the product.

If you cared to actually watch the video:

[ogden]

Then go make your own videos.

What?! Did not expect such "argument" from you.

But if you can't at least understand my intention, even if you don't like it, then 

I said it already - you are doing business. Sometimes engineering excellence have to be traded for clicks. That's OK especially for professional youtuber. BTW existence of argument do not prove that I don't like video. Your videos are fine, they get better over time as well.

If you cared to actually watch the video:

Text overlay that looks like afterthought? - Nah. Sad that you consider critique as an attack rather than idea for your future video(s).

[Alti]

Gentlemen, either we get back on the topic or the topic risks being locked, I am afraid.

Thank you for valuable inputs, regarding requirements a "EEVblog #1284 considered" Smoke Detector (SD) has to meet.

Now we can see Australian requirements for SD are quite stringent, w.r.t upper half of the planet, my interpretation of the presented data indicates that:

SD has to be locally (battery) powered and remotely (mains) powered.

This "and" means the choice is not optional, if either one of power sources fails, the sensor stays functional using second source.
Understandably "locally" and "remotely" means there is no allowance to run the SD from just any two power sources. The case where you power smoke detector from lets say (mains 230VAC wiring) and (24VDC wired to central backup battery) interferes with whole concept. This is because during fire the whole wiring dies first, the SD has to beep till it melts on a ceiling in such scenario.

Also, powering this thing with two batteries (main + auxilary) is not legal because - well, if you forget about SD hanging there, after several years you are left without any SD protection.

Of course the regulation does not prohibit SD having an on-board battery and 24VDC wiring or 4-20mA current loop. Or on-board battery and two independent wirings, etc.


So,  the primary requirement is:

Single point of (power) failure does not compromize core SD functionality.

That is what I think is a good starting point in a discussion about fullfiling SD requirements.