Reverse engineering a big old smoke detector

Current images of fusor efforts, components, etc. Try to continuously update from your name, a current photo using edit function. Title post with your name once only. Change image and text as needed. See first posting for details.
Post Reply
User avatar
Rich Feldman
Posts: 1289
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Reverse engineering a big old smoke detector

Post by Rich Feldman » Sat Mar 14, 2020 10:05 pm

This is to show more detail about a thing closely related to Richard's "interesting electrometer device".
viewtopic.php?f=18&t=13261

Here is an old stock Pyrotronics F5B smoke detector head, similar to the F3/5 head. They are still available for people maintaining old installations, but typically cost well over $100 on ebay.
f5b_photo13.JPG
f5b_photo13.JPG (43.03 KiB) Viewed 1163 times
.
Ionization-type smoke detectors seem to have started in the 1960's. Preventing fires in industrial & institutional settings easily justified the proliferation of 80 microcurie Am-241 alpha particle sources. Those produce enough air ionization that a cold cathode trigger tube can switch the alarm circuit.

I got the pictured device to turn on a neon lamp, within a few seconds of being exposed to smoke from a smoldering popsicle stick. My circuit is shown below, on the right. The original application uses Base Units (sockets) with resistors and a neon indicator lamp, and if any reader has one I'd love to know the circuit details.
f5b_schem.JPG
f5b_schem.JPG (43.85 KiB) Viewed 1163 times
.
These old smoke detectors, just like modern ones with less than 1 uCi of radioisotope, use a reference ionization chamber as the high impedance on top of voltage divider. Cheaper than a many-gigaohm resistor, and it automatically compensates for pressure and temperature and decay of the alpha particle source.

The F5B can be taken apart without tools, but I'm not sure we should do that or even talk about it. I found out that the internal construction is similar to this unfinished drawing, previously posted in Richard's electrometer thread. The ion chambers in schematic above were rendered accordingly.
f5b_v2.png
f5b_v2.png (33.67 KiB) Viewed 1163 times
It came as a surprise that the larger ionization source, which irradiates the smoky chamber, is directly visible through the window-screened ports. Close enough that energetic alpha particles, not just ions, could be escaping from the enclosure. Let's check with some alpha-viewing fluorescent film from "geoelectronics" on ebay (see Low Light Photography thread).
f5b_photo12.JPG
f5b_photo12.JPG (45.71 KiB) Viewed 1163 times
Set shutter time to 30 s, aperture f/1.4, ISO 12800. Lights off, not completely. Go...
f5b_photo11.JPG
f5b_photo11.JPG (54.17 KiB) Viewed 1163 times
Would it be against the law to disassemble a device like this, for closer inspection? Does it matter if no screwdriving, wrenching, or cutting is required? Can anyone substantiate rumors that internal parts could be shedding active material?
All models are wrong; some models are useful. -- George Box

ian_krase
Posts: 628
Joined: Mon Nov 28, 2016 7:48 am
Real name: Ian Krase

Re: Reverse engineering a big old smoke detector

Post by ian_krase » Sat Mar 14, 2020 10:26 pm

Well that is certainly a tempest in a teapot of Nuclear Blue Glow.

User avatar
Richard Hull
Moderator
Posts: 12091
Joined: Fri Jun 15, 2001 1:44 pm
Real name: Richard Hull

Re: Reverse engineering a big old smoke detector

Post by Richard Hull » Sun Mar 15, 2020 4:24 am

So my little tube came from one of these things!?? I wonder how the guy I bought the little vacuum tube from at the hamfest came by it?

This is a rather nasty thing. I would recommend leaving this sleeping dog lie. If you mess with it, you are on your own.

Richard Hull
Progress may have been a good thing once, but it just went on too long. - Yogi Berra
Fusion is the energy of the future....and it always will be
Retired now...Doing only what I want and not what I should...every day is a saturday.

Rex Allers
Posts: 435
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Reverse engineering a big old smoke detector

Post by Rex Allers » Mon Mar 16, 2020 5:33 am

David Prutchi, who has created and shared on the internet and in a book, many physics experiments with custom built apparatus, was the source where I learned about these old smoke detectors.

I thought the page he made might have been taken down but seems not. Here it is:
http://www.diyphysics.com/2013/01/20/80 ... detectors/

So several years back I saw one of these listed on ebay and won it. I was hoping to use one of the strong Am sources to make a neutron source with Be. Then after I won, the ebay seller wanted me to provide details of how I was righteously credentialed to own one and the deal fell through.

At the time I was a member in the yahoo groups for geiger counters and gamma spec. George Dowell "Geo" was the owner of those groups. He posts here from time to time.

I posted in one of those groups bitching about my scuttled ebay transaction. Geo replied that it might be a blessing in disguise because the sources in these things were notorious for shedding radioactive material.

So that is my second-hand source for mentioning the possible problem in an earlier message here. I have no direct experience myself. Maybe Geo will see this and share his knowledge.

-edit-
I just noticed that near the end of the Prutchi page that I linked he warns, "sources have been shown to leak" -- which text is a link, but it now goes to a dead page.
Rex Allers

Rex Allers
Posts: 435
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Reverse engineering a big old smoke detector

Post by Rex Allers » Tue Mar 17, 2020 8:55 pm

A bit of follow-up.

I connected with Geo and he hasn't actually had any direct experience with these old smoke detectors. So I guess what I got from him was second hand.

On Prutchi's page that I shared a link to, he did have some reference in a link near the end of his page but that site is now gone. I did a couple web searches but didn't find any cases of problems or warnings.

So I guess I don't really know if these old detectors have any issues with shedding radioactive material. Too bad that one link is dead. He did have the link tag as "the sources have been shown to leak".

Prutchi's page does have some nice pics of what is inside.
Rex Allers

User avatar
Rich Feldman
Posts: 1289
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Reverse engineering a big old smoke detector

Post by Rich Feldman » Tue Mar 17, 2020 9:58 pm

Am hoping David Prutchi responds to an email I sent. I'd previously found that DIY physics page of his, featuring a Pyrotronics F3/5A detector head. Bet it predates the F5B, which uses the same base (active socket). "Bottom" halves might be identical. F3/5A smoke chamber irradiation comes from a bolt at top center instead of ring near bottom. Reference chamber can is secured with a nut (like on Richard's gadget) instead of a metal stalagmite with internal threading.

I've corresponded with George Dowell, who sent me the fluorescent film pictured above. Speaking of rad detection, I need to finish commissioning or give up on a solid state alpha particle detector (DIAD-II) that can help me check for contamination.

Well since David did it, I will show without fanfare the F5B parts which bear the big and small irradiation sources. Their relative activity is roughly indicated by lighting up that film from Geo. Small round glow is visible to well-adapted eyes, kind of like ordinary comets. Bright ring glow can be seen (without color) after just a few minutes in the dark.
three.png
three.png (205.83 KiB) Viewed 900 times
No reason for anybody to repeat the exercise. This F5B is in working order today, ready to protect lives, and the film stopped glowing when removed from the hot parts.
Last edited by Rich Feldman on Wed Mar 18, 2020 12:38 am, edited 2 times in total.
All models are wrong; some models are useful. -- George Box

Rex Allers
Posts: 435
Joined: Sun Dec 30, 2012 8:39 am
Real name:
Location: San Jose CA

Re: Reverse engineering a big old smoke detector

Post by Rex Allers » Wed Mar 18, 2020 12:26 am

Ah, I'm with you now. It hadn't registered with me that you had an F5B and it was a different design.

The pics you shared are great, especially with the blue glow of where the alpha is.

Thanks for sharing that.
Rex Allers

User avatar
Rich Feldman
Posts: 1289
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Reverse engineering a big old smoke detector

Post by Rich Feldman » Wed Mar 25, 2020 6:29 pm

Got some ion chamber current measurements this morning. Both chambers in series, without disconnecting the untriggered cold cathode tube. Black wire from edge of picture connects power supply chassis to a grounding conductor.
20200325_110836.jpg
.
The shape matches what's shown in textbook. We see saturation current different for forward and reverse voltage.
20200325_110804.jpg
.
Sorry about big pictures, straight from not so smart phone. Could have told the story just as well with 1/10 of the kilobytes.

[edit] Operator doing voltage sweeps needs to be patient. Instrument settling time on 1-pA-resolution range takes getting used to.
And the sensitivity of reading to dV/dt is obvious and strong. Suppose the device under test has a terminal capacitance of 10 pf (another thing that would be interesting to measure).
Then each 50 volt change of bias voltage would bring a current pulse amounting to 500 picoamp-seconds.
All models are wrong; some models are useful. -- George Box

User avatar
Rich Feldman
Posts: 1289
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Reverse engineering a big old smoke detector

Post by Rich Feldman » Thu Mar 26, 2020 5:19 pm

The slope of linear section, with reverse bias, matches a resistance of almost 200 gigaohms. Even higher in the presence of smoke, or as the voltage approaches saturation.

We have seen, here and in Richard's electrometer device thread, that neon-filled trigger tubes can switch whole milliamps according to the ion chambers signal.

I'm looking forward to making similar measurements on a modern ionization-type smoke detector, where the total bias potential is about 9 volts.
Anybody here ever do that? IIRC, at least one of the common smoke-detector IC's has an analog output pin that presents unity-gain-buffered copy of the high-Z input voltage.
My plan has been to pull the ion-chamber voltage divider gently up or down with current from a super-high-megohm resistor, but first step will be to just measure I/V curve of ion chamber stack.
All models are wrong; some models are useful. -- George Box

User avatar
Rich Feldman
Posts: 1289
Joined: Mon Dec 21, 2009 11:59 pm
Real name: Rich Feldman
Location: Santa Clara County, CA, USA

Re: Reverse engineering a big old smoke detector

Post by Rich Feldman » Sun Apr 26, 2020 12:08 am

Bit of progress today with a "modern" household smoke detector,
perhaps adaptable for measuring currents in a scratch-built ionization chamber.
Watched the change in an analog voltage as smoke was applied and removed.

The sensor is a series-connected pair of ionization chambers, both irradiated by a single Am-241 button.
Biased by a 9-volt battery, they form a very high impedance voltage divider, connected to a smoke detector IC like MC14467 in 16-pin DIP package. To minimize surface leakage current, that pin is formed straight out from molded plastic body, and spot welded directly to ion chamber electrode.
rad10.jpg
rad10.jpg (140.98 KiB) Viewed 331 times
This is one of many smoke detector IC's that supports a guard trace on the circuit board, connecting to the neighbor pins on each side and actively driven to the same voltage as the high impedance input. It was a simple matter to spot the "guard trace" connection and bring a wire out to voltmeter.
rad11.jpg
rad11.jpg (7.69 KiB) Viewed 331 times
rad12.jpg
rad12.jpg (136.29 KiB) Viewed 331 times

Long story short: with 9V power applied, the guard trace voltage was about 3.8 volts. It went up when smoke wafted in, and the audible alarm started between 5 and 6 volts. Voltage came back down & beeping stopped soon after smoke was removed. I figured it was worth posting in video form:
https://youtu.be/VLxUsJffWVI
rad13.jpg
rad13.jpg (60.62 KiB) Viewed 331 times
All models are wrong; some models are useful. -- George Box

Post Reply