Skanky DJ

UK DJ playing the finest house music

We’ll all be dust soon enough – for the fun, glory and lasers ´◡`

Smoke or haze with smoke detectors and fire alarms

Smoke detector

A typical UK venue will have smoke detectors fitted and if you’re not in a venue that is set up to handle performance effects you’re likely to find the staff are unable or unwilling to disable smoke detectors in an event room, to allow smoke or haze to be used to make the lighting and lasers look fantastic.  With smoking banned indoors you can’t even rely on your crowd to be the smoke machine any more!  Its a real shame because people are all very used to seeing great lighting effects on TV, at shows and in clubs, all of which uses haze in the air to be seen of course.  But for whatever reason many venue designers and managers just aren’t interested in allowing the same great effects to be possible at their venue. What’s annoying is that smoke detectors are there to detect a fire when people are not there to see it or smell it and active the fire alarm – during an event filled with people and staff there is no need for smoke detectors to be active in the room.

Time to get our geek on…

This post is me getting proper geeky about this stuff, because I’m a tech at heart, a big laser fan and love DJing along with great lighting effects!!

Is there a way round the problem?

There are smoke detectors which are event friendly, which good venues ensure are fitted so that smoke and haze can be used.  These types of detectors are often called “heat rise detectors” but there are also other names used. Or if not permitted they use a fire alarm system and risk assessment that permits their sensors to be disabled temporarily during live events.

Typical smoke detectors fitted though are particle detectors.  Can you get away with using a smoke or haze machine with these?

Firstly we need to understand how these types of detectors work:

Ionization Detectors

Ionization detectors have an ionization chamber and a source of ionizing radiation. The ionization chamber consists of two plates separated by about a centimeter. A voltage is applied to the plates, charging one plate positive and the other plate negative. Alpha particles are constantly released by an ionizing radiation source (a lump of radioactive stuff!) and knock electrons off of the atoms in the air, ionizing the oxygen and nitrogen atoms in the chamber. The positively-charged oxygen and nitrogen atoms are attracted to the negative plate and the electrons are attracted to the positive plate, generating a small, continuous electric current. When smoke enters the ionization chamber, the smoke particles attach to the ions and neutralize them, so they do not reach the plate. The drop in current between the plates triggers the alarm.

Photoelectric Detectors

In some types of photoelectric detector smoke will simply block a light beam and the reduction in light reaching a photocell sets off the alarm. However in the most common type of photoelectric unit light is scattered by smoke particles onto a photocell, initiating an alarm. In this type of detector there is often a T-shaped chamber with a LED that shoots a beam of light across the horizontal bar of the T. A photocell, positioned at the bottom of the vertical base of the T, generates a current when it is exposed to light. Under smoke-free conditions, the light beam crosses the top of the T in an uninterrupted straight line, not striking the photocell positioned at a right angle below the beam. When smoke is present, the light is scattered by smoke particles, and some of the light is directed down the vertical part of the T to strike the photocell. When sufficient light hits the cell, the current triggers the alarm.

Each type of detector has its strengths and weaknesses based on detecting a smoldering fire, or an active flaming fire.  Basically you are unlikely to know what type is fitted.  Also bear in mind that a buildings air conditioning system may also have hidden smoke detection within the air ventilation system.

The above is kind of depressing from the viewpoint of using a smoke machine because both types of detector are simply looking for particles, not some magic specific particle only a fire will produce.

So can you do it.  Well unfortunately there is not a definitive answer.  I’ve used a bog standard el cheapo smoke machine in a couple of medium sized venues fitted with smoke detectors in the past, using a low density smoke fluid and not had a problem (both with the smoke detectors about 8 – 10m high up on the ceiling).  I then went back to one of the venues with a new smoke machine and 30 mins in off went the alarm.  Was it the new smoke machine heating the fluid differently, the different fluid used (still low density but a different brand) or just the movement of air on the night due to differences in the heat in the room.  No idea, but basically no you can’t get away with using a smoke machine unless you have the means to reset the fire alarm quickly and ensure fire prevention safety is retained.  And sorry, this means the majority of “haze” machines also because most haze machines are not really haze machines at all – they use the same types of fluid, heat it up in the same way and just aim for a lower density of smoke.  Basically a smoke machine with a variable output intensity that allows it to run with a low output is often called a haze machine by the manufacturer, but there’s no actual difference in the smoke/haze content itself.

What about a proper haze machine?

We’re talking here about proper cracked oil machines or proper hazers such as the pea soup phantom hazer (regarded by many professionals as the best). Well the answer here is more promising, but unfortunately still not clear cut.  Haze machines are certainly less likely to set off a smoke detector, but they are still perfectly capable of doing it.  The best types of these machines use oil (usually white mineral oil), and the basic reason is that oil hangs in the air much longer than water. However the operation is the basically same for both oil and water types.  Again, its handy to understand how they work:

Cracked oil machines

In simple terms a compressor bubbles air through the oil causing very small droplets to come off at the surface which a fan then blows off into the air.  A classic very widely used model of professional cracked oil machine is the DF-50.

The advantage is that the particle size is much smaller than for a typical smoke machine.  A typical smoke machine water glycol particle is 20 micron. A DF-50 is 1 micron. And this is the reason they are so much better than a smoke machine.  The small particle size means the particles hang in the air much longer (hours) and you need less fluid to produce a great beams through the air effect.

The disadvantage is the cost of the machine is often high and that the oil may collect on surfaces over time (depends on the machine used). You’re not going to notice it after a couple of gigs, but when used regularly the lighting equipment needs cleaning more often to get rid of the the residue that builds up.

The Phantom Hazer

This hazer has 3 big advantages.  The first and biggest one is that it produces the smallest particle size – 0.2 microns.  This is why many professionals love it – the effect it gives is often stunning.  The second is that due to the smaller particle size it doesn’t have the same residue build up issue (they claim).  Finally, before vaporizing it first heats its oil up to around 300ºC which ensures no germs come out with it.

So can they be used?  Well, yes and no.  You will find people who will tell you they’ve used cracked oil machines and never, or very rarely, had a problem.  The problem is that you can’t guarantee you won’t have a problem, and causing your event or even worse a hotel to evacuate is a pretty big deal if you get it wrong!  The reason people often get away with it though is that it is much much easier to control the amount of haze in the air with a hazer. Because the haze hangs for so long and because you need much less of it you can often just run the hazer for a few minutes, taking care to look down some of your light or laser beams, and shut it off as soon as there is enough for them to register.  It can be a decent amount of time before you need to add more, obviously depending on the venue and its air flow.  So, if you take care in how you use it you can often keep it well below the threshold the smoke detectors are looking for.  The catch is you don’t know exactly what that threshold is and as well as varying from detector to detector.

But what about the particle size?

The best you can get is the Phantom Hazer with a particle size of 0.2 microns (0.2µm). That sounds tiny doesn’t it, and it really is.  Its haze is really strange in that you can’t see its there until you shine a bright beam through it.  Surely smoke detectors won’t see something so small? Well I’m not an expert in this area, but based on the very handy Wikipedia page here unfortunately smoke detectors are designed to work with particles that size:

Fire in its early, smouldering stage (before it breaks into flame).

The smoke from the smoldering stage of a fire is typically made up of large combustion particles — between 0.3 and 10.0 µm.

Photoelectric smoke detectors respond faster (typically 30 minutes or more)

Flaming stage of a fire

The smoke from the flaming stage of a fire is typically made up of microscopic combustion particles — between 0.01 and 0.3 µm.

Ionization smoke detectors respond faster (typically 30–60 seconds) in the flaming stage of a fire.

So the Phantom Hazer’s 0.2µm particle size is near or in the range of the typical particle size both types of smoke detector are great at detecting (not that the detectors are actually measuring particle size specifically).

Is there any advantage to the smaller particle size of a Phantom Hazer?

We’re getting very geeky at this point, but the question really intrigues me…

Firstly lets establish the difference in fluid usage between a DF-50 and a Phantom hazer as figures are available for both, so we know what we are talking about.  One of the Phantom Hazer’s own sales claims is that it uses 55% less fluid than the DF-50. This means that the Phanton Hazers smaller particle size does mean that you need more individual particles in the air to produce a similar lighting effect.

A Phantom Hazer particle is 0.2 microns diameter, which equates to a sphere volume of 0.034.

A DF-50 particle is 1 micron diameter, which equates to a sphere volume of 4.19 (doubling the diameter of a sphere equates to much more than double the volume of material so we can’t work just based on diameters to complete this comparison).

If the Phantom Hazer used the same amount of fluid then this would mean for every 1 DF-50 particle, with a volume of 4.19, there would be are around 123 Phantom Hazer particles with a volume of 0.034.

But the Phantom Hazer uses 55% less fluid, so in fact for every 1 DF-50 particle there are around 55 Phantom Hazer particles needed to produce a similar effect.

So, the Phantom Hazer uses 55% less fluid by volume, but produces many more individual particles. How does this matter?

A photoelectric detector is basically doing exactly the same thing as you are doing by wanting the haze in the air.  As light beams pass through the air you want the haze particles to be hit and reflect the light off in every direction.  You want a haze density that gives you enough visibility of the light beams.  A typical photoelectric detector is set to trip when its sensor see’s enough of its own light hit and be reflected off the smoke particles present. So for a photoelectric detector the size of the individual particles isn’t directly important, its just how much of the light is being reflected out within a given area of air.  So using a Phantom Hazer over a DF-50 or similar is unlikely to matter much, its all about just keeping the density low enough to stay below the smoke detectors trip threshold – let your light beams register but only a bit.

An ionization detector works differently however.  The tiny ions (<200 picometers, which is 0.0002um) produced by its radioactive magic travel at less than 1m/second, but much much faster than your haze particles floating through its detection chamber. As each individual smoke particle floats through the chamber it will stop and absorb any ions if they hit it, causing the sense voltage to decrease slightly.  If too many ions are absorbed the sense voltage drops past the smoke detectors trip threshold and it’s alarm time. Does the fact that the haze particles are smaller or there is less fluid being used make a difference here?  Well the difference in the total volume of fluid isn’t the issue because this is just due to the smaller sphere size of the particles.  The issue is that if an ion hits any haze particle it is absorbed and its loss is effectively counted by its voltage not registering on the plates – the ions not making it across the chamber is being counted not the number of smoke particles present. Does 1 big particle stop more ions than 55 small particles?  If you draw 1 big circle and 55 small circles to scale it would appear to:

Smoke particles

But that is a 2D view, whereas the chamber and particles are 3D.  I think once you visualise this in 3D then the question becomes very similar to what happens with light in the photoelectric detector – there’s not much, if any, difference in the odds of an ion making it across based on the 2 different particle sizes.  So it seems reasonable to say here that the smaller particle size is no benefit for this type of detector also, but at the same time not a problem either.


Its hard to find a lot of reliable professional comments on this topic based on real world experience. What I can say from my own direct experience though is that a cracker is much less liklely to trip a smoke sensor if you use it very carefully. My opinion on the best solution would be on any proper oil based haze / cracked oil machine, such as the Phantom Hazer, DF-50 or American DJ Haze Generator (also branded as Antari HZ-100 and MBT HZ100 – a much cheaper less powerful equivalent to the DF-50). I’ve come across quite a few user comments from people saying they’ve have great success at avoiding fire alarm detection using the American DJ Haze generator.  I’ve used the Antari HZ-100 machine at quite a few events with DF-50 oil and the effect it gives is lovely – tiny particles floating in the air instead of big clouds of choking smoke from a smoke machine.

Remember also that even with people smoking indoors now banned, it didn’t used to be and in some worldwide markets a manufacturers smoke detectors are sold into it still won’t be, and smoking indoors doesn’t typically trip smoke detectors.  Also, any detector needs to have a reasonable threshold anyway to avoid triggering from normal air born dust particles etc.  The catch is of course – what is that reasonable threshold?!

Some tips from my usage of the Antari HZ-100

I’ve done quite a few events using the Antari HZ-100 in venues with smoke detectors fitted.  An early lesson learnt was that I needed to buy the timer remote control for the machine.  After becoming distracted by a punter and forgetting the machine was running at an event I turned around to see a large cloud of haze forming a couple of meters away from the nearest smoke detector and cringed for the next little while praying the alarm wouldn’t trigger!  It didn’t, but the lesson of use a timer to add short bursts of haze well learnt.  One other tip was that I replaced the fan in the machine with a more powerful version to help disperse the haze further into the dance floor at floor level, so dancers can help disperse it (and also to help ensure it doesn’t collect nearby to all my kit waiting to be drawn in through their cooling fans).

Once I can see the beams of my lights above / behind me start to register I turn down the bursts.  I keep a decent level of worry up in my head and ensure I keep this ‘just registering’ level of haze constant for the rest of the night by tweaking the timer.  What looks like just there for me looks fantastic for the punters out front who are looking down the beams towards the lights / laser. Its also worth saying the haze effect looks really nice, quite different to smoke machine haze.

I have cocked up once or twice and caused an alarm trigger, basically by getting complacent at a friendly really small venue where I knew if the alarm went off it wasn’t a big deal to reset.

Its not impossible using the cracker won’t trip an alarm at a larger more tricky venue of course even with being super careful, that’s the risk you still have to take, but I’m becoming increasingly convinced its not going to happen when I think back and compare rooms full of smokers or rooms with dust in the air – what I create doesn’t seem significantly more dense. Fingers crossed, touching wood of course and never forgetting that I’m risking some major embarrassment if I lapse my judgment and get it wrong!