Home made underwater video housing
For some time now Iíve been wanting to construct my own UW housing for my JVC DVl 9500 video camera. At the time it seemed that paying $1000 for an Ikelite or
similar housing was just outrageous. Well, after spending countless hours designing and quite a few dollars on CAD software, materials and tools that Ikelite is starting to look good :-)
For some movies recorded using this housing click this link.
At some point I will complete this page with lots of info. For now all I have is a picture of the housing before itís maiden submersion:
Inside is a bag with about a pound of washers to help it sink. As you can see the box is rectangular which is more
difficult to make than using round tubing. The advantage is reduced volume and therefor reduced weight above water
(the lead you have to carry to make it sink). It has a flange with a front plate that are pushed together with 4 stainless steel bolts and wingnuts. An o-ring keeps the water out.
The first attempt was a failure. A tiny bit of glue that had already been identified as a possible source of trouble
indeed turned out to be so. That has since been resolved and the case is waiting for itís next pressure test. The box
is severely overdesigned and should easily withstand 200 fsw pressure, provided the o-ring seal is done correctly.
I designed the case using TurboCAD Professional 7.1. Most of the work was actually done in the 15 day trial period
but I ended up buying the full version anyway.
Here is a hidden line rendering of the case:
The toughest part was rendering the camera and all supporting equipment like the monitor, batteries and control.
Once I modeled this I could figure out the size of the case I needed to build. The tray with all components looks like this:
This is a 3D rendering of my model. The lighting is red and blue to help distinguish the components. The brownish
bars are reed switches that are part of the control mechanism.
In all likelyhood I was too conservative with the modeling of this tray. That means the case is probably bigger than it
has to be. Weíll see when itís all constructed. Once this one is done I could make a smaller one, perhaps one that
somehow uses the built in monitor of the camera. Adding provision for a macro lens (the black cylinder on the right) added a lot of space as well.
Rev 1.0 Finished
Here are two pictures of the case that I took diving in Mexico in October 2001.
Rev 1.1 Finished
I finally added a real record on-off switch. In Rev 1.0 I had to do this by moving a separate magnet close to the reed
switch. That required two hands for the camera. The new switch works by thumb actuation.
The picture on the left shows some of the components of the switch. Itís a simple hinged lever design with a magnet
at the bottom and a spring to keep the switch in the non-actuated position when not in use. Since the camera
switches states (from on to off or vise versa) whenever it gets a Ďrecordí signal itís not a stateful switch like used to
turn a light on. The right picture shows the position when actuated. This moves the magnet over the reed switch inside the case, causing it to make contact and send a signal to the camera.
This shows the power switch. Itís principle is similar to the
record switch, above except that it is stateful. That means the switch needs to remain in the Ďoní position (horizontal, not shown) for the camera to operate. When the switch is off, all
power to the camera is interrupted. This has two benefits:
- The housing can be assembled and closed before going to the dive site. This makes it easier to obtain a clean seal
(no sand) and less likely to get moist air in the housing which could cause condensation.
- Like most cameras my JVC-9500 turns itself off after a few minutes of stand-by. This is done to save the battery
and heads. The problem is that it can only be recovered from by cycling the power. With this switch I can reset the camera in itís entirety.
This is what keeps the water out, the only seal in this design.
Traditionally o-rings are laid in a groove. The problem was making a groove that was smooth enough for the o-ring to seal against. This was solved by fitting a flange on top of the box.
This flange has a perfectly smooth surface whereas the end of the box would be very difficult to get smooth. Now the o-ring had to be secured somehow. This is done by cutting a rim out of
thinner acrylic that was glued on top of the flange. The oring is secured around the rim.
The inside of the box. There are no potential leak locations. The guides keep th tray in place.
In this picture some of the tray components are shown. The relay
is needed because the reed switch can not handle the current required, over 1A. 6 NiMH AA batteries provide power.