Written by Ken Tuesday, 26 August 2008 21:32

Attention: open in a new window. PrintE-mail

ROV Chronicles: Part3

Electronics

The third in our series on designing and building an ROV takes us into the world of electronics. We’ll look at trying to pressure-proof our electronics, and the construction of the first of the motor control boards, complete with soldering.

 

Pressure-proofing

 

As mentioned in the previous instalment, the plan is to do away with the need for a pressure housing by using inherently pressure-proof electronics. The problem is how do you find out which electronics are likely to be damaged by pressure? It’s not something that manufacturers of electronic components tend to mention as part of their product specifications, and most would be aghast at the idea of their components being used at 10 or 100 times atmospheric pressure.

 

There are a few obviously fragile components (some capacitors for instance) but for a better idea, we turned to the professionals. A quick post on the forums at www.rovworld.com returned a host of useful information from ROV technicians and supervisors – people with real-world experience.

 

Capacitors and timing crystals both come in pressure-sensitive varieties
Capacitors and timing crystals both come in pressure-sensitive varieties
The answers were surprising. A lot of electronics are fine, since they are just solid chunks of metal, silicon and other materials. As expected, certain types of capacitor (electrolyte filled and air or paper gap ones specifically) are pressure sensitive. Also interesting was that some quartz timing crystals can’t handle pressure, and even some push-switches can cause problems at depth!

 

There are several solutions for dealing with pressure sensitive components. The most ideal one is to replace them with equivalents that aren’t going to cause problems – luckily an option with most capacitors, since solid state alternatives exist.

 

A second alternative would be to encase individual components in an epoxy or polyester resin, or similar hard-setting material. This is easy enough to do, but there is still a potential pressure at which the component will fail (although depending on how much resin is used, it could be very deep indeed!)

 

The third alternative is to encase the entire board in resin (something that the professionals at ROVworld.com actually suggested) – easy enough to do, but it makes it hard to do repairs and make connections at a later date, and it may cause overheating problems with some components.

 

Where possible we are going to choose pressure-proof components, and if we have to, individual components will be encased in a blob of resin. Hopefully that will suffice for our target depth of 100m.

 

Building the first board

 

Garden-table soldering workshop!
Garden-table soldering workshop!

The first electronic board to be constructed is a general purpose USB interface board. It provides five digital inputs, two analogue inputs, eight digital outputs and, most importantly, two analogue outputs which work both as 0-5v DC and also as Pulse Width Modulation (PWM) outputs, which is the sort of signal used by remote control models.

 

The board was chosen mainly for cost (it’s a general-purpose “experiment” board from Maplin, price £29.99) since other industrial control solutions run to hundreds or thousands of pounds, although ease of programming on a PC was also a consideration (the board comes with lots of example programs!) Given that it only has two analogue outputs, and we want to control at least five motors, it will be necessary to use three of these boards eventually.

 

Using a crocodile clip as a heatsink on the more sensitive components (transistors and LEDs especially!)
Using a crocodile clip as a heatsink on the more sensitive components (transistors and LEDs especially!)
The DinkyKitty labs were a little taken aback when the board arrived – it came in the form of a chunk of green plastic and a bag full of electronic components – some home assembly required! Undaunted, we sat down with a trusty soldering iron and put it all together, with minor supervision from THE dinky kitty, which watched with purring bemusement.

 

At this stage, we have opted to use all the supplied components (including the two suspected of being pressure sensitive) since it will let us carry on with motor testing and software – those components can be replaced before sea trials begin!

 

The first and main snag to the operation came when we realised we didn’t actually have the correct type of USB cable to connect the board to the computer for testing. This entailed a quick trip to the local computer shop, Micro-Plus computers in Oswestry, where we managed to negotiate a good price (thanks Luke!)

 

The (nearly) finished board, ready for testing.
The (nearly) finished board, ready for testing.
The moment of truth – are our rusty soldering skills up to the job? Did we manage to decode the markings on the resistors correctly? Thankfully, everything seems to be in order and the test software on the PC reported the board had connected, and inputs and outputs, as measured with a multimeter, seem to be correct.

 

This means we can get on with the hard part – writing the software!

 

Next time: Software design

 

Click here to discuss this article in our forums