Interesting repair of the day – PART 1: Raymarine C90 W

A customer brings in a Raymarine C90W stating it wouldn’t turn on. They have proven themselves to be very reliable, and very few have come in for repair.  I was initially skeptical, so I tried to power it up and sure enough – it refused to cooperate.

Later, I removed the rear case and found this:


Click on the photo for a larger image.


The internal chassis was like a soggy Florida morning with dew – everywhere. It was nearly time to go home so I decided to leave it overnight to see if some time in the air conditioning might dry it out.

Even with the best seals, atmospheric pressure and temperature changes will cause a unit to “breathe”, and ultimately collect some moisture. Since it’s “fresh” water, it’s no where near as damaging as salt water intrusion, and often the problem resolves itself once the unit has thoroughly dried out.

The next day when I tested it, it powered it up without incident. Since I don’t see many of the “wide” series unit, I thorough it would be a good opportunity to get better acquainted. Several things soon became apparent.

1) Operationally, it behaves much like a Classic C series unit.

2) This one had some quirks.

Abnormal behavior is remarkably common in modern marine electronics – especially with equipment mounted on boats with only one battery.

When the engine is started, the starter motor draws HUNDREDS of amps out of the batteries. The starter motor – like most conventional DC motors has carbon based brushes that maintain physical and electrical contact with the armature contact points – as it spins. There is a brief gap between contact points, and as the armature turns, contact between the brush and one contact point is lost, and contact with another contact point begins.  During this transition, a massive – albeit short duration spark occurs. This spark creates voltage spikes and surges on the entire 12 electrical system, and any electronic device connected to that system is exposed to sheer electrical hell.

The eventual end result is usually memory corruption – and that is usually the cause of the erratic behavior. The only way to resolve this problem is to clear the unit’s memory – a process often called a “master reset”.  The downside is – you loose your data such as waypoints and routes. Sometimes, you can back up this data before hand – sometime you can’t.

Sidebar: Memory corruption usually occurs when the unit is on and THEN you start your engines. Best practice is to start the engines FIRST, THEN turn the electronics on. If you start and stop the engines repeatedly throughout your voyage, turn your electronics OFF – BEFORE starting the engines, and you will likely avoid this problem.

The C90W was just not acting right so I decided to clear it’s memory. I inserted a blank CF memory card into the chart reader and archived the waypoints to it. I put the card in my computer and verified that the archive fie was there. Unfortunately, there was no way to read the data and see if that file would be useable or not. It usually is, but sometimes the unit won’t accept it when you go to restore the waypoint data.

I performed a master reset, updated the software and restored the waypoints. The machine seemed to be working as I expected and I did not notice any erratic behavior this time. I was just about to start the reassembly process when I remembered that I had not yet tested the unit’s ability to read a chart. This unit has built in USA cartography, so an external chart is unlikely to ever to be required. I had used the chart reader circuitry in order to archive and restore the waypoints, as well as update the software – so I knew the circuitry worked. Still – I wanted to be thorough.

I grabbed a test chart – a 16 GB unit I had used just the previous afternoon and inserted it into the unit. I waited for the unit to show some sign that it recognized the chart, and was puzzled when it didn’t. Again – I used this chart less than 15 hours before and it had worked just fine.

Then, I smelled then saw a thin whisp of smoke.



So here I am – on the threshold of completion and success and now a whole new world of hell is opened in front of me. I pulled out the chart and it was quite warm, almost hot to the touch. I immediately checked the chart reader pins and was surprised to see – they were OK.

After a moment or two, I started to get the picture of what happened.

Somehow – the chart failed. It didn’t make any sense, but the heat I felt when I extracted it was a dead giveaway of failure. (Stay tuned for part 2) With its failure, I was confident it had taken out the chart power supply switch on the C90W.

I’ve crossed paths with this sort of failure many times – mostly with the C and E series classic MFD’s. Chart cartridge failure is actually a pretty rare event. A more common  scenario goes like this: The chart is inserted, and somehow gets out of alignment. The operator forcefully presses the card in and causes multiple pins in the units chart reader to bend over, then touch each other. This creates a short circuit that destroys a filter inductor and a FET that acts as a power on / off switch for the chart.

Raymarine requires that you perform a multi-step procedure before removing a chart. This procedure ultimately tuns off the FET – thus powering down the chart cartridge. Once powered down – it can be safely removed from the unit.

Irritated as hell, I extracted the logic board and sure enough, the inductor had visibly melted down. The FET looked OK – but I wasn’t going to take any chances and decided to replace it any way. These are – very – small components and are in close proximity to other very small components.  You almost need a microscope to effectively see them. The replacement process is extremely delicate surgery. Fortunately for me, I’m well experienced with this kind of repair, and was able to get it wrapped up without undue difficulty…


The small 3 legged black device near the pencil lead is the FET, the yellow box next to it is the inductor.

Once the work area was cleaned up and the unit reassembled, I tested the chart reader with ANOTHER chart, and it worked perfectly.

We’ll call this one the “Agonizing repair of the day”!

If you found this interesting, you’ll want to read “Part 2”!



No comments yet

The comments are closed.