Question for the Electronically Savvy

[enovillo]

Hi all,

I have a Garmin GPSMap 545 that worked like a charm. My boat just spent 2 months on the hard in Florida and the plotter stayed in the cockpit under a sunbrella cover.

When I launched the boat and turned on the plotter, I noticed that it started fine but would shut off as soon as I let go of the power button. After a few attempts at normal start up I realized that the only way the plotter stays on is if I continue to hold down the power button. Thinking that it might be moisture on the circuit, I decided to put a C-clamp around the unit to keep the button pressed and left it on for a few hours. That didn't change anything.

The down side of the C-clamp is that while the power button is pressed down, the other buttons do nothing. So the unit is useless as a chartplotter. I know the GPS works fine because the VHF receives GPS data from the plotter.

I took apart the unit hoping that I'd find something obvious mucking up the power button's functionality. But the board and everything else inside look pristine.

Garmin unfortunately doesn't support this unit anymore, so they asked a few questions, tried a couple of standard functional tests, and had no additional suggestions.

I'd like to bypass the power button so that as soon as the unit gets power it turns on. But I don't know enough about electronics to do this.

Can anyone give me some guidance by just looking at the photos of the board?

At this point I have nothing to lose, so I'm more than willing to try anything.

Also, if anybody is trying to offload a working Garmin 545 or 546, I'll be happy to purchase it.

Thanks for any opinions and help!

[DeepFrz]

My bet is that there is a relay somewhere inside that is not latching up. When you push the button do you hear a click? Also check that any connections are clean and well seated.

[Reefmagnet]

Possibly circuit board problem, as the power circuit is a little bit more involved than a simple on/off toggle. I'd firstly check the input voltage is ok with power on and off, reseat all connectors and flush the circuits with isopropanol alcohol (electronic cleaner). Failing that fixing it, perform a master reset and hope for the best (courtesy of google):

With the device powered OFF, press and hold the Home button
Press and hold the Power button
Continue to hold both buttons until the message "Do you really want to erase all user settings?" appears, then release both buttons
Highlight Yes and press the Select button
The device will now restart. When it fully boots, the device will be set back to factory default settings

Good luck!

[Lake-Effect]

To the OP, have you tried checking the basic stuff, like confirming you're getting a full 12v at the plotter, and that it doesn't dip when the plotter powers up?

Originally Posted by DeepFrz

My bet is that there is a relay somewhere inside that is not latching up.

I agree with this, except that the 'relay' is more likely to be implemented in silicon (transistors). If a schematic isn't available, try tracing the ON button to the main board and see what it connects to.

(on preview - seems I've just echoed reefmagnet)

[muddyone]

Not latching would be where I'd start. Reset to default has been known to fix things other than just user settings etc, so is always worth doing.
The way most power-on circuits work is that there is a transistor, or in the past more commonly a relay. The power-on transistor is just an electronic switch. has 3 terminals, 12 volts from the battery, an output to the the rest of the circuitry, and a gate. The gate operates the switch, and is connected to the keypad and to the processor. Once the unit has powered up and you release the button, the processor has taken over from the key and maintains an on voltage to the gate. In this case it sounds like the gate is not getting that voltage from the processor. Possibly a faulty resistor, diode or transistor somewhere between the processor and the power-on transistor. The easiest fix would be to bridge out the power-on transistor source and drain terminals, that is the ones connected to battery and the rest of the circuitry, leaving the gate alone. To repair properly would involve deeper circuit tracing, or a circuit diagram. Post some pics of the main board near where the keypad plugs in and we'll see if the power-on transistor can be identified.
Hope that helps

[Draikhyan]

Originally Posted by muddyone

Not latching would be where I'd start. Reset to default has been known to fix things other than just user settings etc, so is always worth doing.
The way most power-on circuits work is that there is a transistor, or in the past more commonly a relay. The power-on transistor is just an electronic switch. has 3 terminals, 12 volts from the battery, an output to the the rest of the circuitry, and a gate. The gate operates the switch, and is connected to the keypad and to the processor. Once the unit has powered up and you release the button, the processor has taken over from the key and maintains an on voltage to the gate. In this case it sounds like the gate is not getting that voltage from the processor. Possibly a faulty resistor, diode or transistor somewhere between the processor and the power-on transistor. The easiest fix would be to bridge out the power-on transistor source and drain terminals, that is the ones connected to battery and the rest of the circuitry, leaving the gate alone. To repair properly would involve deeper circuit tracing, or a circuit diagram. Post some pics of the main board near where the keypad plugs in and we'll see if the power-on transistor can be identified.
Hope that helps

Those were the days. Transistors, generic components, analog electronics.

Unfortunately these days it's all miniaturized and completely integrated. Companies own the rights to the IC designs and that way ensure that nobody but them can repair malfunctioning devices. "I'm sorry sir, we no longer support that outdated model. But we have a new one that'll work even better! And it's only $599."

I have an old Seafarer 700 sounder somewhere. Oldschool. And something any hobbyist can fix. It's more than 35 years old and has been modified more often than Michael Jacksons nose. But it still works.

But enough of that.

Does the Power LED light up when you press the power button?

The LED is most likely not defective. But if it is, then there's a very small chance that replacing it may fix the problem.

But its most likely a malfunctioning Integrated Circuit. What's the chip in the upper left corner? Signal translation? It could be the problem. But there's no number on it. Could be a Garmin specific chip. No, the problem isn't located at the button and bypassing it won't help.

You will need a Ohm meter and a logic analyzer or oscilloscope to trace the signals from that button to find the origin of the malfunction. And even then it's probably a chip that's owned by Garmin. And Garmin no longer supports this model.

[travellerw]

Alright... I do some EE repair and may be able to help...

This type of power button is called a "Soft" power button, as its usually software that controls power to the device. The power button triggers an interrupt pin on a microcontroller that "wakes up" and powers up the rest of the circuit. Although it could be implemented through transistors or FETs. Here is a good example of a transistor controlled circuit (that would fail like you describe).
The push on push off transistor switch | HACK A WEEK

So... What I would be looking for is a failure of either a capacitor (electrolytic) or a resistor. If there is only a small number of electrolytic caps, I just replace them all. For resistors, I probe with a multimeter and look for strange values. The failed ones are usually pretty obvious. Pretty sure this failure will require some component to be replaced, you won't get away with simply cleaning or drying something.

[enovillo]

Wow, guys, lots of information (some of it I'm barely able to understand). Thanks a lot!

I'll try to address all your questions and comments, and maybe I'll be able to fix it with your help:

1. I don't hear any clicks when the unit powers on.

2. I've cleaned everything with isopropyl alcohol. All connections looked perfectly clean before and they look just as clean now. Nothing changed.

3. The input voltage is a healthy 13.4. One of the diagnostics Garmin had me run is booting in test mode (with the unit turned off, press and hold select and then press power). That's where I read the 13.4v this afternoon. As a side note: all my other electronics continue to work flawlessly and the batteries, though somewhat old, are healthy.

4. The Garmin technician suggested a reset (press and hold Home, press power). Unfortunately, the other buttons don't work while I'm holding the power button, so I'm not able to select "yes" when asked to confirm whether I want to go ahead with the reset. Of course, if I let go of the power button, the unit shuts off.

5. I don't think I can trace the power button to the main board. A flat strip comes out of the buttons board and goes into the main board. I have no way of knowing which cable comes from the power button. I naively thought the pins on the connector might be labelled, but they're not. I've attached a photo of the main board just in case.

6. The power LED does NOT come on when the unit is on.

7. I can measure resistance, voltage, and current. But a logic analyzer and/or oscilloscope are not part of my tool chest (or skill set!). The Garmin technician suggested measuring amperage when powering on and while running, but when I asked him where to measure it he had nothing. He wanted to see if the current dropped sharply, but he wasn't able to tell me where to measure it.

8. travellerw: I think your skills and experience are what I need to fix this. You're not in Jacksonville, Fl, are you?

After re-reading everything you guys have written, I'm inclined to think I'm going to have to bite this bullet and buy a new unit. I would love to keep this one as a spare, but right now it's only good as a lame GPS.

Thanks again for you responses!

[muddyone]

Just taking an experienced guess at the power-on circuitry being faulty, could be that there is level or condition that the processor checks before asserting power-on, or could be something completely different. Anyway, I did a quick search for the type of circuit that I described earlier

http://i.stack.imgur.com/B6bez.png

you can see that by putting a switch or a wire link across Q1 would connect VCC (in this case 12 volts from the battery connection) to MCU power (the rest of the plotter) A rough fix but would let you use the keypad

The capacitor and interrupt described by travellerW is a similar technique used, say if the processor must remain powered up all the time, switching on the power-hungry parts when needed.

[travellerw]

Ya I'm %95 sure it will require a component replaced. Unfortunately I'm not in Florida ( a long way away in cold Canada).

You can trace the power button across the ribbon cable with the continuity function on your multimeter! I would look on the main board for electrolytic caps (little cans sticking up). Visually inspect them, you may see one leaking or bulged. Look at the brand names, if you see names like CapXion the its highly likely to be a failed cap. A cap is pretty easy to replace.

If you do give up, and buy a new one. I would take this one off your hands instead of you throwing it out.

[continuouswave]

Lots of heat in Florida--check for a failed electrolytic capacitor. Usually a very close visual inspection can detect a bad capacitor. Look for signs of damage to the case of the capacitor.

[Cadence]

You have had probably some good info. In practicality I believe insuring your input voltage is not low. When putting it back together make sure any connections are good (there probably aren't many). Reboot the unit.

If that fails do a search for a new unit.

[rwidman]

Originally Posted by enovillo

.......... I'm inclined to think I'm going to have to bite this bullet and buy a new unit...........

That's the best plan. Unless you're a very experienced electronics technician with an actual schematic of the unit and access to replacement parts you're going to have very little chance of repairing this unit.

If Garmin no longer supports it, it's pretty old. There have been a lot of improvements since your unit was designed and manufactured.

BTW: Many of the suggestions posted are way, way off as far as trouble shooting this unit. We're not talking about a toaster here.

[Imprezza72]

Being a circuit level tech for NorthStar a few years back (4) follow muddy's comments been there done that he is on the money...

[Cadence]

Originally Posted by rwidman

That's the best plan. Unless you're a very experienced electronics technician with an actual schematic of the unit and access to replacement parts you're going to have very little chance of repairing this unit.

If Garmin no longer supports it, it's pretty old. There have been a lot of improvements since your unit was designed and manufactured.

BTW: Many of the suggestions posted are way, way off as far as trouble shooting this unit. We're not talking about a toaster here.

You nailed it. Apparently the OP does not have a schematic or an extensive background in electronics. Sam Photofax probably doesn't carry a schematic, if they are still in business? If a shop can be found to look at the unit? If honest would say buy a new unit, it would be cheaper than our time. Lets face it electronics a disposables.