Vehicle transient protections

[GerGa]

Hi,
     I am dealing with a 2011 Suzuki DR650 motorcycle that keeps blowing up my Garmin GPS's.  Luckily Garmin replaces them under warranty, two blew up.  The first GPS lasted a couple rides.  When it bricked, the 2A fuse in the high side of the DC-DC converter blew as well.  I discovered the fuse blew when I tried the 2nd GPS and wanted to ride(impatient) so I put in the nearest 10A fuse and bricked both the GPS and converter on startup.
  Per the cycle enthusiasts, I replaced my old lead acid battery because apparently even though it worked fine, it could be causing bad conditions because it had died before and been recharged.  I keep the new AGM battery on a quality charger that keeps it on float when not in use.

I have read through discussions on here and apparently even low voltage can be damaging.  Here is one thread I have read through: https://www.eevblog.com/forum/projects/vehicle-transients/
I don't have a scope but held my analog meter on the bikes battery while starting it and saw 2.5V to 15V.
I suspect startup could be what is blowing the gps, so I would like to incorporate some relaying to keep the supply off while the bike is off.  Unfortunately that will not be easy because I don't think there is any bus on the bike that is on only when the engine is running.  I can get switched voltage from the key in the "on" position, but that doesn't start the bike.  There is a starter button for cranking the engine, so I could have a relay on that to break the line to the converter, but still would not want the gps switched on as soon as the starter button is released(so maybe incorporate a time delay relay).  I am not sure completely.  I have experience in circuits from school and work, but only a limited amount.

My main question is what sort of protection can I add to shield my GPS from future issues?  I would like to devise a way to keep the gps off until a time delay after the bike is started, but also want to add protection in line with the supply to protect it better.

Here is a schematic of the bikes electrical system and the dc-dc converter for reference.  The bike utilizes a voltage regulator. 
dc converter:

bike schematic:

Any advice is welcome
Thanks

[GerGa]

It is tapped directly off the battery. The converter is 2 wires in, 2 wires out.

[GerGa]

It is tapped directly off the battery. The converter is 2 wires in, 2 wires out.

Strange stuff , the battery should sink anything and thinking of that the inverter should have all that stuff built in too , I'd try installing a dc filter a cheap one off the bay and still tapping into the orange wire off the switch somewhere in the loom , good luck with it .

I forgot to answer the headlight is only switched with the key turned to on.
Thanks

[danadak]

Load dump might be issue -

http://www.ti.com.cn/cn/lit/an/snva681a/snva681a.pdf

http://electronicdesign.com/power/eliminate-those-automotive-load-dump-circuit-protection-headaches

http://www.littelfuse.com/~/media/electronics/application_notes/tvs_diodes/littelfuse_tvs_diode_automotive_circuit_protection_using_automotive_tvs_diodes_application_note.pdf.pdf


Regards, Dana.

[Galenbo]

Strange stuff , the battery should sink anything and thinking of that the inverter should have all that stuff built in too , I'd try installing a dc filter a cheap one off the bay and still tapping into the orange wire off the switch somewhere in the loom , good luck with it .

A parallel battery like here, doesn't filter out higher frequency components, and the response of the device can be a bigger problem than the problem itself.

For problems like this, a good start can be to tap - an + mechanically directly from the battery, instead of chassis and some_wire_on_some_relay.
That gives your device only the problem of the internal resistance of the battery, and not the additional outside_resistances like wires, chassis and multiple connectors.
A surplus advantage is that those outside_resistances are used now as an extra separation between source and problem, instead as additional problem.

It's like connecting something before or after a Low-pass filter.

[GerGa]

Load dump could be too .

That converter gizmo , do you still have it?

It'd be interesting to depot it and look about on the PCB to see whats there GerGa .

I still have it, I wanted to look at it, but didn't have the time to scrape off all the waterproofing silicone.  It's the weekend now so I can get that done soon.  I can't say I'll know everything I'm looking at on the pcb though..

[joseph nicholas]

Use a cheap 555 timer delay on Aliexpress, these can be adjusted.

[jitter]

Popular subject right now, in the Projects, Designs and Technical Stuff section a similar thread is active. Some cross pollination may help both: https://www.eevblog.com/forum/projects/vehicle-transients/.

[GerGa]

Load dump could be too .

That converter gizmo , do you still have it?

It'd be interesting to depot it and look about on the PCB to see whats there GerGa .

I still have it, I wanted to look at it, but didn't have the time to scrape off all the waterproofing silicone.  It's the weekend now so I can get that done soon.  I can't say I'll know everything I'm looking at on the pcb though..

If you do and take a underside shot and topside shot picture (if you can) there will be plenty of eyes to look at it here , no rush , be interesting anyway as there is something suspicious going on , as why does not the tci pop too if its (a spike or whatever) , or the regulator itself , just a thought .

Another thing to , when you have had the converter attached and as mentioned direct to the battery , does that mean the converter is actually on/running all the time ?

The converter was connected directly to the battery, so yes it was on all of the time.  When The first GPS blew, I accidentally left it in the mount powered by the bike for 2 days, it was really hot too.  I rode to a destination a couple miles away, and took the GPS off while I was off the bike.  When I got back to my bike, I didn't even turn the bike on and the GPS would not even power on under it's own battery.  I thought being on for 2 days and being so hot did it, but then the converter blew along with it the next time(with the 10a fuse mind you, which is supposed to be 2a).

Small images of board below, or album of fullsize images here: http://imgur.com/a/B60qi
Top:                                                                                                                                                                    Notice the coil in this picture that I damaged while scraping off the silicone:
   

Bottom:

[GerGa]

But back to your answer about being connected and it is the reason i asked , left on all the time and heat , i really guess heat is the thing that killed it , being potted with nowhere for the heat to escape would be my very 1st guess now with your information .

Actually the first time the gps blew and was left powered by converted in the heat, only the gps blew after a short ride then starting the bike again.  The converter survived but the 2a fuse in the positive feed between converter and bike battery blew.
 I did not discover the 2a converter fuse blown until I plugged in the replacement gps.  All I had in a pinch was a 10a fuse, installed that, the converter powered up, then with the gps plugged in, I started the bike and poof went the gps and converter, while the 10a fuse was uneffected.

As for your comment on the wiring(which i did not quote to save space), there is an empty connector under the headlight cover that is switched with the key in on position.  I can use that, but will need to sit down and look into a more elaborate solution for reasons stated in previous posts above.  Perhaps with a 555 timer that someone mentioned above. 
Thanks

[CJay]

Perhaps with a 555 timer that someone mentioned above. 
Thanks

Don't forget to build in the load dump transient protection for the 555 

[GerGa]

Simply switching the the gps supply on timer by "key on" is OK but not ideal. Even if it is set to 1 or 2 minutes, it still does not ensure the gps is out of the circuit on startup.  The whole point is that my gps goes poof while in the bike circuit.  Yes simpler is better, that's why I bought this bike, but the primary goal is to protect my gps which has consistently blew on each startup.

[GerGa]

I do know how a relay works, but the key on does not initiate startup, as I have outlined in previous posts.  It's on the schematic as well.  I could have a relay coil in the starter button circuit with it's contact opening when the start button is pressed and closed when button released.  At that point I need some sort of time delay because I doubt the bikes electrical system is stable as soon as the start button is released.  In reality the circuit could use multiple relays and timer.  I can't just think up the circuit at once, I need to take the time to do it.
Edit:I should add that yes the circuit design would begin with a relay as you describe, but the design would not end there.

[jitter]

In your application, I would consider a simple on delay relay circuit that could be built with jelly bean components, something like this: