Hi there,

My 450G died on me this weekend. I've had it running for some two years or so with no problems whatsoever. I have the feeling it is a serious hardware problem as it won't beep when I apply the power. Also a reset (holding reset and plugin in power) won't make any difference. Serial cable don't report any chars when booting and the link leds on the interfaces stay off.

Is there a check I can perform to make sure it's really dead as in dead? A visual inspection of the PCB doesn't reveal anything. I also don't smell anything burning when doing a 'nose check' over the PCB. I just read something about faulty capacitors. Is this a common cause?

thank you for your time,
Martijn

did you check if serial port show u something? maybe the problem is ros

did you check if serial port show u something? maybe the problem is ros

Upgraded from v4.6 tot 4.16 some weeks ago. As I stated the serial port has no activity. Never changed the defaults and 9600,8N1 remains dead silent if I powercycle the board. Even with the resetbutton pressed and released whilst plugin the power as it should boot from a failsave I've read.

Are the electrolytic capacitors bulging on the tops or are they still flat as they should be?

If bulging, they need to be replaced.

Tom

Are the electrolytic capacitors bulging on the tops or are they still flat as they should be?

If bulging, they need to be replaced.

Tom

They are indeed bulging and not perfectly flat. I think I'll visit the local radio shack and give it a go. Will report back when I've replaced them.

Thanks Tom

A photo says so much more.

Just paid a visit to the local radio shack and he was positive those electrolytics where bulging. Spent €1.20 for two new ones. Cannot get 560uF but the consensus is that 470uF will suffice just as well. Any thoughts on that?

On another note, I have a 25W or so soldering iron but the guy at the counter was positive I could replace them as there is enough empty space around the mount so heat won't have much influence on the other components.

Will report back.

We have been changing capacitors on our failed RB450s for years, usually to 1000uF ones. The units work perfectly after the operation. Search on the forum, and you can find several topics about this.

Replaced the old capacitors with new 470uFs and it spun into life. Sweet! The Routerboards are an amazing piece of hardware and I would buy one again in a heartbeat if this one fails beyond repair in a near future. Too bad there are series out there that are plagued with the faulty capacitors, but that's some upstream component problem a lot of hardware builders are troubled with.

At work we have some RB 450Gs in service at key positions. We have some spares on the shelve, but I think I'll spend some 10 mins for each board to replace capacitors on forehand. The one at home was the first to fail, but if congo says he's replacing caps for years failure must be imminent.

Indeed there are a lot of pointers on this board about these capacitor problems and ways to replace them. Even came across a rule of thump to get an uF replacement value.

thanks for your time,
Martijn

Congrats on the repair.

We have found that running the older RB450 and RB450G on 12vdc power supplies instead of 24vdc will make them run cooler and the caps last.

Before you go about changing caps on all your RB450G boards, try replacing the power supplies instead.

Tom

Congrats on the repair.

We have found that running the older RB450 and RB450G on 12vdc power supplies instead of 24vdc will make them run cooler and the caps last.

Before you go about changing caps on all your RB450G boards, try replacing the power supplies instead.

Tom

The cause is not Vin or temperature. Those 2 caps are for voltage smoothing on PWM converter. Since the routerboard draws lots of current, we need low-esr caps here. Sanyo OS-CON, NCC polymer or any type of low-esr is preferred here. Regular caps from RadioShack may not last. When I received the RB450G, I replaced them with 2 620uf/6.3V OS-CONs already.

just a note - this can happen only on older RB450Gs, we have switched to a different supplier in the new revisions and this problem doesn't occur anymore.

Had two 450Gs give up the ghost, appears to be the same problem as in this thread.

Any idea where to get 620uf/6.3V OS-CON? I can't seem to find any online. Alternatives?

Had two 450Gs give up the ghost, appears to be the same problem as in this thread.

Any idea where to get 620uf/6.3V OS-CON? I can't seem to find any online. Alternatives?

Look at this website and forums, lots of very valuable data (like where to buy), or soldering guides.
I saved my 22" screen thanks to them (crappy caps in many samsung displays...).

Congrats on the repair.

We have found that running the older RB450 and RB450G on 12vdc power supplies instead of 24vdc will make them run cooler and the caps last.

Before you go about changing caps on all your RB450G boards, try replacing the power supplies instead.

Tom

The cause is not Vin or temperature. Those 2 caps are for voltage smoothing on PWM converter. Since the routerboard draws lots of current, we need low-esr caps here. Sanyo OS-CON, NCC polymer or any type of low-esr is preferred here. Regular caps from RadioShack may not last. When I received the RB450G, I replaced them with 2 620uf/6.3V OS-CONs already.

Sorry but my experience has been different. The caps are cooking from heat. Using a 12vdc supply will lower the heat. Sure, I agree that lower esr on caps is always desirable but if it was required for every design then electrolytic caps would not exist.

The heat is a byproduct. If less heat is given off then less power is being used and the caps will last longer. If the heat level stays below a level that causes destruction of the cap, it will last a long time.

Tom

thx elgo,

any recommend on part# to use on 450G from badcaps? looks like you can actually order from their online store