D-Link router - faulty power brick?

[momaka]

D-Link SMP-T1178 power adapter recapped

Well, good news: I found a 250V, 22 uF Panasonic SUW cap in my junk bin. I think it came from a scrap CRT TV board (I used to bike around and collect parts from busted TVs I saw on the side of the road when the copper scrapers would break them for the deflection coils.) It's only an 85C -rated cap and also considerably larger than the 200V, 22 uF Teapo SEK... but I managed to make it fit.



Before soldering it in, I checked it on my cap meter. Capacitance was in spec: around 23 uF. ESR wasn't great but not bad for a cap of that size either - about 0.75 mOhms. Honestly, I've seen CapXon and other crappy cap brands read much worse in LCD monitors - and those were caps that were much bigger too. So I think this one is going to be fine.

Also, for those of you worried that the fuse could arc over: I put 2 layers of electrical tape and a bit of hot glue between the fuse, main cap, and the PWM IC. I guess heatshrink tubing would have been better, but I would have had to desolder the fuse to put the heatshrink on it, and I was too lazy to do that. So hopefully the electrical tape and hot glue can calm your fears.

how much?
probably more than the tiny difference between a glass fuse and a ceramic one!

Probably true.

[stj]

Re: D-Link router - faulty power brick?

How much does 2 or 3 cm of shrink sleeving plus the time to install it cost?!

how much?
probably more than the tiny difference between a glass fuse and a ceramic one!

[PeteS in CA]

Re: D-Link router - faulty power brick?

that toy fuse is *very* close to the switching i.c. on the dc side!
if it explodes - as glass fuses do, the plasma will bridge the ac live directly into the switching i.c. and blow the whole thing to hell on the primary side - i hope the transformer has good seperation!!!!!!

How much does 2 or 3 cm of shrink sleeving plus the time to install it cost?!

[stj]

Re: D-Link router - faulty power brick?

that toy fuse is *very* close to the switching i.c. on the dc side!
if it explodes - as glass fuses do, the plasma will bridge the ac live directly into the switching i.c. and blow the whole thing to hell on the primary side - i hope the transformer has good seperation!!!!!!

[momaka]

Re: D-Link router - faulty power brick?

That's a flyback design, so the output caps are definitely being stressed.

Yup, that's why I left the 10V, 1000 uF Chemicon KZE in there. I also had 6.3V, 1000 uF KY available at the time, and even these have better specs than the Teapo SC caps original to the SMPS. Not only that, but the KY were shorter than the KZE and would have fit inside the case without having to drill a hole. However, I figured the KZE is just better for a flyback design.

The original issue was the small startup cap, though.

Unfortunately, now after all of these years, it's the primary cap that's gone bad.
So yeah, if it's Teapo, do a full recap regardless of the design/topology. They just don't last. I suppose the only exception is for larger snap-in caps, like in ATX PSUs - at least I haven't seen Teapo fail much there, so long as the PSU doesn't have APFC, of course.

that design is terrible - even if it's electronically good the layout is asking for a flash-over and a sooty boom!

But it's UL-listed!

Actually, take a look at the underside and you will see things aren't that bad. The manufacturer did put cuts in the PCB where there is danger of arc-over. Only possibly slightly more worrisome place is the plug prongs, which stick out a little inside the case and go right above the rectifying diodes. Could probably use a bit of silicone or heatshrink tubing there for extra isolation. Otherwise, it's not bad. Primary and secondary side separation is pretty good and has a legit Y2-class rated cap. Traffo is also relatively chunky, so it probably has enough spacing and good insulation between the layers inside. Not like the super-cheapo adapters that rely on just a single layer of tape (or two) to separate primary from secondary.

Also, another weird quirk that I noticed about this SMPS: the common-mode choke on the primary is on the DC-side (i.e. after the bridge rectifier) with a 2.2 uF electrolytic cap (another Teapo SEK) before the choke. I suppose they did this to save space while still being EMI/RFI compliant??? At least it seems so, since a 2.2 uF electrolytic cap takes a lot less space than a 2.2 uF metal film cap... or even a 0.22 uF metal film for that matter (and much less than one with a X2 rating.)

[stj]

Re: D-Link router - faulty power brick?

that design is terrible - even if it's electronically good the layout is asking for a flash-over and a sooty boom!

[PeteS in CA]

Re: D-Link router - faulty power brick?

...
I suppose the verdict is to replace ALL caps in this power adapter, especially if they are Teapo.
...

That's a flyback design, so the output caps are definitely being stressed.

[momaka]

Re: D-Link router - faulty power brick?

Alright, here are the pictures:




MOSFET Q1 is a 2SK2624, in case anyone ever needs to know.
PWM IC, U1, is a UC3843.
Main input filter cap is a Teapo SEK, 200V, 22 uF.

On that note, I'm really glad this thread got revived and I went to look inside the power adapter.
Why?
- Cheapo Teapo strikes again, that's why. In this case, it was the main input filter 200V, 22 uF cap:


It was bulged, so clearly it's failed. Don't believe me? Here:

Capacitance is almost OK, but look at that ESR. 34 Ohms!

I suppose the verdict is to replace ALL caps in this power adapter, especially if they are Teapo.

More caps to add to my future cap order list. Though, I have so many 5V, 3A power adapters now that I'm wondering if I should even fix this. But I probably will... eventually.

[momaka]

Re: D-Link router - faulty power brick?

Does anyone know the value of the to-220 resistor (R 01) in the smp t1178 power supply?

TO-220 resistor?
From what I remember, and the pictures I've attached previously, there are only two TO-220 devices: a MOSFET on the primary side in an isolated TO-220 case and a schottky rectifier (10A, 45V IIRC) in a regular TO-220 case.

The MOSFET would correspond to the TO-220FP devices on the left of the below image. And the regular TO-220 device above the yellow transformer is the schottky rectifier.


Images originally attached to this post:
https://www.badcaps.net/forum/showpo...09&postcount=8
(Sorry for the bad quality. That post was made way way back when I joined BCN, and I didn't really have much of a proper camera then.)

**** EDIT ****
Well, since my DI-614+ router (and its adapter) wasn't stuck too far into my closet, I dug it out to take a look inside the adapter (since I never really glued it back and only used a wire tie to hold it together - how convenient ).

So here is what I found:
The "R: 01" text on the PCB corresponds to the PCB version. The actual PCB designator for the TO-220 schottky rectifier is D7, and it is a SB1040 (10 Amp, 40V rectifier - my memory wasn't too far off this time, even after 10 years since the repair! )

I'll also grab the MOSFET p/n in a few moments to get better pictures too, since I have it out already... And you will see why. Stay tuned!.

[PeteS in CA]

Re: D-Link router - faulty power brick?

Does anyone know the value of the to-220 resistor (R 01) in the smp t1178 power supply?

Got a picture?

[n3rpc]

Re: D-Link router - faulty power brick?

Does anyone know the value of the to-220 resistor (R 01) in the smp t1178 power supply?

[momaka]

Re: D-Link router - faulty power brick?

No trouble so far whatsoever.

Scratch that. I'm starting to get lots of connection drops on the wireless. Occasionally, the router will restart too and the wired will drop out for a second.

I'm pretty sure this has nothing to do with the power adapter though. I opened the router earlier last week and I think the lone 1000uF Teapo SEK on the input is starting to bulge. Really hard to say, though. It looks flat, but not as flat as other Teapo caps I've seen. It feels a little "springy" too when I try to push down on it's top. It's an 8mm Teapo, so I won't be too surprised if it's bad.

I'll be recapping the router this weekend hopefully. Really annoying when the wireless decides to start dropping every 20 seconds like yesterday. Today, it's been warmer in my room and the wireless has been great, though. I've been using remote desktop for a few hours now without getting any drops.

[budm]

Re: D-Link router - faulty power brick?

Just like 2-WIRE wallwart for AT&T/2-WIRE.
http://s807.photobucket.com/albums/y...OWER%20SUPPLY/

[japlytic]

Re: D-Link router - faulty power brick?

I remember a similar situation where a client's set-top box took 20 minutes to start up - upon opening it, there were bulging capacitors.

[momaka]

Re: D-Link router - faulty power brick?

^ Interesting and good to know. I will definitely keep an eye on mine. My DI-614+ router has been put back in use again. It's been almost 3 weeks now since I put it back in service. No trouble so far whatsoever. The adapter runs much cooler too.

[sebr]

Re: D-Link router - faulty power brick?

Just fixed another JTA0302F-E by replacing the 47uF, 25V startup cap.
I also advise you to check the main cap, as on two of these PSU I fixed this was slightly bulged and decreased capacity. Yes, I know, very rare to have main caps failing but here it seems to be a problem. So better replace it in advance if you plan to keep this device.

[momaka]

Re: Jentec JTA0302F-E for the D-Link DGS-1008D

king of ghetto

Thanks
You should see what I have been cooking up in the last few days . I just took some pics today so I'll post them in the ghetto mod thread tomorrow or Sunday.

[goodpsusearch]

Re: Jentec JTA0302F-E for the D-Link DGS-1008D

Congrats on the fix!
Yeah, I've found that those small start-up caps fail quite often. That's why I keep a few of them in stock now.

Also, here is what my power adapter looks like now after I finished it:
https://www.badcaps.net/forum/attach...1&d=1327456835
Basically, the 10v, 1000uF Chemicon KZE cap was a bit too tall, so I drilled a hole in the case to accommodate for that. I did have some 6.3V, 1000uF Chemicon KY caps that would have fit, but I thought that this would be cooler (pun intended).
Held together with wire ties (because I don't like gluing things).

king of ghetto

[budm]

Re: D-Link router - faulty power brick?

At least it uses SMP IC, not like 2-WIRE AC adapter.
http://s807.photobucket.com/albums/y...OWER%20SUPPLY/
PVC pipe glue works well.

[TELVM]

Re: Jentec JTA0302F-E for the D-Link DGS-1008D

... Basically, the 10v, 1000uF Chemicon KZE cap was a bit too tall, so I drilled a hole in the case to accommodate for that. I did have some 6.3V, 1000uF Chemicon KY caps that would have fit, but I thought that this would be cooler (pun intended) ... Held together with wire ties (because I don't like gluing things).