Repair of 2413S PLM When the Power Supply Fails

[danu1964]

If you purchased a 2413S PLM Dual Band you will most likely experience a power supply failure sometime around 2 years of operation.  Some have seen it sooner and some later.  I think I'm on my third unit now and simply decided to understand what the failure mode was on these units.

 

The failure mode this repair is referring to is exhibited by no communications getting onto the power line and the Green LED is not lit or might be dimly lit.  Sometimes unplugging the unit for a day and plugging it back in will allow for operation for a limited amount of time (maybe 30 minutes) and then the problem repeats itself.

 

After tearing down my last unit there wasn't any obvious blown or burnt parts visible on the PCB.  I've designed a reasonable number of power supplies in my day and was quite disappointed in seeing a 250Vdc capacitor hanging on the rectifiers straight off the AC line.  Keep in mind, this capacitor is the main bulk cap and must handle a normal voltage of around 170Vdc.  At first, this would seem ok as there is 80Vdc of headroom left.  However, since it is not recommended to plug this unit into a surge protected outlet strip we now see our first potential failure mode.

 

Likewise, the secondary side of the power supply was found to have low quality filter capacitors reported in these forums by BrianH.  These appear to be the primary problem resulting in low voltage going into the 5V regulator.

 

So, I decided to just replace all the caps with high quality caps and specifically replaced the high voltage primary side capacitor with one rated for 400V which will have a better chance of surviving surges and spikes generated on the AC line.

 

I was able to purchase all the caps from Mouser.

 

DON'T TRY THIS AT HOME UNLESS YOU KNOW WHAT YOU ARE DOING!!  PLEASE DON'T POST ME ASKING FOR STEP BY STEP INSTRUCTIONS, ETC., AS THIS WOULD INDICATE YOU DIDN'T READ THE PREVIOUS SENTENCE.

 

 

Original Part                                                New Part#
Replace YICCON 6.8uF 250V cap with      ESX106M400AH4AA 10uF 400 volt
Replace FUJICON 100uF 25V cap with     UTT1E101MPD 100uF 25 volt
Replace SAMCON 10uF 35v cap with        EKY-500ETD100ME11D 10uF 50 Volt (2 of these)
Replace FUJICON 10uF 16V cap with        UTS1C100MDD 10uF 16 volt

 

Mine has been running for two weeks after changing out these caps.  Keep in mind that sometimes the capacitor failures could cause other components to fail.  In my case all is well.

 

I spent in the neighborhood of $2 for the replacement parts and 3X that for shipping, and a little time at the bench.

[Brian H]

Thank you for the list of parts you used.

 

Anyone with 2413S with the the older PCB. The two original 10uF 35 V caps. Will be on a reworked  to add a coil and the second 10uF cap. That are on the newer PCB.

They will have long leads and kind of tacked soldered onto the original PCB pads. With the coil between the + side of the caps and a blue wire to the back of the PCB. Never looked at the back of the PCB but imagine a run cut to add the extra filter parts to the unregulated 12 volt supply.

One mine has 10uF 25 volt ones in it. Since the unregulated 12 volts is closer to twenty volts. 25 volt caps is not a large safety margin. The other one has the 35 volt ones in it.  Could be a size thing as the 25 volt one is smaller and fit better sticking up off of the PCB.

If you where a good neat solderer. You maybe able to change the caps on the older revision PCB.

My two Hardware 1.0 have the rework but I can't say exactly where the later PCB was used. I know Hardware 1.5 has the later PCB.

 

I doubt anyone has the version Smartlabs has in the FCC Database photos. It has one 10uF cap on the 12 volt supply. No coil or second cap.

[apostolakisl]

Thanks for posting the BOM,  makes it much easier.

[Panda88]

Can anybody provide a link to the thread with suggested modifications - my 2413 just died - it must be pretty old as it has survived more than 3 years

 

Thanks

Chris

[Brian H]

I believe this is the only thread with the exact information. The part numbers on the right side of the original post are the Mouser Part Numbers.

There is No photos or step by step instructions that I have seen.

Check the hardware revision sticker on the back.

I know my V1.5 and 1.B have the PCB with all the capacitors on it.

My older V1.0 has the older PCB and one of the capacitors and small coil tack soldered to the PCB and a Blue Wire running to the back of the PCB.

From the later PCB Silk Screened on the PCB:

C3 is the 6.8uF 250V

C11 is the 100uf 25V

C7 and C13 10uF 35V

C8 is the 10uF 16V

On the older PCB I believe C13 is the one tack soldered to C7. As I did not see a C13 Screened on the PCB.

[larryllix]

...   I've designed a reasonable number of power supplies in my day and was quite disappointed in seeing a 250Vdc capacitor hanging on the rectifiers straight off the AC line.  Keep in mind, this capacitor is the main bulk cap and must handle a normal voltage of around 170Vdc.  At first, this would seem ok as there is 80Vdc of headroom left.  However, since it is not recommended to plug this unit into a surge protected outlet strip we now see our first potential failure mode.

 

Regardless of voltage ratings, capacitors have a current limit rating, and this configuration cannot be expected to last long slapping 170vdc directly into any voltage capacitor without a ballast impedance. Upon replacement a good practice  would be to have a discharge direction diode parallellel by a few ohm resistor, both in  series with the replacement capacitor. 

 

The low voltage capacitors usually get away with this huge initial current charge inrush potential  as wire leads and foil impedances typically limit currents to less than 100 amperes for microseconds. On a higher voltage application the PCB  foil and wire impedances become insignificant to the initial chaarging inrush current.

 

Nice going!!

[Brian H]

There is a 27 Ohm 2 watt  resistor in series with the AC Line Input to the bridge rectifier feeding the high voltage DC side of the switching supply. That should limit the start up current into the capacitor.

When powered down. The switcher should keep running until it discharges the cap and stops running.

[Brian H]

Though I have not seen any reports of the Hardware Revision 2 {Hardware 1 was on the older 2412 base PCB and an added Daughter Card}  2443 Access Points going belly up.

It has the same Main Board as the 2413S PLM with out the Serial Daughter Card and different firmware.

So the power supply components should be the same.

[Panda88]

Tried the fix andit recovered my broken unit

[danu1964]

Tried the fix andit recovered my broken unit

Awesome!  Thanks for reporting back about your success.

[LeeFleishman]

Just completed the re-cap of a 2413S that died in May of this year. As is typical, this PLM failed approx 2 years after installation at a customer site. I used the specific caps suggested in this thread (procured from Mouser). Very happy to report that the unit is restored to working order.

 

I will re-cap 2 additional dead PLMs (also replaced for clients this year), plus the one that's been in service in my own home since Dec 2012 (so as to - hopefully - prevent its "upcoming" demise).

 

I'll post the results of these 3 additional "resurrections" upon completion. 

 

Thanks to all who provided the specifics of this solution.

 

Lee Fleishman

INControl Home Automation

Thousand Oaks, CA

 

[LeeFleishman]

Successfully re-capped two additional failed 2413S PLMs, plus the working one in my own home system (as I said I would in my earlier post). The two additional failed units are once again in proper working order. The PLM from my home system is (as expected) also working properly following the re-capping. Now, only time will tell if replacing the electrolytic caps in the power supply section of the PLM will indeed reduce the outrageously high failure-rate of the 2413S PLMs. 

 

At the very least, being able to "resurrect" a failed PLM, and putting it back in service in the ISY-based system in which it was previously working (as opposed to replacing it with a new one) has one very attractive benefit (assuming no other failure besides the power supply caps)...don't have to perform the "Restore Modem" process, which can take several hours if the system includes a large number (100+) of devices. 

 

Again, thanks (and kudos!) to danu1964 and Brian H !!

 

Lee Fleishman

INControl Home Automation

Thousand Oaks, CA

 

[Steven]

Thanks so much for writing this. I took my failed 2413S apart last night, and didn't see anything obviously wrong, but I was thinking about replacing the capacitor, so I googled "2413S repair" and found your article. I bought some capacitors today at the local electronics store (living in Silicon Valley has its advantages!) and replaced two of them (I didn't buy all the correct ones). Now the PLM is working great. Thanks again.

[Brian H]

Where the replacement you used for  C7 and C13 a Low ESR types?

[Steven]

I'll have to check on the ESR, I don't think it was indicated on the package. But they were labeled as "high temperature", which should help them last longer.

 

I told Smart Home about the repair I made, and they responded:
 

Thank you for contacting Insteon Support.

 

I have forwarded this information on the the head of the product department to take it into account for future revisions.

[Teken]

I'll have to check on the ESR, I don't think it was indicated on the package. But they were labeled as "high temperature", which should help them last longer.

 

I told Smart Home about the repair I made, and they responded:

 

 

I think its great they replied back and indicated it would be sent up the engineering ladder. I hope very much the new PLM that I received has better quality components with in.

 

But, history has proven *Hope don't float* so I won't hold my breath in this area.

[LeeFleishman]

Just sat down to re-cap another 2413S. This time, it's a brand new one slated for a new system I'll be installing tomorrow for a new client. My thinking is that re-capping now may prevent the all-too-common failure 2+ years from now (couldn't hurt, right?).

 

The label on this new PLM indicates that it is rev V1.C, with the date code 1425.

Upon dis-assembly and inspection, I found that:

**C7 and C13 are now 100uF, 35V, Fujicon (as opposed to 10uF, 35V, Samcon);

**C3 is still 6.8uF, 250V but it is no longer Yihcon brand (now it's "CA" brand, as best as I can tell);

**C8 and C11 are both unchanged from earlier rev units (as specified by danu1964).

 

I suspect the changes at C7 and C13 is SmartLabs' attempt to mitigate the 2413S's high failure rate. However, I don't know if simply increasing the capacitance from 10uF to 100uF will in fact "do the trick" (do these new caps address the ESR issue?). On that basis, I plan to go ahead with the re-cap, using the 10uF, 50V caps suggested by danu1964.

 

Additionally, the change in brand of C3 (with a cap of the same 250V rating) does not address the headroom issue on the primary side of the power supply. Changing C3 to a 400V rated cap just makes good sense. 

 

All the best,

Lee

INControl Home Automation

Thousand Oaks, CA

 

 

 

 

 

Just sat down to re-cap another 2413S. This time, it's a brand new one slated for a new system I'll be installing tomorrow for a new client. My thinking is that re-capping now may prevent the all-too-common failure 2+ years from now (couldn't hurt, right?).

 

The label on this new PLM indicates that it is rev V1.C, with the date code 1425.

Upon dis-assembly and inspection, I found that:

**C7 and C13 are now 100uF, 35V, Fujicon (as opposed to 10uF, 35V, Samcon);

**C3 is still 6.8uF, 250V but it is no longer Yihcon brand (now it's "CA" brand, as best as I can tell);

**C8 and C11 are both unchanged from earlier rev units (as specified by danu1964).

 

I suspect the changes at C7 and C13 is SmartLabs' attempt to mitigate the 2413S's high failure rate. However, I don't know if simply increasing the capacitance from 10uF to 100uF will in fact "do the trick" (do these new caps address the ESR issue?). On that basis, I plan to go ahead with the re-cap, using the 10uF, 50V caps suggested by danu1964.

 

Additionally, the change in brand of C3 (with a cap of the same 250V rating) does not address the headroom issue on the primary side of the power supply. Changing C3 to a 400V rated cap just makes good sense. 

 

All the best,

Lee

INControl Home Automation

Thousand Oaks, CA

[Teken]

Just sat down to re-cap another 2413S. This time, it's a brand new one slated for a new system I'll be installing tomorrow for a new client. My thinking is that re-capping now may prevent the all-too-common failure 2+ years from now (couldn't hurt, right?).

 

The label on this new PLM indicates that it is rev V1.C, with the date code 1425.

Upon dis-assembly and inspection, I found that:

**C7 and C13 are now 100uF, 35V, Fujicon (as opposed to 10uF, 35V, Samcon);

**C3 is still 6.8uF, 250V but it is no longer Yihcon brand (now it's "CA" brand, as best as I can tell);

**C8 and C11 are both unchanged from earlier rev units (as specified by danu1964).

 

I suspect the changes at C7 and C13 is SmartLabs' attempt to mitigate the 2413S's high failure rate. However, I don't know if simply increasing the capacitance from 10uF to 100uF will in fact "do the trick" (do these new caps address the ESR issue?). On that basis, I plan to go ahead with the re-cap, using the 10uF, 50V caps suggested by danu1964.

 

Additionally, the change in brand of C3 (with a cap of the same 250V rating) does not address the headroom issue on the primary side of the power supply. Changing C3 to a 400V rated cap just makes good sense. 

 

All the best,

Lee

INControl Home Automation

Thousand Oaks, CA

 

 

 

 

 

Just sat down to re-cap another 2413S. This time, it's a brand new one slated for a new system I'll be installing tomorrow for a new client. My thinking is that re-capping now may prevent the all-too-common failure 2+ years from now (couldn't hurt, right?).

 

The label on this new PLM indicates that it is rev V1.C, with the date code 1425.

Upon dis-assembly and inspection, I found that:

**C7 and C13 are now 100uF, 35V, Fujicon (as opposed to 10uF, 35V, Samcon);

**C3 is still 6.8uF, 250V but it is no longer Yihcon brand (now it's "CA" brand, as best as I can tell);

**C8 and C11 are both unchanged from earlier rev units (as specified by danu1964).

 

I suspect the changes at C7 and C13 is SmartLabs' attempt to mitigate the 2413S's high failure rate. However, I don't know if simply increasing the capacitance from 10uF to 100uF will in fact "do the trick" (do these new caps address the ESR issue?). On that basis, I plan to go ahead with the re-cap, using the 10uF, 50V caps suggested by danu1964.

 

Additionally, the change in brand of C3 (with a cap of the same 250V rating) does not address the headroom issue on the primary side of the power supply. Changing C3 to a 400V rated cap just makes good sense. 

 

All the best,

Lee

INControl Home Automation

Thousand Oaks, CA

 

Hello Lee,

 

I can confirm all of what you have stated in the 2413S PLM I recently received from Smarthome. Thank you very much for taking the time to follow up and provide more insight on this topic.

[Brian H]

Thanks for the updated capacitor information.

Fujicon maybe a better brand of capacitor. I will have to look on the Badcaps forums and see what they say about Fujicon.

 

I know C13 is right off of the rectified 200KHz voltage output from the small transformer in it.

Both C7 and C13 should be designed for switching power supplies.

 

Only way to know if C7 and C13 are low ESR would be to look their type.

As there are many types. Like high temperature, general purpose, low impedance and high reliability to name a few.

 

If you can see the type on the capacitor below is a link to the Fujicon data page.

 

http://www.fujicon.com/en/radial_lead_snap_in_type.php

[Steven]

My repair didn't last - my 2413S is dead again.  It could have been because I didn't replace all the capacitors. I'll have to take it apart again.

[danu1964]

My repair didn't last - my 2413S is dead again.  It could have been because I didn't replace all the capacitors. I'll have to take it apart again.

Hi Steven,

 

You will want to order all the capacitors I have listed above from Mouser.  I suspect you would receive them quickly as I believe Mouser is in California.  You will spend more on shipping than you will the parts.

 

I researched each capacitor to make sure the ESR specifications were as good or better than the original recommendations made by the manufacturer of the switchmode power supply chip.  I also made sure to select the longest lifetime capacitors available.  Temperature rating doesn't guarantee a low ESR capacitor.

 

 

Best Wishes!

[Brian H]

I see that Lee found different brand capacitors in his new 2413S. Also C7 and C13 are changed to 100uf 35V capacitors.

[apostolakisl]

 

I researched each capacitor to make sure the ESR specifications were as good or better than the original recommendations made by the manufacturer of the switchmode power supply chip.  I also made sure to select the longest lifetime capacitors available.  Temperature rating doesn't guarantee a low ESR capacitor.

 

 

 

 

Thanks for your research.  Mine is still up and running for a couple weeks.

 

Somebody should just order like 20 or 30 sets of those capacitors and make a little repair kit to sell.  They could easily mail in a padded envelope for pennies instead of the $8 or so Mouser charges.

 

My big conundrum is whether to replace the caps on my new one which is currently running my system.  "If it ain't broke, don't fix it" keeps coming to mind.  Maybe I'll buy the caps, put them in a baggy and tape them to the unit, ready to go for that day.

[Brian H]

If your PLM is real new. It may have the updated brand capacitors and larger capacity C7/C13 in it already.

Maybe a peek under the top cover may reveal something.

[apostolakisl]

If your PLM is real new. It may have the updated brand capacitors and larger capacity C7/C13 in it already.

Maybe a peek under the top cover may reveal something.

 

I'll take a look.  But . . .. when I went to buy it, they were our of stock, and my system was dead in the water.  So I went hunting and found another source that had one in stock.  Now, I am wondering if the reason they were out of stock is that they were waiting for the new revised version and I unwittingly bought old stock from a 3rd party.