Repair of 2413S PLM When the Power Supply Fails

[junkycosmos]

Ordered last week direct from SH, arrived today

PLM v2.4  build code 3817

(38th week of 2017 was September 18, 2017)

 Back of main board has stickers 2413S-9E  and RFDB-D2

Front of main board has  P/N 200-87902000-025-9  v1  Date 20160822

[healeydriver]

My PLM died a couple of weeks ago.  Version 1.7,  I just received my new PLM, it is version 2.4.   I've discovered that I do not need to plug the PLM onto the wild/raw circuits.  Instead I have it plugged into my battery backup system which powers all the other security products in my house.   (We suffer from periodic power outages and have a backup generator....however, there is a small window of no power when the generator fires itself up.)   The battery backup ensures that there are no surges or outages as the household shifts over to generator power or shifts back to line power.   I thought, from the PLM manual and also advice I've read here, that I could not plug the PLM into a surge protection device, but instead had to be on the main line in order to ensure good communications with the Insteon switches, outlets, etc. scattered elsewhere throughout the house.   However, when I decided to run an experiment and plug the PLM into the battery backup, I discovered that my ISY 994i was communicating fine with about 1/2 my household insteon devices.    On a hunch, I decided to leave the PLM in the battery backup and to plug in an access point (a lot cheaper than the PLM) into the wild circuit..... Lo and Behold!!!.... My ISY is happy with every other Insteon device.   apparently the PLM's RF signal is good enough to get into the household network of most of the Insteon devices, and the access point connects with the rest.   While I intend to do the cap replacement fix to my version 1.7 PLM, I'm hoping to have my new version 2.4 PLM live a long and happy life protected by the battery powered UPS. … Just a FYI... 

[MarkJames]

Interesting.

I hope that works and that the failure isn't just poor design.  Having had Insteon devices from day zero I have a box full of failed PLM's, Filters, KPLs, SwitchLinc relays, dimmers, outdoor switchlincs and on and on.

 

They're definitely far more robust now than their early products.  The dual band are a big improvement too.

 

Good luck!

 

[Brian H]

When you put it onto the UPS. It filtered out all the Insteon power line signals to and from the PLM. You where using Insteon RF only.

Adding the 2443 Access Point for added Rf to power line signaling.  Got the Insteon power line signals onto the power lines.

[MikeD]

1 hour ago, Brian H said:

When you put it onto the UPS. It filtered out all the Insteon power line signals to and from the PLM. You where using Insteon RF only.

Adding the 2443 Access Point for added Rf to power line signaling.  Got the Insteon power line signals onto the power lines.

This is exactly how I am configured, PLM powered by my UPS. Been configured like this for several years. The only thing you loose by this configuration is 1 hop, since you are using an Access Point (or a SwitchLinc in my case) as the first hop, which is not a problem if you have a robust Insteon network. Lately I have been installing Z-Wave devices instead of Insteon. My only issue has been (not) implementing multi-channel devices since I have not made the jump to v5.xx yet. 

[ScottAvery]

second plm dead and Amazon doesn't have any replacements - this repair requires more than a radio shack soldering iron, I take it?

[Brian H]

The PCB is double sided and some of the capacitors have a plated through hole. Linking the top and bottom sides. So care has to be used not to tear the plating out. Also the negative connections may connect to a heavy ground plane in the center of the board.

A good soldering iron and good solder is needed. Things like a solder sucker or solder wick helps.

If you are into electronics a nice desoldering machine is nice. Well I can dream anyway. I don't have one either. ?

[junkycosmos]

 

I took a small risk and ordered another PLM 2 weeks ago from SH direct.    Sure enough it showed up as a v2.4, same date code 3817 and on box "supports 1000" links, meanwhile the sales page on SH still shows 

• 2,000+ links (compared to around 1000+ links previously)
• Added a beep sounder
• Red/Green LED light
• Supports faster reading and writing of the PLM's database by another controller
• Much lighter weight

I called today to return it and the answer was sure, but I pay return shipping.   It always amazes me how SH has their default customer policies set.  After complaint and direct reminder that I was shipped something other than I bought they are now mailing me a shipping label.

 

I do wonder what else they changed to drive the specs on site of

• 2,000+ links (compared to around 1000+ links previously)
• Supports faster reading and writing of the PLM's database by another controller
• Much lighter weight

 

Also I was told on the phone today "we have been getting these calls a bit, seems the web site is wrong and current version only supports 1,000 links."    Meanwhile email I had with support 2 weeks ago noted::

Thank you for contacting Insteon Support.
We got an update from our resources stating that the latest revision of the 2413S is Rev # 2.5
The changes from 2.4 to 2.5 are very minimal.

 

So ... seems they still need to use up the v2.4 units... 

 

Curious to hear when anyone gets a 2.5 unit in their hand

 

[Brian H]

I know at 2.4. The serial port Daughter Board was completely redesigned.

RS232 interface chip changed to one with better ESD ratings and both serial signals to the output jack. Now have what looks like some surge protection on them.

To get more Link Database size. The memory chip on the Daughter Board would have to be a larger capacity one.

Except for the larger {2000} Link Database. The rest of the features look like what they said comparing the old 2412S to the 2413S.

Even my original 2413S V1.0 has a beeper,  Red/Green LED, is lighter and faster memory over the 2412S. My V2.4 also has the same 24FC64 memory chip in it as the V1.0.

[jakekooser]

Here's a spin on this.  I was thinking of shifting away from the ISY994i to a PC running Ubuntu with Home Assistant on it, so I ordered the 2413U USB version, couldn't make it work, then opened it up, swapped in my old PLM from the one I was using (failed on Friday night), put it back together, plugged it in and right back in business.  I'll recap the old one, and another dead one that's been here on my desk for a while, but just letting people know it's an option...

[junkycosmos]

hi Jakekooser

so you removed the daughter board of your old 2413S (serial) into a 2413 USB or do I have that mixed up?

[Brian H]

If you swapped the Serial Daughter board from the old 2413S in to the new 2413U.  It probably should work fine.

Only thing is if a program  reads the PLM ID.  It could see it as a 2413U PLM as the subcategory ID is different between the 2413U and 2413S.

If the old PLM was a 2412S then the memory on the daughter board is a slower speed and probably the >2000 link size.

[Kirboid]

A soldering tip (no pun intended) that may help those who want to repair their PLM. I have not repaired a PLM, but this works on many other devices I have repaired over the years. Commercially manufactured boards typically use a higher temperature low lead or lead free solder. This makes is more difficult to desolder without a dedicated desoldering station. To desolder, as you heat the component, add a small amount of leaded/lower temperature solder. This solder will transfer the heat better to the board trace and then mix with the factory solder, making it much easier vacuum or wick away. This reduces the overall amount of heat applied to the board and can help avoid lifted traces.

[jakekooser]

On 9/2/2018 at 10:38 PM, junkycosmos said:

hi Jakekooser

so you removed the daughter board of your old 2413S (serial) into a 2413 USB or do I have that mixed up?

Exactly, I placed the old 2413S daughter board into the 2413U.  No problems (that I have detected) so far, and everything just works again.  I don't have a huge amount of devices, so I don't think there is any issue with PLM ID.  I'm not too concerned about speed, as it mainly acts as a conduit between Alexa, Home Assistant, and Insteon.

[Brian H]

I don't think the Category and Subcategory  are used by the ISY994i. The HouseLinc software read it and uses the proper communications protocol that model PLM or HUB would use.

[MichaelS]

Just wanted to say thanks to all for this great info.  My v1.B went bad, and I ordered the parts from Digikey. Just installed them, and bam, a working 2413S.

I ended up going with a Rubycon cap for C7 & C13.  It is a smaller design for the 100UF 50v, which enabled me to install them upright.  There were stated to be long life 10,000hour power supply caps. The part number is: 50ZLJ100MT78X11.5

Thanks again!!

[bigb]

Thanks for this thread!

My 6+ year old  V1.7 PLM died a few weeks ago.  It was always on a UPS, in a cool ambient temp environment, and we also have a whole house surge suppressor.  I did always think that the PLM seemed to run warm. 

I replaced just C7 with a 100uf 50V generic cap I had laying around, and it jumped back to life. 

Here are the readings on the old C7, 10uF 35V cap:  140 ohm ESR   6.7 uF     Vloss=18%  (totally trashed)

Here are readings of a new generic 6.8 uF 50v cap:   0.72 ohm ESR   7.5 uF     Vloss=0.5%

And here are readings from the the generic 100 uF 50v cap I put in:   0.73 ohm ESR   93 uF     Vloss=1.1%

 

[ngocvin]

Thank you for this thread!

My PLM died about 3 years ago and I never had time to look into it.  At first I thought it was my ISY994/IR Pro.  The firmware version on my ISY would error out and not do anything if the PLM is dead.  So, last summer, I decided to look into the problem.  I updated the firmware on the ISY and realized that it was the PLM.  I have no status LED lit.  I took my PLM apart and did not notice any blown caps.  The back of the PCB seems like something leaked.  See pics.  Anyhow, I ran across this thread.  Here is what I replaced:

Replace YICCON 6.8uF 250V cap with ESX106M400AH4AA     KEMET             10uf 400v

Replace FUJICON 100uF 25V cap with UTT1E101MPD           Nichicon             100uF 25 volt

Replace SAMCON 10uF 35v cap with EKY-500ETD100ME11D   Chemi-Con     10uF 50V (2x)

UTT1C100MDD1TP was not available/on back order.

Anyhow, my PLM is working again!

Thank you.

[Brian H]

That white smeary film. I probably residue left over from a PCB board cleaning after wave soldering.

Some of mine also have a similar film on their PC Boards..

[Harold]

I looked at the specs for the 2413 today on the Insteon website. Both interface versions show as 1000+ for links. That + could be the "2000" not yet disclosed.

I am going to call them and see if they have anything useful to say about V 2.5. The steady breathing you hear is me not holding my breath.

Just out of curiosity, has anyone tried a class action suite against these guys. They have , for a long time, been selling a device well known to be unfit for purpose. This behavior is unacceptable. They have had long term knowledge of faulty design. The solution is readily understood and simple for the manufacturer to fix. They have not generally made the necessary changes in a timely manner and not made a recall to satisfy users of a defective product. The obvious and clearly observable  defect is caused by the use of lower price components (generally capacitors)  not meeting required specifications  that are clearly obvious to an even moderately adept engineer. These electronics should not experience the limited lifetime. It is at odds with electronic device reliability.

They are effectively stealing from the customers by requiring repeated purchases of faulty devices once a user is committed to the (small to quite large) proprietary product installation. Stopping the (I presume) agreement with UD to prevent the manufacture of a UD version of the device. Which would have most likely provided a superior device. Again, a specific ploy to consider only device income without consideration of people purchasing system devices to implement complex and extensive device  home implementations. These people have spent significant money to be locked into a single device that is the only interface to their products, and regularly fails.

[larryllix]

19 minutes ago, Harold said:

I looked at the specs for the 2413 today on the Insteon website. Both interface versions show as 1000+ for links. That + could be the "2000" not yet disclosed.

I am going to call them and see if they have anything useful to say about V 2.5. The steady breathing you hear is me not holding my breath.

Just out of curiosity, has anyone tried a class action suite against these guys. They have , for a long time, been selling a device well known to be unfit for purpose. This behavior is unacceptable. They have had long term knowledge of faulty design. The solution is readily understood and simple for the manufacturer to fix. They have not generally made the necessary changes in a timely manner and not made a recall to satisfy users of a defective product. The obvious and clearly observable  defect is caused by the use of lower price components (generally capacitors)  not meeting required specifications  that are clearly obvious to an even moderately adept engineer. These electronics should not experience the limited lifetime. It is at odds with electronic device reliability.

They are effectively stealing from the customers by requiring repeated purchases of faulty devices once a user is committed to the (small to quite large) proprietary product installation. Stopping the (I presume) agreement with UD to prevent the manufacture of a UD version of the device. Which would have most likely provided a superior device. Again, a specific ploy to consider only device income without consideration of people purchasing system devices to implement complex and extensive device  home implementations. These people have spent significant money to be locked into a single device that is the only interface to their products, and regularly fails.

They guarantee it for two years against manufacturing defects (I am betting). They will stand behind that. There is nothing else to demand and there weren't many other places to go. That is changing with many more protocols out there but IMHO Insteon still has the best hardware protocol for security.

[Brian H]

They have over the years. Made changes to the design. Larger sized capacitors and higher working voltages. In the power supply.

V2.4 has a redesigned Serial Port Daughter Board with a better Serial Post RS232 interface chip. That has a better ESD  specification. Along with a surge network  to protect the serial signal lines. As I had seen a post that the RS232 chip was sometimes failing and spikes were the possible issue.

So they are trying to improve the 2413S. Even with some advanced life test in the labs. You still may have issues as normally it took over two years to see them.

As for the 2000 Link Database and V2.5 mentioned on the Insteon Forums. I suspect it was a very premature set of information or management changed their minds. That seem to still follow the Smarthome  philosophy being  continued by the new owners.

[Harold]

My point was, the 2413 is a simple device. Should we expect it to last for more than a year or two. I would. I am at the point where I have to buy the same device for the fourth time. Roughly $320 total for the very same device. Absolutely required for the thousands of dollars spent buying the basic system devices. With no alternative to buy since they have stopped UD from making a similar device. And I would certainly believe that UD would make a much better product. Would we accept our televisions to need replacement every couple of years. A vastly more complex electronic device.

Your car, your phone, your home heating system, your computer; and on and on. All vastly more complex than a trivial line interface that has spent years not getting seriously better quickly. It is just annoying. And think about how much money they have made by repeatedly selling the same device that has not been made reliable. It is inexcusable.

[Brian H]

I understand where you are coming from.

The older model 2412S PLM with an old fashioned linear power supply. Had a longer life and some users still have them in use now. Only issue was they had a slower processor and memory. So some links took too long to find if you had many modules and scenes. They where also power line only. So some used an Access Point or Range Extender in the pass through outlet on the front

From what I have seen. Smartenit module built on the base PLM main board. Went back to a basic 2412 main board with its linear power supply. You can guess same short lives with the base 2413 main boards.

[Harold]

There is a 2413S 2.5. But you can't specify a 2.5. You may get a 2.4 or a 2.5. They, of course, said that a .4 is just as good as a .5. I would not be surprised if I get a .4 delivered by the Easter Bunny.

I just ordered one on the phone. There is a Presidents Day sale expiring today at SmartHome. The site they are using for the sale coupon is being blocked by several pieces of protective software in my PC and router. A discussion ensued. A helpful person manually credited me for the sale price.