I'd reply to one of the many existing threads about this but can't because they've all been archived...which is why there are so many threads.

I believe 2.5 is the last version made by Smarthome before they went belly up. It's been since I got it but 3 times in 2 weeks tells me I'm in trouble. 

Has anyone had a 2.5 memory failure yet? Is there any point in rebuilding the power supply? I heard a rumor that new ones might be available again soon.

the new insteon has PLM's in stock now, both serial and USB.  Serial is required for 994, USB is the better choice for Polisy/eisy.

https://shop.insteon.com/collections/interfaces

18 minutes ago, MrBill said:

the new insteon has PLM's in stock now, both serial and USB.  Serial is required for 994, USB is the better choice for Polisy/eisy.

https://shop.insteon.com/collections/interfaces

 

Thanks. I ordered a new one. I also ordered multiple sets of caps for the 2.4-2.6 versions. I guess cap failure on their power supply redesign was a thing, just not a soon as the old design. I still have my previous failed PLM which appears to be a 2.4.

The caps look to be through hole, so they should be pretty easy to replace. I won't even attempt surface mount.

I believe 2.6 was sold shortly before the closure. Don't know if any revision above 2.6 is now in use.

Besides the caps being questionable. I had see on Smarthome about some having static discharge and the two serial signal being damaged.

I imagine some may have had connections to the two TTL level signals on the PLM with possible trouble.

V2.5 has the new serial port daughter board. With a better ESD specification. Surge suppression network on the two serial lines to the outside world. C7 and C13 where changed to a Fujicon TY series capacitor with a better high frequency rating for switching supplies.

They are through hole mounted capacitors. Just use some care. As some of them have a plating connection to both the top and bottom circuit runs. I accidentally didn't have the solder molten enough and tore out the plating between the top and bottom runs. Now I have a blue fly wire between a top to bottom connection.

The PLMs now shipping are hardware version 2.6

Ordered yesterday. Shipped today. Should be here by Tuesday.

Does anyone know if I can swap a board from my new 2.6 to my 2.5? Which board has the address of the PLM? The big one with the caps or the small one with the RJ45 jack? Hopefully the small one?

Yes you can but the replacement serial board with the link and scene memory. Would be empty and you would have to reprogram it.

The Serial Daughter board has the memory for the links and scenes. It is the small one with the single electrolytic capacitor and the RJ45 socket to the outside world.

The main board has the firmware revision, six digit Insteon ID, Category and Subcategory ID in it. The one with many electrolytic caps on it.

Edited May 9, 20232 yr by Brian H
Add information

8 hours ago, Brian H said:

Yes you can but the replacement serial board with the link and scene memory. Would be empty and you would have to reprogram it.

The Serial Daughter board has the memory for the links and scenes. It is the small one with the single electrolytic capacitor and the RJ45 socket to the outside world.

The main board has the firmware revision, six digit Insteon ID, Category and Subcategory ID in it. The one with many electrolytic caps on it.

 

So it sounds like this is a "replace PLM" process rather than just a restore and would require placing all battery operated Insteon devices into programming mode to update the links since the board with the caps is also the board that has the Insteon address for the PLM?

As you know. If you swap boards between the 2.5 and 2.6. A replace PLM is needed. With either combination. In the PRO you can disable automatic writes to battery devices and then do one at a time. Correct. The main board has the six digit Insteon ID and if you change the PLM or just the new main board with old serial board. The modules you are controlling will not process the commands if the six digit ID is different.

One of the symptoms of a failing power supply. Is low link database count or a blank database. Sometimes you can reload it and it may last for a short time.

I got my replacement caps today from an ebay seller. Spent a few extra bucks for the curation of them rather than going straight to Digikey. Now I'm not so sure it was the right thing to day. 

Two of the caps have significantly bumped voltages which I thought was an issue for powerline signal reliability:

C3 - 6.8 µF 250 v (replacement is 450 v).
C11 - 100 µF 25 v (replacement is 50 v)

Is this ok?

C3 is the filter capacitor for the AC Line side of the supply. 450V should be fine and was used by many in the very long thread. Maybe taller but still should not hit anything when mounted.

C11 is the filter capacitor for the Zero Crossing sensor and 50V should also be fine. As long as it physically fits and should if the kit was designed for the 2413S PLM.

The two that matter the most are C7 and C13 the two in the switching supplies main DC output to run the whole system.  Smartlabs tried many different capacitors in those positions. My original V1.0 actually has 10uF 25V with fly wire rework because the originally didn't have the Pi filter on the output. Versions have had 10uF 35V, 100uF 35V and 100uf 50V. In V2.3 they went with 100uF 50V Fujicon TY series capacitors that where supposed to be for switching supplies.

V2.5. Now 100uF/50V Fujicon TY capacitor. Parallel to the main board. Maybe with some component fastening glue holding them. Because of the physical size too large to fit directly on the board.

Edited May 10, 20232 yr by Brian H
Correct an error

The 2413S,2413U and V2 of the Access Points. All use the same base main board. Except V1.0 that had an older designed board with some runs cut and fly wires added to update them for a Pi filter output.

They have circuit runs on both sides of the board. Some connected with plating between the top and bottom sides. The capacitors for some places. Have the plated through holes. Use care not to tear the plating out of the holes and killing the unit.

C3 and C11 are in power line derived dc supplies and the higher voltage should not effect operations.

Edited May 10, 20232 yr by Brian H
Add something