Brian H Posted January 25, 2018 Posted January 25, 2018 Thanks for the information. I would not think they would have V2.3 units still in the warehouse but I got my V2.3 when it was the current model from Amazon 04/21/17.
slcasner Posted January 30, 2018 Posted January 30, 2018 I have done my first PLM capacitor replacement repair, on a 2413S V1.7 1222 that wasn't communicating reliably but still showed a green LED, with mixed results. I ordered from DigiKey: Digi-Key PN Mfr Part Number Manufacturer Description 399-6538-ND ESG106M400AH4AA KEMET 10uf 400v replaces 6.8uf 250v 493-3276-ND UTT1E101MPD Nichicon 100uf 25v replaces 100uf 25v 565-1598-ND EKZM500ELL101MHB5D United Chemi-Con 100uf 50v replaces 10uf 35v 493-10390-1-ND UTT1C100MDD1TP Nichicon 10uf 16v replaces 10uf 16v The assembly process was relatively straightforward. The 10uF 400V cap is much larger than the original but actually fits the hole spacing on the PCB whereas the original cap's spacing was narrower. I put in the larger 100uF 50V caps lying down, as Smarthome does in the newer revs, but I observed that I needed to push C7 to be parallel to the bottom edge to avoid hitting the modular connector for the cable (photo below). I think these caps might fit better standing up. When I put this repaired PLM in place of the operational one to test, the green LED still lights. After doing Restore Modem I can control some devices but not others. In fact I got a "can't communicate" for one keypad that I tried; resetting the keypad restored communication, but I still could not control it. In the log I see: Mon 01/29/2018 01:35:17 PM : [iNST-TX-I1 ] 02 62 1E 02 E7 0F 11 FF Mon 01/29/2018 01:35:17 PM : [iNST-ACK ] 02 62 1E.02.E7 0F 11 FF 06 LTONRR (FF) Mon 01/29/2018 01:35:17 PM : [iNST-SRX ] 02 50 1E.02.E7 1E.47.4F AB 11 FF LTONRR (FF) Mon 01/29/2018 01:35:17 PM : [std-Direct Nack] 1E.02.E7-->1E.47.4F, Max Hops=3, Hops Left=2 This Nack is sent right away every time I try to command the keypad on or off, so it is not a timeout problem. What does this mean? When I did the Restore Modem operation I was concerned because the progress register advanced only to 6% before the popup disappeared. But when I switched back to the working PLM and did Restore Modem again, the behavior was similar (perhaps 7%). But the log of the restore operation is identical and the result of Show PLM Links Table was identical for the two PLMs. Since I can successfully control some devices and not others, perhaps the repaired PLM is only able to communicate with RF or powerline but not both, and the devices that I can't control were depending on the other mode? Is there a way to investigate that hypothesis? Other debugging suggestions welcome.
Brian H Posted January 31, 2018 Posted January 31, 2018 (edited) You may want to do a Tools Device Link Table test on the module. Then a compare to see if it is what the ISY994i thinks should be in it. The Restore Modem (PLM) should have written the new PLM's ID into the modules. Edited January 31, 2018 by Brian H
slcasner Posted February 2, 2018 Posted February 2, 2018 I mentioned that the output of Show PLM Links Table was identical for the two PLMs. Is your suggestion to check the PLM's ID shows up in the Device Link Table for the devices that are failing to be controlled? Is there a reference to tell me how to interpret the logs in the Event Viewer and the Links Tables? BTW, I bought an inexpensive LCR tester that tells me the EKXM500ELL101MHB5D caps I bought for C7 and C13 are low ESR as claimed: 0.02 ohms. One of the two 10uF 35V originals that I removed shows an ESR of 41 ohms, but I don't know how much damage may have been caused by the desoldering process.
Brian H Posted February 2, 2018 Posted February 2, 2018 Yes if the PLM ID is still the old one in the device. You would get a NAK from a I2CS enabled module. Response back from the problem module. Mon 01/29/2018 01:35:17 PM : [iNST-SRX ] 02 50 1E.02.E7 1E.47.4F AB 11 FF LTONRR (FF) So it is responding to the command.
slcasner Posted February 3, 2018 Posted February 3, 2018 Hmmm, the suggestion to check the device like table is a good one, except that ISY can't read the device link table, either, which I guess is not surprising. After putting in the questionable PLM and doing Restore Modem, I get the following when trying to read the keypad's link table: Fri 02/02/2018 05:05:44 PM : [iNST-TX-I2CS] 02 62 1E 02 E7 1F 2F 00 00 00 0F FF 01 00 00 00 00 00 00 00 00 C2 Fri 02/02/2018 05:05:44 PM : [iNST-ACK ] 02 62 1E.02.E7 1F 2F 00 00 00 0F FF 01 00 00 00 00 00 00 00 00 C2 06 (00) Fri 02/02/2018 05:05:45 PM : [iNST-SRX ] 02 50 1E.02.E7 1E.47.4F AB 2F FF (FF) Fri 02/02/2018 05:05:45 PM : [std-Direct Nack] 1E.02.E7-->1E.47.4F, Max Hops=3, Hops Left=2 Fri 02/02/2018 05:05:54 PM : [iNST-TX-I2CS] 02 62 1E 02 E7 1F 2F 00 00 00 0F FF 01 00 00 00 00 00 00 00 00 C2 Fri 02/02/2018 05:05:54 PM : [iNST-ACK ] 02 62 1E.02.E7 1F 2F 00 00 00 0F FF 01 00 00 00 00 00 00 00 00 C2 06 (00) Fri 02/02/2018 05:05:54 PM : [iNST-SRX ] 02 50 1E.02.E7 1E.47.4F AB 2F FF (FF) Fri 02/02/2018 05:05:54 PM : [std-Direct Nack] 1E.02.E7-->1E.47.4F, Max Hops=3, Hops Left=2 Fri 02/02/2018 05:06:03 PM : [iNST-TX-I2CS] 02 62 1E 02 E7 1F 2F 00 00 00 0F FF 01 00 00 00 00 00 00 00 00 C2 Fri 02/02/2018 05:06:03 PM : [iNST-ACK ] 02 62 1E.02.E7 1F 2F 00 00 00 0F FF 01 00 00 00 00 00 00 00 00 C2 06 (00) Fri 02/02/2018 05:06:04 PM : [iNST-SRX ] 02 50 1E.02.E7 1E.47.4F AB 2F FF (FF) Fri 02/02/2018 05:06:04 PM : [std-Direct Nack] 1E.02.E7-->1E.47.4F, Max Hops=3, Hops Left=2 I should make clear that the problem is not just this keypad; there are several devices I can't control with the problem PLM. If I compare the event logs for the Restore Modem operations on the good and bad PLMs, they are identical. Is there some other step that is supposed to be done to tell all the modules about the new PLM?
Brian H Posted February 3, 2018 Posted February 3, 2018 (edited) Powering down the old PLM and ISY994i Powering up the new PLM letting it initialize. Power on the ISY994i so it detects the new PLM ID. Restore Modem (PLM) should go out and change all the old PLM ID number to the new PLM ID number. In the modules. I have seen some say. Restoring on modules may also be needed for some of them. I am not too good with decoding Insteon messages. I know an I2CS module will NAK back with an FF in command 2. If the senders Insteon ID is not in the modules Link Database. That maybe why trying to read the modules Link Database is failing as the new PLM ID is not in the modules Link Database. Edited February 3, 2018 by Brian H
slcasner Posted February 4, 2018 Posted February 4, 2018 Thanks for your help. You hit on the right answer, although I coincidentally found it myself earlier that day after searching the forum for information about PLM replacement. I had not powered off the ISY because I did not realize that was required (it has been a while since I replaced this PLM before). That is why the devices had the wrong PLM ID and why the Restore Modem operation finished at only 7% on the progress register. When I did the full recommended procedure (#1) then the Restore Modem operation took much longer to rewrite all the devices. I have relied on the User's Guide document as my primary reference. Why doesn't it describe the PLM replacement procedure? It doesn't even explain the Restore Modem command. Better yet, why doesn't Restore Modem cause the ISY to go read the PLM ID first? The repaired PLM has now been operating fine for a day, so it looks like the capacitor replacement was successful. I'll leave it in place to see if there are any indications of the instability that caused me to replace it before.
sun-sense Posted February 15, 2018 Posted February 15, 2018 Sorry for the delay in replying. Too much going on, wife got sick as well. My simple replace and test didn't work 100%. What a surprise I found the wiki entry for the correct replacement procedure. It's about half way down the page under :- Restore Modem (PLM) Replace Modem (PLM) https://wiki.universal-devices.com/index.php?title=ISY-99i/ISY-26_INSTEON:File_Menu
sun-sense Posted February 16, 2018 Posted February 16, 2018 I just checked amazon again. Insteon PowerLinc Modem (PLM) Dual-Band Serial Interface, 2413S 3.2 out of 5 stars 130 customer reviews | 14 answered questions Note: This item is only available from third-party sellers (see all offers). Available from these sellers. Yep. Not available direct from Amazon anymore.
Brian H Posted February 16, 2018 Posted February 16, 2018 I saw the same thing when I wanted to get a new 2413S from Amazon. I few days later they again had some. Stocked, Sold and Shipped By Amazon. It is almost like from time to time. They push the Amazon merchants stores products. Unfortunately they may not have the present, latest hardware version being shipped.
sun-sense Posted February 16, 2018 Posted February 16, 2018 (edited) I tried Smarthome looking for v2.4. Here is their response. NOTE: How they ignore the v2.4 and and state the latest is v2.3 and above. I won't be buying there. From: James Reed Sent: 2/16/2018 Subject: Ref: PLM 2413S Hi Martin, Thank you for contacting Smart Home. With regards to your email, the latest version of PLM starts with Version 2.3 and above. So, we in smarthome website we always ship the newer version of items. Have a good day! Regards, James. -----Original Message----- From: 000001422808 MARTIN xxxxxx Sent: 2/15/2018 Subject: Ref: PLM 2413S Ref: PLM 2413S Will I be getting version V2.4 that came out in October (if not before) 2017. I have just returned an older version to Amazon, that was 14 months old even though it was advertized as new. I would like one that is less than 3-months old. This will be my third replacement. Thanks...Martin. Edited February 16, 2018 by sun-sense
Brian H Posted February 16, 2018 Posted February 16, 2018 Was it Amazon or an Amazon Store? Amazon probably will have the latest but it depends on if any where still in stock for awhile. I have seen both 2.3 and 2.4 reported. Amazon Stores may have anything. Depending on how many they sold and when they last purchased from Smarthome.
slcasner Posted February 25, 2018 Posted February 25, 2018 Sadly, I have to conclude that my capacitor replacement repair was not a complete success, but that may be because the problem with the PLM was not a failure of the power supply in the first place. This PLM never got into the state where the LED was red or dark. I replaced it with a new one in 2016 because programs were occasionally failing to do what they were supposed to do, that is, the ISY was sometimes unable to control some devices. Power cycling the PLM seemed to help for a while. Now, after replacing the capacitors in the unreliable PLM it seems to mostly work, but after a time the admin console was popping up error messages about being able to communicate with one or another device. When I swapped back in the newer PLM, it was able to communicate with all devices. Going back to the repaired one, it took multiple executions of the Restore Modem operation to get through to all the devices, and there still was one that it could not reach. I tried this with the PLM plugged into outlets in two different rooms (but not sure if they were on different halves of the panel) to try to make sure I could get through to all the devices so they would be linked to the PLM. So, even though this repaired PLM mostly worked in my system (all the timed lights still worked), it could not communicate as reliably as the newer one. Perhaps some component other than the capacitors has failed or degraded. Perhaps RF is not working well (I did put the antennas back into their original positions after the repair). I have more sets of capacitors, so if the newer unit fails I will try again.
ehjay Posted March 17, 2018 Posted March 17, 2018 For informational purposes, here are the caps I bought from Mouser to fix my EZIO device. I don't know if they meet all the requirements. I only replaced the 50V caps and they've only been in service a few months. 661-EKZM500D101MHB5D UCC Aluminum Electrolytic Capacitors - Leaded 100uF 50 Volt 647-UPW2G100MHD1TO Nichicon Aluminum Electrolytic Capacitors - Leaded 10uf 400V
MarkJames Posted March 30, 2018 Posted March 30, 2018 A little late to the game but I just got my parts to re-cap an EZFlora. My soldering skills are good but my de-soldering skills are atrocious. I've lifted a couple of traces on the board. I have a second one that I tried to figure out the answer from but it eludes me. I need to confirm where the - side of C7 goes and the - side of C8 goes Thanks! mark
Teken Posted March 31, 2018 Posted March 31, 2018 10 hours ago, MarkJames said: A little late to the game but I just got my parts to re-cap an EZFlora. My soldering skills are good but my de-soldering skills are atrocious. I've lifted a couple of traces on the board. I have a second one that I tried to figure out the answer from but it eludes me. I need to confirm where the - side of C7 goes and the - side of C8 goes Thanks! mark
MarkJames Posted March 31, 2018 Posted March 31, 2018 lol - ok - my bad. I didn't explain myself well. Thanks Teken, but what I meant was on the PCB what does the - side of each of those caps run to? I've lifted the trace on the PCB and so I can't solder it to the board anymore. I'm going to have to jumper the - leg on each of them to wherever it has to go next and I don't know where that is. Thanks! mark
Brian H Posted March 31, 2018 Posted March 31, 2018 (edited) Your EZFlora is new enough to be on the base dual band 2413 main board and not on the earlier power line only 2412 main board? I will see if I can find out some information for you. Minus side of C7 and C8. Edited March 31, 2018 by Brian H Fix Typo
Brian H Posted March 31, 2018 Posted March 31, 2018 The 2413 main board is a multi-layered PCB. With a ground plane in the center layer of the PCB. It can absorb heat from a soldering iron. Making some of the capacitor minus connections dificult to remove. I pulled one of my plated through holes myself and had to use some fine insulated wire jumpers. Depending on your soldering skills. You can jump the C8 and C7 minus connections to the C13 minus connection. Or you can use Pin #1 of the RF ICP programming connector. It is the pin with the square silk screening around it and marked RF ICP. It is the 5 position set of pads near C7 and C8. Near the edge on the PCB.
MarkJames Posted March 31, 2018 Posted March 31, 2018 Thanks Brian, I'll try that this evening and post back. Actually now that you mention soldering to C13 - I should ask about that because C13 - was a problem for me as well. It it appeared to me to be connected to one end of the surface mount component labeled C14 adjacent to it and so I jumped them together. Was I correct to do so? If not I have no C13 - connection either lol. I'm glad you explained why this happened (ground plane in the center). I've desoldered quite a few things over the years and have generally had a higher success rate than I had with this one lol. Mark
Brian H Posted March 31, 2018 Posted March 31, 2018 (edited) Minus C13 is soldered to the ground layer in the PCB. If there was a top run from minus C13 to one side of C14. It should be providing the ground end of C14 with the top run. Make sure the jumper is not to the plus side of C13. As C13 and C14 are in parallel on the unregulated 12 volts. Normally 18-19 volts. Jumping minus C13 to the end of C14 top run with a wire. It should be fine. You can use the square pad on the connector mentioned earlier. As the common ground for the possible missing minus connections and a good place to use your meter to verify the minus connections. There was a reversed engineering schematic of the power supply in one of the messages. Of page #12 of this thread. Edited March 31, 2018 by Brian H Find Page Nunmber,
MarkJames Posted April 1, 2018 Posted April 1, 2018 Thanks Brian, I'll give it a go in the morning and hopefully I'll see a green LED again Much appreciated, mark
MarkJames Posted April 2, 2018 Posted April 2, 2018 Sadly no love from my dead EZFlora. Ahh well - it was worth a try. Thanks much for all the help. mark
Recommended Posts