PLM Repair Service through Insteon

[sdynak]

Didn't see this posted anywhere apologies if it was.

I noticed while browsing products on the Insteon site. I know there are folks on E-Bay offering this but going through Insteon might not be a bad option to consider while we wait on the replacement units to arrive for anyone needing it sooner. 

 

PLM_Repair_Service

[Techman]

@sdynak

Thanks for posting this. It's well hidden on the Insteon site. I looked and couldn't find it.

[sdynak]

6 minutes ago, Techman said:

@sdynak

Thanks for posting this. It's well hidden on the Insteon site. I looked and couldn't find it.

My pleasure.. I found it looking at the Hub Subscription Service (just browsing.. I would never use their hub lol).. it was showing as something I might be interested in at the bottom :). 

[larryllix]

Not for Canada unfortunately. It would likely be cheaper for Insteon to just produce and ship a new one than to pay shipping both ways plus $50 to repair.

Sounds like about the $120 CAD mark by the time it's done.

Sent from my SM-G781W using Tapatalk

[NDSSM]

I lurk around every now and then, this may even be my first post. 

It's been an honor to see all the suggestions for my service on eBay. 
I thank all of you that have suggested my service, and those of you that have purchased my products and services. 
I don't plan on stopping anytime soon despite repairing Insteon gear becoming a full time job in itself. 
If Insteon themselves ends up taking over, I'll just be less busy in the electronics repair world. 
I'm happy to see Insteon succeed, and even happier to be part of making your smart homes continue to function. 

With that said, 
It'll be interesting to see if they continue to use the same, mediocre at best, 1000 hour Fujicon capacitors.
It'll also be interesting to hear back from those that use this service, and how many Insteon deems unrepairable. 

I often find a multitude of failed components in these devices...
I'd say a solid 75% of the time there are other component failures especially when communication issues arise,
thus why I offer various services. 

$50 is cheap, especially after paying someone to rebuild them, paying shipping, taxes, etc.  
The RJ45 connectors alone are near $5 each, even in bulk.  
I dont see any warranty's on the work either. 

Thanks again everyone! You know where to find me :-) 

-Chris @ NDSSM (Nilachi Data Systems / Sydney Micro) 

[Brian H]

Hardware revision 2.3 changed the caps to a Fujicon TY type and is designed for switching supplies and what looks like a slightly better time rating. A MAX232EI serial port chip. With better ESD ratings. Though sticking with Fujicon may not be the best choice.

Revision 2.4 and above. Also has a new serial port daughter board. That also has a signal protection network on the two serial signals to and from the PLM.

I actually have a V1.0 with 10uF/25V caps for C7 and C13  and reworked to add the Pi filter components. That they look like they didn't put into the original design. I have seen many different caps for C7 and C13. As they unsuccessfully tried to fix the power supply issues.

I too have seen network jack connections issues reported.

Edited December 7, 2022 by Brian H
Fixed statement

[oskrypuch]

@NDSSM And, will you ship (back) to Canada?

* Orest

[mango]

@oskrypuch to be cost effective for Canada I recommend: https://www.kijiji.ca/v-general-electronics/mississauga-peel-region/insteon-repair-service/1629894953

He's fixed up quite a few of my devices.

@NDSSM not wanting to take anything away from your US services I hope okay.

[larryllix]

1 hour ago, mango said:

@oskrypuch to be cost effective for Canada I recommend: https://www.kijiji.ca/v-general-electronics/mississauga-peel-region/insteon-repair-service/1629894953

He's fixed up quite a few of my devices.

@NDSSM not wanting to take anything away from your US services I hope okay.

 

Since a good portion of us live close to this repair guy, we could ship to, and then drive over to pick it up again. Would be several days faster  and also save a shipping fee. Square One would be a side trek making it more worth while. Tell the wifely you are taking her shopping.

Maybe the guy would try same day service with a shopping mall gap in the timing??

Edited December 7, 2022 by larryllix

[asbril]

8 minutes ago, larryllix said:

 

Since a good portion of us live close to this repair guy, we could ship to, and then drive over to pick it up again. Would be several days faster  and also save a shipping fee. Square One would be a side trek making it more worth while. Tell the wifely you are taking her shopping.

Maybe the guy would try same day service with a shopping mall gap in the timing??

Expecting that your wifely will not outspend the saved shipping fee in the shopping mall ?????   

[larryllix]

Just now, asbril said:

Expecting that your wifely will not outspend the saved shipping fee in the shopping mall ?????   

Oh Jeeezzzz! I never thought of that!  I love the support in this group.

[NDSSM]

8 hours ago, Brian H said:

Hardware revision 2.3 changed the caps to a Fujicon TY type and is designed for switching supplies and what looks like a slightly better time rating. A MAX232EI serial port chip. With better ESD ratings. Though sticking with Fujicon may not be the best choice.

Revision 2.4 and above. Also has a new serial port daughter board. That also has a signal protection network on the two serial signals to and from the PLM.

I actually have a V1.0 with 10uF/25V caps for C7 and C13  and reworked to add the Pi filter components. That they look like they didn't put into the original design. I have seen many different caps for C7 and C13. As they unsuccessfully tried to fix the power supply issues.

I too have seen network jack connections issues reported.

The Fujicon TY's aren't fairing very well either. They are only rated for 1000 hours at their peak temp (previous Fujicons were 500 hours iirc), and I find that number to be a bit bloated. Most I yank out test significantly out of spec.   I bench tested these as well as other popular caps being used in other devices... the results were sad. I did a test run with cheap ChongX caps for the heck of it, and the PLM's lasted 3 months. ChongX is another one that promises "high reliability" and has gained popularity in factories... yet I've lost count how many I've seen dead, including dead shorts taking out transformers in 5 and 12v power adapters. 

It's surprising, but I've repaired more of the revised daughter boards than any other, including fried memory chips and zapped serial controllers. I've never seen an official document regarding "better ESD protection" (nor am I doubting you), but on my end, its as if ESD protection ended up being worse. My assumption had been this change was made to allow the use of non-electrolytic caps. 

The unnoticed elephant is C8.... Focus was never put on C8... these are almost always 50% out of spec and showing very high resistance. This resistant puts stress on the voltage regulator (U3), and that stress ends up putting more stress on the 20v rail. I've replaced a ton of regulators because of failed C8's. The trace design under C8 prior to is position being revised was also troublesome, if the trace becomes too hot, I've seen the trace burn and fuse to the grounding plane. 

Many seem to ignore the electrolytic on the daughter boards too. Dont.  

There also issues with the communication side of the circuitry that even the last revisions did not address. I have my own design and fix for these issues.  

When I look at the changes over the years on these PLM's... it makes me shake my head.  Many of the revisions made could have been avoided by just using quality components elsewhere from start. 

 

[larryllix]

8 minutes ago, NDSSM said:

The Fujicon TY's aren't fairing very well either. They are only rated for 1000 hours at their peak temp (previous Fujicons were 500 hours iirc), and I find that number to be a bit bloated. Most I yank out test significantly out of spec.   I bench tested these as well as other popular caps being used in other devices... the results were sad. I did a test run with cheap ChongX caps for the heck of it, and the PLM's lasted 3 months. ChongX is another one that promises "high reliability" and has gained popularity in factories... yet I've lost count how many I've seen dead, including dead shorts taking out transformers in 5 and 12v power adapters. 

It's surprising, but I've repaired more of the revised daughter boards than any other, including fried memory chips and zapped serial controllers. I've never seen an official document regarding "better ESD protection" (nor am I doubting you), but on my end, its as if ESD protection ended up being worse. My assumption had been this change was made to allow the use of non-electrolytic caps. 

The unnoticed elephant is C8.... Focus was never put on C8... these are almost always 50% out of spec and showing very high resistance. This resistant puts stress on the voltage regulator (U3), and that stress ends up putting more stress on the 20v rail. I've replaced a ton of regulators because of failed C8's. The trace design under C8 prior to is position being revised was also troublesome, if the trace becomes too hot, I've seen the trace burn and fuse to the grounding plane. 

Many seem to ignore the electrolytic on the daughter boards too. Dont.  

There also issues with the communication side of the circuitry that even the last revisions did not address. I have my own design and fix for these issues.  

When I look at the changes over the years on these PLM's... it makes me shake my head.  Many of the revisions made could have been avoided by just using quality components elsewhere from start. 

 

Sounds like a barrel connector should be installed and use external power supply adapters to feed it that can be replaced every few years and protect/isolate the grid power from the electronics.

Oh yeah, the grid signals would require some passthrough...hmmmmm.

[NDSSM]

3 hours ago, oskrypuch said:

@NDSSM And, will you ship (back) to Canada?

* Orest

Yes

[NDSSM]

27 minutes ago, larryllix said:

Sounds like a barrel connector should be installed and use external power supply adapters to feed it that can be replaced every few years and protect/isolate the grid power from the electronics.

Oh yeah, the grid signals would require some passthrough...hmmmmm.

I was actually working on splitting the 5v section to a separate power device, but the idea got shelved. Havent had the time. 

[larryllix]

11 hours ago, NDSSM said:

The Fujicon TY's aren't fairing very well either. They are only rated for 1000 hours at their peak temp (previous Fujicons were 500 hours iirc), and I find that number to be a bit bloated. Most I yank out test significantly out of spec.   I bench tested these as well as other popular caps being used in other devices... the results were sad. I did a test run with cheap ChongX caps for the heck of it, and the PLM's lasted 3 months. ChongX is another one that promises "high reliability" and has gained popularity in factories... yet I've lost count how many I've seen dead, including dead shorts taking out transformers in 5 and 12v power adapters. 

It's surprising, but I've repaired more of the revised daughter boards than any other, including fried memory chips and zapped serial controllers. I've never seen an official document regarding "better ESD protection" (nor am I doubting you), but on my end, its as if ESD protection ended up being worse. My assumption had been this change was made to allow the use of non-electrolytic caps. 

The unnoticed elephant is C8.... Focus was never put on C8... these are almost always 50% out of spec and showing very high resistance. This resistant puts stress on the voltage regulator (U3), and that stress ends up putting more stress on the 20v rail. I've replaced a ton of regulators because of failed C8's. The trace design under C8 prior to is position being revised was also troublesome, if the trace becomes too hot, I've seen the trace burn and fuse to the grounding plane. 

Many seem to ignore the electrolytic on the daughter boards too. Dont.  

There also issues with the communication side of the circuitry that even the last revisions did not address. I have my own design and fix for these issues.  

When I look at the changes over the years on these PLM's... it makes me shake my head.  Many of the revisions made could have been avoided by just using quality components elsewhere from start. 

 

Sound slike the classic non-technical management micro-managing things. When things get tough they make it cheaper and eventually the product is crap. Wrong path.

[apostolakisl]

18 hours ago, larryllix said:

Sounds like a barrel connector should be installed and use external power supply adapters to feed it that can be replaced every few years and protect/isolate the grid power from the electronics.

Oh yeah, the grid signals would require some passthrough...hmmmmm.

Or maybe a breakout board with all the caps that fail on it.  Mount it outside the device and then you can just unplug it and plug a new one in every couple years.

[Brian H]

The Fujucon TY sheet I have. Says 2000 to 7000 Hr load life at 105C. Still not the most  reputable cap in my opinion.

Your finding other bad components. Would explain why some still had problems after a capacitor replacement. While others started working again.

Edited December 9, 2022 by Brian H
Add a statement

[PinchRoller]

I have attached the spec sheet for factory-installed Fujicon TY Series caps discussed in this post.  The 100uF/50V is used at C7 and C11 since 2016.  When Hubs and PLM are repaired by Insteon C7, C13, C11 use Rubycon 50YXJ100MT78X11.5 100uF/50V, 7,000Hrs @ 105°C

Capacitor - Electrolytic Fujicon TY Series High Ripple Current, High Reliability (6.3-50VDC) Specs for [100uF-50v, TY1H101M-RBF11WP00]; 560mA Ripple at 100kHz, 6K Hour life at D8.5 x 11.pdf

Edited December 17, 2022 by PinchRoller
added more information

[Brian H]

I actually have a revision 1.0. Has (2) 10uF/25VDC caps in it. Along with the rework to add the second capacitor and inductive coil. For a Pi output filter. Original design looks like they originally tried just the one 10uF/25VDC. I have seen many different capacitor sizes used to try for a fix.

I also used a well known capacitor brand to fix my 2413S and V2 Access Points using the same main board.

V2.4 has the improved serial board in it. With a chip with a better ESD rating and an added network to clamp the signals to a safe level.

[NDSSM]

Even Fujicon spec sheets make me cringe.  
 

4 hours ago, PinchRoller said:

I have attached the spec sheet for factory-installed Fujicon TY Series caps discussed in this post.  The 100uF/50V is used at C7 and C11 since 2016.  When Hubs and PLM are repaired by Insteon C7, C13, C11 use Rubycon 50YXJ100MT78X11.5 100uF/50V, 7,000Hrs @ 105°C

Capacitor - Electrolytic Fujicon TY Series High Ripple Current, High Reliability (6.3-50VDC) Specs for [100uF-50v, TY1H101M-RBF11WP00]; 560mA Ripple at 100kHz, 6K Hour life at D8.5 x 11.pdf 218.28 kB · 1 download

 

[NDSSM]

The 100uf/50v is the  Fujicon ty1h101m-rbf11wp00 and is indeed rated for *5000 hours.
I dont have much faith in their rating, or their formula.  I havent found a Fujicon calculator like other
manufacturers have to get a realistic life for application expectancy either.

And again, its too bad no attention was paid to all of the capacitors. The C8 as I referenced above is almost always out of spec (RM series, 1000 hour life @ load...which it often is at), and then there is the lovely, cheap, "CI" branded C3, which should have been rated at least 100 volts higher...if not 200, and really the only electrolytic cap on the board that needed a voltage upgrade. 
 

[Brian H]

All the capacitors where covered in another extremely long thread here. Some where discontinued and others found. As parts supplies dried up especially during the hight of the pandemic.

https://forum.universal-devices.com/topic/13866-repair-of-2413s-plm-when-the-power-supply-fails/

Edited December 18, 2022 by Brian H

[NDSSM]

On 12/18/2022 at 6:22 AM, Brian H said:

All the capacitors where covered in another extremely long thread here. Some where discontinued and others found. As parts supplies dried up especially during the hight of the pandemic.

https://forum.universal-devices.com/topic/13866-repair-of-2413s-plm-when-the-power-supply-fails/

To be more clear, my comment is directed to Insteon themselves, not the great people of UD or other forums. 
From revision 1 to 2.6   C3, C11, and C8 have gone ignored among other components, by Insteon. 

[Brian H]

Hopefully the new management. Will give the whole 2413S/U a good look at. Parts wise. Over the previous owners band aid tries to fix it. With different sized capacitors all of dubious quality.