Repair of 2413S PLM When the Power Supply Fails

[Teken]

Insteon has more talent than people realize. Unfortunately in many ways they were shackled due to the lack of focus from leadership. While I too wish they utilized 3rd party more, I understood the hesitation and lack of interest in having other companies churn out insteon products (look at how fractured zwave is). I think this will change in time as the interest of new ownership isn't to keep everything in house. 

 

They really didnt let business slip away. To a certain degree yes but overall No. From the outside looking in, its easy to call someone out on something. However from the inside the story is much much different (look at Trump. Forever criticizing Obama but then finds out what that life is really like once he took office). 

 

As they restructure under new leadership, I think you will see a difference. It may not be what you want personally (they will always go towards their market despite what power users want), but in time you will see a better company. 

 

LOL ~ Always the optimist . . .    

[lilyoyo1]

LOL ~ Always the optimist . . .    

We gotta give the people both sides. We're yin and yang

[Will49]

Sorry for hijacking this topic, but I'm trying to repair an Insteon 2450 low voltage controller. Basically the supply rail that drives the relay is pretty unstable for some reason and doesn't always have enough voltage to pull in the relay. It varies between 13 and about 17 volts. Below 14 or so it won't pull in the relay. I've replaced all of the electrolytic caps on the board and checked the values of several others, as well as swapped the relay for the "senstive" version, but still can't get this solved. The voltage varies from day to day so sometimes it works, and sometimes it doesn't. I haven't traced out the circut that derives the relay supply voltage and haven't been able to find any kind of schematic online.

 

Is anyone familiar with this board (or others that use a relay) and how it derives the relay power supply, and any hints of what else to look for?

 

Thanks in advance.

[Brian H]

See this thread for the 2450.

https://forum.universal-devices.com/topic/22771-iolinc/?hl=2450&do=findComment&comment=222761

 

The original C7 was way too small.

It has a switching power supply. Use one rated for a switching supply not a general purpose one.

A general purpose may have a short life or possibly be unstable.

The power supply is close to 21-22 volts out. Even with the 220uF/50V cap.

Low side of the relay coils voltage  tolerance.

Edited November 10, 2017 by Brian H

[fisix]

Just an update.

 

I went ahead and purchased the selection of caps from Mouser, and installed them in my V1.5 1130 PLM.  The LED hadn't dimmed in a way that I could recognize, but given the general failure rate, I went ahead and replaced them because I'm having problems that sound a lot like PLM issues.  After replacing the caps and Restoring Devices, everything is working well.  I won't know for sure that it was the PLM that was the issue (and therefore needed/benefited from the cap replacement) for at least a few weeks.  If the system is reliable for more than a few weeks, it'll be safe to add my PLM to the list of failed/repaired.

 

I still have the old caps, so I guess I could try to measure their capacitance...

Please add my PLM to the list of PLMs successfully repaired using the replacement caps (mine were from Mouser) identified in this thread.

As noted above, my PLM is a  V1.5 1130 PLM.  I've been running for more than 2 weeks without a problem, after replacing the caps.

 

Thanks everyone for your help and for this thread.

[Brian H]

Good to hear of another successful repair.

[LFMc]

About to replace mine on a 2413U. I previously had a 2413S which I think I already chucked in the bin... Wishing I had saved it now. The 2413U is a v1.B, purchased in 2013, so almost 4 years... Not too bad. Will report back shortly.

 

I also ordered 10 of everything... Probably should have ordered 30 of the 

EKZM500ETD101MHB5D, but oh well...

 

I'm thinking I can kit these up as well and just do US mail ($0.42) for anyone else who might need them to save on the shipping.

 

Hey Bill,

 

Just had my first PLM total failure. It locked up about two months ago and a reset fixed it, but now it is lights out. It is V1.6 and date coded 1214, so I almost got 3 years out of it. (it went into service about 6-9 months after I bought it)

 

If you still have the kits, PM me and let me know what you would take for per set for two or three sets. I'm in Dallas, TX, so figure postage too. 

 

I am not in a hurry as I bought a spare PLM a year ago and got my system back up quick. 

 

Thanks!

Edited November 21, 2017 by LFMc

[johnstonf]

Sounds like maybe a few of us are ready to buy a set, if anyone in the USA would be willing to do it.

I'm in Canada, glad to pay postage.

Fred

[hendryjl]

So, I can't express how valuable this thread has been. It's given me all I needed to know to repair my 2413S, but also gave me the inclination to pop open my smartenit 5010G that had taken off on me about two months earlier and realize it's got the exact same set of capacitors. I'll be replacing all 5 on each board shortly and will get back with "hopefully" my success story to add to the list. I went ahead and ordered a brand new 2413S to get myself up and running more quickly, and to have a spare. I will also report back when that comes in with what revision I get and what caps are on it.

 

So, I know I'm "saying" i will do all this stuff before it's done, but I'm hopeful I can get some advice on loan until I give my repayment... I have the most recent parts list from the page on post 333 and other places, but here's my question... I know these are "much better" components from a quality perspective, and from a design perspective, but are they "the best" ones. I don't mind to take the time to repair these two boards, primarily because I don't want them ending up in a landfill and being replaced by new ones, but I want to do it once, or at least, as few times as possible If we ignore cost for a second, and consider any of the caps on Digikey or Mouser as an option, are there any better choices that might cost a bit more? I'd rather spend a little more to extend the time until I need to set up my soldering station again.

 

I hope that request if coming across right. I just want to put in the components that will have the longest predicted life, as my time is worth spending a bit more on the right component. If the ones already listed meet that criteria, I'm good with that too.

 

Again, thanks to all the contributors on this long lived and informative thread!!

Edited December 12, 2017 by hendryjl

[Brian H]

The Smartenit {Simplehomenet} devices are built on a base 2413 module.  Purchased from Smartlabs. Older ones on the 2412 base module.

 

I have seen reports of hardware revision 2.4 lately. Mine is a 2.3.

 

There are a many reputable capacitor manufacturers carried by both Mouser and DigiKey. I can look up a few if you want.

All the manufacturers mentioned in this thread should be reputable ones,

 

C7 and C13 are the two in the unregulated 12 volt power supply. Normally around 20 volts.

They are the two that should be rated for a switching power supply. Different definitions by between manufacturers. Low ESR or Low Impedance is frequently pointed out. If your C7 and C13 are in older hardware. They was a different value in them. The present ones in my V2.3 is 100uF/50V. Maybe even bigger now, We have used 100uF/50V in them.

 

Temperature rating of 105C or higher, Rated life of 2000 Hours or more. I have seen 5000 or even 6000 hours. Every 10C below the maximum rating exponentially extends their life.

 

You may want to look at their size. I have seen some values with two sizes. One taller with a smaller diameter and the other a shorter one and a larger diameter.

Edited December 12, 2017 by Brian H

[hendryjl]

Temperature rating of 105C or higher, Rated life of 2000 Hours or more. I have seen 5000 or even 6000 hours. Every 10C below the maximum rating exponentially extends their life.

 

You may want to look at their size. I have seen some values with two sizes. One taller with a smaller diameter and the other a shorter one and a larger diameter.

 

Yeah, I'm familiar with this part of things from the reading. That got me thinking... I don't care about aesthetics. both devices are connected in place people won't see them, and one is in a semi-temperature controlled attached garage where it might get a little hotter than ideal. I am fine with some case modification too. I was thinking about mounting a fan like this one (https://www.digikey.com/product-detail/en/sunon-fans/MF25101V2-1000U-A99/259-1833-ND/7805272) to the outside surface of the case, centered the AC connection, exhausting so that it brings air up and over the caps. intake venting would be provided by drilling some holes along the bottom corner edges of the case nearest the caps between the two boards..I would think that the low 360mw draw shouldn't affect the rest of the circuit. I'll probe around a bit to find a good place to tap in. If you don't see any issues with that plan, I think I'll give it a shot. I know that might be overkill, but I would think it would produce a better result than the option some have suggested of just leaving the front of the case off all together as it would force air around the caps.

 

Also, I'd love to take you up on the offer to ID what you'd consider the "premium option" on the caps. It's ok if there are physical constraints of the case that would be violated. Again, I am fine with a little case modding to accommodate.

[Brian H]

Regulated 12 volts maybe a issue.

The output of the power supply is unregulated 12 volts. Closer to about 20 volts.

The regulators are 5 volt logic. One on the main board and one on the serial port daughter board.

The unregulated is on the 8 pin socket and 8 pin header between the main and daughter boards or one of the unused F1 fuse pads on the daughter board.

 

In the older 2412S F1 was installed and the voltage was on the network style serial port connector of the PLM.

Though I suspect its power supply current rating was also more. As it had a power transformer and large filter caps. Not a switcher like the 2413 boards.

Edited December 12, 2017 by Brian H

[Brian H]

Post #282 of this thread. Has a partial schematic on the 2413 main base board. AC input and power supply area.

 

Post #302 has the specification sheet attached for the LNK345GN switching supply chip. If you have troubles finding it.

 

I also found this link in one of the messages here.

It was a 2413U so the daughter board is the USB one not the Serial Port one. Main board should be the same.

https://1projectaweek.com/blog/2017/6/2/repairing-a-failed-insteon-powerlinc-power-line-modem-plm-2413u

Edited December 12, 2017 by Brian H

[hendryjl]

Those F1 fuse pads look like an easy tap point on the serial board. too bad they don't exist on the EZRF_X10. I will make some measurements on it once I have the repair done to see where I've got something useful, then use some resistors to pull down the 20 (or whatever I end up observing) down to the acceptable range of the fan. I'll go with the list of parts in port 282 unless you (or someone else) have any improvements on it.

[Brian H]

C13 is the output side of the Pi filter. So its pads are one option.

[hendryjl]

Any idea what you could safely pull if you tapped the +5V side? I thought about AC fans, but they just don't come small enough, and the unregulated 12V gives me pause... it just seems easier to go 5V. Do you think you could safely pull an added .12A for a 600mW fan off that switching IC?

[Brian H]

I don't know about the safety margin designed in to the switching supply. That outputs around 20 Volts DC.

 

I looked at a 2443, Hardware V2.5 Access Point. That I rebuilt. Same main board no daughter board. Easier to look at.

The 5 volt regulator is on the main board. U3 Surface Mounted 78L05L. 78L05 has a maximum current rating of 100mA.

~ 20VDC input 5VDC output.

 

There is another one on the Serial Daughter board. I would expect the Smartenit model also has a 5 volt logic supply on it. To take the ~ 20VDC and convert  to 5 volt logic.

Edited December 13, 2017 by Brian H

[ehjay]

Another successful repair here.  Strangely, my PLM (usb) has been going strong for many years (5+), but it was my EZIO device that failed.  I had tossed it into the recycle pile when I came across this thread and learned that this device has the same base board as the PLM.  I only replaced C7 and C13 (10uf originally) with 100uf 50V caps.  I measured C7 at about 7uf after I took it out.  Glad to have my sensor back (which was monitoring my alarm panel).  I hope the repair job lasts a lot longer than the original.

Edited December 13, 2017 by ehjay

[Brian H]

Glad you got it working.

If you had an ESR tester.

Good chance C7 and C13 now have high ESR readings. Along with the lower capacity.

[stmrocket]

Repaired a 2413s last March and just finished the same repair on an EZX10RF that had completely died.  Same parts and same repair.  My parents' EZX10RF is dead as well and on the docket for repair over Christmas.

[Brian H]

If the EZX10RF is a older hardware model. It was built on the base power line only 2412 main board.

The capacitors would be different. As it has a power transformer style power supply.

Hope yours is on a base 2413 main board so you can rebuild it.

[stmrocket]

The one I did last night was based on a 2413 so the parts were identical.  My parents' one is newer than mine by a couple years so, hopefully, it will be 2413 bases as well...

[Brian H]

If it is newer than yours. It should be a 2413 base board.

Mine from when they where Simplehomenet. Are on the older 2412 base board.

[hendryjl]

Here's part 1 of my promise... the details on all my units. I listed this over on the PLM version tracking thread.

 

ISY     Unit       PLM1         PLM2      PLM3      Sourced                  Was ALL ON Witnessed

v9E    4x:xx:xx   2413S       2.3           4616      Amazon 12/2017     No, only in service for a couple of days so far

???    2x:xx:xx   2314S       V1.B        1317      Not sure                   No, in 3-4 years of use.

???    2x:xx:xx   2443         V2.B        1304      Not sure                   No, in 3-4 years of use. This is a SmartEnit #5010G Rev 1C 1714

 

I'm going to be repairing the capacitors on the bottom two units once they come in. I am also going to be modifying the cases to add fan based ventilation in the process. I'll send some pics along.

 

-Jim

[hendryjl]

Ok, my OCD is on overdrive trying to pick out these capacitors. I'm in the right ballpark on all of my options, and I'm finding that when I end up between a couple of options, they separate themselves into ones that allow more ripple current versus ones with the lowest impedance/ESR. Which factor is more important, and does that change depending on which capacitor we're talking about (it's position/function in the circuit?)

Edited December 15, 2017 by hendryjl