Some tests done on a problem 2413S PLM

[Brian H]

Had a chance to do some inspecting of my 2413S Hw 1.5, Fw v.98.

Removed the PCB late last week and didn't see anything obvious.

Not the cleanest. The soldering was OK, I guess. As I know the new no lead, ROHS solder never looks shiny. Like the old fashioned lead stuff does.

 

I noted the manufacturer, case size and value of the electrolytic caps in the board.

 

Reassembled it and left it powered up in an unused outlet.

Last night the LED was Off. Let it sit unplugged for an hour and then the LED was dim when powered up. The communications test made the LED flash.

This morning after cooling down. The +12 volt DC was 8 volts DC and slowly ramped up to about 16.8 volts DC. I believe it is closer to 19.5 volts DC

in my older 2413S Hw 1.0, Fw. v.92

I suspect the electrolytic capacitors or a voltage reference used for feedback into the LNK354GN power supply IC.

I have never seen the brand name electrolytic caps before. I am not sure if they are even in the Known Bad Caps List on the badcaps.net web site.

[gatchel]

I'm sure the caps are not Mil-Spec. Keep us posted.

[Brian H]

Well I am sure the one with the YIHCON Type RK is not.

 

Been letting it powered up and the +12 on the unused F1 fuse pad is +9.92 Volts DC when I just checked it.

I was going to jump F1's pads so I could read it on the serial connector. Just like the 2412S had the power on the serial jack.

[apostolakisl]

I'm just going to wait for you to tell me which one(s) to replace.

 

[arw01]

+1, I lost two older style keypads in the last 60 days. So that brings my failure on insteon to 8-9 over 3-4 years, pretty much anything I bought used has gone bad. Replacing that much stuff has taken my payback period on energy savings out a couple extra years, but I figure if I can get 4-5 years I'm even from here and then it's all gravy.

 

Just buying on the 20% off sales, if any of those fail outside of warranty I'll mail them to you.

 

Alan

[Brian H]

I am going to try to safely run it out of the case. So I can test the circuit on the bottom of the PCB.

If there is a big FLASH then I guess I slipped on the AC side of the supply.

[Brian H]

I finally got to changing the capacitors in the low voltage side of the PLM's power supply.

I could not get the one 6.8uf 250 V DC for the Line Side DC input to the switcher.

The raw 12 volts is now around 21.4 V DC and seem steady.

Going to let it run for a week or two and observe it from time to time.

If it is OK. I will post more on what I did.

[larryllix]

I finally got to changing the capacitors in the low voltage side of the PLM's power supply.

I could not get the one 6.8uf 250 V DC for the Line Side DC input to the switcher.

The raw 12 volts is now around 21.4 V DC and seem steady.

Going to let it run for a week or two and observe it from time to time.

If it is OK. I will post more on what I did.

What was the voltage rating on the cap running at 21.4 Vdc you replaced?

 

Put some electrical tape over that 120Vac side. Better safe than sorry.

[Brian H]

This is for the later revision PCB with all the components on the board.

 

Earlier version had rework to add the extra cap and coil in the PI filter

One of the caps and coil where hanging off the board with a Blue Fly-wire to the back of the board.

 

The unregulated +12 Volts DC has a PI filter.

C7 and C13. Both where Samcon 10uf/35V. My vendor didn't have a 35 volt so I used a Nichicon PM 10uf/50V.

 

C8 was a Fujicon RM 10uf/16V. RM is the extra small case style.

C11 was a Fujicon RK 100uf/25V

C3 is a YIHCON RK 6.8uf/250V and is the one I did not change. As none of the vendors I deal with had them in stock.

 

Since the PLM in question has run for over a week before getting flaky. I an not saying quite yet that the caps where the problem.

 

Though I suspect C7 and C13 being Low ESR and are directly connected to the diode on the secondary of the transformer. Driven by a 200kHz frequency.

Where at least part of the problem.

 

I may solder a small wire between the F1 pads on the Daughter Card so I can measure the raw +12 volts on the Serial Jack of the PLM and keep the cover screwed on.