Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[Boiler]

I recently became interested in Firmware revisions when I noticed that my two CM15a units responded differently to Identical macros downloaded from AHP.  My "old reliable" unit would not trigger a macro when the transceived input was disabled.  When I plugged my newer CM15a in,  and downloaded the identical file,  everything seemed to work.  I chalked it up to a bad interface and went along fat, dumb, and happy.

The problem is, everything else hasn't worked.  The newer interface simply acts differently (I have a couple of posts documenting some of these differences).  Some of the fixes that I had in place for the old interface are no longer required.  Bottom line is, my AHX file had not changed since February, plugging in a newer CM15a fixed one problem but caused several others (I'm still working through these). 

I now believe that my old interface is fine and that it's firmware differences that broke my system when I swapped interfaces.  I am currently running the old unit with my the AHX file from Feb of this year (yes, I keep backups on my server).

I've never seen a Firmware revision history posted anywhere (This may be a closely guarded secret).  I thought it might be useful to document the firmware revisions in use by the forum members.  If we can collect enough revision history, and observations, we may be able to clear up some of the mysterious behavior that people run into.

Device   Model   Date Code   Firmware   Notes
Transeiver Module   TM751   01H35   P10485B   Date code and Firmware Version on back of device
Transeiver Module   RR501   02D16   P10283E   Date code and Firmware Version on back of device
Controller   CM15a   04I38   P10792E   Date code on back of Device - Firmware on Controller IC
Controller   CM15a   06L50   P10792M   Date code on back of Device - Firmware on Controller IC

What I believe to be the Firmware revision (P numbers above) is easily located on the back of most of my plug-in devices.  X10 pulled a fast one on the CM15a - the "P#" is attached to the controller IC inside the unit (See photos).  The top of the unit (where the batteries are located) is rather easy to remove after taking out the 4 screws on the backside.  There is plenty of wire length to prevent yanking the antenna and battery connections off the motherboard, but be careful none the less.

Newbies - While I would appreciate information on the latest revision levels, understand that removing the cover will void the warranty and could cause ESD damage (electrostatic discharge damage)  to the controller.  Don't do this if you're unfamiliar with ESD device handling procedures.

Back to the table above - If I'm interpreting things correctly, my "old" CM15a was manufactured in '04 (about right) and has a Firmware revision "E".  The new controller was made in '06 with a Firmware revision "M".  That's a lot of revisions! If we can collect enough information, maybe we can figure out what they all mean.

TTA: Sorry about the horrible formatting - I can't seem to get the tables to work.

Firmware Revision History By Module

Device           |_	Model |_	Date Code |_	Firmware |_	Notes
Transeiver Module	TM751	01H35	P10485B	Date code and Firmware Version on back of device
Transeiver Module	RR501	02D16	P10283E	Date code and Firmware Version on back of device
Controller	CM15A	04I38	P10792E	Date code on back of Device - Firmware on Controller IC
Controller	CM15A	06L50	P10792M	Date code on back of Device - Firmware on Controller IC

[TTA Edit: Formatted table.]

[Brian H]

I will give my CM15As a look. I did notice in the photo that one of yours had the controller chip in a socket. I believe one or more of mine also are in a socket.
I believe the base number starting with the P is a part number the revision is the letter suffix.

04J41 Soldered in P10792E
04J44 Soldered in P10792F
05C10 Socketed   P10792M
05C18 Socketed   P10792M
Both 05 date code units also have a different RF Receiver as those two do not have a large tunable coil on them that both 04 units do.

[Boiler]

Brian,

I can confirm that my '06 unit does not have the tunable coil either.  It's interesting that your '04 units are soldered.  Both of mine are socketed.  The socket obviously adds cost (and reduces reliability).  I wonder why they switched back.

I've been poking around the Cypress site reading up on the controller (CYC63723x-PSC).  I found the following "errata" that refers to a problem with the two GPIO inputs and false interrupts.

http://download.cypress.com/publishedcontent/publish/design_resources/errata_update/contents/encore_cy7c637xx_errata_rev_____silicon_rev_a__13.pdf

From what I can tell from the schematic, one of the inputs is used for Rf and the second is used for the powerline crossing detector.  The errata refers to "individual software application fixes".  This could explain a bit about the quirkiness of the CM15a.  It'd bee a H@$$ of a note if the "software" problems we've been complaining about were due to a controller hardware problem.

On a positive note, Cypress has a newer version of the chip with Flash memory (how nice would it be to be able to download new firmware?).  Unfortunately the pin-out has changed as has the memory and stack allocation.  Firmware would be a re-write.  The emulators for the new chip seem to be reasonably priced - makes a person stop and think.

http://download.cypress.com.edgesuite.net/design_resources/datasheets/contents/cy7c638xx_8.pdf

[Brian H]

I have seen posts on powerline noise in some cases messing up the CM15A. My 04 and 05 units do not have bypass caps where normally would be used. Can't say about 06 and 07 units.
Post indicated adding from VCC to VDD on U1 and U2 along with across C11 and C12 helps in the crashes.
Socket on the U1 chip maybe was for easier firmware updates as it could just be exchanged if a problem was addressed.  My 04J41 to 04J44 changed from E to F. Both soldered so harder to do a production upgrade if already soldered.

[Walt2]

For my CM15A...

• There is a round sticker on the back that reads "04B09"
• There is a hand-written note inside the battery compartment that reads "Rev May 21"
• Inside there is no sticker on the CY7C63723-PC, and that device is in a socket

What version firmware do I have?   

[Boiler]

For my CM15A...

• There is a round sticker on the back that reads "04B09"
• There is a hand-written note inside the battery compartment that reads "Rev May 21"
• Inside there is no sticker on the CY7C63723-PC, and that device is in a socket

What version firmware do I have?   

Sorry Walt - your beta unit doesn't conform the the "standard X10" firmware convention.  From what I can tell the "04B09" indicates:

• 04 - year of manufacture
• 09 - week of manufacture
• B - month of manufacture (updated 9/8/07)

These assignments appear to jive with the purchase dates of my various components.

Are you still actively using this unit?  Is it reliable (particularly with RF)? 

I caused myself a lot of grief a while ago when I accidentally swapped my two units - I had been using the "old" unit.  All kinds of strange things going on (similar to some our our newbie posts) until I recognized what I had done.  I have since started morphing my installation over to the new interface - it seems a lot more robust (particularly with RF).



 {Boiler - updated date code year/month/week decoding}

[Walt2]

Sorry Walt - your beta unit doesn't conform the the "standard X10" firmware convention.  From what I can tell the "04B09" indicates:

• 04 - year of manufacture
• 09 - week of manufacture
• B - no clue (internal batch #??)

These assignments appear to jive with the purchase dates of my various components.

I don't know for sure, but I always thought the letter was the month.  So, "04B09" would be Feb 09, 2004. 

I have many other X10 items with the two-digit-number, letter, two-digit-number sticker.

Are you still actively using this unit?  Is it reliable (particularly with RF)? 

I caused myself a lot of grief a while ago when I accidentally swapped my two units - I had been using the "old" unit.  All kinds of strange things going on (similar to some our our newbie posts) until I recognized what I had done.  I have since started morphing my installation over to the new interface - it seems a lot more robust (particularly with RF).

Yea, its RF really stinks.  That is why I have it completely turned off for all house codes.  Or I thought I did, until I read Pucks postings over in another thread.   However, it does seem to be actually fully off.  I mean, I don't seem to be getting any duplicate RF commands showing up in the Activity Monitor.

[KDR]

I have 2 CM15A's

Unit one date code 06A01, Chip is soldered in, Firmware P10792M - AHP shows receiving RF and triggers a Macro but no power line signal unless the house code is checked to transceived.

Unit two date code 06C11, Chip is soldered in. Firmware P10792M - AHP shows receiving RF and triggers a Macro but no power line signal unless the house code is checked to transceived.

^{----------------KDR}

[Boiler]

I don't know for sure, but I always thought the letter was the month.  So, "04B09" would be Feb 09, 2004. 

I have many other X10 items with the two-digit-number, letter, two-digit-number sticker.

Doesn't work with some of my date codes.  My 2004 controller is "04I38".  That would be Sept 38, 2004.

OK, I found a "week of the year" calculator.

http://www.onlineconversion.com/day_week_number.htm

"04B09" looks like the 9th week of Feb '04. 

Yea, its RF really stinks.  That is why I have it completely turned off for all house codes.  Or I thought I did, until I read Pucks postings over in another thread.   However, it does seem to be actually fully off.  I mean, I don't seem to be getting any duplicate RF commands showing up in the Activity Monitor.

This agrees with my '04 unit.  Easily "confused" by RF and when I disable a transceived house code it shuts down everything.  No power line activity or macro response.

[Boiler]

I have 2 CM15A's

Unit one date code 06A01, Chip is soldered in, Firmware P10792M - AHP shows receiving RF and triggers a Macro but no power line signal unless the house code is checked to transceived.

Unit two date code 06C11, Chip is soldered in. Firmware P10792M - AHP shows receiving RF and triggers a Macro but no power line signal unless the house code is checked to transceived.
^{----------------KDR}

KDR,
I had missed your response initially.  Thank you for providing the information on your units.  We (myself and one of the CO's) hope to put together a data base of firmware/date codes to try to correlate some of the differences in the revisions.

I also noticed that both of your controllers were soldered (as were Brian H's units from '05).  I find it very curious that my 12/06 unit is socketed, yet has the same firmware revision. 

The only thing I can figure is that X10 was anticipating a firmware revision (that I just missed) and socketed units to allow re-fitting of hardware already in flow.  I have a new '07 CM15a that I'm in the process of checking out - I'm dying to look inside for the Rev on this unit.

[Boiler]

Finally got around to opening up my "new" CM15a

And...(drum roll)...Identical to my '06 unit.  What a let down.

No hardware changes that I can see.  Same firmware.

Date Code: 07A04, Firmware: P10792M

I guess the good news is we should need to worry about firmware updates causing new problems.

[Boiler]

New VS Old WS467 Comparison Table -

Took the liberty of reformatting Charles Sullivan's white paper on the New WS467 into a table format.  Charles' original paper can be viewed from this thread:

Thread Link:The "NEW" LM465/WS467 (Early 2007) Don't Work Like the "OLD"

This "new unit" continues to be a source of aggravation for new and old users since it will not operate properly with the standard WS467 interface in ActiveHomePro.

Currently, the easiest method of interfacing this switch is to re-identify it as a LM14a (AHP will then use extended code direct dim commands to operate the unit).
The use of alternate AHP interfaces is discussed in the thread above as well.

As noted in Charles' paper (and the tables below), there are a couple of items that are viewed as firmware "bugs"

My purpose for posting here is to track the status of these units as time progresses - I.E. will X10 Fix these bugs and re-release without our knowledge?

The earliest reported Date Code for these New Units is 07A05 (keyboardknome) and after (small round label on switch).
If anyone receives a "new model" WS467 that does not operate per the table below - Please reply (PM or post) with a description of the difference and the unit Date Code.

If I've erred or omitted something in the tables below, please PM and I will update.

Boiler

[KDR]

Very nice comparison layout Boiler. Very Helpful data.

^{----------------KDR}

[Puck]

Thanks Boiler and Charles, those are very useful charts and info on the new lamp modules. I just checked and one of my spare lamp modules is an "07A05" model, so I'll have to play with it now and check out the different features.

P.S. Looking at this thread again reminded me that I still need to look at my CM15A versions.

[-Bill- (of wgjohns.com)]

Yes, thanks Boiler and Charles, very helpful indeed!