Print

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

[JeffVolp]

I think we are talking apples and oranges here.  There are two different "clocks" in the CM15A.  One is the internal hardware clock in the microcontroller that steps through the firmware.  The other is the firmware "real-time" clock that determines the time of day needed by the various timers.

The Cypress data sheet says that hardware clock is adjusted from the USB port.  In that case they are "tuning" the hardware clock so it runs within 1.5% of its design center frequency.  That hardware clock is the reference for the firmware real-time clock when the CM15A is not powered from 60Hz.  What isn't certain is whether that hardware clock is still used as the reference for the real-time clock even when the unit is powered from 60Hz.  If that is the case, it would explain some of the clock drift problems reported.

When the CM15A is attached to a computer running ActiveHome Pro, the CM15A firmware real-time clock will be periodically updated to remain in sync with the PC real-time clock.

Jeff

[Noam]

I think we are talking apples and oranges here.  There are two different "clocks" in the CM15A.  One is the internal hardware clock in the microcontroller that steps through the firmware.  The other is the firmware "real-time" clock that determines the time of day needed by the various timers.

The Cypress data sheet says that hardware clock is adjusted from the USB port.  In that case they are "tuning" the hardware clock so it runs within 1.5% of its design center frequency.  That hardware clock is the reference for the firmware real-time clock when the CM15A is not powered from 60Hz.  What isn't certain is whether that hardware clock is still used as the reference for the real-time clock even when the unit is powered from 60Hz.  If that is the case, it would explain some of the clock drift problems reported.

When the CM15A is attached to a computer running ActiveHome Pro, the CM15A firmware real-time clock will be periodically updated to remain in sync with the PC real-time clock.

Jeff

Jeff -
When are YOU going to build a replacement for the CM15A, one that actually is designed properly, and works like it should? ;-) Based on my wonderful experience with the XTB-IIR, I can only imagine what you'd come up with!

[JeffVolp]

When are YOU going to build a replacement for the CM15A...

I'm afraid that will never happen.  I'm not interested in the controller end because there are just too many cans of worms to deal with.  There is a lot of good high-end software out there that is PC based, and uses the CM11A for a powerline interface.  Since that unit is long discontinued, and is becoming difficult to find, I may be offering an alternative for that market in the future.

Jeff

[Dave4720]

Since that unit is long discontinued, and is becoming difficult to find, I may be offering an alternative for that market in the future.

Where do I sent my pre-order check?

[JeffVolp]

Where do I sent my pre-order check?

Thank you, but we are a long way from that.  I do my development during the slow summer months when it is too hot here to do much outside.

Jeff

[Noam]

When are YOU going to build a replacement for the CM15A...

I'm afraid that will never happen.  ...
Jeff

I'm sorry to hear that, but I really was only kidding.

Since that unit is long discontinued, and is becoming difficult to find, I may be offering an alternative for that market in the future.

Jeff

What about a "hybrid" type unit, one that combines the best of both worlds. The stability of the CM11A, but with USB support (maybe BOTH USB and Serial - perhaps Ethernet, too?) and RF? It is getting hard to find PCs with a serial port these days, and for many people it is easier to connect over CAT5, especially if there is a good distance between the PC and the controller.

I'll keep an eye out for future updates from you on the subject....

[Knightrider]

Noam hit the nail on the head.  I've been considering asking for a unit that works on cat5 myself.  Maybe even a repeater that outputs on cat 5 to link buildings on different transformers.

Alas, the market would be small, and I'm still waiting for the last project shelved by JVDE due to the lack of interest.

[JeffVolp]

What about a "hybrid" type unit, one that combines the best of both worlds. The stability of the CM11A, but with USB support (maybe BOTH USB and Serial - perhaps Ethernet, too?) and RF? It is getting hard to find PCs with a serial port these days, and for many people it is easier to connect over CAT5, especially if there is a good distance between the PC and the controller.

I started looking at supporting the CM11A "real time" protocol because a number of my customers are running high-end automation software on their PC, but still use the CM11A as their powerline interface.  There have also been people looking for a standard serial interface rather than the proprietary TW523 interface.  I have been thinking about one unit to address both of these issues.

Microchip has released a new chip that provides all the resources I now use, but also includes an asynchronous serial port.  That provides the pathway to this new product.

I am not interested in breaking new ground, only offering better ways of supporting what is already out there.  So, USB is not in the cards – at least not in the foreseeable future.  And, WGL already has the RF solution.

Jeff

[JeffVolp]

I've been considering asking for a unit that works on cat5 myself.  Maybe even a repeater that outputs on cat 5 to link buildings on different transformers.

The TW523 protocol is best run through flat telephone cable.  However, it is also possible to use CAT5 if you dedicate a twisted pair to each of the 4 signals.  The reason is to keep the distributed capacitance down.

I had looked at the possibility of interconnecting two XTB-IIRs like can be done with the ACT CR234.  The digital ports of two or more units can be wired together to transfer the commands between them.  While the firmware will do the job as it is now, the second unit would only transmit the repeated half of each doublet.  Should there be demand for such a product, it is certainly possible to modify the firmware for more optimal use in a large interconnected installation.

Jeff

[Noam]

The CM15A was introduced back in 2004. Other than perhaps some minor firmware changes along the way (none ever confirmed), X10 hasn't changed the product at all in that time, and I haven't seen any indication of plans to introduce a fancier replacement for it any time soon.

[JeffVolp]

The CM15A was introduced back in 2004.

Yes, that is one of the reasons why I bought the Ocelot back in 03.  I had been using a beta CM14A, but that was never released.  I would have opted for the CM15A if it had been available when we built this house.

The last time I checked, some of the PC-based automation systems were still not supporting the CM15A because X10 had only released limited information on that unit.  I believe they all support the CM11A.

Jeff

[Brian H]

Yes there where some firmware changes. One I found was how RF signals from a HR12A Palm Pad where handled.
My earliest is P10792E. Date Code:04J41
P10792F Date Code:044J44

Latest I have P10792M. Date Codes: 05C10; 05D18 and 08B09
Same basic programmed part number. Just a different suffix.

[dave w]

The last time I checked, some of the PC-based automation systems were still not supporting the CM15A because X10 had only released limited information on that unit.  I believe they all support the CM11A.

FWIW I know "Homeseer2" for PC, and "Xtensions" for the MAC both support CM15A (neither supports downloading to the CM15A).

[pconroy]

The last time I checked, some of the PC-based automation systems were still not supporting the CM15A because X10 had only released limited information on that unit.  I believe they all support the CM11A.

FWIW I know "Homeseer2" for PC, and "Xtensions" for the MAC both support CM15A (neither supports downloading to the CM15A).

that reminds me to check their blog and see if they've responded to my SDK question.
...

<time passes>

...

Nope.