Insteon Developments

[SmartLabsMike]

Well that is good news. I was told in the SDK [maybe a week ago] that it was not going to be sold to end users. Also that means my Simplehomenet unit is most likely NOT a PLM but is a 2414S with some replacement coreapp

It will ship with a PLM.

 

[BLH]

Well unless the PLM is in a 2414S case and just has a special coreapp in it. That is not what I got.
May not matter as it failed in about an hour. LED went out and is smells like burning transformer. Replacement promptly on its way from Simplehomenet.

 

[Xpendable]

Despite what you may have heard, the PLM is unfortunately not any faster than the PLC.  It's every bit as slow.  I know, because I have worked with 2 different revisions of the PLM.  You actually have to add a 10 ms delay between each byte you send the PLM, or transmission will fail.  It's really annoying as a software developer to have to implement software flow control to only send 1 byte of data every 10 ms, but that's the limitation we are given with the PLM.

The requirement of a 10ms delay between bytes has been eliminated for the released product.

While some may find it hard to believe, we listened to the feedback of our development partners and adapted the code to accept an entire standard and extended message via the serial port.

SJ

Glad to hear that the 10 ms delay between sending bytes to the PLM is no longer required. This was limiting communication to about 800 baud. I'll have to swap out my prototype for a newer version.

 

[BLH]

Well I hope this isn't a new trend.
Two other people on the Simplehomenet Forums have reported the exact problem I had. PLM ran for maybe a half an hour then stoped working. Smelling like burning electronics and no longer functioning.

 

[Oman]

As far as the 10ms delay between bytes being sent to the PLM go - it is worse with the PLC (and the issue is now gone with the PLM). The turnaround time when dealing directly with the PLM is about 50% lower than with the PLC/SDM solution. Total uptime is significantly higher.
Jon

I did comparative stress tests with 2 different PLC's & 2 different PLM's. The PLM was not even the slightest bit faster than the PLC (even with the PLC's added layer of the SDM). You would think that there would be a big difference in performance becaused of the added layer of the SDM, but that's actually not what I saw. Unfortunately, I believe the real bottleneck is either the actual chip inside the PLM & PLC, or the programming of the chip. It's slow enough that the added layer of the SDM really doesn't seem to make any difference. At least that's what I got from my own tests. Hence the reason I decided not to add support for the PLM in LightShowMaster.

I do agree that stability is much improved in the PLM.

I have been doing my testing going from .NET2 through COM Interop to the SDM, then a serial (and one USB) 2414. My PLM testing is native .NET2 Serial support to the PLM. From transmit to ack to transmit on non-hop devices I'm seeing a pretty decent speed improvement. It is possible that other interfaces into the SDM could see different results. The biggest reason for me to use the PLM though is getting rid of SALad and the SDM. The SDM still has some long-term stability issues in some environments and I'm still having the TimerCoreApp just "go bad" even on the newest firmware - although not nearly as often as it used to. The more the system gets simplified the better the uptime gets.

I have confirmation that SmartHome will be making the PLMs available to end-users, if not directly then through several other distributors. I did not want to make a change like this without making absolutely sure that SmartHome would be behind it all the way.

Jon

 

[johnnynine]

For the record the reason I went with Insteon over UPB is primarily due to the UPB delay issue. Cost was a factor as well but if it weren't for the delay, I'd probably have ponied up the extra $ for UPB.

Randy, how significant is the upb delay? I notice a delay in Insteon when operating a switch at the local light that only amounts to a fraction of a second, but it is noticeable. How much more pronounced is the delay with upb?

 

[WayneW]

Randy, how significant is the upb delay?

0.75 second, if I recall correctly. Arguably it is UPB's #1 complaint, some find it totally unacceptable, some deal with it.

 

[Steve]

Yes, the current batch of SAI switches is 750ms. I'm not even going to attempt to say anything other than you need to just try it for yourself. It is a VERY subjective thing.