Aprilaire 8870 Power

[IVB]

Thx for the suggestion. I have a distro panel, paid the $$, so I might as well use it. I'm supposed to be getting an Elk serial expander today to go via it into the PC for other reasons, so i'll try that.

it will appear to seat and then partially pull away from the socket board.

That has been my problem MANY times, damn it's irritating cuz it's not like those wires are easy to see.

 

[pgray007]

I recently connected my 8870 to 8811 to ELK and got everything worked once I corrected some wiring issues on the 8870. Initially I did not include the wiring from furnace "C" to 8870 "C" and 8870 "R" to 8870 "RC" ? These are shown as a dashed line on page 16 of the installation manual. Once I added those connections everything worked. Have you included those wires?

Jumpers are all good, since teh 8870's work fine as standalone thermos. I am going to bite the bullet and get a distro panel. It does seem to be some kind of termination issue since the lights indicate something should be happening. I'll report back.

 

[IVB]

Thx for the suggestion. I have a distro panel, paid the $$, so I might as well use it. I'm supposed to be getting an Elk serial expander today to go via it into the PC for other reasons, so i'll try that.

it will appear to seat and then partially pull away from the socket board.

That has been my problem MANY times, damn it's irritating cuz it's not like those wires are easy to see.

Well, it wasn't the PC or serial port, as I just plugged in the M1XSP and put the aprilaire into that and it still no workie.

I swear i've checked the other terminations 3x, but that was months ago so i'll check them out again.

 

[pgray007]

Does the 8818 panel need power to function? I ask because I already have power to the thermos, so it would be nice to add the 8818 in without having to rewire anything.

IVB, are you using the 8818? Are we the only two people in the world that are unable to figure out how to make the 8870's talk to the Elk?

 

[IVB]

I am using the distro panel, i think that's the 8818 right? I think it needs power for it, but I dunno for sure.

I spent another hour today rechecking connections, they seemed fine, but still no workie. I even pulled out my T2 wiring and ran it directly to the 8811, to no avail. That thermo used to work though, so now i'm thinking I screwed something else up. I didn't ring out the wires to make sure there's no cuts/etc as I ran out of time, I think that's the next step for me.

 

[hult]

Remember, you are not locked into using the 8811. There are many other companies out there making protocol converters.

The 8811 provides for fan-in and fan-out of the RS-xxx signals with separate driver IC's for (IIRC -- it's been a while since I opened it up) three channels.

Do you know of any other reasonably-priced converter that provides signal isolation/buffering and uses a DC power supply?

(The 8811/Proverter needs VAC power from the wallwart which is rectified for the +/- DC supplies internally which complicates backup from batteries. During an AC power failure, the thermostats including temperature measurement go down too).

TIA ... Marc

 

[pgray007]

Well I got my distribution panel all wired up today and... no luck, although the panel does help debugging a bit. Here are my symptoms if anyone can help.

The "B" (transmit) data light on the panel and 8811 blinks whenever I enter text via hyperterminal or when connected to the Elk. The "A" light, indicating receive, never blinks on the distro panel or on the 8811, so it looks like any commands sent out on the Aprilaire bus are "lost."

This is what was happening to me before, without the distro panel in the picture. I would send SN? commands and never get a response.

I've quadruple-checked my wiring, so I am not sure where the fault could lie. The symptoms are the same with each of my two thermos, and using the distro panel I was able to test each in isolation quite easily.

Would the receive light not blink if there was a problem with the connection to the 8811, or the connection from the 8811 to the PC/Elk, or does this indicate a more insidious problem somewhere between the panel and the thermos? I replaced the 8811 but not its cable, and that is the only thing I can not easily check since I don't have a compatible replacement lying around. I am hoping it is as easy as a bum cable.

Thermos get power from the distro panel and operate in standalone mode just fine. Any suggestions are greatly appreciated!

 

[gacevich]

just installed two 8870's in standalone mode so that i knew they worked before tying into home automation. upstairs works fine. down stairs will turn on heat and cool at proper setpoints with fan. however, the fan only mode will not work. that is, when i press the fan button the "fan on" indicator comes up and a relay click can be heard at the 8870 but the fan only does not come on. fan only mode worked with the prior "dumb" thermostat (even though wife now thinks the dumb one was smarter and it cost less!). double checked wiring and it is according to specs and the same as upstairs unit. looked for troubleshooting guide on aprilaire site but no joy. any thoughts on additional troubleshooting procedures? i am afraid 8870 may be faulty but want to be sure before trying to return it for another one. thanks.

 

[gacevich]

Over the past weekend i got my elk M1G installed..whew. Lots of little issues to deal with but the most worrisome was the integration of two aprilaire 8870's each going directly into the 8810 like pgray007 was looking to do. i can confirm that this configuration will work and take commands from ELK RP. i ran into several issues along the way, but checking, rechecking and perseverence paid off. sometimes i wasn't sure what solved the issue at hand but, for the moment, the system works.

one of the recurring themes was the tempermental nature of 8870 wiring. if u don't get it in the connectors just right, it won't work. takes lots of fiddling but now that it is working i loathe the thought of ever having to futz with it. thanks to all here at CT who provided guidance along the way.

now on to other challenges.

 

[will40]

Over the past weekend i got my elk M1G installed..whew. Lots of little issues to deal with but the most worrisome was the integration of two aprilaire 8870's each going directly into the 8810 like pgray007 was looking to do. i can confirm that this configuration will work and take commands from ELK RP. i ran into several issues along the way, but checking, rechecking and perseverence paid off. sometimes i wasn't sure what solved the issue at hand but, for the moment, the system works.

one of the recurring themes was the tempermental nature of 8870 wiring. if u don't get it in the connectors just right, it won't work. takes lots of fiddling but now that it is working i loathe the thought of ever having to futz with it. thanks to all here at CT who provided guidance along the way.

now on to other challenges.

sorry for dragging this thread up from the past as there seems to be some painful memories for some.

I too will be heading down this new 8870-8811-M1XSP road this week and was hoping for some help as from the sounds of things I may need it.

I will be replacing two AC/Heat thermostats with Aprilaire 8870's, running the CAT 5's to the ELK panel today. I will have the (two transformer) boiler with AC setup and do not plan at this point to install the distribution panel. Just a couple of questions before I get started:

1. I plan to power the 8870's from the AC unit transformer as I have the "C" conductors present at the existing thermostats. Is this ok?

2. Communications: I have read how others have used the flying leads from the 8811 and parallel up both thermostats to the leads, I also see that there are two RS-485 (RJ11) jacks on the adaptor as well. Should I parallel on the flying leads or crimp a jack on each 8870 cable and plug-into each port?

Thanks for any help or comments. I am sure you will be hearing from me soon.

Bill

 

[IVB]

Run them in parallel.

 

[fcwilt]

The distribution panel is entirely passive

There is one IC on the circuit board. It is in the area of the "common" connection terminal strip. Any idea what it does?

 

[hult]

The distribution panel is entirely passive

There is one IC on the circuit board. It is in the area of the "common" connection terminal strip. Any idea what it does?

Good catch -- especially for something that has been in the air/error for nine months ... ;-)

The immediate discussion was of the power supply, which is passively distributed (there is no "power gain") through slide switches. The LEDs at each just indicate whether there is power or not but I suppose that some folks would consider LEDs "active devices" because they are semiconductors -- even if they don't increase power.

The RS-xxx signals are also passively distributed (not actively buffered one from another) on the board to each of the eight supported 8870's. If you short together any of the 16 signal pairs all eight TXs or RXs as (the case may be) will go down.

With respect to the IC, I assumed it was to drive the communication status/debugging LEDs. But it is possible that it is a buffer at the input of the board for TX+, TX-, RX+, RX-. (Aprilaire calls these A+, A-, B+, B-) What are the markings on IC ?

The A and B LEDs (which are separate from the power LEDs) should flash when there are communications. They can turn solid ON if there is a problem. To identify which channel/output is causing the issue, the switches are opened for each channel one by one until the light flashes again.

HTH ... Marc

 

[fcwilt]

The distribution panel is entirely passive

There is one IC on the circuit board. It is in the area of the "common" connection terminal strip. Any idea what it does?

Good catch -- especially for something that has been in the air/error for nine months ... ;-)

The immediate discussion was of the power supply, which is passively distributed (there is no "power gain") through slide switches. The LEDs at each just indicate whether there is power or not but I suppose that some folks would consider LEDs "active devices" because they are semiconductors -- even if they don't increase power.

The RS-xxx signals are also passively distributed (not actively buffered one from another) on the board to each of the eight supported 8870's. If you short together any of the 16 signal pairs all eight TXs or RXs as (the case may be) will go down.

With respect to the IC, I assumed it was to drive the communication status/debugging LEDs. But it is possible that it is a buffer at the input of the board for TX+, TX-, RX+, RX-. (Aprilaire calls these A+, A-, B+, B-) What are the markings on IC ?

The A and B LEDs (which are separate from the power LEDs) should flash when there are communications. They can turn solid ON if there is a problem. To identify which channel/output is causing the issue, the switches are opened for each channel one by one until the light flashes again.

HTH ... Marc

It seems to be an LM393 which is listed by National Semi. as a dual comparator. So it would make sense that the IC is used to drive the A and B LEDs. If I had more energy I would trace the board but I don't.

Did they OP ever get his system up and running?

Today I fired up an 8870/8818 setup and hooked it up to my HomeLogic HA system and it worked. Perhaps I will try hooking it to my M1. Seemed pretty straightforward so I am curious what the problem turned out to be.

 

[hult]

The distribution panel is entirely passive

There is one IC on the circuit board. It is in the area of the "common" connection terminal strip. Any idea what it does?

Good catch -- especially for something that has been in the air/error for nine months ... ;-)

The immediate discussion was of the power supply, which is passively distributed (there is no "power gain") through slide switches. The LEDs at each just indicate whether there is power or not but I suppose that some folks would consider LEDs "active devices" because they are semiconductors -- even if they don't increase power.

The RS-xxx signals are also passively distributed (not actively buffered one from another) on the board to each of the eight supported 8870's. If you short together any of the 16 signal pairs all eight TXs or RXs as (the case may be) will go down.

With respect to the IC, I assumed it was to drive the communication status/debugging LEDs. But it is possible that it is a buffer at the input of the board for TX+, TX-, RX+, RX-. (Aprilaire calls these A+, A-, B+, B-) What are the markings on IC ?

The A and B LEDs (which are separate from the power LEDs) should flash when there are communications. They can turn solid ON if there is a problem. To identify which channel/output is causing the issue, the switches are opened for each channel one by one until the light flashes again.

HTH ... Marc

It seems to be an LM393 which is listed by National Semi. as a dual comparator. So it would make sense that the IC is used to drive the A and B LEDs. If I had more energy I would trace the board but I don't.

So the LM393 is almost certainly being used as an LED driver (as I originally assumed) to reduce loading of the data lines to minor/negligible.

... Marc