> I have a two story home with a completely separate HVAC system for each
> floor.
That's a variation of what I have in my house - one HVAC controlling
the north side, another controlling upper *and* lower floor on the
south side - but the first HVAC is doing a better job of cooling the
first floor, as a side effect.
> The second floor HVAC is under-sized. Maybe 30 days per-year it
> can't keep the upstairs cool enough.
That looks about right, maybe just a *little* undersized - the system
is sized for "design temperature", not peak.
Out of curiosity, did you run the heat load calculation on your house?
My old one had a 4 ton heatpump installed, HVAC salesmen were telling
me I need 5 ton, I ran the heat load calculation and found that 3 ton
would've been sufficient as it was, and 2.5 would've been sufficient
if I installed attic insulation and replaced single pane windows with
double pane.
> The first floor HVAC doesn't keep my office (with 4 computers) cool enough,
A purely practical consideration - I was amazed at how much more power
efficient and cold computers are becoming. My old workstation was
consuming 300-350 W, new one, ten times faster, is under 100 W. For
less than $300 (not including the case) - that is with Core i5-2500
and 8G RAM. You can easily save that much in your power bill in a few
months, and the room will run cooler.
Another thing - in your climate, you might get very good results with
just ventilation. This is what I did for my server room under the
staircase:
http://www.flickr.com/photos/tkachenko/4726646146/in/set-72157624212628229
Detailed drawing is at
http://sketchup.google.com/3dwarehouse/details?mid=57077d3648762c4bfcf6ee47fc453c58
- if you have a router table or someone with access to it, you'll have
it done in no time (hint: template). Surprisingly, it is very quiet -
with fans I installed.
> and one bedroom full of afternoon sun is also too warm.
Deja vu :) By the way, when I replaced a single pane window in my
hottest second floor southern exposure bedroom with a double pane with
reflective coating, it has become the coldest room in the house. Even
being on the second floor.
> My idea is to run a trunk line from the upstairs HVAC to the downstairs
> trunk. Install baffles as such that DZ (if capable) can use the smaller A/C
> and fan of the upstairs unit to maintain temperature in the overly-warm
> rooms downstairs, and use the downstairs HVAC to help maintain the
> temperature upstairs.
>
> Is control scheme possible?
Yes, I think so.
> Any pointers on implementing it, if so?
In any case, it would be a good idea to start with instrumentation -
actual situation is never what you perceive to be happening. When you
see what exactly is happening, you'll have more food for thought.
In best case, you'll be able to get away with just using dampers to
control each floor separately and not needing to modify the ductwork.
--vt
Out of curiosity, did you run the heat load calculation on your house?
> The first floor HVAC doesn't keep my office (with 4 computers) cool enough,
A purely practical consideration - I was amazed at how much more power
efficient and cold computers are becoming.
Deja vu :) By the way, when I replaced a single pane window in my
hottest second floor southern exposure bedroom with a double pane with
reflective coating, it has become the coldest room in the house. Even
being on the second floor.
> My idea is to run a trunk line from the upstairs HVAC to the downstairs
> trunk. Install baffles as such that DZ (if capable) can use the smaller A/C
> and fan of the upstairs unit to maintain temperature in the overly-warm
> rooms downstairs, and use the downstairs HVAC to help maintain the
> temperature upstairs.
>
> Is control scheme possible?Yes, I think so.
> Any pointers on implementing it, if so?
In any case, it would be a good idea to start with instrumentation -
actual situation is never what you perceive to be happening. When you
see what exactly is happening, you'll have more food for thought.In best case, you'll be able to get away with just using dampers to
control each floor separately and not needing to modify the ductwork.--vt
> So, anyway, it's just been a short time since I started thinking serious
> about this and haven't done a heat load calc yet. Are there any good free
> tools available to do that online these days? It's been years since I
> looked.
Don't know about any free tools, but used http://www.hvac-calc.com/
for myself - and would say that you're getting a hell of a bang for
$50.
> Yeah, I agree with the idea of getting instrumentation first. I guess I
> just need to get that 1-wire USB adapter thingy, some CAT5, and some 1-wire
> temp sensors and that's it?
If you decide to go the wired way - yes. Might want to start playing
with a few XBees - you can move those around and get some ideas as to
how to place permanent 1-Wire sensors better.
XBee sensors are cheap (as far as TCO is concerned), somewhat error
prone (there are unexplained peaks, most probably of RF interference
nature) and less precise (10 bit resolution vs. 12 bit for 1-Wire).
1-Wire are more expensive to install if you need to break walls, but
may turn out to be less expensive if you need lots of them. They are
also more precise - but I don't know how much practical difference
does it make. I run both, didn't notice any detrimental effect on
control quality. Also, power consumption will be less - though I
haven't power optimized XBee sensors yet.
> But then again, I've only been seriously thinking about this for like 30
> hours, so what do I know!? :D
:D
> I'm just excited about the idea not only for comfort reasons, but, like I
> said here: https://plus.google.com/117177689300294532641/posts/jgYMXZPNTjG,
> it's so *nerd cool*! I expect to be posting in and reading this group a
> lot.
Oh, and about that idea you expressed there... Take a ticket after
Darren and hold off for a couple of weeks :D
--vt
1-Wire are more expensive to install if you need to break walls, but
may turn out to be less expensive if you need lots of them. They are
also more precise - but I don't know how much practical difference
does it make. I run both, didn't notice any detrimental effect on
control quality. Also, power consumption will be less - though I
haven't power optimized XBee sensors yet.
Oh, and about that idea you expressed there... Take a ticket after
Darren and hold off for a couple of weeks :D--vt
> In fact, I may not have to route Cat-5 anywhere. My house has at least 2
> runs of Cat-5 ran to jacks in nearly every room ran to a structured wiring
> panel
> (https://picasaweb.google.com/117177689300294532641/October92011#5661621300099461634).
> I haven't looked into how to wire those sensors yet, so that may or may not
> work. But anyway, that's for another day.
Food for thought: http://www.flickr.com/photos/tkachenko/sets/72157624447094338/
--vt
> One thing I'm a little confused about is how many sensors is too many,
I'd say one per room is just about right. There are situations where
that is not enough -
- when a big room is influenced by more than one HVAC
- when a high room has stratification with a few degrees difference
(example: two story rooms with a loft on the second floor)
- when a room behaves differently under heating and cooling (one of my
rooms is like this, sensor at 14" from the floor lags for HOURS in
heating mode, when people in the room are already half baked) - also
related to stratification
> and when you get more than that amount, what exactly the solution is.
I don't think you'll ever get there :)
> Here (http://www.homeclimatecontrol.com/deliverables/supported-hardware) it says
> "6 Channel Master Hub (use this if you're planning on a lot of sensors and
> actuators);" but I'm not sure what "a lot" is. Also, the link to that
> Master Hub is broken, and a Google search doesn't show any suppliers for
> that part number.
Thanks for the update. Here's a link as it is today:
http://www.hobby-boards.com/store/products/4-Channel-Hub.html
Wonder when small suppliers will learn to keep old product information
for data sheets :/
Note that this is not a 6 channel hub, but 4 channel. I'd say there is
a good reason the number of channels was reduced - 6 is overkill for
most 1-Wire applications, maybe even 4 is.
The original idea was that 1-Wire network is a bus, not star, but I
was able to dodge this with using the same CAT5 cable for feed *and*
return (see the note on the photo here:
http://www.flickr.com/photos/tkachenko/4810146029/), so effectively a
physical star configuration becomes an electric bus. Moreover, for
limited applications you can get away even with an electrical star,
but that's asking for trouble - test it first. If you add a punchdown
block to your setup - that'll be a piece of cake to test and change.
--vt
>> Here (http://www.homeclimatecontrol.com/deliverables/supported-hardware) it says
>> "6 Channel Master Hub (use this if you're planning on a lot of sensors and
>> actuators);" but I'm not sure what "a lot" is. Also, the link to that
>> Master Hub is broken, and a Google search doesn't show any suppliers for
>> that part number.
>
> Thanks for the update. Here's a link as it is today:
> http://www.hobby-boards.com/store/products/4-Channel-Hub.html
>
> Wonder when small suppliers will learn to keep old product information
> for data sheets :/
>
> Note that this is not a 6 channel hub, but 4 channel. I'd say there is
> a good reason the number of channels was reduced - 6 is overkill for
> most 1-Wire applications, maybe even 4 is.
Further analysis suggests that this product may lack an important
component that was present in the 6 channel version - the 1-Wire
adapter. Given the lack of documentation and "sold out" status - I'd
pass.
I guess you saw warnings about LinkUSB here as well.
Information on Supported Hardware page has been updated.
--vt