DX Geothermal Pond 4 ton Heat Pump

[jeff5may]

Originally Posted by MEMPHIS91

Just now starting to look into electric expansion valves,
New Carel Electronic Expansion Valve E2V24BWAE0 3 8" 3 8" ODF | eBay

Pretty sure the price is right, but what the heck controls them?
Proprietary electronic devices. The newer superSEER, hyper-heating, remote control, networking units all use different proprietary devices.
Is it worth it?
Probably too new to be cheap yet, but the elements of luck and timing are always there.
Electric Expansion Valve Control

Also should I start looking for a variable speed indoor fan? Or is the fixed speed I have now ok?

I would definitely set up something simple first. As time ticks on, these bleeding-edge products will become more standardized and reliable once the manufacturers iron out the field issues. A fixed-speed air handler can be made variable-speed pretty easily. For example, I use a light dimmer and a relay to do defrost on the window units I morph into heaters. When the unit is actively defrosting, the defrost relay puts the dimmer into the blower circuit. I adjust the blower to do 20 or 30 rpm so the indoor coil can supply enough heat to completely defrost the outdoor coil in a few minutes or less.

[Fordguy64]

Yea I wouldn't worry about the exv at this point.. Get the thing up and running and work the bugs out of it and then you can try this and squeeze every bit out of it.. Just my .02

[MEMPHIS91]

Jeff, yes the controller looks to be the most expensive part. I wonder how hard it would be to design a controller?
Ok I would at least like to have the system set up to easily add the exvs, I'm thinking a couple extra ports for pressure senors and temp?

So parts list that I still need,
-2 x 4 ton txvs (should these be external equalized?)
-2 x 7/8" check valves
-2 x 1/2 sight glasses
-1 x reversing valve
-1 x high pressure switch (what psi? 250?)
-1 x low pressure switch (what psi?)
-1 x 25' line set
-1 x thermostat to control the VFD
-200' 1" pvc
-Tons of pipe insulation
-I have this water pump, will it work? WILO 4090765 Star S21Fx CI 3 Speed Circulator - Power Water Pumps - Amazon.com
Dimmer and relay for indoor fan
-Water temp sensors for incoming and out going water temp (any suggestions on model?)
-I already have the wire and breaker box.

Compressor and coaxial coil on the way already

[jeff5may]

Originally Posted by MEMPHIS91

Jeff, yes the controller looks to be the most expensive part. I wonder how hard it would be to design a controller?
Depends on whether you're a software engineer or not.
Ok I would at least like to have the system set up to easily add the exvs, I'm thinking a couple extra ports for pressure senors and temp?
An EXV is bi-directional, you only need one. The access valves can be used for discharge and suction pressure sensing. Just make sure to buy sensors that have the valve depressor built in.
So parts list that I still need,
-2 x 4 ton txvs (should these be external equalized?)
For your coax coil, it doesn't need to be. For the air handler HX, it depends. If it has a flow distributor that drops pressure, then yes. If it has a cap tube or fixed orifice with some kind of distributor, then maybe. It could be left alone possibly.
-2 x 7/8" check valves
Size of your liquid line
-2 x 1/2 sight glasses
-1 x reversing valve
-1 x high pressure switch (what psi? 250?)
R22: about 300 psi. R410: about 450 psi.
-1 x low pressure switch (what psi?)About 25 psi
-1 x 25' line set
-1 x thermostat to control the VFD
-200' 1" pvc
-Tons of pipe insulation
-I have this water pump, will it work? WILO 4090765 Star S21Fx CI 3 Speed Circulator - Power Water Pumps - Amazon.com
Looks like a Taco pump. Probably will not self-prime.
Dimmer and relay for indoor fan
-Water temp sensors for incoming and out going water temp (any suggestions on model?)
-I already have the wire and breaker box.

Compressor and coaxial coil on the way already

Your planned control scheme will dictate how the thermostat does what it does. Most residential heat pump thermostats have direct control over next to nothing. Some will run a small blower on their own, nearly all will run valves and relays and contactors and the like. Typically, they are wired to separate control boards/relays/contactors/logic devices that sense the thermostat demand and react however they are supposed to.

The basic thermostats have 1 stage of heating, 1 stage of cooling, with manual changeover from heating to cooling cycle. The more elaborate models have multiple stages of temperature control and automatic changeover. The deluxe models have tons and tons of features, like wifi or ethernet connectivity, motion sensors, predictive demand sensing, programmable heating/cooling schedules, data logging, and the list goes on and on and on. As far as the air handler and outdoor unit are concerned, they push the same signals out as the basic thermostats do.

Here's an industry-standard chart:

Here is a highly common wiring scheme:

This diagram accounts for 98% or better of all residential units. Main difference in above: Left image has furnace plus A/C, right has heat pump. Both diagrams have extra heating stages. Less than 1% of residential systems have more than a single stage of cooling.

Your first decision you will need to make in your control scheme is this: Do you want a standard-looking thermostat on the wall? This is not a trivial decision. Visitors, family members and house guests will be able to operate something standard. If you have some strange interface to operate the HVAC system, everyone will be calling on you for advice on how to do what (if you're lucky). Worse, they may try to operate the control without your guidance, causing who knows what mayhem.

[jeff5may]

On the other side of the user interface, the possibilities are wide open. You can stick with the industry-standard wiring schemes, to make troubleshooting or retrofitting easy in the future. Like I said in the last post, 99% of home units have 1 zone, 1 stage of cooling, and 2 or less stages of heating, with relay logic control. Regardless of how luxuriously the wall thermostat operates, there are 5 or less active wires that dictate control of the rest of the system: heat, cool, indoor fan, reversing valve, 2nd stage or emergency heat. The rest of the system obeys these few wires.

The modern manufactured units have all kinds of stuff built in. The variable-speed air handlers have controllers built into them now. The more elaborate ones have zone and capacity control, automatic fan throttling, data and error logging, serial- or web-based communication hardware, and the like built in. The modern outdoor units communicate with the air handler (via a digital interface) to enable precision control of everything. There are two main trends in this direction: the mini-split units have nearly all of the control and sensing functionality built into the outdoor unit, while the full-sized splits have a standard (5-wire) thermostat connection that nearly always connects to the air handler. Some split systems integrate all of the fuzzy logic control into the air handler, and the outdoor unit is basically a slave with sensors attached. The others split the command and control between the thermostat, the air handler, and the outdoor unit, which are more modular and autonomous in functionality. Naturally, the more distributed and smart the individual modules are, the more flexible the installation options are out in the field. A super-smart air handler or outdoor unit can be connected to something dumber and still retain a lot of added features.

AC_Hacker and I started a discussion a year or so ago, in hopes of developing a general-purpose heat pump controller. For the first month or so, we made good progress towards this goal. We had a somewhat short list of basic functions that the unit would do. After some collaboration between members, the list of desired functions grew and grew and grew to the point that it was impractical to use an arduino uno or teensy stamp based dev board due to the limited processing power, onboard memory, and pin count. Members with intimate knowledge (of mainly software) picked and chose from the list, ran off to their labs and workshops, and came back with all manner of controllers that were tailored to their specific sets of conditions, desired functions, and skill levels. Very few finished controllers were universal or generic in nature.

I consider the thread as a successful work in progress that is not closed yet, as the discussion has already yielded at least half a dozen working controllers that the individual makers employed to fulfill their specific purposes in the real world. Much information has been shared, and there are tons and tons of hardware and software building blocks to pilfer through, as well as a wide range of abstract design philosophies to consider, contained in the thread. A basic, casual ecorenovator can read through the discussion, buy maybe $50 or less worth of hardware, and copy some code snippets or entire programs, and begin to develop their own solution.

I made a simple, uno-based controller, that has a generic LCD keypad shield, a single one-wire temperature sensor, and a 4-wire (no 2nd stage heating) TTL output, that can be rigged to your choice of relay shield. It can be assembled, plugged into a USB cable, fed my program, and will operate as a plain jane wall thermostat. It earned its worth controlling a hacked window unit until I retired it. Single setpoint, auto or manual changeover, rigs up like the diagram on the right.

I don't think AC is satisfied with the outcome of the thread so far. He was expecting a single board to be developed, that could be fed a program, rigged up to what have you, and be generic enough to control everything. As the to-do list grew, things like data-logging, frost sensing, digital communications, and remote control were mandated that made it very difficult to fit a program containing all this functionality into the space available. Many members who made devices with lots of functions switched to an arduino mega board, due to its larger memory and pincount, to fit everything in.

I encourage you to read through the thread, as it will help you make decisions on how to assume control of the parts of your system and give you some ideas on how to get them all to play well with each other. I am beginning to work on another window unit hack, and plan to integrate more control functions (mainly defrost control, maybe data logging) into the controller I made on the thread. As always, I will share what I accomplish, so others can pick up their own projects and run with them. Hopefully my piddling will at least tell people what doesn't work.

[MEMPHIS91]

Wow tons and tons of info......
-So what is so expensive about the exvs? Is it the controller? Because if I only need one of them that is not too expensive. I can't really find a kit with valve and controller so that is probably the main reason to wait.
-Ok so I'll just put a few extra access valves in when I build it.
-Ok so 1 x internally equalized and one external because right now the indoor coils have a distributor to 5 cap tubes.
-Any suggestion on the pump then? Mine is a lot like a Taco
-Gotcha, so basically it my choice on how I set up the system as to what thermostat I should need. Since this I am going to need to do some fine tuning and need to know exactly what is going on, data logging may be a good idea. I keep looking into them and see, I did read that thread, it took me almost 2 weeks of reading here and there. I was for sure overwhelmed with the amount of info there. I would love to just do the plug and play stuff, but that be way to expensive for what I am wanting. I am hoping that the VFD will give me some good control as well.
I was thinking crazy things like reading the outside temp and water temp to make the compressor start at a certain speed in order to match needed capacity.
This reply is rushed, I'll do more work and reply back with more concrete info.
Thanks Jeff

[jeff5may]

Here's the thing about electronic expansion valves: They don't all work the same way, and you are relying on an electric over mechanical control layer to regulate how your system acts. If you pony up the money for a turnkey rig, it will do what it was designed to (hopefully well).

There are 2 main types of EXV's: pwm and stepper.

The pwm type work like a super-fast solenoid valve, and I imagine they have some kind of damper built into them.You feed them a square wave, and they conform to the duty cycle of open / closed. Who knows whether they are normally open or normally closed? Who knows how slow or fast the damper is? Not me.

The stepper type are a modified version of a brushless dc motor. They work kind of like the 3-phase compressor you bought, and require a specific motor driver like the VFD you are selecting. They have to be told to move. I don't imagine the EXV have a position sensor built into them, so at the power-up events they have to be told to close fully. This initializes their position. After that, the position must be kept track of in software somehow. During operation, the position is tightly controlled by some kind of high-precision algorithm.

The EXV-metered systems in industry are meant to be engineered and specified to the particular application. Parker and Emerson both sell the EXV's in a range of capacities. They are meant to be mated to a controller that is picked out, a sensor array that is picked out, cables that are picked out, and everything is ordered as individual components. The order is quoted, and ships as assembled as the engineer spec's out. Not cheap at all.

The OEM way this happens includes this previous step and has more options. The design team gets a few basic controller boxes and sensors to do research and development with. Licensing contracts and agreements are made, the manufacturer gets a software development toolbox, and design parameters are hammered out. While running trials, first with simulation software and then in the lab, the OEM decides how to integrate the EXV software into the microcomputer that was spec'd to run the machine.

During trials, the design team twiddles and tweaks with hardware and software while the marketing and accounting teams are projecting the future and counting costs. While prototypes are being tortured and certified to conform to safety and operating standards, everyone involved is figuring out who gets paid to do what during manufacturing. In the end, Parker or Emerson or whoever the source is gets a bulk order for raw goods, plus a cut for their intellectual property. Service and spare parts orders are contracted out the same way.

As far as we are concerned, the only useful universal controllers to be had are going to be the industrial type, as the OEM's hold on tight to their design team boxes. These industrial boxes are made to be set by sharp fridgie techs and mechanical engineers. Not quite user-friendly, but it wouldn't be a major undertaking to get the thing to do what you need it to do reliably.

The stuff inside the residential units may be able to be used by salvaging all the controls, sensors, valves, etc. from a unit that has developed a leak or some other "minor" problem and transplanting the guts (or repaired unit) into something else. Finding out where individual components came from, and what they need to run with, is nothing like robbing a cap tube or TXV from an "inferior" older unit, as the EXV drive hardware and control software is integrated, embedded, and grafted into a master control that serves a purpose larger than just EXV control. Not many people have been able to successfully command or "jinx" these OEM units.

Staying with the Emerson / Alco line, there are two basic lines:

Super Universal, expensive:
097707 - Emerson Flow Controls 097707 - EC3-X33 Stand-Alone Universal Superheat Controller
Descriptive brochure:
http://estore.rrd.com/emerson/download.ashx?did=873
Plain jane driver module:
EXD-U-00

Cheaper version: good luck finding one.

http://www.emersonclimate.com/europe..._EXD_HP1-2.pdf

Parker / Sporlan, Siemens, Danfoss, Carel, etc. all make their own takes of these controllers. They all seem to like RS485 / ModBus serial communications. Not surprising, due to the industry standard.

[MEMPHIS91]

Jeff, So basically it is the control that is a problem. And designing the controller would close to impossible with out time and money.
So Exv's are out of the question until they become more common, so my best route is 2 txvs. One internal and one external equalized.
I can program the VFD for control of the "stages", should I do it like starting out run at 40% and if the temp doesn't change in a certain time then boost to 55% and so on til 100%?
I really like the idea of a controller that measures outdoor temp and water temp and starts the compressor at a certain capacity based of the difference between the two temps. That doesn't sound to hard to add later if I wanted to either.
Look what showed up today.


Here is the data sheet on it. http://www.hvacusa.com/media/datashe...466-signed.pdf

And what else showed up in a crazy awesomely done box. THANKS BUFFALO BILL PATRICK!!!

[MEMPHIS91]

So I have been searching forever now for a self priming pump. Normal closed loop circulators don't need to be self priming, but this is open loop. But why would I ever loose primp?
Anyway, best I could find is something like this. Not sure what size because this looks way to big. 1 5HP 1" Shallow Well Garden Sprinkler Jet Water Pump Booster Self Priming 1100W | eBay

Also found this AMT 1 1/2" Cast Iron Self-Priming Centrifugal Pump - 2822-95

Any tips?

[MEMPHIS91]

Been doing some research on TXVs. I'm looking at balanced port TXVs, just because this is variable speed. I believe they will work much better according to what I have read. Any ideas?
Here are some articles.
Choosing the correct TXV

http://www.emersonclimate.com/Docume...re-article.pdf

http://www.ebay.com/itm/NEW-Carrier-...-/121415905907

http://dms.hvacpartners.com/docs/100...atx-hso-01.pdf