Morphing Air Conditioner into Autocascade System

[mytekcontrols]

I have almost completed all the plumbing changes required for the Buffer Valve addition, and have also fabricated, and installed an access port manifold.

My problem is that the pressure gauges I was using are basically cr_p!!! The readings were off by 10 psi from when I had initially calibrated them. What I need is at least one reliable gauge (suction side) that is fairly small, and can be easily panel mounted. Anybody got a good source?

P.S. I'll be posting some pictures of the plumbing changes momentarily (need to switch computers).

[n00b 0f l337]

Gonna betray a source for you Michael, hope you enjoy it
(Jk its just ebay)
http://cgi.ebay.com/ws/eBayISAPI.dll...MEWA:IT&ih=005
Hows that?

[mytekcontrols]

You are the MAN!!!!
Thanks Adam that's a great deal

As promised, here are some pics on the latest changes:

1. Access Port Manifold (unpainted)
2. Access Port Manifold (painted)
3. Access Port Manifold connected
4. Buffer Valve and associated Pressure Switch
5. High Pressure Cut-Out switch

Now just got to wire it all up

[n00b 0f l337]

Looking very sexy there.

[teyber]

VERY nice idea about the access port manifold!

[kayl]

nice work Michael, thats an awsome idea for the access valves hey.
WIth the suction line access port where do you braze that into so it doesnt freeze over? the expansion tank?


WHat the purpose of the buffer valve?

[mytekcontrols]

WIth the suction line access port where do you braze that into so it doesnt freeze over?

Good point! I forgot about that.
Hmmm... my connection is almost 6" away from the suction line... could be a problem. I can insulate the line, but you are right, there will probably still be enough conduction to probably make it sweat (I doubt that it will freeze). I might need to also use some armaflex tape on the painted steel plate, and use small spacers between it and the front panel that it will eventually get mounted to. A little sweating on the exposed end of the access fitting and brass cap should be ok.

...the expansion tank?

Actually one of the access ports does go to the expansion tank as well. Both are needed due to the pressure restriction of the expansion tank captube (couldn't expect to pump a very good vacuum through 12 feet of 0.031" ID captube).

I guess if I really had a freezing problem, a piece of stainless steel tubing could be soldered between the suction line and the manifold port connection. Or a small power resistor (used as a heater) clamped to the backside of the manifold mounting plate and connected to my power would also work. At 120 VAC, a 3K 5 watt resistor would probably do it.

I'll just have to see how it goes (a bit of trial and error).

WHat the purpose of the buffer valve?

If the discharge pressure gets too high during start-up/cool-down, it dumps some of the refrigerant gas into the expansion tank to lower the amount of refrigerant circulating in the system. This reduces the discharge pressure, and then slowly meters the refrigerant back into the system via the expansion tank captube. Since the gas source is taken after the 1st phase separator, we leave the bulk of the highest boiling refrigerants (as separated condensates), to continue the work of cooling the succeeding stages, so that eventually enough of a phase change has occurred in the other circulating refrigerants to keep the discharge at a more reasonable pressure.

[mytekcontrols]

Here is a drawing showing the construction details of the Access Port Manifold (Dimensions are in INCHES --- all holes are aligned to horizontal CenterLine):

[mytekcontrols]

Suction Gauge is on order (thanks again to Adam). When this comes in I'll be able to complete the pressure circuit, and put the unit on a vacuum pump.

Until then, I have a bunch of wiring to do. And of course I'll need to think this over a bit, so as to keep everything neat and to the point (no sense in rushing through this, especially since the rest of the unit is coming together so well). Still gotta find a good location for the RUN Capacitor. Also need to install a 10 amp 120 VAC relay between the compressor and the HPCO, since it is only rated for about 3 amps. The 400 Psig HPCO was a great find, due to its very small size, and inexpensive pricing. I got this from Under-The-Ice.com (thanks Ron).

My next test will be with R-600 and R-290 as my liquids. I will probably stick with the same ratio and amounts of the other refrigerant gases as used in the last charge. I'll post the results and the new charge specs.

Here's a pic of the 30" - 0 - 300 psig gauge I ordered from Frosty-Freeze A/C Products Company (looks to be a good quality ASHCROFT gauge, not bad for under $5.00):

[n00b 0f l337]

I'd put the capacitor right below the fan motor, and again no problem, least I can do

[Polizei]

If the discharge pressure gets too high during start-up/cool-down, it dumps some of the refrigerant gas into the expansion tank to lower the amount of refrigerant circulating in the system. This reduces the discharge pressure, and then slowly meters the refrigerant back into the system via the expansion tank captube. Since the gas source is taken after the 1st phase separator, we leave the bulk of the highest boiling refrigerants (as separated condensates), to continue the work of cooling the succeeding stages, so that eventually enough of a phase change has occurred in the other circulating refrigerants to keep the discharge at a more reasonable pressure.

I love explainations like these. Well said.

I knew why you used the valve, but not how or where. Great idea.

[n00b 0f l337]

Seems rather common in these systems, at least newer ones, I love the idea, its protection of sorts, like a burst relief valve, however less dangerous And I think we saw CryoTek do this as well. Allows HX's to chillax out.

[mytekcontrols]

I'd put the capacitor right below the fan motor...

Good idea, but since it would be a bear to gain access to this area now, I think a good compromise would be to put it above the fan motor instead.

Quote:
Originally Posted by mytekcontrols
"If the discharge pressure gets too high during start-up/cool-down, it dumps some of the refrigerant gas into the expansion tank to lower the amount of refrigerant circulating in the system. This reduces the discharge pressure, and then slowly meters the refrigerant back into the system via the expansion tank captube. Since the gas source is taken after the 1st phase separator, we leave the bulk of the highest boiling refrigerants (as separated condensates), to continue the work of cooling the succeeding stages, so that eventually enough of a phase change has occurred in the other circulating refrigerants to keep the discharge at a more reasonable pressure."

I love explainations like these. Well said.

Thanks
My goal is to get responses like yours. That way I know I am hopefully going down the right path in conveying my thoughts as clearly as possible.

I knew why you used the valve, but not how or where. Great idea.

Seems rather common in these systems, at least newer ones, I love the idea, its protection of sorts, like a burst relief valve, however less dangerous And I think we saw CryoTek do this as well. Allows HX's to chillax out.

Yep! Quite common in some of the Polycold units, especially the larger models where it helps keep the size of the expansion tank(s) more reasonable. I think the units that CryoTek builds use a different system, since they use no expansion tanks what-so-ever. However I do believe you are right about him discussing it on this forum. I wish he would get back on here, because I would love to see him join in on some of these conversations.

[n00b 0f l337]

So ya still havent revealed it, but what are you using this little thing for? Do you need a CPU evaporator setup? If so I'd gladdy braze one up and send it your way.

[kayl]

Good point! I forgot about that.
Hmmm... my connection is almost 6" away from the suction line... could be a problem. I can insulate the line, but you are right, there will probably still be enough conduction to probably make it sweat (I doubt that it will freeze). I might need to also use some armaflex tape on the painted steel plate, and use small spacers between it and the front panel that it will eventually get mounted to. A little sweating on the exposed end of the access fitting and brass cap should be ok.


Actually one of the access ports does go to the expansion tank as well. Both are needed due to the pressure restriction of the expansion tank captube (couldn't expect to pump a very good vacuum through 12 feet of 0.031" ID captube).

I guess if I really had a freezing problem, a piece of stainless steel tubing could be soldered between the suction line and the manifold port connection. Or a small power resistor (used as a heater) clamped to the backside of the manifold mounting plate and connected to my power would also work. At 120 VAC, a 3K 5 watt resistor would probably do it.

I'll just have to see how it goes (a bit of trial and error).


If the discharge pressure gets too high during start-up/cool-down, it dumps some of the refrigerant gas into the expansion tank to lower the amount of refrigerant circulating in the system. This reduces the discharge pressure, and then slowly meters the refrigerant back into the system via the expansion tank captube. Since the gas source is taken after the 1st phase separator, we leave the bulk of the highest boiling refrigerants (as separated condensates), to continue the work of cooling the succeeding stages, so that eventually enough of a phase change has occurred in the other circulating refrigerants to keep the discharge at a more reasonable pressure.

That makes sense and would help alot.

Sweating schraider valve will be fine i think. When looking at your
drawings i was interested why there was a access valve on your expansion
tank and though you might have used that to charge and tune the system. I
know when using rotarys with cascades and tunning mine freeze over big
time and you cant remove the refrigerant gauge hose unit will you have
turned the system off and left it for ages. If i take it off even when its
almost defrosted my schraider valve leaks gas which stuffs my charge up.

So i was trying to decide for the cascade im building now to have 2 low
side access points line in your autoC and use the one on the suction line
for vacuuming and then use the one on the expansion tank for charging??
You think thats ok?? or any suggestions. I like you idea of bring all the
access valves together at the one point, will make tunning much easier
with the case all done up.

[mytekcontrols]

So i was trying to decide for the cascade im building now to have 2 low
side access points line in your autoC and use the one on the suction line
for vacuuming and then use the one on the expansion tank for charging??
You think thats ok?? or any suggestions.

Yes that would work, but you would of course have to allow for the delay in balancing out the gases when charging through the expansion tank, and into the system via the expansion tank captube.

I take it by your description, that the access port connected directly to your suction would not be a part of your manifold. And when the vacuum process was completed, it would be capped up and insulated, thereby leaving only 2 externally accessible ports (expansion tank & discharge). Was this what you had in mind?

I could see how the upper stages in a standard cascade would be problematic when it came to freezing suction side ports. In my AutoC I wont have nearly the problem, since my suction is not going to be much below 0 C (low load), and probably around 5 C under full load. So I think the resistor heater idea would most likely solve any sweating, or slight freezing problems that might be encountered in this situation. The resistor can be one of those metal encased types bolted to the backside of the manifold plate with a little heatsink compound, located close to the suction access port. This could then be connected to the compressor power source, only coming on when the compressor is running.

Although if you have the space, a foot or so of small o.d. stainless steel (3/16" or 1/4") tubing would provide enough isolation, due to the much poorer heat transfer as compared to copper tubing. This would allow you to bring all 3 connections out to your manifold as I have done, and not have any freezing problems. Of course the first 8 to 10 inches where it connects to the suction will need to be insulated.

Photo of suggested Power Resistor type for use as manifold anti-condensation heater (select appropriate wattage for application):

[godmod]

Hey Michael, any progress?

[mytekcontrols]

Progress??? What's that?

Yes there has been some, but not as much as I would have liked. Basically I am still buttoning up several loose ends:

• Insulated compressor suction line (boy was this a bitc_h to do with every thing else in the way)
• Patched foam insulation where I had changed some of the plumbing
• Finalized test load by foam insulating inside of a 4" ABS tube with end caps
• In the process of doing final electrical wiring

Hopefully will be recharging this week, and then test.

[teyber]

So ya still havent revealed it, but what are you using this little thing for? Do you need a CPU evaporator setup? If so I'd gladdy braze one up and send it your way.

hey mate i am also curious... what evap are you using?

[mytekcontrols]

Teyber --- I'm using a coil with a radiant heat source as my dummy load.

Here's a look at the test set-up...

If you look at my avatar you'll also see a representation of a large test load made for up to 5000 watts. I am simply using a very small version of this same design. The copper tube coil is wrapped in aluminum foil to reflect any radiated heat that is trying to escape, back into the coil. The heater (large ceramic power resistor) sits directly centered within the coil.

Initially I was using fiberglass insulation, but as of this weekend, I have placed it within an air tight ABS enclosure, and foam insulated it to about 1.25" thickness all the way around. This should get me pretty close to a zero load condition when not applying any heat with the power resistor (no static losses).

Of course this is just a dummy load, and only meant to characterize the heat load capabilities of the AC-2 Chiller.

Sorry for not posting any pictures, but I left my camera at home, and the shop camera seems to have disappeared. First thing tomorrow I'll snap some pics of the test load as it stands right now.

Good news is that I got everything done on my to-do list, so in the next couple of days I should be testing again. This time with the following charge:

4 oz R-600
4 oz R-290
85 psi R-170
85 psi R-14
12 psi Argon

This is an ozone friendly mix.

[n00b 0f l337]

Good
Hope you enjoy the r290
I wish I could pass you a cylinder of Ethylene (ethene) r1150 though. I think you'd love it.

[mytekcontrols]

I wish I could pass you a cylinder of Ethylene (ethene) r1150 though. I think you'd love it.

Can't you just email it to me?

It would be nice to try at some point, since R-14 is so darn expensive, and it is a green house gas. Although if it could be bought in the small quantities required for my unit it wouldn't be to unreasonable. Too bad the gas suppliers don't offer a system like the fill your own drinking water dispensers you see in front of the grocery stores.

[n00b 0f l337]

Well r1150 is easy to get, and you can get small quantities. I have 11lb and its going to last a while me thinks.

[teyber]

mytek:
local airgas carries it, not sure on tank $ or regulator, but for ME at least they carried 10, 35 and 200 cubic foot tanks. 10 was 180$, 35 cuft was 225$, and 200cuft was 310$...

ethylene regulator runs around 750$, so i think people just use hydrogen regulators with a cga350 adapter (feel free to correct me here).

Regards

[mytekcontrols]

Sorry I haven't been updating for a while, but I got sick last week, and it took a bit longer to run some of the tests.

The good news is that I was able to run several different charge combinations. The bad news is that the hydrocarbon liquids (R600 & R290) don't seem to play very well together in my autocascade. I tried using just the R290 as my lower stage refrigerant, but it just wasn't getting the first stage cold enough to do much condensing of the next refrigerant in the chain (R170). After nearly an hour, the first phase separator was only running at +10.4C. So I added R600 a 1/2 an ounce at a time. This did get things running better, but not anywhere near as good as using HCFC's for the liquids. Also I started seeing pressure and temperature oscillations, plus it wouldn't hold a load worth a darn.

After trying 3 different HC tests, I decided to give up on this for a while, and instead switched back to using R123 as my first liquid, but substituted R410a as my 2nd liquid (R22 HFC replacement). This gave me much better performance and stability, and nearly matched my R123/R22 charge results. The only problem is that I got c_ocky and added 5 psi more Argon this time, which didn't buy me anything besides higher pressures and slightly warmer temperatures, although I was able to load it up to 200 watts without the evap return going to hell (except the pressures were getting very high just before I shut down 70/365, middle of evap = -89C).

I will have to do some more tests with different ratios of the R123 and R410a, so as to tweak it in for the best results. This combination is better then using R22, since the GWP and ODP are much lower (virtually non-existent). Actually I really don't understand why R123 is to be phased out in 2015. It really is a very good refrigerant, and has an extremely low atmospheric lifetime as well. Anyway getting rid of R22 will allow new systems to be built based upon this design up to 2015.

Now if only small amounts of R123 could be purchased

Buffer Valve Results:

This worked very well indeed. I did tweak it up a bit higher to buffer at 370 psi on the discharge. Below you can actually see it buffering when I warm started the system. It did this quite a few times (just over 20 times), but after 15 minutes the pressures dropped significantly, and the buffer valve no longer needed to open.