My first Evaporator Design :) 'The Daisy'

So finally breaking that cherry and making my own Evaporator.

I really would prefer a maze design but the plates I want are too thin, so too many passes on a drill mill, too short a run for CNC, just too much cost.

But I think this will keep up.

Pics of the bits...
http://img254.imageshack.us/img254/1821/img0359g.jpg
http://img571.imageshack.us/img571/4273/img0360t.jpg
http://img853.imageshack.us/img853/1995/img0361cu.jpg
http://img59.imageshack.us/img59/390/img0362y.jpg
http://img543.imageshack.us/img543/6277/img0364h.jpg
http://img826.imageshack.us/img826/382/img0365oid.jpg
http://img213.imageshack.us/img213/4788/img0366ng.jpg
http://img192.imageshack.us/img192/1279/img0367w.jpg
http://img842.imageshack.us/img842/559/img0368do.jpg
http://img213.imageshack.us/img213/6828/img0370bt.jpg
http://img18.imageshack.us/img18/6560/img0371gb.jpg
http://img805.imageshack.us/img805/7489/img0358rq.jpg

Dimensions are 1.5" and 3/4" + 1/8 plate for the top.

Holes are 3/8", so the column plates end up 1mm to 4mm along with the outside being around 5mm thick.

Drilled it to a 3.5mm base with a home made bottoming drill bit (I'm pretty good at remaking drill bits) so it's between 3.5mm and 4mm base with a lot of columns or plates or whatever you want to call them.

It's 6 chambers. Slots at the tops of each for gas flow, and a plate installed/brazed for liquid input into the central hole.

Liquid flow into the chambers is a 1/8" hole each, and it's angle drilled so the liquid spray wants to go down and swirl first. Expansion will take care of it mostly, but may as well use velocity and direct it too.

Similar to the old Vapo/bowman/baker types for simplicity. Also using some variation of Dimas.

But the main thing is, all I need is a few wooden jigs, a drill press, dremel (I used a cutoff wheel but will set up a chopsaw with a jig for the o-ring slot.

No machine shop.

And pics of it brazed up...

http://img19.imageshack.us/img19/7979/img0373vf.jpg
Not a mirror yet 'til it's on and tested. It's a semi mirror finish, I find it more ideal for heat transfer than a full mirror but it's way cooler to send someone a unit with a mirror finish of course :D
http://img703.imageshack.us/img703/2223/img0374fo.jpg
http://img708.imageshack.us/img708/9474/img0375lp.jpg
Made the mount and plate for it from a thicker end cap and alu plate. Pretty basic, fast to make. The edging for the plate to hold down was awkward without a lathe but totally do-able. Saw, bench sander, hand file, fine paper. Took 10 min all told.
http://img841.imageshack.us/img841/1862/img0376p.jpg
http://img407.imageshack.us/img407/51/img0377e.jpg
http://img580.imageshack.us/img580/3526/img0372c.jpg

Hoping it works nice. I have tons of data from other evaps, so will try it and see. If it's as good (should be) or better(never know) then I'll just start making them when I need.

It's the time factor I guess. Cost isn't as much of an issue since the Under-the-Ice shop's chilly1 evaps aren't bad, but shipping, time to ship, etc. Just nice to have stuff here.

So I bought enough precut blocks to make 7 or 8. One'll be a throwaway most likely, just playing with different shapes, but the rest may actually be good.

The only took 2 hours to design and make the block and mount. Now I have a system, likely an hour :) Will also be prettier too, now I don't have to cut and recut.

So I hope it's good. I think it should be.

Will likely test it on the unit I'm making now, rotary single stage. I put a Dimas evap on it already, but now that I have one ready, may just braze another flex line and see how it works :)


Gray

Heh just as you post this I'm posting about picking up a Kingpin pot base to turn into one in the same style. And PM'ing gomeler about doing it with an internal manifold.

Good to know we're on the same page.
My guess is it will work quite well. Consider picking up a tap to add surface area to those holes. Other option is to extend the holes from the internal manifold center with some 1/8" tube, so that refrigerant hopefully doesn't come out of hole and shoot right up suction.

oh not gomeler! lol

darnit! i got some designs...but i can't draw 3d... and i'm an engineer lol

LoL.
Keeping up with Gray Mole is like trying to catch the devil these days ;)
It's a solid evap design. Looking forward to results.

I've got a foot of 2.5" diameter stock to play with....

How'd you manage the snap ring?

Nice one. Hope it is performing well.

For a nice mirror finish I used fine-grit sandpaper (1500-2500), finally using some kind of chrome polish.

Well it has been years I've been playing around, probably should have made one ages ago.

But this is the best I could come up with that I could do without a machine shop. I'll know later on if it's decent.

Oh, I'm using some rubber rings like the older chilly1 blocks did. I just used my cutoff (air) wheel to grind the first one. Later I'll make a jig for that and set up a saw. Much easier.

I braze that little plate in ;) so no chance of backflow. The plate wedges in really well, and I hook the captube to make sure it's the right height. I preassemble, then pull the top off and purge really heavy while the evap's open. While it's still hot, slide the top in place and braze that.

So brazing this one's a pain but not the end of the world.

I'm a bit nervous about threading the holes. I took the grinder to them to roughen them a little but a full thread could catch oil. I don't know if it would but if if oil slowly built up, could start to hinder heat transfer.

I normally just lap the blocks with 250 / 400 / 600 / 1200 and finish with some polishing. I thought about going higher with the paper but it's hard to keep that flat on the block.

Thanks Guys :)

Next time (when I run out of these) I may go with 1.75" (easy to get) or 1.625" (not often in stock) but I have a feeling these should be a bit better on the edges of the cpu heatspreader. I've been finding the steppers a bit weak to the outer edges and the intel 6 core stuff is pretty 'wide' under their spreaders.

Gray

Is there no one producing evaporators anymore? Cryostar?

Cryostar was one batch of about 10... then gone.
Chilly1's can still be bought, and Piotres has steppers, so does Kayl.

But nothing cheap n easy or really "different" since 2004.

Cryostar was made for GPU's wasn't it ?
Sure it had a nice design but not enough mass imo.

I am very happy with Piotres 42mm stepper. It holds huge loads on my Phase unit built by Ron, temps hardly warm at all even on very heavy loads.

For a Cascade his 50mm stepper would be great. Nice and big and lots of mass.

how big is that evap, is it a larger one like we were talking about with regards to coverage on 6+ cores? looks pretty big :D

it's gone together well, back to the old days with the drilling rather than machining!

Great job Gray!!

Dont you worry Martle.
I think we need to just get the push back in this sport. I was looking through the old stuff, back in 2005, there was a cascade A WEEK being built. EVERY WEEK. FOR 9 MONTHS.

We don't have that anymore. We've lost the creative spirit, and some of us have just gotten old (or moved to new locations).

One month till I get a 63mm evap or two done. I'm going to need some people who want to test them I think.

Well like I said, I'll go to 42mm at some point soon. This will cover the 6 core just barely and the biggest weakness of steppers or spirals has always been a lack of any mass to the outside edge. Piotres with the big stuff has gotten around that really well though.

I threw it on the system I'm making and wow!

On the new load tester with the probe that's in the middle of the cartridge, I've only tested a few evaps.

Duniek's 40mm. Old chilly1, new chilly1, and one Kayl.

So far with those, the Duniek was 42 degree difference. The old chilly1 was 50 degrees, new was 40 degrees, Kayl's 41 degrees. So really, around 40 degrees at 350w was the best of what I tested.

This one's sitting at 36 degrees difference :D

I didn't vac as much as I usually do, so just threw in some propane. Haven't done the captube mod either.

At 350w the evap temp was -20c. The resistor internal temp was +15c.

The actual temps aren't the issue, just the difference in temps is the goal.

So I've been talking to my brother about making some jigs up to get the chambers closer to what I'd like, so hopefully can get that difference lower, but 30 is my goal if I can get there, and really at 350w that's about the best case I think.

So I'm happy so far. Considering how easy and cheap this design is to make (other than that stupid little plate for the captube) I'm impressed. I had some thoughts about what would work better and I was right for a change.

Gray

Did you cross drill at the base in order to get feeds into each chamber?
It looks like you have feeder holes in the bottom of each chamber and notches along the top to suction out is that correct?

I think you might find it sucks out too quickly have you tried it yet?

Hey Scott :)

Yup, results for temp difference are just above your post. I didn't cross drill, the holes are angle drilled from the inside to the base of each chamber, angling down to get them spraying toward the base.

Pics are awful as always, but if you look close you'll see 'em.

Works great :D best I've tested so far with the newer load tester. The internal sensor (k probe) on the resistor gives a decent comparitive analysis.

It's only accurate to tests on this tester though. Unless someone had the same block/same cartridge it's meaningless.

Wish I could test more, but now I've got a plan I can just make them. Kinda wish I still had the @itor head for testing. That's one I never did try really high load.

I'm going to install another probe in the load block. it has a hole for a surface probe.

But so far it's great, I figured it would keep up :)

Gray

So I've done the captube mod, and a quick insulation job. Nothing fancy, just to get the results.

Took a few pics to give an idea of the temps, and it kind of shows a bit of the mod improvement too.

I didn't take the gas out when I soldered the captube since it's a lower temp than it takes to melt braze anyways, done it quite a few times now.

So the charge is the same as the previous temp results. Only difference is the cap mod.

Some pics...

http://img828.imageshack.us/img828/7043/img0378j.jpg
http://img716.imageshack.us/img716/8567/img0383ly.jpg
http://img594.imageshack.us/img594/3608/img0379u.jpg
http://img651.imageshack.us/img651/8075/img0382pz.jpg
system uninsulated, running no load

Haven't redone the charge, and it runs warmer with it like it is without load. Main reason is that it's not really 'no load' anymore since the cap mod preloads the system. Doesn't really detrimentally affect anything except a mariginal increase in extremely low load applications.

When you run a system without the mod, you get a really low temp with no load. When you add even 40w or so, it goes up to where it works in 'normal' operation.

So usually what I find is a system that's running say, -110c unloaded (cascade) it goes to -100c with even a tiny load. Then from there to 350w, it's something like -90 or -80c, based on the compressor's ability to maintain LP.

So the first 10 is really an artificial result. Not as dramatic in a SS but you get the idea.

So the load test after the mod....

http://img43.imageshack.us/img43/9808/img0380bw.jpg

Just a 350w test on this post.

It's still around 35 degrees difference, but with the mod done it's running 5c colder, even with the gas charge that's a bit on the high side.

So instead of -50ish unloaded, and -20ish at 350w, it's running -45ish unloaded, and -25ish at 350w.

From 30 degree delta on the evap to a 20 degree delta on the evap, no load to full load.

Anyways, the point of this thread, the evap. Around 35 degrees from the evap temp to the internal resistor temp. Not quite a perfect evap to cpu temp setup since a cpu has next to no mass so will 'block' less heat transfer, but it's just a way to see what the different evaps do for heat transfer. good way to tune a system too. Get the best temp you can on that resistor, and you'll have the best cpu temp you can manage.

Gray

No pictures of dry uninsulated? Or of capillary area?
What size rotary?

Insulation's not permanent so I'll strip it down again. I'll have to anyways for a couple things :)

But yeah, specs on this one...

Matsu:banana::banana::banana::banana:a 5/8hp or so, will grab the number. It's a small footprint version, 40lra so a bigger one.
7.5' of .031" captube
Condensor is custom, I chopped and rebrazed a 12k btu one into about 2/3rds. Whole thing fits the mach2 baseplate, have to heighten the rest of the frame and make panels.
Will have a pair of Scythe 120mm fans. Need to make the plate for that too, was just running with them against the condensor for the test.

Otherwise, just 3/8 return to 1/2" pipe for the captube soldered area and back to the accumulator.

Wasn't much getting into that part of it here, I'm going to make a thread on the unit too.

I'll post a pic of it dry without the insulation in a few :)

Gray

Ok, well more pics...

http://img825.imageshack.us/img825/180/img0111h.jpg
http://img641.imageshack.us/img641/4884/img0109ik.jpg
http://img38.imageshack.us/img38/7446/img0110yf.jpg
http://img684.imageshack.us/img684/7089/img0112ukv.jpg

So yeah, compressor's bigger but the small footprint smaller size. So it's a large version but has the usual slight weakness of an identical larger type.

Smaller ones tend to be a bit quieter though. And it fits the Mach2 base, which the bigger ones don't.

Info on the Matsu's is sketchy, but 40lra and likely the 12.6cc displacement in the numbering system.

Gray

http://www.panasonic.com/industrial/...11U126A-6B.pdf

Guess it's either a Panasonic rebranded, or Panasonic rebrands the Matsu's. either way, a spec sheet :D

Says it's 7300btu's or so. So around 5/8hp is right I guess :) Was wrong on the numbering system though. It's a 10.3cc displacement on the sheet.

Gray

Cool. No high side valve?

Nope. Most of the time I don't on SS stuff anymore, unless I'm testing something new. Condensing temp tells me enough and even that I don't need to know all that much anymore.

HS valves are the most prone to leakage, that's the only real reason I don't use them much.

In cascades I'll often not have one on the 1st stage either. Same reason, just don't need it.

2nd stage I'll have them for testing no matter what. Can't afford to not know what the HS pressure is doing.

I see a lot of the guys with no HS valves on SS. For one, they're a pointless cost when it's not really that critical. Second, the risk of them leaking if the cap loosens over time with heat and vibration.

Could seal the cap too, but like I said, after making a hundred or more units, you get to know what works. HS pressure just isn't something I need to know for sure on a SS.

Edit: I should say that anyone new to this stuff, you really should have HS valves. You need to learn about the pressures and confirm HS reactions, pressure spiking, and anything else related to it. Getting to know HS pressure is as important as LS pressure. When you get to know how the compressors work, what their range is from high to low, etc. then you'll get to the point where you likely won't use them either. For yourself, they great. For a unit you send away, they're a bit of a liability.


Gray

Heh on small singles I gave them up as well unless the filter/drier had it built in. Then after tuning I'd just pinch it off with my pinch off tool.
But on rotarys never know and like to know.

Cool though, unit looks nice.

On a single stage, whether Rotary or Recip, I find only 2 things truly relevant to safe operation.

1. Condensing temp under load.

If your system is capable of maintaining close to ambient condensing and your ambient is reasonable, then there is no serious issue with any gas within the realm of 'normal' application. R410 is beyond that, so would take careful consideration and require pressure safeties. Otherwise, even 402a would be completely reasonable when maintaining even 5 degrees above ambient condensing, when your ambient is 30 degrees or lower.

Above that, professional discretion is always used. Condensing temperatures are no replacement for pressure readings when you live on the edge.

2. Compressor operating temperature under load.

There are specifications for the limit of safe operation and they are higher than I would ever consider allowing a compressor to operate at for any length of time. For example, a Rotary (this one, in fact) can operate to a maximum winding temp of 130c. That's higher than I consider reasonable, and beyond that the heat dump is detrimental to optimal evaporating temps, especially where cascades are concerned.

I see 80c as the high limit. I also see 60 to 70c as reasonable, and 50c to be optimal. On most of the 1st stages of the cascades using Rotaries and SS with them, tuning with best condensing temps and good return cooling gives no higher than 60c under load. Some as low as 45c. That's the shell temp though. I would think the winding temp would (or could) be marginally higher since the return (at least in the systems I'm making this way) has the mass and temp to maintain a lower temp. I haven't seen a compressor with a temp sensor on the windings inside.

So, if the condensing temp is at ambient at the filter, and the compressor shell temp is under 80c (preferably by a lot) then a HS valve is pointless in my systems, since it gives little information I haven't already gathered in previous systems.

But that's just me.


Gray

Great news on the evap!

I use steppers and sometimes Chilly1's still so its nice to see a new design that works.