High load Vapo mod Guide

[Gray Mole]

So really, any of the older units could be redone with this mod and have it work for high load.

This one specifically is a 'vapo' build, though there's not a lot left of the original parts.

The compressor's now a NL9 instead of the original NL5.5 and the temps are better because of that, but the theory applies universally to any retail or custom cooler that's specced out for a 200w load.

So, for the mod, you'll need....

A vapo, mach2, or any cooler (of course)

2' of 3/8" copper tube
2' of 3/8" id tube insulation (13mm wall recommended, 10mm wall ok 6.5mm wall a bit thin, but still works better than not having insulation)
7' of .031" captube (unless the unit uses that or larger size already)
1 additional condensor, similar to the original one in the system.

Examples. Both around 120mm for ease of install.
1 additional fan, if you don't have a rear fan in the cooler (vapo's do)
1 new access valve (required)
1 additional new access valve (recommended)
Swagers or copper joiners in 1/4 and 3/8
or
3-4" of 1/2" tube (optional)
Some Rivets to refit the top of the case chassis (if applicable)
1 new filter dryer (highly recommended)
So the cooler's original configuration has a bit of space wasted, and there is no room for wasted space, so you will likely have to cut the parts on the original build. Recover the gas, and chop it up with long pipe 'tails' to make it easier to rebraze.

Lay the parts out so that you have a clear idea of the length of the new build, and that it will fit the case if you're doing a Vapo/Mach2 mod. Some custom cases have more length, and open builds won't matter, but it's a tight fit in a case, so it's better to know the length and size for sure.

When you know you're good for that, recut your pipe for the new brazing, and either use a swager to rework your pipe, use joiners for it, or (in this example) you can just cut very short sections of 3/8" pipe to join the 1/4".

For the 3/8" line (or 5/16 in some coolers) you will need the same gear, or you can chop short pieces of 1/2" tube to make joiners. Up to you.

So set up the lines and if you're doing this out of the case, use something to set the condensors to the approximate height you'll want them. Saves bending the pipe later, though 1/4" bends pretty easy. The condensor tubes are a lot thinner though, so it's better to just have it close to where it needs to be and not have to bend it much on the final install.

If you have it in the case, and the condensors are premounted then it's easy to braze in place. Just depends if you have the case apart.


So when you're brazed on the condensors (purge while brazing) you should have the compressor to one of the condensor's top inlets, then the bottom of that condensor into the inlet of the next, then to the filter dryer.



It's not critical which condensor you go to first. That is more about how your install is, where the best cool air will come from, how your fans run etc.


I strongly recommend that the air comes in both ends, so you should switch the rear fan to 'in' instead of 'out'. That way they both get cool air, so you'll get the best capacity out of them.

Vapo Cases have side venting so you don't need to cut blowholes. You can, if the air seems restricted (you'll get a lot of wind rush noise if it is) but it should be adequate.


For cases with no holes, you'll have to do something about it. On this one I just cut 2 80mm holes, and rivetted on some fancy 80mm fan guards. Open holes and good airflow, so I didn't add fans there. You can if you like, but more so if you can find room


So that's the condensor portion of the mod.

The suction line is a bit short, and the accumulator on Vapo/Mach is a bit small, so adding some volume there is recommended or you can get floodback at very low load.

So, you'll need to see where and how you have room for that.

Bend the 3/8" tubing into line that will replace the original suction basically adding as much length to the build that will still fit the case. I added around 18" of pipe to this one.

As long as it fits, and doesn't jam up into anything, it's fine. So you'll need to chop the suction, and 'plumb in' this new piece. Insulate before or after, depending on your skill at not burning the insulation

The captube mod can be anything that works out to 300+ watts, but in this case I use .031". It's common and easy to get. Any size is ok really, but .028" and smaller sizes end up pretty short. Vapo and mach original captube is usually .028 as far as I know.

So either you live with that, or replace the captube.

If you're going to just work with the original stuff, the results are a bit annoying, longer cooldown and such. And the slhx mods become a requirement at that point, the cap just needs the help.

For the replacement of the captube you can just 'tag' in the new stuff.

Measure your captube to around 7' of .031" if that's what you're using.

Cut the captube 4-6" away from the evaporator.

Fit the new captube to it, and cut a short piece of 1/4" pipe.

Crimp it around the 2 captubes, careful not to shut them, and slide them against each other, then pull 1mm apart so you're not blocking any flow, then keep pinching the flat of the 1/4" 'tag' until it's sealed pretty well.



Braze, but purge as you go, so that the flow isn't blocked when you braze, and of course, no crud in the tubes.

After you're done, you'll have a short piece of smaller tube, but it's not a huge concern. You can go shorter on the piece from the evap, but that little extra length gives you a second chance if you plug the captube. You can come down to 6.5' of new captube if you're going longer on the original stuff.


You can use any length you like, but this mod used this length and I know it works well for 375w.


Run your captube down the line as normal, wrap on the flex, then however you want/need on the solid line.

Braze into your filter dryer as normal. You may want to do that captube mod while you have the line chopped out. Easier to insulate if you want to use non split insulation. You can split it and either use tape to refit, or rubber cement works well to join the split insulation too.

If you're using the .028" captube, you'll have to come down to 5.5' at least to hold high load. 5' is really the limit of how short any captube should be, and any shorter can give serious grief. I haven't tested 5' of 28 and don't recommend it much, but if it's what you have, you can try it, and see how much capacity you can get out of it. you should really uncoil the captube, measure as close as you can, and chop it down. If it just won't give a good result you can always replace the captube with another size, as stated.

If you're modding a Mach2 with it's original head, I don't know it's capable of working with high load. I've never tested it and it's pretty awful really so I do recommend a new evaporator. If you're replacing the evap, then just running a new length of captube is easy.

On existing evaps of good quality and original Vapo evaps (that are actually quite good and should work to 300w), tagging in the new captube is a much easier option than drilling and replacing. there's just too much risk of copper shavings in the evap, and cutting open the evap and reworking it is really more hassle than it worth. 4" of 28 is similar enough to 6" of 31 (give or take) so you'll end up with around 7.5' of total .031 captube when you're done, and that's working for me as a 350w (give or take) capacity mod.

So you're going to look something like this when it's done.

Once you're assembled and brazed, you could consider some slhx options.

There's usually no room to put a real suction line heat exchanger in a case, but you can enhance the heat transfer of the captube with some soft solder, or in many cases, just tape and thermal paste.

I've set up accumulators like this.



Or you could look at this variation of captube on the solid suction return pipe.


Or if you aren't a good plumber, or don't have soft solder and don't want to braze the captube to the return pipe, taping with paste is also an option, if you have a nice big tube of cheap paste. Brazing or soldering gives much better heat transfer though.

So that's the brazing done, the slhx mod in there, the new condensor added.



Or this, with an insulated slhx in a small package.





Vac and charge.

I recommend a load tester to confirm, but if it's very close to what I've done with this one, it should end up over 300w.

On charging at no load and ambient of 22c, I charged it with a static pressure of around 40psi.

Then running with the fans and insulation cut back to monitor, I ran it unloaded, slowly charging until I was frosting to, but not into, the compressor return.



This is pushing just a bit, prefer if it's not frosting the compressor unloaded, but gives an idea. This one wasn't flooding back though, thanks to the accumulator/hx mod enhancing the heat exchange and making sure it's just vapor coming back. Frost doesn't mean liquid, just means cold. But it's more ideal if the frost is just at the compressor instead of freezing it just a bit.



That gives you a 'ghetto' method of charging to your max capacity that won't flood back.
If you are flooding back you'll hear the 'popping and clicking' noise that goes with it so it's pretty easy to tell.

I did that on this one, and I got a little over 375w of capacity, which should be adequate for 6 core full load benching. Small changes to the numbers I've listed should give a very similar capacity, but size differences in the compressor, or changes to the length of suction return from what I've stated with increase/decrease your capacity. I prefer a system that won't flood back even with no load, but some think that's ok if it's just a 'benching system' and will always have load when it's running.

I'll leave that up to you, and will only recommend that if you don't quite have the capacity you need with a maximum charge that does not flood back, you may have to add more suction volume (longer pipe, larger or additional accumulator) or further enhance suction line heat exchange.

That's the works. Takes a lot longer to explain than to do the mod, and even if you don't have a load tester, you can do the final charge to max, and get a good result.

You'll know for sure if you have enough capacity to run your 6 core cpu when it's attached, and if you're looking for the 'best' result you can tune it right on the cpu. Just be careful if you're slowly adding/removing gas so you don't spike the temperature on load and kill the cpu.

No 'guarantee' that this mod works, but it works for me, every time, to get 300 to 375w of total capacity using the captube and condensor mod, with the higher capacity range more likely doing the captube heat exchange mod along with it.

Hope it helps whether pro builder or amateur, the mod is ideal and the vapo/mach2 coolers can still perform as fairly quiet 24/7 units.

I've added some pics with compressor temps at time of load tests, to give an idea of compressor temp and strain.

Started by running 3 hours at 375w, then changed every half hour down to the next lower wattage to ensure full temps on running tests.

Temp results for the 1/4hp unit were -20 evap 350w -40 0w
Temp results for the 1/3hp unit were -25 evap 375w -45 0w






Both tested with R290. Ymmv using other gasses. 290 works well since the condensors are smaller and 290 is light so head pressures and condensor heat is lower. Better condensors would make colder gasses easier to work with.

This mod with the condensor I used gave about 3-4 degrees over ambient on the filter (won't bother with subcooling figures, for those that don't understand them) and the closer to ambient condensing temp, the better your system will work with higher capacity and colder evap temps. So a colder gas may look better, but in actual use it may perform poorly. This is how this one works with this gas.


Gray


Edit, a few more pics added, will try to finish up asap, and keep replacing with better pics as I go.

If anyone has some good pics of their version of this kind of thing, feel free to post No threadjacking here, anyone willing to share their tricks (or secrets, if you think about it that way) is more than welcome to throw in.

[Gray Mole]

Added a few pictures, starting to get a decent amount that the words might make sense.

Added load test results, with the laser temp reading the compressor too. Gives an idea of load and strain, also indicating the compressor is at full temp running.

Did the load testing 'backwards' from high load for 3 hours, then stepping down. Then, if anything, the temp picture is going to be warmer at a given load, instead of colder, because it's creeping down instead of creeping up.


Welcome to add pics and info if you have a way you like better for upping the load on the Vapo and Prometeia stuff, or just for old custom units that need a serious upgrade for load.

Gray

[sdumper]

Send it to me and let me load test it for you.

[Gray Mole]

Hey Scott

Bit of a pricey shipping cost there but it'd be cool to get a second test I guess.

I'm setting up the new load test station soon, so I'll have another setup to confirm things too.

Don't know if I'll buy the same power meter, or go with a KillaWatt.

There's another one at the place I bought this one, that's set up really nice, but it's 100 bucks, so a bit more extravagant.

But it's cool though, thanks for the offer, but the load testing stuff I have here is fine. The block itself is fugly, but it's insulated so well that I'm getting as close to zero heat loss as I can with aluminum resistors.









Those are some pictures of the construction of my current load block.

The resistors are fixed with thermal epoxy to the copper plate, and it was in a vise overnight to ensure full heat transfer.

Otherwise, the power meter just tells me what my load is, the volts, amps etc.

Do you have any pics of your test setup Scott? Curious what you're using, if it's more accurate than mine I'd be tempted to try to get something similar set up


Gray

[Gray Mole]

Oh, just a little disclaimer here.

I'm not posting this to get people to use me for this mod. Faaar from it

I used to a lot of modding and custom stuff, but now (whenever possible) I just build a unit here and there and it ends up in the Classy's.

This is just to try to help guys get through the mods, and give a bit of info as to what I've found that makes a good result in what I've built.

The details of the mod, the captube changes and condensor stuff, that works not only in a mod to an old unit, but you can also pick out some of the details of the build for your own custom work.

Having a large condensor is a big improvement, and staying close to whatever ambient you have at the filter will really lower the high pressure, and therefore bring the low pressure down as much as the unit can go, which gives the best temps you can get from whatever your build is. Best method is to just tune to the high load and run for a few hours, and then start seeing where you stand at the filter. If you can stay around ambient, or even 5 to 10 over at very high load, it's going to keep the high pressure AND return pressure lower.

The captube stuff, or using a real suction line heat exchanger, can do a lot to get better results at high load, but working out exactly how much heat exchange, and for the purpose you're after (whether to enhance the vapor reaction in the return, or to enhance the subcooling in the captube itself, or a combination of the two) is up to you, and as you keep playing you'll notice that you'll find a great balance of it. the spiral wrap does more for the return vapor and load response/capacity and the single string method over a long suction pipe tends to be more of a captube subcooling boost.

So if you take from this guide what you need for whichever you're working on, you may find that some of the tweaks work well to make it work better, even from scratch.

Just like overclocking and benching a PC, all the little tweaks you can find for refrigeration can really make a unit perform, and there's so much more information for PC cooling units now that they can be much better than they used to be, just try to do as much research as you can.

Spend a week tuning and retuning a unit with as much of the type of things I'm trying to do, and add some of your own. Some things will fall flat or just do no good (I've had a lot of things go that way ) and others will make really nice 'jumps' in the unit's performance, and that's what you carry to the next build with you.



Gray

[Gray Mole]

I forgot to mention, there is the choice to just buy a much larger/better condensor instead of adding a second.

That choice is a mixed bag. There are some 1kw (1000 watts) condensors available that you can squeeze in but the rating may or may not end up working as well as this.

The reason for that is, just like a water cooling radiator, the capacity is based on a certain amount of airflow.

A condensor rated to 1000w is always for a given amount of airflow, and less will reduce that, sometimes dramatically.

If you want a very quiet cooler, especially where the Vapo (NF5.5 Danfoss) is concerned, you've got a really quiet compressor, and probably want a set of quiet/low flow fans to go with it.

If your condensor is rated to 1kw at 150cfm, for instance, but you run a quiet fan at 50cfm, you're going to see a reduction that's somewhat linear. Even if the numbers aren't perfect but close, that only net's you 350w or so of capacity on your condensor.

When you consider that the compressor is throwing maybe 100 to 150w in there (just a guess based on what I see for results) then you're left with a 200w max on the evap, without seeing a major increase in the temp at the filter.

That's a big part of where the twin condensor set can work better. The original condensor isn't 'terrible' per say, but with low speed fans, 250w is about where you stand. So that leaves 150w at best before the temp goes up, sometimes less.

Adding a second condensor gives a full boost you don't have to take the compressor heat dump into account for. Whatever you add, you get.

So even a smaller one, like the one there, should add 150 to 200w more capacity with low airflow.

That gives you around 300 to 350w before that all important condensing temp goes up. That's where you can keep the low side/suction pressure lower, and always get the best the compressor can give.

I did notice in the unit posted tests, the compressor temp spiked a lot over 350w, and the condensing temp was only 2-3c over ambient with 350w, where it was more like 8-10 at 375w.

Gives you a real way to measure where you stand for condensor capacity, by it's effectiveness in the system. The actual numbers you work with are important, but if you notice that temp rising significantly, you know your condensor capacity is 'giving out' for the load.

So you have the choice of adding more to the capacity, or increasing the airflow to make the condensor more effective. Once you get to the condensor's full strength (usually where they've rated it's max airflow) then you've hit that wall, but having a 150cfm fan isn't what you want on a low noise 24/7 cooler.

Hope that clears up a bit of why I took this direction.


Gray

[runmc]

Send it to me and let me load test it for you.

Scott!!!

[Gray Mole]

Oh yeah, I wanted to mention something and really, it's why I started looking at why my coolers, well...they sucked at high load before.

Ashrae states that a NL11F or NF9 or really most of the 1/3hp compressors were rating at -23.3 and 375w. Some higher some lower, but really, at that load, with 134a, it should be theoretically possible to get that capacity running at that temp with that gas.

It's a scientifically proven fact.

So I started to look, and the last 3 years while I was away and working on commercial stuff, I found out.

So working out that the compressor is working 375w at -25c, but with R290, means I'm still not quite there, but I'm getting close to the best this compressor can give.

R290 doesn't have the mass of 134a, so I can't just work out the LP and apply it directly to temp, which would net -30 or so. It's too light for that, so I could say that maybe -27, maybe -28c would be the 'perfect world' result.

I'd need a better condensor setup for sure, these are a tad small to get the condensing temp down to ambient 100% of the time at that load.

But I think this is getting close to where it's about right.

If a person could get enough room in a vapo case (proabably have to chop and change a bit) and have 2 larger condensors, enough to get the full power with quiet fans (at least in a 24/7 cooler) then you could get a fair bit better.

I think -30 with the right gas would be possible at 350w to 375w. Considering -23 is around 2psig with 134a, change to R507 and keep 2psi and you could get -44 (we're talking perfect world here, though potentially a small loss of compressor efficiency is possible with higher pressure gasses that aren't specified to a compressor) but -40c should be that 'target' to keep fighting for.

Finding an absolutely perfect captube, and upsizing the condensors to get that ideal temp at the filter, as well as enhancing the catube condensing efficiency to assist with that, is where it's a challenge. But it should be possible, and I'd like to keep trying.

My numbers and ratings could be off slightly, I keep seeing different Ashrae ratings for the same compressor from different suppliers and Danfoss's site kinda sucks a bit for the fine details, but if a compressor is able to do a specified temp with 134a, it should be possible to match or beat that with a colder gas.

It's not wishful thinking, it's really quite possible. Just need to figure out a few more tweaks to get the captube perfect to the load, the condensor power over the requirement for load, and the volume and heat exchange ideal for the full range of load.

Everywhere you lose efficiency in the system is going to cost some degrees. Let's try our best to find them


Gray

[Ian[Frozen]]

Well the timing is great! I'm about to attack my old VaporChill this week hopefully. It has an NF9FX installed however due to a lack of capacity and as it turns out a small condenser leak is incapable of cooling the higher load of an i7 whose temps skyrocket beyond a certain clock speed (seems almost regardless of vcore)..

So the old condenser is destined for the trash then a higher capacity condenser (thank Ron) and a change of gas..

From this great read it looks like I should extend the cap tube as well??

[sdumper]

---
Sorry Ron

[Gray Mole]

Well the timing is great! I'm about to attack my old VaporChill this week hopefully. It has an NF9FX installed however due to a lack of capacity and as it turns out a small condenser leak is incapable of cooling the higher load of an i7 whose temps skyrocket beyond a certain clock speed (seems almost regardless of vcore)..

So the old condenser is destined for the trash then a higher capacity condenser (thank Ron) and a change of gas..

From this great read it looks like I should extend the cap tube as well??

Yep should work pretty close since thats the larger of the 2 units

If you have the original .028 captube then it's better to replace and with this it's 7.5' of .031.

That's with r290 and if you used a higher pressure gas like 507 then leaving 6" of 28 and adding 8' of 31 may work better to allow you to tune it to your own best length since different gasses like different lengths but just settling on 7.5' of 31 or the equivalent in another size should do fine.

10.6 of .042 is what Scott said he liked and ron was at around 9' of 40 so just reduce whichever one to compensate for the 1/2' of 28 you're joining to.

What consensus are you adding on yours? Or are you going with a larger single? Better the condensor capacity you have, better this mod works overall but remember you need as much suction volume as you can pack in there so you aren't flooding back with low load! Soldering the cap helps but you really need that volume increase.

Mod works best with all these done together whether it's my way or your own variation of it

When you're done, post it here so we can see how it worked for you

[Ian[Frozen]]

Condenser upgrade (replacement) with this
http://under-the-ice.com/popup_image...ID=144&image=2
Seems like condenser replacements to fit a confined space are hard to come by

H'mmmm, liquid flooding I can see being an issue (for the compressor) so was wondering about how to construct an accumulator?? Or simply wrap the suction line with the capillary tube as illustrated here...?

[Gray Mole]

Both if you can manage

Capo and Mach units both have a small accumulator and a 12-16" addition of 3/8" on the suction is great to take care of the flooding with the added gas volume.

That condenser will be ok just getvgood air through it. If you could squeeze both that and the original (very tight) you'd be able to run very quiet fans

[Ian[Frozen]]

Well how about we just ditch the air cooled condenser altogether and fabricate a liquid cooled condenser since I already have an abundance of chilled water cooling capacity, an excess from the video card chiller... One of these would be the most logical step up in cooling capacity http://www.xtremesystems.org/forums/...29&postcount=2 and easy to fabricate..

[Gray Mole]

Yup, it's all in the space you have and how small you make it but that's a thought.

There's a guy selling smaller plate exchangers in the US, link's in that other Plate HX sizing thread.

They're just small enough to fit a Mach case I think, but too tall for a Vapo.

That, or buying a 3-4" pipe and doing up a coil (like a desuperheater) inside would work well too.

As long as you have the capacity to keep the subcooling high at the filter, you're set. Maintaining that close to ambient makes a system perform well.


Gray

[Ian[Frozen]]

I've been looking at the construction methods of commercial desuperheaters and by comparison I'm thinking they may have a higher surface area for heat transfer given same length of tubing, a home grown version may need extra length to compensate yet should not have much difference in coolant pressure drop by comparison.

Ensuring refrigerant condensation in approximately the mid-length point of the coil appears to be the key then in sufficient sub-cooling at the coil exit. Given my noobish experience here I think I'm going to have to wingit!! So on the basis of these great articles by bazx and the inspiration from Gray Mole, I'm thinking of an extra two coil loops and connect the coolant in a counter flow manor.

Now i got to go buy some more tubing

[Gray Mole]

Now i got to go buy some more tubing

Never said it free just cheap.

Go get a few free water coolers and clean up the 1/4" copper they wrap around the water tank.

People are usually selling them for 10 bucks or they're free.

You'll get a small compressor to play around with too

[Ian[Frozen]]

Forgive my noob thinking here, but how come we don't tap some of the cooling from the suction line to help sub-cooling efficiency of the condenser and in the process prevent liquid return to the compressor??

Just thinking out loud here...

[Gray Mole]

Yeah, a small suction line heat exchanger, same idea as in a cascade.

I posted that as an alternative. That's what's under the insulation in the original 'twin turbo' thread. There's a picture of it torn open in that thread. Still haven't finished pics on this one.

Tough to squeeze that in, but do-able for sure.

I used 3/8 in 7/8 I think? About 6" long. Great method of increasing potential refrigerant charge and capacity

Gray

[Ian[Frozen]]

Wait... I guess wrapping the cap tube around an accumulator does the same perhaps?

[Gray Mole]

That one, I added it to the guide up there too.

And yeah, that's part of the reason for the soldering or brazing of the captube.

Adds subcooling to the refrigerant in the captube, enhances superheat in the suction return.

Basically helps it condense and makes it boil off faster and more completely.

A captube soldered to the suction is a suction line heat exchanger. Just less pronounced an effect.

Wrapping without anything to 'connect' it does a small amount but imho just not quite enough to really make the difference.

I'ts NOT a replacement for suction volume though, it just makes better use of the volume you have, so that's something you really need to do along with the slhx mod.


Gray

[Ian[Frozen]]

cool thx

Plumbing store is closed now so will have to go buy tubing tomorrow and then stop procrastinating...

[Ian[Frozen]]

ok another noob question.. Are there contents of a dryer which can be damaged during brazing???

[Gray Mole]

Depends on your brazing

Short answer is no. Obviously if you're purging you won't be dumping a bunch of crud into it, and it takes a lot of heat to burn out the dessicant. Actual numbers on heat damage to a dessicant are available from the manufacturers, but they make it so you can braze it.

So no, brazing isn't going to damage them.

[Ian[Frozen]]

Thx again.
We'll see if I can remember how to braze... Not ever having brazed up a dryer before I just didn't know if we had to wrap the thing in wet rags in order to protect the contents...