We three Kings...

[Gray Mole]

Well I'd mentioned in some other threads I was doing a few Mach2 mods, so I figured in keeping with the season, I'd name the thread aptly :rolleyes:

All 3 were basically the same mods, Desuperheat coils, Chilly1 head(new revision) and 402a.

That's as far as the similarities went, as only one of them was straightforward. I'll just do a piccy set on that one, as it's probably the best example of the mod anyways, but the overview of each one is...

1. Came to me a little busted, the threads had stripped out and the head had been pried away from the plate, so the metal stock socket bracket was stuck to the stock head. Suction line was cracked either in shipping or when the head was pried away. Got a new suction line, and the rest. He also requested the ability to run it reasonably quiet, so it got a heavy addition of R290 and 404a with the 402a, so the blend was approx. 50% 402a, 20% 404a, 30% 290. It worked pretty well, and the idle/load was around -56/-41 on the fluke, showing about -59 on the display under load. This one was tuned and loadtested at 1800rpm fan speed set with the mach's controller, so not silent by any means, but still a nice temp and reasonable noise levels.

He's already gotten it back, but is going away over Xmas, so he's only had it on and in Bios, and the indicated temp was about -30 on his DFI. I'll have to wait a couple of weeks for the real results.

2. Came to me bought used by the current owner. Said the USB wouldn't connect, and I confirmed that it was bust. So instead of the expensive Control board replacement, we went with rewiring to 240v fans and a fan header for signal to the control board inputs to trick the control board into believing it had fans. The fans used were Papst 90cfm jobbies, same dimensions as the stock Panaflo's but a little louder than they would be on full speed, but with more airflow. Otherwise the same mods, but with a full 402a charge with a couple squirts of R290 to make sure the oil flows nicely. Was the best of the bunch temp-wise, at idle/load of -61/-46 on the fluke, and -81/-66 on the display. A little fan power goes a long way ;)

3. Was already 507 regassed, but wasn't happy enough with the temps, so decided to get Chilly's delicious evap and 402a. Was easiest of all, with no complications. Got the same full 402a/R290 charge. Temps are as shown in the pics, at idle/load of -60/-45, display of -81/-65. Fans on 'Turbo' I think.

All were loadtested with my resistor, and I've gone through being unsure of it's exact heatload in other threads, but comparing it against a few processors, it's higher than the 3000venice, about equal to the FX-55, and not as hot as the 4800 X2, all very clocked and under load. I estimate the 150w mark, but like I said, not 100% sure. No load tuning was to a superheat of about 10-15 degrees and the estimated overall capacity is about 190 to 200w for each unit.

The piccy's of one of the units... :toast:

http://www.imgbox.co.uk/uploads/5edc685004.jpg (http://www.imgbox.co.uk)

http://www.imgbox.co.uk/uploads/40a2dfeab6.jpg (http://www.imgbox.co.uk)

http://www.imgbox.co.uk/uploads/59f6bd445e.jpg (http://www.imgbox.co.uk)

http://www.imgbox.co.uk/uploads/bb702321d3.jpg (http://www.imgbox.co.uk)

http://www.imgbox.co.uk/uploads/685abdcd44.jpg (http://www.imgbox.co.uk)

http://www.imgbox.co.uk/uploads/d540d3cadf.jpg (http://www.imgbox.co.uk)

Yeah, I'm hoping for a new camera for Xmas :D

I wanted to add a thanks to Chilly1 for the heads, I should have done a few pics on the brazing up, but I was a little rushed to get all 3 done so they would be able to get 'em back before Xmas. The new head is amazing, and that's after working with the old revision, which was great. This one's far better. Brazing it is tricky, but well worth the effort, cos I doubt the results would be this good without it. Cheers, Reg!

Merry Xmas everyone! :toast:

[johann]

Some good work there Gray, keep it up

[wdrzal]

Should have 3 HAPPY Kings !!!!!! good work!!!!!!

[jinu117]

Looking good Grey :)

[MeltedDuron]

very nice Gary :toast:

[catkicker]

Gray Nice work. With mixing different refrigerants such as 50% 402a, 20% 404a, 30% 290. are there issues with oil and refrigerant compatabilities?
Looking at charts it shows 402a with AB or MO and 404 with POE and I know 290 is ok with all oils. I notice alot of people adding a squirt of 290 is that soley for the purpose of improving oil flow?

[Gray Mole]

Thanks guys :)

I hope they all love 'em as much as I enjoyed modding them.

The R290 in the 2 'Pure' 402a units is solely to improve oil flow. There's not enough to hit the performance of them in there and I just don't like to risk 100% 402a as the oil in the Mach's is POE and it's not going to carry in 402a as well as I'd like. It still should carry it a little with the 290 component of 402a, but at 2% it's a little less than I like, so I put a little extra and forget about it.

In the unit that's using lower fan speed, it's mostly to pull the pressure down so it's safer to run at the lower speed. 290 has a decent NBP, but it's mass is half of what 402a or 404a is, so I used a lot more to bring it down. I used a mix of 402a and 290 in my Mach1, but I found that it brought the boiling point and capacity down further, so in this one I used some 404a to try to get the capacity back up without the pressures of 402a. Blending is trial and error at the best of times unless you're a very well experienced phase guy, and for the likes of Chilly1 who's blended a lot and knows the gasses and how they mix it's a lot easier, but I've ended up with a decent mix that's holding the load, at a decent temp, with a pressure/temp on the discharge that's manageable with low fan speeds. Probably just got lucky ;) I don't recommend blending gasses for anyone who's not all that experienced unless you're willing to risk having to dump the whole charge into your 'recycling' bottle and start over, and that can get expensive in wasted gas and recycling fees.

Cheers

Gray

[kayl]

thats some nice work there ya doing for the UK garry, keep it up

[Firzen]

great work gray :)
the desupheater is very nice .. did u use a 8mm pipe there ?

[Gray Mole]

Thanks Kayl :) Hopefully the guys that use 'em show the results!

Hey Firzen :) I used 1/4" pipe, metric sizing isnt' something I've found out here in the UK, but that's about 6.5mm pipe or so. The standard discharge is just slightly smaller so it shouldn't hold anything back I wouldn't think. I don't know if I could get quite a tight enough radius to fit 5/16"/8mm pipe where I did, and this one seems to be doing the job ok. it was hard enough to make the 1/4" pipe bend that tightly!

Cheers

Gray

[placebo]

Yeah, I'm hoping for a new camera for Xmas :D

Merry Xmas everyone! :toast:

Yeah, been wishing you a new camera for a loooong time.... lol ;)

Good stuff/info with the mods m8! :)

[SoddemFX]

Great work Gray :)

What and where are you using for the flexible suction lines?

Tom

[Gray Mole]

Hey Placebo :D remember my skills are just as bad as the camera so even a new one may not get much better pics lol

Hey Soddem :) thanks, I'm really happy how they came out!

The stock replacements I get from Chilly1, as I'm not sure where else to get 'em to be honest. It would be nice to find something local, but I haven't found any yet.

I know that Johann can get it, so you could ask him about what he'd get you some for.

For braided I use the ones from BES unless I'm planning an order from Chilly1, they're not really that long at 600mm max, but doubling up the 450mm would work I suppose. The ends are annoying but for less than £10 they're not bad I guess.

Cheers

Gray

[Icy]

geez £10 is soo cheap :D Btw great work on your systems. They look amazing. How long is your desuper heater and how tall is it? as im kinda used to seeing them short and fat lol

[chilly1]

Great work ... Keep it up..

[axion]

Looks good :D

[piotres]

Yes new cam will be nice :hm:

LOL that 1/4' desuperheater's pipes look so thick that I have thought that they're 8mm too :wierd:

Permanent question ;) : how long captube in version with chilly1 block ??

Good work :)

Regards
Peter

[Gray Mole]

yeah yeah everyone go ahead, diss my crappy old digicam :p:

The pipes go just a little oval when they're wound that tight, which is why they look a little fat, but not enough to reduce flow any, thankfully.

The Captube is about 2-3 inches shorter than stock as that's about what got taken out in the process. I don't know offhand what the captube size and length is on a standard Mach2GT, but assuming it's .028" and meant to be good for about 200w with 404a, I'd hazard a guess at about 7'. I've heard that 9' or so is the stock length, but it seems like Gary Lloyds captube sizing guide would put that at about 165w or so capacity IF .028" were being used.

If someone knows for sure then it would be nice to know, but that wasn't really all that much of an issue to me.

I think for the next units, I'll be replacing the captube on all of them with around 11' of .031" captube, and wrapping the captube more around the suction line, and possibly replacing the suction with 1/2" as well. May end up with better results that way, and then I know for sure where I stand with the captube.

Gray

[placebo]

yeah yeah everyone go ahead, diss my crappy old digicam :p:

The pipes go just a little oval when they're wound that tight, which is why they look a little fat, but not enough to reduce flow any, thankfully.

The Captube is about 2-3 inches shorter than stock as that's about what got taken out in the process. I don't know offhand what the captube size and length is on a standard Mach2GT, but assuming it's .028" and meant to be good for about 200w with 404a, I'd hazard a guess at about 7'. I've heard that 9' or so is the stock length, but it seems like Gary Lloyds captube sizing guide would put that at about 165w or so capacity IF .028" were being used.

If someone knows for sure then it would be nice to know, but that wasn't really all that much of an issue to me.

I think for the next units, I'll be replacing the captube on all of them with around 11' of .031" captube, and wrapping the captube more around the suction line, and possibly replacing the suction with 1/2" as well. May end up with better results that way, and then I know for sure where I stand with the captube.

Gray

92 inches of .028 cap tube... ;)

[placebo]

Btw, what's the manufacture of your current camera? Next time I go out and by mine I will stay away from those guys, for sure...! ;):D

[Gray Mole]

Cool, so after these mods, about 7.5' then.

I noticed that the no-load return temp was a lot higher, roughly -10 or so, on the units that were stock GT. When I was done I'd put them at about -50 or so to get them to pretty much max capacity for the system. 10 degrees of superheat on the return would be closer to -55 I think, but I didn't want to run with any liquid return, and I figured that would be close enough, considering they aren't charged quite that much at stock anyway.

They say that 200w should be stock GT max capacity. 7.6' of .028" captube puts 'em roughly 190w or so. Not sure how they test them though, so maybe they are set up with a load tester.

Anyways, I finally figured out where the 'Macro' setting was on the camera :am: so maybe I'll get some new pics...I'm such a retard :rolleyes:


Gray

[Gray Mole]

Figured I'd add a couple of pics that weren't quite so bad lol

http://www.imgbox.co.uk/uploads/13230bdc57.jpg (http://www.imgbox.co.uk)

http://www.imgbox.co.uk/uploads/32d3371edb.jpg (http://www.imgbox.co.uk)

http://www.imgbox.co.uk/uploads/f34a7b3a26.jpg (http://www.imgbox.co.uk)

http://www.imgbox.co.uk/uploads/4956ddefad.jpg (http://www.imgbox.co.uk)

And my Fluke tester chose NOW to kill the battery so I couldn't match up the temps with it, and loadtesting seemed silly as it would just be the display temp anyways, but at least the mod is a little clearer anyways.

Cheers

Gray

[Gray Mole]

Not to recycle the thread, but I finally got a little feedback from one of the 3 units that I made.

I ended up doing a 240v fan conversion on it as well, as he wanted 'the best at all cost' so that made the 2 units full 402a, chilly evap, and 90cfm fan conversion.

Still no word on results with the other 2, but I don't think either one's had a chance yet.

Using a week 13 FX-55 that could do 3.5ghz on the 507'd unit he had me convert, he was able to get just over 3.6ghz, so I'm pretty happy that it did make a difference for him. Running 1.7vcore and 3.6ghz.

It's the reported temps that surprised me. He said that the DFI SLI-DR that he's using was reporting -17 under load, prime95 load. He also said he'd had to use the stock heater, as the backplate even with insulation was getting condensation on it.

Display temps were around -73 full load, and I tested all 3 units to about 20c warmer than display, so holding -53 or thereabouts under load.

I'm pretty happy to hear about it, you so often don't get the feedback you'd like to get on how they perform, and what improvement is had with the improved temps, but I'm impressed.

Just thought I'd share that with you :D

Gray

[catkicker]

Wtg!

[quintus]

what tha......
-80? is that a nl11f? if yes i trow my nle9f directly in the bin....
nice

[Gray Mole]

Heh

When you see the display temps on a mach2 you've got to take them with a grain of salt that's for sure. After the probe was relocated on the Chilly1 head, I found that almost exactly 20 degrees was the difference between reported, and actual temps measured with the Fluke, at least in this temp range.

So -80 is more like -60 for what really going on.

The Mach1 display was a lot more accurate from what I've found, only about 8-10 degrees difference that I've found there.

The difference between the regular fans on turbo, and the 240v Papst fans I replaced them with, was about 3 degrees colder, but obviously when under partial load it's making all the difference.

This unit I modded in the pics is the very same one he got, and that was before the fan mod was done, so he ended up with about -83 no load, and -73 load temps with his FX-55. The translates roughly to about -63 no load and -53 load temps if you had a real temp probe with a good temp reader like a Fluke or UEI.

I think that the Chilly1 head has a lot to do with the performance as well. The newest revision is truely amazing for both idle and load temps, and with the 402a having good condensing efficiency the evaporator is able to do it's job well, and Reggie's evaporator does that job amazingly well.

I don't think you need to throw that NL9 in the bin just yet though. While I personally think that greater displacement wins most of the time, the difference in frequency between UK and US power means that they will operate roughly on par with each other, so you should be able to get similar results with that compressor as long as the rest of your system is working well, and your evaporator is very good.

Gray

[jinu117]

One thing people forget to think about is that with bigger compressor usually extra heat on the loop. You do need condenser to deal with the extra heat. Something like NEK (embraco) or NTY(danfoss) compressors would cool exceptionally well since they introduce much less heat on the loop but... Found out those cost about $80-100 more of counter part. I might grab few for really luxury models of my V2000 units.