First 404 DD, but can't go below -40C. Need help :)

[meke]

I prepare my first single stage, details are;

1/2 hp tecumseh 404 (CAE2424Z)
404 gas (don't know the amount in the system)
1/2 hp compact condenser
1 mt 1/2" flex suction line
1,5 mt 0,7 mm cap. tube.
home made evap.

I prepare the system as shown in the photo and let them stay 30 hg vacuum for 2 hours, no leaks at that moment, so I charge with 404.


http://img.photobucket.com/albums/v217/meke_oocay/223K/first_try1.jpg


The test results are;

Room temp = 27°C
Block temp = 27°C @ t
Block temp = -42°C @ t + 15 mins.(idle)

Suction line/before comp. = (15 hg, -0,4 bar, -58°C 404) the blue hose in the photo.

edit: funny that the blue hose was in the wrong place, the photo was originally taken when I charge the system, :p: the measurements were taken from the correct place, of course, right behind that point. :slap:

Suction line temp = 15°C measured with digital thermo.

High line/before dryer = (187 psi, 13 bar, 28°C 404) the yellow hose in the photo.

Highline temp = 24°C measured with digital thermo.

The compressor itself is around 40°C and the gas out line is around 44°C

Also, I insulate the flex and the cap tube is inside the insulation.
Block gets -42 in 15 minutes but nothing more.


So, I need your advices, is my system;
1-undercharged?
2-overcharged?
3-something wrong that I don't realize?
4-wrong measurement?

take care,
meke

[pythagoras]

What makes you think theres something wrong with the block temp?

Regards

John.

[meke]

Chilly1 already warns me that more than -40°C in the evap is suprise, but I'm self motivated :D

I need -45 @ 210 w load but this makes -20 as far as I measure.

[pythagoras]

-20C is about right for this system at 210 watts. -45 is unrealistic. Listen to Chilly1, he knows what hes talking about ;)

Regards

John.

[meke]

I belive I can get around -40 at load, this is why I choose 1/2 hp.
but feel that system needs some fine tuning.

any advice?

[wdrzal]

are you running 15 inches of hg ? (vacuum) A pic of manifold gauges while operating would be great.

you said -15 psi did you mean -15hg

[meke]

yes, the suction line shows -15 psi vacuum.

here is the photo of the manifold set. It's some value starter kit, but works fine.

remember that photos below are taken while the system is shut down for 12 hours, its not working at that moment.

the whole set:

http://img.photobucket.com/albums/v217/meke_oocay/223K/manifoldset.jpg

close up:

http://img.photobucket.com/albums/v217/meke_oocay/223K/manifoldsetclose.jpg

[wdrzal]

I wanted too see them when system was operating to make sure you are reading correct scale. your gauges read in inches of hg . not neg psi, thats what got me confused a little.

[kayl]

if you are running 15hg, -15psig? (reads hg for correct reading values)then you need alot more gas you want about 0psi (correct reading values in psig).
the iner most one reads psi ?? thats weid.
but do what wdrzal said get a good pic of gauges when at lowest temp.

just for a ruff guide want low side gauge here at load and high side 150-200
at idle low side will suck vacumm about -40c's hopefully
thats a nice compressor/phase chage system,
should be able to get -40c's with tunning and correct gas charge.

also is it me or hasnt the system equaled here in those last few pics.
how long after switch off is that pic, might have a obstruction some where in the system

[wdrzal]

also look closely at his high side gauge,its a compond gauge too, never seen a high side gauge that read vacuum.

[pythagoras]

Kayl, both guages are reading the same pressure when off, about 90 psig. So I guess its equalised.

Regards

John.

[meke]

I wanted too see them when system was operating to make sure you are reading correct scale. your gauges read in inches of hg . not neg psi, thats what got me confused a little.

you are right, after 0, in turns to Hg, not negative psi. It's my fault. :hitself:

[meke]

Kayl, both guages are reading the same pressure when off, about 90 psig. So I guess its equalised.

Regards

John.

Yes you are right.

I check with my eyes on the gauges, both reads around 86 psi.

I will start back and get the readings for your comments.

:toast:

[meke]

Ok, I run the system again.

here are the working conditions;

it gets +23 to -42C in 15 minutes.

the high line is: 10 hg (-0.4 bar)

the suction line is: 177 psi (12 bar)

http://img.photobucket.com/albums/v217/meke_oocay/223K/high_low-42C.jpg

http://img.photobucket.com/albums/v217/meke_oocay/223K/high_low-42C2.jpg


edit: so @kayl, thanks for the advice, you offer to "add more gas to reach the 0 hg in the suction line", but when ever I add some more, the high line gets around 200 psi.
I still need advice to measure the suction around 0 hg, should I increase the cap lenght from 1.5 m to 1.8 or 2 m?

[wdrzal]

you have the high and suction backwards, also it is "-10hg" unless you say "10 hg vacuum" either way is correct. but it is "understood" saying 10hg in refrigeration you are in vacuum.because +10 hg is not used

[wdrzal]

is that under load or not,sounds like cap tube is to long not to short ,seems lke tube is too restrictive.how did you cut your cap tube?

[pythagoras]

There are two reasons the suction side pressures are too low.

1. Not enough refrigerant.

2. Cap tube too long.

First you need to find out if one is the case, the only way I know to do this is to have a sight glass in the liquid line and check that its full, no bubbles. If it is then you can start to length the cap line.

Regards

John.

[wdrzal]

he has 200 psi high side john when kayl had him add refrigerant

[meke]

@wdrzal, ok I learn to read the manometers, thanks.

I cut the cap tube by a sharp knife, but slowly. I see the inner part clean.

@John, I have two ideas:

1-my refrigerant charge is not enough, and I'm using my nearly empty gas cylinder to add some more gas in to the suction line. Should I use some liquid full gas cylinder to reach to 0 hg?

2-my cap tube may be resistive, I'm afraid I cannot make triple evac. very well, so some black smoke may be still in the system. And additionally, I learned that my cap tube may be 0.8 mm ID, but Gary's estimation makes this 288 cm lenght, I use 150 cm.
If I increase the lenght to 288, then I will increase the resistivity, right?

Thank you for your support.

[wdrzal]

I think you cap tubes blocked,or filter

[pythagoras]

Yes increasing the captube will increase the high side pressure because of resistivity. The easiest way to tune a captube system is with a sight glass. Can you install one?

Walt may be persuaded to write a guide ;) . He just helped me clarify my thoughts on this.

Regards

John.

[meke]

I will do this:

-replace the filter and the captube.
-triple evac. with r11 or n2 something.
-recharge with liquid.
-share the results.

This will make more clear referance to discuss over.

[wdrzal]

first get some gas and add some too see if suction pressure comes up, what size is 7mm in thousands. I'm not used to metric you also need to pic you rig so we can see if there are any other problems

your original temps are good just you are in too deep vacuum,don't cut it apart yet, add some gas first

[wdrzal]

i just re-read all your posts is this correct -42c evap temp and -58c suction line temp?

what is that comming down right below the blue hose in the first pic you posted

[kayl]

ok.
ya need to insulate the suction line.
when you turn the system off how long does it take for it to equalize at the refrigerant gauge, this can tell if theres a restriction.

i think that capillary line is too restrictive as well. 1.5m , maybe try 2.4-2.7m (starting at 2.7m).
when too short get lots of flash gas, and can seen higher pressures (high side).

then try 3psig and should get frost all way back to compressor and not worry about high side pressure for now.
then can bleed some refrigerant back out back to 0psi and high side pressure should get lower same time.

if working ok now try about 4Hg, should see close -50c with that unless something else wrong.

[kayl]

one thing also make sure temp probes attach well and have some insulation covering probe to ensure it reads correctly.

post 14 isnt that bad results.
when you leave it run for an hour things will be different again
insulate it and put a probe on suction line just before the compressor 6" and insulate that probe really well as well.
if not get correct readings can be leading us a stray.

[meke]

i just re-read all your posts is this correct -42c evap temp and -58c suction line temp?

what is that comming down right below the blue hose in the first pic you posted

blue hose was connected to the suction of the compressor, not the suction line.
check this picture, you will get better.
http://img.photobucket.com/albums/v217/meke_oocay/223K/suctionline.jpg

the thing coming down right below the blue hose is the correct suction valve, the measurements are taken while the blue hose is connected to that "thing" :rolleyes:

-58C suction temp is what I read on the manometer, but the correct reading is 177 psi.

[meke]

Now, I cahnged the captube.

I measure somehow and the ID is 0.8 mm.

Estimation calculates that I have to use 2.88 Meters, I prepare it.

Change the drier also, and the suction pressure comes to "0", the high line is 200 psi.

But still -41C and I am starting to bang my head to walls :upset:


Any comments?

[wdrzal]

how cold do you think a single stage will get ? is is that with load?

[meke]

no, idle.

with load it makes -20 -25C

I wonder this will make -50 idle and -40 load. :confused:

[wdrzal]

-46c is the boiling point of 404a at 0 bar ,your ambiant temp is 27c=80f thats 10f hotter than average,humidity also affects capicity.pc ice units hes selling are -43 and i bet the ambiant is around 70f,I think some tunning will help but your expectations are high. what is the liquid line temp?

[meke]

the liquid line temp is 44C before the condenser and 25 before drier.

suction is around 0 hg, high is around 200 psi.

I can accept the minimum -42C for idle, but what about keeping it there @ load?

[wdrzal]

if you have a amp meter,I would add refrigerant a little at a time allowing system to stabilize for 10 min.before adding more, until you either carry the load or get to the RLA of your compressor.try this first .adding a few psi to suction pressure will increase capacity.

[meke]

if you have a amp meter,I would add refrigerant a little at a time allowing system to stabilize for 10 min.before adding more, until you either carry the load or get to the RLA of your compressor.try this first .adding a few psi to suction pressure will increase capacity.


Thank you for your support. :toast:

I always thought that frozen suction line is not good for comp., afew psi to suction makes a good white snowy suction flex :cool:

[wdrzal]

as long as the frost stops 6 inches from compressor,you'll be ok, you must insulate that suction line,you are losing capacity there. a uninsulated suction will asorb heat.

if your line is not insulated do this before adding any refrigerant

[meke]

no its very well insulated.

I took the photo before insulate.

I will charge a little more tonight, althought today is my wedding anniversary, I hope my wife will understand my psychology:fight:


:bday2:

[wdrzal]

Happy Anniversary :party:

[meke]

Thank you :)


The highline must be measured from the exit of condenser-before drier, right?

When I change the condenser (fix a smaller one) I braze the valve before condenser, this is just for evaluation, but I'm afraid now it's measuring a bit higher, or am I wrong?


Also, will there be any effect on starting the load before starting the cooler?

I mean, I connect the block to a heat source, and the block was 80C when I start the system yesterday night. I think the compressor didn't like this situation :D

[wdrzal]

if using a cpu for load it can be damaged if you start pc first,if using another load it really won't matter.it will just take system a while longer to catch up. why smaller condensor?,unlees to fit in case. in your higher ambiant temps you would benifit from larger condensor.the line from compressor to condensor is hot gas, temps will be hotter.valve goes on liquid line.thats between condensor and metering device.

[kayl]

meke single stages at load are alot harder than ppl lead on, well good temps at load any ways.
-41c idle is quiet good really first time, it also depends what cpu overclocking as well and the quality of insulation on suction lines, evap and how you insulated the motherboard can have a large effect as well.
now ya need to play with charges etc learning things as you practice loading it.
try loading for awhile, then let it recover and try adding a little refrigerant little by little, let it settle again then loading it again. i would only worry about flooding the compressor with too much gas if ya see a big patch of ice forming around the suction line where it enters the compressor, or if can here a dripping sound. then ya have too much. then try letting a very small amount out until find a happy sweet spot. It may take a few trys of adding and removing to get it at the righ charge.
once you have done the best you can with charges you could try shortening the capillary line by about 10-20cm and repeat again.
If you attach a temperature gauges on the suction line as well like I mentioned in past post in this thread that can also help you in deciding whether to add or remove refrigerant.

[meke]

thank you both Kayl and Wdrzal :toast:

I will keep these advices and hope to release my rig this weekend. :rolleyes:

[kayl]

thats cool look forward to some more pics. :toast:


Thank you :)

Also, will there be any effect on starting the load before starting the cooler?

I

dont do this it wont get cold.
i found booting under -20c to be a disaster, at -10c or so my cooler just gets stuck there not getting colder, and above 0c is differently not good it will never get cold with these small systems and such little gas.

[Ssilencer]

Are you sure your temp probe is making really good contact and is well insulated, is designed for this temps or the batteries are not low?
How about evac? you said that it was not very good the first time, and now? buy a sight glass, be sure that you dont have any moisture or air in the system.

[Ssilencer]

thats cool look forward to some more pics. :toast:



dont do this it wont get cold.
i found booting under -20c to be a disaster, at -10c or so my cooler just gets stuck there not getting colder, and above 0c is differently not good it will never get cold with these small systems and such little gas.

Maybe I dont understand what are you talking Kayl, mines starts pc when the evap is at -20c and the evaps still going down fast.

[meke]

the load that I mention is a small electric hotplate, its rated 350W, I fix my evap over it and try to keep the hotplate around +120 C to simulate my 560 prescott @ 5.3 Ghz(burning 3.6 @ 5.3 Ghz, my DFI shows +120C under -56 C DICE :D )

Although this hotplate is 350W, my block is poor little kitten against this bulldog, but a good way to tame the whole rig :slobber:

When I first meet them all, I forgot the hotplate open and suddenly evap copper reach +110C

then shut down the hotplate and start the system.

for nearly 1 hour, rig cannot come to itself and could not take down the evap any minus degrees celcius.

And the compressor come to +65 C,

don't :slapass: to me, I am just evaluating whats goin on. :rolleyes:

[kayl]

Maybe I dont understand what are you talking Kayl, mines starts pc when the evap is at -20c and the evaps still going down fast.


sorry i should of said wait at least -20c but one more minute and ya at load boot temp. just an example for nebies safe side :stick:
i like to boot cold -40c or more

[wdrzal]

when heating the evap ahead of turning on the unit you are not only heating the evap but the vapor inside which will transfer the heat thru conduction to warm the rest of the parts. even when rig is off. Giving the load a big headstart makes it hard to recover from. in your case when your not maintaining load real well yet, I suspect the mass flow is a little low,this makes recovering even harder. While this won't harm anything,you just waste a lot of time waiting for system to cool down. let it cool down first then add load.

[Malachia]

then, with 2.88m of 0.8mm you have -42 with no lload and how much with load??? btw, my way to mod mach is this, i charge them until suction is 0bar with no load, then i place it on a processor and i charge the sistem in the way to have 0/-0.1 bar with a default or little overclocked CPU(around 100W..), and this work really good!!!

[meke]

Now I braze lots of the pieces again and change some design.

It's still -41C idle, so I give up to reach some lower degrees.

I concentrate on the load temps,

last night I spend 7 hours to make it run with performance.

here are the results;

2.8 m 0.8mm captube,
new drier,
again big condenser,
404a liquid charge.

-41 C idle
-31 C @ TEC 150W with small evap

HL = 270 psi
SL = 0.4 bar (6 psi)

comp. gets too hot but cannot go over -31 with the small evap.

Any advice?

pics. are;

http://img.photobucket.com/albums/v217/meke_oocay/223K/V2/V2.jpg

http://img.photobucket.com/albums/v217/meke_oocay/223K/V2/V2-41.jpg
http://img.photobucket.com/albums/v217/meke_oocay/223K/V2/evap150w.jpg
http://img.photobucket.com/albums/v217/meke_oocay/223K/V2/Manifold-41.jpg

[Malachia]

try it on a cpu...

[pythagoras]

Suction side is 6 psig which is about 22 psia, you could try letting a little refrigerant out until you get 0psig or 15 psia and see if it can handle the load there.

Regards

John.

[meke]

I will but sleepless anyway.

I hear the birds singing and suddenly recognize that the twilight :p

[meke]

It can handle.

I was adding some more if it can reach any below -31 but no way. :(

I will discharge some

[meke]

ya need to play with charges etc learning things as you practice loading it.
try loading for awhile, then let it recover and try adding a little refrigerant little by little, let it settle again then loading it again. i would only worry about flooding the compressor with too much gas if ya see a big patch of ice forming around the suction line where it enters the compressor, or if can here a dripping sound. then ya have too much. then try letting a very small amount out until find a happy sweet spot. It may take a few trys of adding and removing to get it at the righ charge.
once you have done the best you can with charges you could try shortening the capillary line by about 10-20cm and repeat again.
If you attach a temperature gauges on the suction line as well like I mentioned in past post in this thread that can also help you in deciding whether to add or remove refrigerant.


So kayl, I guess I have to shorten my captube, right?

I will change my block also :mad:

[meke]

Guys,

never use cheap or unrated equipment.

I already lost 2 weeks to tune some f***in reason, but I'm happy now.

Check what my old thermo reads and also compare what the rated equipment reads. :slapass:

http://img.photobucket.com/albums/v217/meke_oocay/223K/chy-50.jpg



wait for the 3.6 prescott @ 5.x Ghz OC results. :banana:

[wdrzal]

You must always check thermometers in a ice bath
(0c) to see if they are calibrated correctly.

you can also check in boiling distilled water.(100c) at sea level


If after these checks your meter dosn't read correctly it could be one of these reasons
1: meter out of calibration
2:meter not set to proper range/type
3:wrong type wire for the type.

in industry there are many types of wire that are used to make thermocouples ,you MUST use the correct calibration range for the given type.

[pythagoras]

Also dont believe the cny 502 just because it gives you -50C ;) Stick a probe from each in an ice water bath. That will tell you which one is more accurate.

Regards

John.

[_HL4E_HalfLife_]

ur compressor is getting too hot because (in my opinion) ur cap tube is too long and ur discharge pressure of 270Psi is too high I think.

[hatemi]

That old probe/meter showd +3c in icewater IIRC.

[wdrzal]

be sure you have a least 80% ice. the recommened ice bath is 90% crushed ice 10% water in a vacuum thermos for insulation,let stand for 5 min. before reading.

[meke]

The difference between the two thermo's are:

3 C around 0C
5 C around -10 to -15
7 C around -15 to -30
10 C around -30 to -50

Of course I connect the new probes to the old thermo
and the old probe to the new thermo.

results are same with the %1-2 difference

[meke]

That old probe/meter showd +3c in icewater IIRC.

I must thank hatemi for his help in this thermo issue. :toast:

[wdrzal]

the farther away from zero the larger the error, if you correct the 3c to 0c .the 10c will not be 7c off at -30c to -50c. it may only be a degree or two.

[Ssilencer]

Are you sure your temp probe is making really good contact and is well insulated, is designed for this temps or the batteries are not low?

Need to say more?
:D

[meke]

Need to say more?
:D

I remember your post that time and take care of it.

My sensors was checked again and they were good contact and isolation.

But my thermo is still working and batteries are ok.
the design issue is:

I evaluate some dice connection 2 months ago, with the same thermo.

whether it was showing -56C there, I never thought that this thermo was showing something different. At least about the range issue, because i saw that it showed -56C.

But now I understand that dice experience was not -56, was also at least -63 that time, or -65 may be.

Ok, I'm now walking through the steps of the huge Phase Change Experiment road, I belive I already pass some 1-2 steps :D

[Tonic]

How does it hold at load?

[meke]

I prepare my rig yesterday night.

I isolate the block and mount them on my 3.6 Prescott (560 aka. oven:))

I will post the pictures and SS, but;

3.6 @ 1.4 V
block -48 C
die -43 C
Speedfan -35 C

4.8 @ 1.8 V (just to try for the heat load)
block -40 C
die -24 C
Speedfan -21 C

I was tuning the gas charge with the loads @ 4.8 Ghz after Spi and Prime95 tests, unfortunately an accident occur and I broke one pipe near the condenser. :mad:

I shut down everything and save as much gas as I can.

I will braze tonight and regas.

Seems I fix a pretty good phase change DD, at least for the first time. :banana:

[hatemi]

Woot:D The results are very very nice.

[meke]

Woot:D The results are very very nice.

thx hatemi, hope to finish this weekend and paste some good results :cool:

[piotres]

Guys,

never use cheap or unrated equipment.

I already lost 2 weeks to tune some f***in reason, but I'm happy now.

Check what my old thermo reads and also compare what the rated equipment reads. :slapass:

http://img.photobucket.com/albums/v217/meke_oocay/223K/chy-50.jpg



wait for the 3.6 prescott @ 5.x Ghz OC results. :banana:

Ohh yes reason of that "bad" temps was really stupid .. :( :( ...I only can sympathize yopu that you had worked so long , before you find the reason..cap tube change, drier change ..ehh :rolleyes:

But now I see that you cooler works like devil... awesome temp for first p-c :toast: ...

best regards
peter

[meke]

thank you peter

[Ssilencer]

Yes, impresive job.

[meke]

no, idle.

with load it makes -20 -25C

I wonder this will make -50 idle and -40 load. :confused:


btw I proof my beliefs to myself.

If I gave up, never understand the potential of this first try and may be lost my desire.

I must try harder :comp10:

[mesyn191]

For a first try you did a pretty good job man, way better than most phase change nOObs!!

If you decide to do a cascade post pics!!

[meke]

thx Ssilencer & mesyn191

hopefully this summer I am planning to build one cascade.

[meke]

just a startup with the small evap for vga.

I prepare a rig for 3.6 prescott.

This photo was taken just the spi 1 M finished.

3.6 @ 4.8 (266 x 18)

http://img.photobucket.com/albums/v217/meke_oocay/223K/V3/27sec-37.jpg

more will come with the insulation pictures soon. :fact: