PDA

View Full Version : Promethia Problem



Holst
05-29-2003, 08:11 AM
I thought this might be an issue...

With my current high OC my CPU is putting out some mega wattage even when the system is in reset...

What happening is that the evap is being heated by the CPU to 40*c before the compressor can get upto pressure and start cooling.
So my system makes a loud allarm, turns off and wont post :(

So far I can let the evap cool and turn on again, after a couple of trys there is enough pressure and cooling to power up..

Or I can use a 12v PSU to cool the system then plug into the mobo and post normally.

Neither is a good solution..

Any tips to get it up and running without messin about as much?

I allready have my fans on full power during power up (although they go slow once im up and running as I hate noise)

I nevr liked this holding in reset thing, why cant they just include a seperate 12v PSU in the unit?
Does the Rev 2.0 work in the same way ?

WxChaser
05-29-2003, 08:27 AM
All I can think of is to check the way it's seated on your CPU again.

Good luck! Let us know how you make out.

LORD
05-29-2003, 08:27 AM
Unless your system is in a very warm environment. I really don't think that it should reach 40 Degrees Celsius during power up. Which is preset to stop booting. You can change this though, P26 I believe.

You may have something blocking the inlet. This may cause a problem... However I think you would have already checked this.

It could be that your Prom has a gas problem. Like too little. These compressors really do carry no more than a Bees fart when it comes to the amount of R134a or what ever other variant may be used in there.

JCviggen
05-29-2003, 08:46 AM
There is no way around this.

There was a "power conditioner" tool once which chip-con are no longer offering. the Mach II prometeia has this built in. It basically cuts the 12V so the CPU isnt getting power until things get turned on.

There's no problem with the prom..... just too much heat at start-up. Best thing to do is power it off just before it hits 40C... then wait a minute , turn it back on and it should drop just fine. It needs a little time to get the pressure built up. Dont wait too long, one minute should do it

Holst
05-29-2003, 09:59 AM
Its just a very hot day today.

As soon as I power up the temp is 26*c and it gets to 40/c within 30 seconds, way before the compressor has powered up.

Ive rigged up a couple of leads and swiches from a 12v PSU so that I can swich on the prommy let it cool down to 0*c, swich it onto the PC PSU and im ready to go.

With this sort of ambient temp prommy is brilliant, I can still get great clocks even with 25*c ambient.
Now its running its fantastic, its just getting it going.

Ill email chip con and ask what they think, if they cant help ill have to make somehting myself.

Kunaak
05-29-2003, 10:15 AM
the easiest way around this is to clear the CMOS and begin from scratch, big deal really, takes a minute to get bios as you like it.

QuadDamage
05-29-2003, 11:36 AM
do this. When you hit power on and temps go up, turn it off, wait for 15 secs and power back on. If temps still go up, repeat this again and i can assure you it will work just fine.

muzz
05-29-2003, 12:20 PM
I'll have to remember this when I get mine in 5-6 days QD.
Thanks for the tip.

Jeff
05-29-2003, 12:25 PM
You and Holst are very lucky people muzz! :slobber: ;)

muzz
05-29-2003, 12:39 PM
Yep I agree....... but Jeff you know how long I've been waiting for 1 of those things........

Now I need to invest in a 9800P ........:D
I already have an NF7 and 1800 on the way..... figured it was time to try something a little different.

Congrats on the 3g Barton Holst!!!!:toast:

Jeff
05-29-2003, 01:06 PM
Originally posted by muzz
Yep I agree....... but Jeff you know how long I've been waiting for 1 of those things........

Yeah I know. ;) But you will have waited less than me so I can relate. :D

Definitely keep us posted on your progress/experimenting/fun. ;)

zakelwe
05-29-2003, 01:22 PM
Originally posted by Jeff
Yeah I know. ;) But you will have waited less than me so I can relate. :D

Definitely keep us posted on your progress/experimenting/fun. ;)

I think Jeff should stick to air cooling and be an inspiration to the rest of us air cooled peeps ;)

Regards

Andy

muzz
05-29-2003, 01:51 PM
Yep Jeff does great on air.. but he cheats he uses the swifty/pelt HSF........:D
Still does a great job.:toast:

I'll let ya know Jeff.

Holst
05-29-2003, 02:23 PM
Ive found a reasonable solution.

I got a 240v 12v PSU and a 3 pole swich.

In the middle position they system is off.

Push it up and the 12v PSU is connected to the prommy cooling it.

Once the temp is going down I swich the swich to the bottom position and connect the prommy to my PSU, then I press the power button and boot using the normal sequence.

This is pretty easy to do and its pretty foolproof (foolproof is important for me :P )

Your gona love the prommy Muzz, it took me a few days to learn its wuirks but its kicking ass now :D

muzz
05-29-2003, 02:27 PM
Yes Holst I WILL love the prommy...... and I cannot wait to get it.
Tough watching everyone else have all the fun.

saaya
05-29-2003, 02:31 PM
??? i thougt the prommy cools your system down first and only powers up your system when the cpu is minus 20° or smthn like that :confused:

Holst
05-29-2003, 02:43 PM
Its supposed too saaya by holding the system in reset whyle it cools down.

Unfortunaely my 3gig barton gives out so much heat that even in reset it gets to the 40*c temperature limit that the prommy forces.

This should only happen to those with big voltage and overclocks, unfortunatley that means almopst everybody with a prommy and a new AMD chip :(

Jeff
05-29-2003, 03:42 PM
Originally posted by muzz
Yep Jeff does great on air.. but he cheats he uses the swifty/pelt HSF........:D

:eek: ;)

saaya
05-29-2003, 03:54 PM
so its keeping the system in reset all that time? cant you change it to keep the system off until a nice temp is reached?

Holst
05-29-2003, 11:46 PM
Originally posted by saaya
so its keeping the system in reset all that time? cant you change it to keep the system off until a nice temp is reached?

Thats how it keeps it off by using reset.

saaya
05-30-2003, 12:39 AM
well it keeps it in reset, there IS voltage going through the cpu, if not it wouldnt heat up. so instead of sending the reset signal all the time until the temp is low, cant you change that so it only sends the start signal when the low temp is reached?

Holst
05-30-2003, 03:17 AM
Nope, not without making something.

Im working on making a box that will turn the prommys restet signals into swiches that "press" the power button.
Im trying to do it without using any logic.. but ive not come up with something that will work yet.

muzz
05-30-2003, 03:42 AM
You could make a delay timer Holst , like they do when building Tube Linear amplifiers ( where they don't startup until all circuits are ready).
It obviously wouldn't be AS good, but set it for 2 minutes, and all should be well no?

Otherwise you would have to setup a temp switch of some sort.
I guess I will probably be dealing with this soon myself..... I am going to start looking for something.

SpicyHuevos
05-30-2003, 06:20 AM
There is a much easier way
Just unplug the atx from board make a little jumper out of wire
jump the green wire and the black wire on atx plug prommie starts up without board running.
Let it cool then plug the atx back in and turn on and youre golden

Holst
05-30-2003, 06:57 AM
But that means unplugging stuff every time Spicey, I want to press one button and have it start on its own.

Muzz... if we use a 12v supply to power the prommy we can use the current reset swich to controll another circuit that will boot the PC.. Ill have to get a multimeter on the prommy and find out if the reset connector just goes to 0 ohms when its in reset, if it does then i should be able to use a relay to swich a timer that will boot the PC when reset it open... thats the theory anyway I just need to work out the specifics.

Holst
05-30-2003, 06:59 AM
Hmmm... turning off will be a problem then. I suppose I could put the whole thing onto another 240v relay that runs on the PSU 12v, but is bypassed when power is off or reset is closed...

This is more complicated than I thought. Im starting to understand why chip-con went for there more simple solution.

muzz
05-30-2003, 07:10 AM
Well I won't have mine for a few yet, but I will be looking into setting up a little xsfmr/diode/Cap/timer circuit to accomplish this ( breadboard).
I believe it will be fairly simple once I see whats happening.

Holst
05-30-2003, 07:13 AM
I hope you can work it out as im not much cop at making electronics stuff from scratch..

muzz
05-30-2003, 07:23 AM
I think it will be a fairly simple circuit that I will be able to just draw up in word...
I just thought of something....
What if you put the circuit between the leads going to the mobo, and the mobo itself, so that when you turn on the power to the mobo it will start the timer/relay and you should be able to set the timing by R/C values.
So the prommy would start, but the mobo wouldn't for whatever value/time you choose.
It can be done, just finding the particulars is all I would think.

Holst
05-30-2003, 07:28 AM
Sounds good if you can get it working.

muzz
05-30-2003, 07:44 AM
Amateur Radio enthusiasists have been adding delay circuits for years, to combat inrush current.
I'm sure we'll come up with something, I have a few books I need to look through........ haven't looked in them for years.

muzz
05-30-2003, 08:02 AM
I will have to take off the cover to my case, as I am not sure of the voltage applied to the board ( which would be the input voltage).

There are commercially available timer/relays, here is a page listing a boatload.

http://relays.tycoelectronics.com/relays/agastatdatasheets.stm#dis

saaya
05-30-2003, 09:26 AM
i dont think it would be very complicated for chipcon to build the prommie in that way so it keeps the system off until the temp is down. sux that you have to mod a 600$ cooling system...

Holst
05-31-2003, 02:26 PM
Thnkas for the link muzz, with one of those babys I should be able to make it fully automated.

I dont suppose you know any pricing on those units ?

muzz
05-31-2003, 03:15 PM
Unfortunately I don't Holst, but if ya google it ya may be able to find out ez enough.
I won't have my unit for prolly 2 weeks, so I will be out of the loop till then.
GL bud, lmk how ya make out.

I'm sure ya can make something to do this EZ enough if that is too expensive.......
Nice that its already done for ya though.:D

Holst
05-31-2003, 04:05 PM
Ive allready made a solution using a simple swich, but id like to fully automate it for fun.

muzz
05-31-2003, 04:14 PM
Of course......... my WC setup was rigged to turn on with the PC ( the first setup with my old PSU) worked like a champ.
The things you can do with a simple relay and some wire........:D

toofast
05-31-2003, 10:06 PM
By far the easiest way is to clock down when you shut it down . :confused:

Holst you take good care of my baby ;)

p.s dont put too many Epox mobo`s in there she dont like em :p

JeffPH
05-31-2003, 10:12 PM
Originally posted by toofast
By far the easiest way is to clock down when you shut it down . :confused:

Holst you take good care of my baby ;)

p.s dont put too many Epox mobo`s in there she dont like em :p

hehehe

Nohto
06-01-2003, 06:17 PM
Holst, Muzz, all you have to do is use a 12v power supply. A small transformer installed on the inside of the prometeia. You can pick up a 3A trans at radio shack for $15 US. The trans only needs to run the two fans inside the Prom and close the Ice cube relay. You hook up a switch to turn on your Prom and then you turn on your PC when your temp gets to the desired level. This also solves the reset problem where you have to turn off the computer on hard lockups and by doing so, it in turn shuts off the Prometeia, which everyone knows is bad for the compressor.

Liquid3D
06-01-2003, 08:04 PM
I didn't know it was bad for the compressor? I've had my Prometeia for almost two months, and haven't used it because I've had contact issues. However when I hear, toofast make that statment about Epox mobo's I find it oddly coincidental both board I've struggled with are the 8RDA, and 8K9A. So I was going to wait for my P4 2.4C, Abit IS7 combo, and get her up and running then. I'm getting great freezer temps, but the CPU is at 15C, which I can accomplish on a Fall night, open window, Sweater, and my SLK800/Vantec Tornado. toofast, could you be specific as to the Epox issue?

BTW back on Topic, my manual states any shutdown, requires 30min wait period before restart. Don't know if this pertains to specific problem, however; hasn't Chip-Con made a "modification" circuit plugged into her?

Holst
06-02-2003, 03:03 AM
Thats exactly what ive done nohto :D

Are you using a 8rda+ liquid3d as my prommy works fine.

There were some problems with older boards like 8kha+ where they wouldnt accept the reset signal and tried to boot before the evap was cool, but thats not a problem on the newer boards.

I think the problems toofast was refering too is his supprising ability to kill epox boards, not any specific incompatibility. (my prommy was second hand from toofast, so it was more of a joke than anything)

RoydRage
06-02-2003, 07:30 AM
JC Is Correct...

I have expeirienced the Same thing...

And have found that letting it sit for a minute and re-trying... Usually works... And it's funny... It does not always happen, and I'm not sure it has to do souly with room temperature... (although, of course that would affect it), because it had the trouble this very cold winter... And I always keep the room that my Prome/Lian-Li Pro was in Very cold...

Is this a problem with the Mach II also?

RoydRage

Liquid3D
06-02-2003, 10:05 AM
Originally posted by Holst
...Are you using a 8rda+ liquid3d as my prommy works fine...
I think the problems toofast was refering too is his supprising ability to kill epox boards, not any specific incompatibility. (my prommy was second hand from toofast, so it was more of a joke than anything) I thought it had an "inside" feel, but I jumped on it because for some reason I have such tension on the mircrofrezer, that when mounting it, rotational resistence from the tube's "copper" insert hose, actually turns the microfreezer head about 3mm perindicular to the case corner? No the mount isn't on backwards, nor is the brackett (which would then force it to be backwards). I'm still confused. It was the 8RDA which caused the most trouble. I'm the only person who has experienced this cranial concoction.

DisposableHero
06-02-2003, 03:41 PM
this has never happend to me with my 2750 2.05vcore...

it'll get to -33 though and try to boot up but my stupid 8rda+ will go FF since its been in reset so long.. i hate that

toofast
06-02-2003, 07:45 PM
Originally posted by Liquid3D
I didn't know it was bad for the compressor? I've had my Prometeia for almost two months, and haven't used it because I've had contact issues. However when I hear, toofast make that statment about Epox mobo's I find it oddly coincidental both board I've struggled with are the 8RDA, and 8K9A. So I was going to wait for my P4 2.4C, Abit IS7 combo, and get her up and running then. I'm getting great freezer temps, but the CPU is at 15C, which I can accomplish on a Fall night, open window, Sweater, and my SLK800/Vantec Tornado. toofast, could you be specific as to the Epox issue?

BTW back on Topic, my manual states any shutdown, requires 30min wait period before restart. Don't know if this pertains to specific problem, however; hasn't Chip-Con made a "modification" circuit plugged into her?


I think the problems toofast was refering too is his supprising ability to kill epox boards, not any specific incompatibility. (my prommy was second hand from toofast, so it was more of a joke than anything)

Holst is correct i just kill the epox tooeasy i dunno why maybe they dont like playing rough :D

I`ll use these next time :rolleyes:

DisposableHero
06-02-2003, 09:13 PM
hey holst.. could you possibly show pics and a little writeup about wwhat you did please! thx

Holst
06-03-2003, 03:22 AM
Unfortunatley i dont have a camera any more, but I can write a bit about my prommy and the mods ive made so far.


I didnt want to write anything too much untill id got the hang of it a bit more. Read Icee's review there is allot of good info there.

I recon liquid3d's problem is his evap being so twisted, you may have to bend it even more at the base to get a good mount.

daddy_fizz
06-03-2003, 07:45 AM
hey, well i'm new here but i'll be getting my prommy in a week or so i was doing a little research. The motherboard i am getting (asus P4P800) has the same problem with the reset...

I e-mailed chip-con and they pointed me here for one possible solution...

I was hoping (as someone else said) that someone could make a little diagrams/instructions for a simpleton like me so i can install this switch and enjoy it right away :D



thanks

~Fizz

Holst
06-03-2003, 07:54 AM
Ill break out MS paint for you tonight :D

Liquid3D
06-04-2003, 07:30 PM
Originally posted by Holst
...I recon liquid3d's problem is his evap being so twisted, you may have to bend it even more at the base to get a good mount. Thank you. I'm slightly perplexed as to why so few may have had this problem? This is not a "minor" amount of "clockwise" tension when attempting to mount the Mfrzr, but enough that I truly have difficulty countering it to keep the screws aligned with their perspective mounts. And this is no way a criticism of Chip-Conn their design is superior to the other (only) option. I still have the Prometeia damaged in shipping (awaiting claim) and am going to try a comparison. Holst your suggestion of "bending" the tubing at the base had crossed my mind, however; the slightest over-torque, may create a hairline fracture, or worse "snap" the brittle solder which affixes it at it's base. I have tried applying a "slight" amount of bend in it, but per this fear relented.

I must ask this question again. Since this tubing is so inflexible, yet is designed to adapt to both horzonatlly/vertically mounted Sockets; has no one felt a significant clockwise tension, which must be countered to align their MFrzr? And has no one noticed the 2mm-4mm amount of play in the Socket-A bracket? Had these not been significant concerns, I can assure you I wouldn't have a Prometeia siting here for 60-days while I go deaf from a Vantec Tornado decibal meter tester. My resolution to switch to what I feel will be a more secure, Intel (bracket) system awaits my Abit mobo in the next few days. This is an ad hoc solution at best, and deprives me of the versatlility AMD's, multiplier/FSB overclocking offers, and AMD's best OC core ever.

AKRedneck
06-04-2003, 10:01 PM
Originally posted by Liquid3D
....I must ask this question again. Since this tubing is so inflexible, yet is designed to adapt to both horzonatlly/vertically mounted Sockets; has no one felt a significant clockwise tension, which must be countered to align their MFrzr? And has no one noticed the 2mm-4mm amount of play in the Socket-A bracket? Had these not been significant concerns, I can assure you I wouldn't have a Prometeia siting here for 60-days while I go deaf from a Vantec Tornado decibal meter tester. My resolution to switch to what I feel will be a more secure, Intel (bracket) system awaits my Abit mobo in the next few days. This is an ad hoc solution at best, and deprives me of the versatlility AMD's, multiplier/FSB overclocking offers, and AMD's best OC core ever. [/B]

While I can appreciate your hesitation installing the microfreezer, I think you may be taking it to the xtreme...

When I first did a trial fit on the microfreezer, there was alot of pressure or clockwise torque trying to get the head to slip onto the posts. I recommend you shape the vertical section of evap tubing by putting in an offset bend to get the head lined up. It takes a few overbends (gradual bends..no pinched tubing) to get the thing to finally sit in the right place.. Just continue to keep applying pressue on the tubing untill finally a permanent bend sets in. After putting pressure on it a few times to get the bend, I finally achieved a fit where the microfreezer would just slip right onto the posts... Trust me, it works.

As for the slop in the cpu bracket, that's not a big deal.. not at all. If anything, I would say the slop only aids in alignment and allows for some slight twist or rotation.

Have you done a trial fit-up to check for contact using some paste?

Once you've done this a couple times, a cpu change can take as little as 20min. It's really quite foolproof. The worst part is separating the damn sealstring when changing out a CPU... Lemme know if ya got any more ?'s..be glad to help.
:D

Holst
06-05-2003, 03:30 AM
ALso I had to trim ALLOT off of the CPU shim to get it to sit down propperly.

As you have a Tbread not a barton this shouldnt matter as much but it might be worth checking that the shim isnt hooked up on any caps on the CPU package. I alsmost didnt use mine at all as I nigh on cut it in half.

Play in the mount is normall, once your mounted that slack is taken up by the evap pushing against the core.

Ill try to get some pics taken of my install on 8rda+ so that you can see how I bent the hoze. It is quite stiff to bend but if you get the bending in the middle not at the evap end you shouldnt be able to crack or fold it too easily.

What case are you using? maybee your CPU is too high up..

Can you take a pic of your entire system so we can see the problem better.

Liquid3D
06-05-2003, 04:54 AM
Thank you guy's I can't tell you how much that means to me. I was isolated without getting any real feedback. I guess I'm overly cautious due to being disabled, as I'm financially indigent, and every item, must be treated like it's irreplacable.

I think the tubing is more malleable, then I'd previously thought. It's the Chip-con case, which is sort of tall. I'm waiting to borrow a dremel, or some tool to cut my Cheiftech (which is a MUCHJ better case). As you can see from the photo below, it is sort of maxed on it's reach, and this is the 8K9A, the 8RDA had a slightly higher and veritcal Socket. I can switch out thwe CPU's fairly quickly. it takes me abot 40min, depending upon whether i have to change out the mobo too.

Sometimes, it's just reassuring to hear "real" world experience. And I have checked the seal by applying paste, mounting, and removing to verify it. I did write Chip-Con and they informed me about the difference in mobo thermistor accuracy. And with the heating element, behind the CPU, wouldn't this heat the mobo's tem even more so? The microfreezer showed -45C, and MBM-5 and BIOS showed (+)10C - 15C? The Pentium kit would seem to have better contact though.

Holst
06-05-2003, 06:14 AM
Can you prod the hoze up a bit.

Im using 8rda+ so i had to rotate my evap around 90* so the hoze is all curved.

Yours does look a bit streched compared to mine.

This pic is a bit like what mine looks like

QuadDamage
06-05-2003, 09:09 AM
muzz,

insulation is a key to success.

daddy_fizz
06-10-2003, 01:28 PM
well i got my prommie today...

i am waiting to install it until i can find a solution to my reset problem.

my motherboard (Asus P4P800) is one that is on Chip-Con's list of motherboards that won't hold a reset, and power up right away...

i was still hoping someone could help me with a switch solution or some way to cut power to the cpu until the prommie is nice and cool and then turn power back on to the cpu and start her up...

at the least could anybody tell me where i could get a Prometeia Power conditioner?? that would solve my problems but i can't find it online anywhere...

~Fizz

r00t3d
08-10-2003, 03:22 AM
I am sure I mentioned the fix for Mach I reset problem, but just in case here it is again:

You can get around this by using 2 power supplies. One for booting the cooling and one for the PC.

You connect all the wires from the PC-case to the motherboard as you would normally do when using a standard cooler. Then instead of connecting the Prometeia to the 12V power line from the PC-case power supply you connect it to the spare power supply. In order to start the extra power supply you need to connect two of the pins on it's ATX connector (just like it's descried on page 7 in the manual) Then when you start the PC, you start the Prometeia by turning on the power to the 2.nd power supply and wait for it to reach -30, then you turn on the pc normally.

The advantages with this method, is that the system usually boots a little faster and it's safer for the CPU if you use a very high VCore and CPU frequency.

Just some old model power supply for booting the Prometeia will be just fine as it barely requires any power for running the 2 fans and display as the compressor itself are power from the mains circuits.


:D

Liquid3D
08-10-2003, 07:23 PM
Sorry this is off-topic but I have to ask because were on this page here. OK when I got my new unit about five months ago I'd wish I'd read your review more thoroughly. I had my AMD in the Prommie at default and the system showed temps (MBM-5) of 10C (+). I'd tried to re-position that dam micro-freezer several times, but it was just under too much counter pressure. Meaning, whether I used my 8K9A (horzintal Ziff) or the 8RDA (vertical Ziff) there was so much counter twisitng on the hose that the microfreezer was not contacting the core as well as it should. I checked the first Promiie that arrived, which was broken in (it was damaged in shipping that's how I got a new one) and it's hose was much more flexible. SO I guess my hose was stiff being new.

Still my on-die temp was high for default. Anyway iwaited for my 2.4C it came, I benched it on air, as I felt I should use the Thermalright SLK900U they sent (by the way this has to be the best heatsink on the market along with MCX4000, and Alpha PAL). When I tried it I has excellent contact with the P4, started it up, the Prommie dropped to -33C, system shut down, and I haven't been able to start it again? The P4 was at deafult, and the BIOS alarms were going off while the Prommie cooled down, but I had that happen with the 8RDA as well.

My question: Is there a circuit breaker in the Compressor, which may have tripped? Is there a fuse which may have blown? The unit has never been dmaged, jolted, or anything, it just died. I tried the bypass test metyhod as described in their manual, which worked when I tested the unit upon it's arrival, but it won't work nopw. I don't hear a peep from the compressor? ANy idea's?