Crystalfontz 635 Review

[Joe]

Post questions, answers, correct me, flame me, but what ever you do read this:

http://www.procooling.com/index.php?...icles&disp=139



Enjoy the review!

[BillA]

come on Joe

good sensor specs, ok
-> but give us some nominal 'out of the box' accuracy #s
(real data, compare to something - cal your Extech ?)
'cal' in software (offset ?) or hardware (trim pot ?)

if it is going to measure we need to assess how well it does so
yea, I saw the bigger screen

[Joe]

pH did some tests but nothing calibrated, I wanted to do something but I have nothing thats nearly ready to test those kind of numbers.

I dont have any TC's for the extech to do any comparison if I wanted to. I can only work with what I have, I have an IR gun thats been beat up nicely and I dont trust for any accurate work, and I dont have a calibrator for the Extech.

I will probably run tests like you want down the road once I get that stuff, but I am not going out of pocket for test gear for the site. The site needs to pay for it. (just good business sense, since I am not running the site to go further in debt)

[BillA]

want a loan of a Fluke 2190A for a month or so ?
just back from cal

[Joe]

Sure, I am not going to pass up an offer to get some real numbers for these devices!

[Joe]

Bill you read this thread:
http://www.crystalfontz.com/forum/sh...?threadid=2142

Very interesting, I like the using grouped sensors to read a single source for cross reference.

[crystalfontz]

Joe:

Thanks for putting up the review.

The fan connector specs are in the SCAB data sheet, which is separate from the CFA-635 data sheet.

CFA-635 Data Sheet

SCAB Data Sheet

[Joe]

ahh thanks Crystalfontz (whether thats you Brian or Brent )

I will add the SCAB Data sheet to article

[jaydee]

Originally Posted by Joe

Bill you read this thread:
http://www.crystalfontz.com/forum/sh...?threadid=2142

Very interesting, I like the using grouped sensors to read a single source for cross reference.

It is almost necessary actually with this and the CF633. The probes are just not reliable for accuracy as they are. I like the group of 3 approach. Do that for every measurement and then group all your groups of 3 and calibrate in a water bath/stirrer of some sort. IMO that is the only way to test with these and be somewhat happy with the results.

Once you get everything calibrated and what not you can make your adjustments in a spread sheet and plot your data however.

Anyway that is were I was headed until I got a good thermometer and phased out the CF633 for temp readings.

The CF633 and 635 are cool products.

[crystalfontz]

How accurate of a measurement do you need?

I have never seen one of the Dallas probes read outside of its specs. We enforce the CRC on the DOW bus and over the serial link, so there are no undetected errors there.

I am not saying that there are not more accurate sensors--I'm sure there are--but I am curious what kind of an accuracy target you are after.

[jaydee]

Originally Posted by crystalfontz

How accurate of a measurement do you need?

I have never seen one of the Dallas probes read outside of its specs. We enforce the CRC on the DOW bus and over the serial link, so there are no undetected errors there.

I am not saying that there are not more accurate sensors--I'm sure there are--but I am curious what kind of an accuracy target you are after.

I am not a good one to ask that question as my needs are far more picky than most that will use your product.

This unit from ColeParmer is what I replaced the CF633 probes with and consider it the minimum spec wise for my needs as far as temp measurements go. http://www.coleparmer.com/catalog/pr...=1&sel=0850216

I also needed significantly faster response time. The CF633 was a little slow when temps change.


I got 8 probes with my CF633 and all but one was within .5C. The other was off around 5C and did not scale the same as the others. The other 7 would raise 5C and this one would raise like 2C. I can only assume it was damaged.

I found 3 probes to read almost identical and those are the one's I used. I wish I kept my logs I would post a graph of them when I calibrated them.

All in all though I really like the CF633 and it will go into the water cooled PC I am working on.

[crystalfontz]

Yep, the Dallas probes do not hold a candle to that Cole-Parmer unit!

If you have a bad probe from us, you can write our support and we will replace it. Or, since you are local, bring it by. Call first though, we are not set up for walk-in traffic.

[Joe]

the testing I hope to do on the probes coming up should be interesting.

Also temp response would be slower because the thermal diode on the DOW sensors are surrounded by resin thats pretty thick.

[BillA]

cf
you can get very different responses to your question
some here will attempt a differential measurement of an increment almost below the units resolution (coolant in-out), obviously they will be frustrated

I have not seen the units' inst., is a method described to determine the (gross) offset between sensors ?
(trim pot to 'zero' ?)
beyond that, which could be done at several temps if desired, an external reference is required which is outside of most users 'feel good' needs

[jaydee]

Originally Posted by Joe

the testing I hope to do on the probes coming up should be interesting.

Also temp response would be slower because the thermal diode on the DOW sensors are surrounded by resin thats pretty thick.

Yes, I took pH's advice and sanded my probes down and they respond a little quicker.

Also to anyone thinking about testing with the CF633/635 I would suggest getting the DS18B20 probes as opposed to the DS1822 probes that come stock with the unit.

I am not sure if CrystalFontz sells the DS18B20 probes or not.

[Joe]

Did you read the CFA-635 review Jaydee? I show clearly that it comes with the DA18B20 probes.

[jaydee]

Originally Posted by Joe

Did you read the CFA-635 review Jaydee? I show clearly that it comes with the DA18B20 probes.

I skimmed it. Have not had time to read throughly. You have indeed showed it it clearly! Nice pic.

I would like clarification however that all 635's come stock with those probes and not just a review sample or special requests for those probes. I would also like to know if the 633's now come with those or the 18B22's. I am pretty sure mine are 18B22's. Not sure if I still have one with the numbers on it.

[jaydee]

Also if you get a chance Joe take all the probes you have and put them in a water bath of some kind and log the results and post a graph or even just the log file.

This is the only one of mine I could find on the forum. All my old logs are on another computer if they still exist.

http://forums.procooling.com/vbb/att...7&d=1112226776

[Joe]

I am getting the test for this setup, I have it mostly there, and I should have some results to post by the end of the week or this weekend.

[sprucesquad]

Originally Posted by BillA

come on Joe

good sensor specs, ok
-> but give us some nominal 'out of the box' accuracy #s
(real data, compare to something - cal your Extech ?)
'cal' in software (offset ?) or hardware (trim pot ?)

if it is going to measure we need to assess how well it does so
yea, I saw the bigger screen

What actually should the 'out of the box' accuracy be?
www.sprucesquad.com