Fan experimentation, home made GFB
While messing around and reading at OC Forums someone started a thread asking about stacking fans. After many answers to the OP, the consensus was that it wasn't worth the trouble to even try unless you could meet all the parameters that Delta has followed with the design of their GFB series of fans. These parameters consist of 3 main things:
1. opposite rotation of the impeller assemblies. This improves efficiency and counteracts harmonics problems.
2. different blade counts for the individual impellers. This is to counteract harmonics problems
3. different rpm ratings of the individual impellers. This is also to counteract harmonics problems.
All three of these steps are also used in design of wind tunnels too.
I own a Delta GFB1212VHW 120 X 76 mm fan already and it is a tremendously effective, but loud fan. I have never run across a fan that concentrates its airflow as well as it and it has almost no "dead zone" even close to the outflow hub. The drawback to the GFB is the noise and the fact that to control noise and speed you have to get each impeller wired individually for use on a regualr fan controller or buy a monster sized fan controller that can handle almost 32 watts of power draw.
While looking around when preparing an answer for that thread, I happened to notice that the Delta EFB1212LE 3 blade fan, which is a 120 X 38 mm fan, turns in opposite rotation from your normal 120 X 38 mm fans such as the Panaflo fans. That gave me the idea to combine it with another fan such as the Panaflo FBA12G12L1BX 120 X 38 mm fan that has 7 blades on its impeller. The Delta is rated at 76 cfm and 34 dB of noise and the Panaflo is rated at 69 cfm at 30 dB of noise. Since I was ordering a Thermalright Ultra 120 eXtreme anyways I went ahead and ordered the Panaflo from Sidewinder with the U-120 eX and then ordered the Delta from Performance PCs. I got them in on the same day and used some 6-32 bolts and nuts to bolt the 2 fans together with the Delta feeding the Panaflo. They lined up real well and the assembly looked good, so it was now time to test the compound fan out. I hooked the fan to an old AT PSU I keep around for fan testing and powered up and it was a success. The homemade compound fan showed much the same flow characteristics as the GFB I own but with a lot less noise and lower total airlflow. The airflow out of the compound fan was still concentrated and with nearly no "dead zone" and you could actually still feel air movement at a 6 foot distance from the fan. The initial test was a success, so now it was time to test the compound fan in real world usage. I went ahead and pulled my main rig down, which is an E6600 on an Asus P5W DH Deluxe board. I was showing loaded temps with this machines in the 60-64 C range while cooling it with a Thermaltake Big Typhoon with a Panaflo M1A fan on it. I went ahead and installed the U-120 eXtreme and hooked my newly built compound fan to the wires and the U-120's fan wires held the big compound fan assembly very well, considering the mass of the compound fan. While I wouldn't want to transport my computer in a car with the fan on the heatsink, it feels perfectly secure for usage in a stationary environment. After installing the mobo/HSF assembly back into the case I went ahead and hooked the fan connectors to 2 of the fan headers on the mobo and then fired the computer back up. After checking in BIOS to make sure that everything was looking OK on heat I went ahead and booted back into Windows. Then new U-120/compound fan made quite a difference in cooling too. My loaded temps dropped from the 60-64 C range all the way down to the 47-51 C range, all at the same overclock. And the compound fan is no noisier than the old BT/M1A combination was either. So it is mission success with the home made compound fan. And unlike the GFB I own it can be powered off mobo fan headers without worry of burning anything out, because neither fan draws more than 2 1/4 watts each.
1. opposite rotation of the impeller assemblies. This improves efficiency and counteracts harmonics problems.
2. different blade counts for the individual impellers. This is to counteract harmonics problems
3. different rpm ratings of the individual impellers. This is also to counteract harmonics problems.
All three of these steps are also used in design of wind tunnels too.
I own a Delta GFB1212VHW 120 X 76 mm fan already and it is a tremendously effective, but loud fan. I have never run across a fan that concentrates its airflow as well as it and it has almost no "dead zone" even close to the outflow hub. The drawback to the GFB is the noise and the fact that to control noise and speed you have to get each impeller wired individually for use on a regualr fan controller or buy a monster sized fan controller that can handle almost 32 watts of power draw.
While looking around when preparing an answer for that thread, I happened to notice that the Delta EFB1212LE 3 blade fan, which is a 120 X 38 mm fan, turns in opposite rotation from your normal 120 X 38 mm fans such as the Panaflo fans. That gave me the idea to combine it with another fan such as the Panaflo FBA12G12L1BX 120 X 38 mm fan that has 7 blades on its impeller. The Delta is rated at 76 cfm and 34 dB of noise and the Panaflo is rated at 69 cfm at 30 dB of noise. Since I was ordering a Thermalright Ultra 120 eXtreme anyways I went ahead and ordered the Panaflo from Sidewinder with the U-120 eX and then ordered the Delta from Performance PCs. I got them in on the same day and used some 6-32 bolts and nuts to bolt the 2 fans together with the Delta feeding the Panaflo. They lined up real well and the assembly looked good, so it was now time to test the compound fan out. I hooked the fan to an old AT PSU I keep around for fan testing and powered up and it was a success. The homemade compound fan showed much the same flow characteristics as the GFB I own but with a lot less noise and lower total airlflow. The airflow out of the compound fan was still concentrated and with nearly no "dead zone" and you could actually still feel air movement at a 6 foot distance from the fan. The initial test was a success, so now it was time to test the compound fan in real world usage. I went ahead and pulled my main rig down, which is an E6600 on an Asus P5W DH Deluxe board. I was showing loaded temps with this machines in the 60-64 C range while cooling it with a Thermaltake Big Typhoon with a Panaflo M1A fan on it. I went ahead and installed the U-120 eXtreme and hooked my newly built compound fan to the wires and the U-120's fan wires held the big compound fan assembly very well, considering the mass of the compound fan. While I wouldn't want to transport my computer in a car with the fan on the heatsink, it feels perfectly secure for usage in a stationary environment. After installing the mobo/HSF assembly back into the case I went ahead and hooked the fan connectors to 2 of the fan headers on the mobo and then fired the computer back up. After checking in BIOS to make sure that everything was looking OK on heat I went ahead and booted back into Windows. Then new U-120/compound fan made quite a difference in cooling too. My loaded temps dropped from the 60-64 C range all the way down to the 47-51 C range, all at the same overclock. And the compound fan is no noisier than the old BT/M1A combination was either. So it is mission success with the home made compound fan. And unlike the GFB I own it can be powered off mobo fan headers without worry of burning anything out, because neither fan draws more than 2 1/4 watts each.
0
Comments
EDIT: Here's a shot of it in action on my computer.