Pro/Forums - View Single Post - Auto-reversing pump?

RoboTech · 10-25-2002, 11:10 AM

Hi guys,

This is a very interesting topic…

(I haven’t seen it talked about before)

A while back when I first started tinkering with water-cooling I noticed that all of the pumps I was using incorporated the same basic design. They all had very loose impeller-to-case clearances and center outlets. In addition, all but the E-1250 had the impeller mounted on the rotor shaft so that it could slip through at least 120 deg. of shaft rotation before hitting a hard stop and becoming directly coupled to the motor rotation. After a little web research and experimenting with my various pumps I have come to several conclusions.

1. Most pumps used in water-cooled PC applications were originally designed for aquarium and/or garden pond applications. They have relatively high flow but very low discharge pressure. They are also designed to be relatively inexpensive but long lasting pumps.

2. The basic design of the Eheim 1048 and 1250, ViaAqua1300 and Danner Supreme2 (these are the pumps I personally have) all use single phase, two-pole synchronous AC motors. For simplicity and low cost, the motor stator consists of two windings (poles) wired together and connected to the AC mains. The stator assembly is potted in a waterproof material. The rotor consists of a two-pole permanent magnet. One side of the rotor is magnetic North and the other side is magnetic South. The rotor has plastic bearings (lubricated and cooled by water), which turn on a plastic shaft held captive in the pump housing between two rubber bushings. The plastic impeller wheel is fitted directly onto the rotor shaft. (Technically these are NOT mag-drive motor/pumps. The motor has a permanent magnet rotor, but the impeller is not coupled to the motor shaft with magnets, as they are in a true mag-drive.)

3. Single phase, two-pole AC motors are relatively easy and inexpensive to manufacture and they lend themselves very well to wet environment applications. They have very low starting torque and will randomly start rotating in different directions each time they start up depending on the resting position of the rotor and angle of the AC waveform when the stator coils are energized. More expensive AC synchronous motors used in industrial applications incorporate multiple poles or a resistance-capacitance (R-C) network or some other timing circuit to shift one or more stator windings out of phase, which will force the motor to always turn in one direction or the other.

4. Our PC pond pumps don’t have any starter circuitry (keep it $imple). Instead the pump case is designed to work with the rotor turning in either CW or CCW direction – center outlets. The down side to this is that it makes the pump very inefficient. As the water-filled impeller turns, water is thrown outwards due to centrifugal force. But the water is also rotating, so ideally a good pump design will capture both the centrifugal flow and inertial flow of the water. This is why more expensive pumps have a volute with a tangential discharge. Volenti modified his pump from a center discharge (centripetal flow only) to a tangential discharge (centripetal AND inertial flow) with GREAT results!

5.The clearances between the impeller and pump casing walls is very loose in a pond pump so that it can ingest small amounts of dirt and gravel and keep on pumping. Because these clearances are so large, they also provide an alternate path for water flow inside the pump. A good percentage of the water sucked in and spun up to speed just leaks back to the inlet and never leaves the pump. This is why this type of pump has such a low discharge head – put a little resistance to the flow of water exiting the pump and the water will just quit flowing out, and instead will flow (leak) back to the center of the impeller.

6. The plastic impeller wheel is mounted on the rotor shaft so that it can slip through at least 120 deg. of shaft rotation before hitting a hard stop and becoming directly coupled to the motor rotation. The Eheim 1250 is the only pump I have that has the impeller permanently mounted to the rotor shaft (no slipping). Many people believe this is to protect the impeller against a foreign object being sucked into the pump. I disagree. I believe the purpose is to allow the rotor to start turning (without any drag from the impeller), to insure the motor can start. Remember, this type of motor has a very low starting torque. Once rotation starts (even 120 deg.) inertial kicks in and off it goes, right up to full speed. When the rotor is at full speed and pumping water, the water is pushing back against the impeller blades, resisting the motor shaft rotation and holding the impeller against the hard stop. The only way the impeller can slip on the rotor shaft at this point is if it spins FASTER than the rotor. If a piece of gravel gets ingested and wedges between an impeller blade and the case wall the impeller can’t slip backwards (it can only slip forwards). My theory is that loose clearances protect the pump and impeller slipping allows easier startups.

7. Because all of these pumps have their stator windings and wiring connections potted inside the plastic case, it does not easily allow adding circuitry to the motor to force it to start in the same direction each time. Somebody smarter than me will have to figure that one out…

Note: I am by no means an expert on any of this stuff and don’t pretend to be. I am not trying to lecture anyone, just sharing some info that I think might be of interest. I’m a noob around here so hope no one takes this the wrong way. And yes, I cut and pasted from my notes…

RoboTech

Edit: Changed "radial" flow to tangential flow in section 4.

10-25-2002, 11:10 AM	#21
RoboTech Cooling Savant Join Date: Sep 2002 Location: Cincinnati, OH Posts: 229	Hi guys, This is a very interesting topic… (I haven’t seen it talked about before) A while back when I first started tinkering with water-cooling I noticed that all of the pumps I was using incorporated the same basic design. They all had very loose impeller-to-case clearances and center outlets. In addition, all but the E-1250 had the impeller mounted on the rotor shaft so that it could slip through at least 120 deg. of shaft rotation before hitting a hard stop and becoming directly coupled to the motor rotation. After a little web research and experimenting with my various pumps I have come to several conclusions. 1. Most pumps used in water-cooled PC applications were originally designed for aquarium and/or garden pond applications. They have relatively high flow but very low discharge pressure. They are also designed to be relatively inexpensive but long lasting pumps. 2. The basic design of the Eheim 1048 and 1250, ViaAqua1300 and Danner Supreme2 (these are the pumps I personally have) all use single phase, two-pole synchronous AC motors. For simplicity and low cost, the motor stator consists of two windings (poles) wired together and connected to the AC mains. The stator assembly is potted in a waterproof material. The rotor consists of a two-pole permanent magnet. One side of the rotor is magnetic North and the other side is magnetic South. The rotor has plastic bearings (lubricated and cooled by water), which turn on a plastic shaft held captive in the pump housing between two rubber bushings. The plastic impeller wheel is fitted directly onto the rotor shaft. (Technically these are NOT mag-drive motor/pumps. The motor has a permanent magnet rotor, but the impeller is not coupled to the motor shaft with magnets, as they are in a true mag-drive.) 3. Single phase, two-pole AC motors are relatively easy and inexpensive to manufacture and they lend themselves very well to wet environment applications. They have very low starting torque and will randomly start rotating in different directions each time they start up depending on the resting position of the rotor and angle of the AC waveform when the stator coils are energized. More expensive AC synchronous motors used in industrial applications incorporate multiple poles or a resistance-capacitance (R-C) network or some other timing circuit to shift one or more stator windings out of phase, which will force the motor to always turn in one direction or the other. 4. Our PC pond pumps don’t have any starter circuitry (keep it $imple). Instead the pump case is designed to work with the rotor turning in either CW or CCW direction – center outlets. The down side to this is that it makes the pump very inefficient. As the water-filled impeller turns, water is thrown outwards due to centrifugal force. But the water is also rotating, so ideally a good pump design will capture both the centrifugal flow and inertial flow of the water. This is why more expensive pumps have a volute with a tangential discharge. Volenti modified his pump from a center discharge (centripetal flow only) to a tangential discharge (centripetal AND inertial flow) with GREAT results! 5.The clearances between the impeller and pump casing walls is very loose in a pond pump so that it can ingest small amounts of dirt and gravel and keep on pumping. Because these clearances are so large, they also provide an alternate path for water flow inside the pump. A good percentage of the water sucked in and spun up to speed just leaks back to the inlet and never leaves the pump. This is why this type of pump has such a low discharge head – put a little resistance to the flow of water exiting the pump and the water will just quit flowing out, and instead will flow (leak) back to the center of the impeller. 6. The plastic impeller wheel is mounted on the rotor shaft so that it can slip through at least 120 deg. of shaft rotation before hitting a hard stop and becoming directly coupled to the motor rotation. The Eheim 1250 is the only pump I have that has the impeller permanently mounted to the rotor shaft (no slipping). Many people believe this is to protect the impeller against a foreign object being sucked into the pump. I disagree. I believe the purpose is to allow the rotor to start turning (without any drag from the impeller), to insure the motor can start. Remember, this type of motor has a very low starting torque. Once rotation starts (even 120 deg.) inertial kicks in and off it goes, right up to full speed. When the rotor is at full speed and pumping water, the water is pushing back against the impeller blades, resisting the motor shaft rotation and holding the impeller against the hard stop. The only way the impeller can slip on the rotor shaft at this point is if it spins FASTER than the rotor. If a piece of gravel gets ingested and wedges between an impeller blade and the case wall the impeller can’t slip backwards (it can only slip forwards). My theory is that loose clearances protect the pump and impeller slipping allows easier startups. 7. Because all of these pumps have their stator windings and wiring connections potted inside the plastic case, it does not easily allow adding circuitry to the motor to force it to start in the same direction each time. Somebody smarter than me will have to figure that one out… Note: I am by no means an expert on any of this stuff and don’t pretend to be. I am not trying to lecture anyone, just sharing some info that I think might be of interest. I’m a noob around here so hope no one takes this the wrong way. And yes, I cut and pasted from my notes… RoboTech Edit: Changed "radial" flow to tangential flow in section 4. Last edited by RoboTech; 10-25-2002 at 04:33 PM.