Help with Water Pump

Status
Not open for further replies.
Synopsis:
Frugal Salty Joe (SJ) is building a salt water reef tank with live Coral. He built the tank too into a wall in his house.
SJ bought some 24 VDC pumps (underwater propellers, basically) off of ebay that ran off a 240 VAC to 24 VAC wall wart.
He wants to use the 24 VDC pumps to "make waves" and "simulate a tide"
SJ enlisted ETO for some help building a controller.

Initially, the problem seemed simple. A little CMOS Voodoo and a FET switch for 2 pumps at a time and your done.

alec_t bit the bullet and designed version 1 of the tide controller.
This is basically two pump pairs that run for an extended period of time and then switch. The pumps that are off, "flick" or turn on briefly so the fish won't get blown away when they turn on full force.

SJ built the controller and it worked for a few days to a week or so and one motor motor catastrophically died.
SJ dissected the motor (hammer) and sent it to me for some other analysis. The motor was powered by a 24 VDC, 6-10 amp supply with no protection, not even a fuse per pump. So, it let out the "magic smoke". There may have been fuse(s), but I think it didn't blow.

So, we found that it was a brushless DC motor and the failure mode is somewhat unknown, but one coil was shorted and solder melted where it should not have melted.

I think it uses a Melexis Hall IC and P-channel FET, and that the operation might be specified for 18 V max, but the application note says 5V to 24 VDC is permissible.. I found the inductance and resistance of the coils which allowed a better simulation. The OEMC, or OEM controller doesn't run the pumps for very long at full speed and the OEMC has some current limiting because it uses a Switchmode regulator with Imax of 3A and a peak current of 5A.

So, alec_t worked on another design with help, primarily critiques, from the peanut gallery ronv, me, and ()oblivion which requires a separate motor controller with protection for each motor.

There is a need for a FEED MODE" which slows the pumps for 15 minutes, so the discussion has a brief interlude trying to come up mechanisms for doing the feed mode. From modifying the OEMC, to a different switchmode regulator and finally alec_t's simulated and physically made design that is now an article at ETO.

Initially, the idea is to use four modified OEMC's to provide 18 VDC for the next test. I guess, we are not sure yet, if Alex's new control mode will be used for the upcoming test or not.

So, SJ went on vacation and came back and is procuring parts as we speak, so SJ can try again to see if the pumps can survive.

SJ has been asking good questions and he's not an electronics hobbyist, but does know which end of the soldering iron is hot and seems to be doing a great job.

So, we will be waiting for SJ to build a controller.
Don't know if it ill be with modified OEMC's or alec's new controller mode that reduces the average current applied to the motor, bit we are in agreement that the lowest voltage, probably 18 VDC or less is preferred.

So, we basically sit in "Hurry up and wait mode" until SJ builds something and can put it to the test.

If it doesn't work, we do have a "Plan B" which none of us like which would require replacing epoxied electronics in the motor.

The "tide sequencer" worked in terms of turning off and on the motors properly, but the motors didn't survive. There still is the design of a "wave maker" to deal with and a "feed mode" timer. Trivial in comparison as to what's been going on.

The thread is still on topic "Help with water pump"

Guys: make any corrections you may see fit. I wasn't participating in this thread in the beginning, but I then saw some struggling, so I offered help.

PS: This thread, I think, inspired me to look at aquarium filtering solutions for laundry lint. The device I chose was extremely effective, but I need to do some major plumbing to move it toward the back of the laundry sink instead of the front. On another related note, I manged to create an extremely usable silt filter for a residential drainage system that captures 99% of the silt by weight. At present, the working prototype has no bypass, but one can be added. Incremental cost without the bypass, $10, without shipping of parts. The prototype, about $80. I had to buy 500 thingys to get 3 and had to buy 5' to get 1.5" of the stuff.
 
Last edited:
Nice synopsis, KISS. Handy to have a road map every so often. It tallies with my recollections of the saga.
 
LOL, it's cool brother. I don't expect people to remember one blurb from ~8 moons ago. I just wanted to point it for self gratification... LOL
 
@KISS
Re the pots.
In this thread the 'SetMaxSpeed' trimmer is used only for setting the average voltage across the pump at ~18V. Simulation shows about 5k in series with the fixed 10k resistor does the trick; hence a 10k trimmer to allow a small range of adjustment. Since the trimmer is in series with the 10k and I anticipate the trimmer will be at ~mid-point I figure a linear one is more suitable than a log one. (But it makes no real odds if it's linear or log.)
In the Article I was allowing for a more versatile, wider range of adjustment without fixed resistors in series with the pots, to allow a greater choice of both max and min speeds in various applications. For a wide adjustment range a log pot is arguably more suitable than a linear one.

@Joe
It's my understanding (from some other thread on this ETO site) that the anti-moisture precautions are for commercial setups where components (any type) get baked and re-flow soldered (steam generated inside/under a component could be problematical). The concensus was that they can be ignored for hobby use where components are soldered manually.

@all
I'm suspicious about the second link in post #928.
It may just be concidence, but when I clicked on the link ZoneAlarm (Internet security app) warned that one of its security services had been shut down.
 
Last edited:

If certain electronic components absorb moisture, they can "pop corn", which literately means that the moisture locked inside boils rapidly enough to POP the component. The same way pop corn pops. It *can* be a concern for hobby use, but as stated it's generally more about the re-flow process than anything else. The problem is easy to fix either way. Simply keep them in the sealed package with the moisture absorbent packet until ready to solder them in place. If you open the package and the moisture indicator says it has been wet, or you left it open for a long time, you can literately bake your parts in an oven on low heat for an hour or two. If you have access to a food dehydrator, you can use that too. With this method it's best to leave it in for longer, such as over night.

Edit: If you have opened the bag already. You can put all the parts back in the bag with the moisture absorber, suck out all the air, and make shift seal it back up with tape. When you open it again, the moisture indicator should say it went back down.

Alec_t: What's a mail box? said:
I'm suspicious about the second link in post #928.
It may just be concidence, but when I clicked on the link ZoneAlarm (Internet security app) warned that one of its security services had been shut down.

Hummmmm interesting.... Comodo suit doesn't seem to care much... :shrug:... It could be the fancy map app toward the bottom of the listings. Also... those fuse comes from Bangkok Thailand anyway and will take a while to get shipped. I would just go to the local hardware store and get an assorted set Joe.

(P.S. The title giving is just for kicks, feel free to label me if you like.)
 
Last edited:
Thanks guys. So the moisture problem is only when the fuse is soldered? That would be good because without a doubt they will be exposed to high humidity, at least sometimes. The package is still sealed.

I got all the parts and will start to work this weekend. A mistake I won't make again is soldering while tired.

I plan to build four copies from post 812 to drive two pairs of pumps. If need be, could this setup drive two sets of three pumps?

The second link is from Ebay. I have McAfee plus a couple anti-spy/malware programs-everything seems OK from here, but you never know.
 
If need be, could this setup drive two sets of three pumps?
Shouldn't require much of a tweak for that. Of course, you'll need 6 PDMs rather than 4. But get 1 working first .
Re the second link. ZoneAlarm (or Windoze) was perhaps having an off-day if no-one else had a problem with it.
 
@All
No problem with the second link. I'm running Microsoft Security Essentials (free) and Mallwarebytes (monimal - updates for life). I only had to add a few EXCEPTIONS for Malwarebytes. Skype is one of them. Another is free program that allows mounting of CD images.

@Joe
Do you know the difference between a linear and a LOG pot? LOG pots are generally used for volume controls because gain and perception of loudness is not a linear function.

@ 2 pairs of 3 pumps
If they need 5 A to start, then you'll need a 15 A power supply which you have. If worse come to worse, the pump start times could be staggered slightly or get a bigger supply.
 
Last edited:
Synopsis:
SJ dissected the motor (hammer) and sent it to me for some other analysis.
Not just a hammer-a hammer and a torch. That's the only correction I could find.

SJ has been asking good questions and he's not an electronics hobbyist, but does know which end of the soldering iron is hot and seems to be doing a great job.
Thanks, that's nice of you to say.


Nice synopsis, KISS. Handy to have a road map every so often. It tallies with my recollections of the saga.
+1

@Joe
Do you know the difference between a linear and a LOG pot? LOG pots are generally used for volume controls because gain and perception of loudness is not a linear function.


No. I had not even noticed. Thanks for pointing out that difference. the 50K pot I have for the Alec 812 is logarithmic. Should I get a 50K linear? What about the logic circuit-should those pots be log or linear?

The only writing I could find on my 50K log pots is A 50K. Does the A mean that it is log?
 
Last edited:
Easy way to check: Center the adjustment and measure the wiper resistance to any end. It will be way off from 25K if LOG. My guess, it will be about 8 or 42K.

Most pots are linear.

From Wikipedia: https://en.wikipedia.org/wiki/Potentiometer
Resistance–position relationship: "taper"

The relationship between slider position and resistance, known as the "taper" or "law", is controlled by the manufacturer. In principle any relationship is possible, but for most purposes linear or logarithmic (aka "audio taper") potentiometers are sufficient.

A letter code may be used to identify which taper is used, but the letter code definitions are not standardized. Newer potentiometers will usually be marked with an 'A' for logarithmic taper or a 'B' for linear taper. Older potentiometers may be marked with an 'A' for linear taper, a 'C' for logarithmic taper or a 'F' for anti-logarithmic taper. When a percentage is referenced with a non-linear taper is relates to the resistance value at the mid-point of the shaft rotation. A 10% log taper would therefore measure 10% of the total resistance at the mid point of the rotation; i.e. 10% log taper on a 10K ohm potentiometer would yield 1K at the mid point. The higher the percentage the steeper the log curve[2]

I'm not sure which end is up
 
Last edited:
I shouldn't worry, Joe, whether your pots are log or linear. It's just that sometimes the 'useful' bit of the pot adjustment is rather cramped towards one end, so a log pot can be used to spread things out and make adjustment to find the 'sweet spot' less critical.
Does the A mean that it is log?
I'll display my ignorance here and say I don't know. It could well stand for 'Audio taper' (another name for log). On the other hand.....
Anyone else know?

Edit: KISS's post arrived while I was composing mine. So someone else DOES know
 
Last edited:
()blivion, thanks for your points, although I'm not trying to be negatively critical -- I was simply curious as to how things got so abnormally long. I'm all for the thread.

KISS, thanks for the synopsis.

I notice that about a full page of subsequent posts happened since my post! Wild!

Oops! The wife tells me I have a life and there are things I need to be doing. I'm outta here.
 
Status
Not open for further replies.
Cookies are required to use this site. You must accept them to continue using the site. Learn more…