Help with Water Pump

[salty joe]

Thanks guys-really appreciate you chiming in. 12V it is for the Vt on the filter pumps.

Do you think it would be a good idea to pigtail the 24V and power ground lines from every power supply that came with the new pumps before sending 24V and PG to the controller? I'm afraid if I use 24V from only one new power supply, everything will shut down if that particular power supply fails.

https://www.electro-tech-online.com/attachments/tidalsystem-gif.71768/

 

[alec_t]

Good idea re backup. However, you can't directly connect all 24V lines together.....the supplies may fight each other. Instead, you could connect them via respective diodes (1N4148 will do) to the 24V terminal of the controller so that you have backup. In other words you would have the cathodes of 4 diodes connected to the controller 24V input; the anodes of the diodes would be supplied respectively by the four power supplies. The power ground lines are already connected together in the schematic.

 

[salty joe]

Very cool, thanks.

 

[alec_t]

I don't know about cool, but it's just occurred to me that with all these pumps it would only need a powerful heater to convert your tank to a luxury whirlpool spa tub . Of course, it could prove painful if you inadvertently sat on the coral!

 

[KeepItSimpleStupid]

1n4148? https://www.electro-tech-online.com/custompdfs/2013/05/1N4148_1N4448.pdf I max is only 450 mA.

Isn't joe dealing with more or am I missing something?

 

[alec_t]

Isn't joe dealing with more or am I missing something?

The 24V input to the timer/controller is only used to derive the 12V supply for the logic ICs, so only a few mA needed.

 

[salty joe]

I don't know about cool, but it's just occurred to me that with all these pumps it would only need a powerful heater to convert your tank to a luxury whirlpool spa tub . Of course, it could prove painful if you inadvertently sat on the coral!

There should be some serious water movement. I anticipate tuning the pumps down a lttle bit.

If the light sensor is not installed on the new controller, the pump runs for ½ sec, hesitates, and then runs. That’s with the new controller hooked up with our controller. With the light sensor installed, everything works fine. I don’t remember that happening with only the new controller and no light sensor. Would jumping the connections to the light sensor make the new controller think that it is not dark? (When it’s dark, the pump runs on low) Anyway, very minor glitch.
Here is the final configuration.
Two pairs of the new pumps in wave mode and two pairs of the new pumps in tidal mode. I connected eight diodes to tie all eight 24V lines together-works beautifully. Thank you Alec for helping me get backup for the pumps.
Two old pumps on a dedicated 21V power supply for the filter pumps. I’d like to back these two pumps up with a battery.
One old pump on another 21V power supply.

Is it OK to supply the PDMs for the three old pumps with 12V and signal ground from our controller for the eight new pumps?

I’m getting the jack plugs for the new power supplies soldered on the PDMs, giving them a test run, I’m halfway there. I have all the PDMs built, but one needs de-bugged.
Thanks guys for all your help.

 

[KeepItSimpleStupid]

It might be that the light sensor is resistive or it could be a phototransistor.

You might try putting an ohmmeter across the sensor. If it's polarity insensitive then it's probably resistive and could be replaced with a resistor.

If it's a photo-transistor, then things could get a bit more complicated. Are their any numbers on the light sensor?

I'd suggest replacing it with an optocoupler or a resistor.

Dark is likely less current, so when it's removed, it probably thinks its dark.

Are you sure your not making a Jacuzzi?

 

[alec_t]

The 1/2 sec glitch is a puzzle. Perhaps the micro in the new controller checks at power-up to see if the sensor is present. Does the sensor have only 2 pins on its connector (which would suggest it's just resistive)?

Would jumping the connections to the light sensor make the new controller think that it is not dark?

Directly shorting them is not recommended. Bridging them with a suitable resistor may be ok. We need to find out a bit more about the sensor first.

all eight 24V lines together-works beautifully

Now that's what I call real backup!

Is it OK to supply the PDMs for the three old pumps with 12V and signal ground from our controller for the eight new pumps?

Sure.

a Jacuzzi?

That's the word I was looking for

 

[salty joe]

Here's a picture of the light sensor. When it has light shining on it, my DMM says it has 1.4K ohm resistance. In the dark it registers around 27K. So it looks like replacing it with a 1.4K resistor to defeat the slow night mode might be the way to go, like you said KISS. Thanks KISS..

 

[KeepItSimpleStupid]

Definitely a resistive sensor. Like these: https://www.adafruit.com/products/161#Description

So, resistive substitution will work.

 

[salty joe]

When everything is hooked up to a pair of PDMs as shown in the link except all 24V lines are fed into the anodes of 8 1N4148 diodes, (8 pumps, 8 24V lines, 8 diodes)a single power supply runs both pumps. The cathodes of the 8 diodes are connected together to feed our controller.

When everything is hooked up to a pair of PDMs as shown in the link, https://www.electro-tech-online.com/attachments/tidalsystem-gif.71768/ the only pump that runs is the one connected to the PDM that has 24V going to our controller. That’s with two power supplies hooked up.
In both cases, the alarm shuts down pumps independently, and there are no apparent heat issues.

I got the same results with all PDM pairs.
I found one deviation from the PDM-Mk15. I moved the IC and left C2 connected across R11 and R12. I added a 0.1 uf cap across the IC.

I am at a loss as to what’s going on. If you guys could take a look at this, I’d be forever grateful. If you want the PDM-Mk15 schematic, please let me know. I’ll get on another computer to find it. The "search this thread" function does not work on my computer.

 

[alec_t]

I'm posting this from a hotel in Italy. Will look into things when I get back to my home PC, if someone doesn't beat me to it!

 

[salty joe]

Italy-wow that sounds great! Thanks and enjoy your break.

 

[alec_t]

Hi, I'm back! Need a rest now to get over the hol .

When everything is hooked up to a pair of PDMs as shown in the link except all 24V lines are fed into the anodes of 8 1N4148 diodes, (8 pumps, 8 24V lines, 8 diodes)a single power supply runs both pumps

Is that 8 pumps or 2 you've got hooked up?

a single power supply runs both pumps

Each new pump should have its own OEM supply.

the only pump that runs is the one connected to the PDM that has 24V going to our controller

The OEM supplies, not the PDM, should provide the 24V to our controller via the diodes. Here's how I see it:-

I found one deviation from the PDM-Mk15. I moved the IC and left C2 connected across R11 and R12. I added a 0.1 uf cap across the IC.

That's fine.

 

[salty joe]

Thanks Alec, I'll go over it with a fine toothed comb after work today.

 

[salty joe]

Hey Alec, hope you had a great time, welcome back, I’m glad you’re back.
I have been working with only a pair of PDMs and two new pumps. When hooked up as shown in your link, the pumps work as expected. Each pump is driven by its own PS. If the PS that has 24V running to our controller is unplugged, both pumps die, so I tried the backup.
I connected the cathodes of 8 diodes and ran that to our controller’s 24V. I connected the 24V from both PS to individual anodes. A single PS plugged into either PDM makes both pumps run. Is that OK?

 

[alec_t]

A single PS plugged into either PDM makes both pumps run.

Something wrong there. If one PS (I assume you mean the new PS which came with the new pump) is trying to drive 2 pumps it is overloaded. Presumably it has internal protection to prevent damage. I don't see how, if the setup is as in post #1915, 24V gets from a single PS to more than one PDM (and hence via a respective OEM controller to a respective pump)?

If the PS that has 24V running to our controller is unplugged, both pumps die, so I tried the backup.

I'm puzzled by your reference to the PS. If you mean one of the old PS's then that should be removed; the setup shown uses only the new PS's, with their outputs 'OR'ed by the 1N4148 diodes to provide a backed-up 24V to the voltage regulator in our controller.

 

[salty joe]

Sorry for the PS confusion. I am working with two new PS, two PDMs, our controller, two new controllers and two new pumps.

The PDMs have separate 24V and power ground lines. They share signal ground, 12V, and Vt lines.

With everything hooked up according to post 1915, each pump seems to run on its own PS. If the PS that does not have 24V connected to our controller is unplugged, only that pump stops. If the PS that has 24V connected to our controller is unplugged, both pumps stop.
The only change made was to take 24V from each PS and connect them to individual anodes of 1N4148, then running a wire from the connected cathodes to our controller. Then, as soon as a PS is connected to either PDM, both pumps run.

 

[alec_t]

If the PS that does not have 24V connected to our controller is unplugged, only that pump stops.

This is where I'm getting confused. Don't all the PS's have 24V connected (via diodes) to our controller?