Help with Water Pump

[alec_t]

When I hooked the pumps to my truck battery and unplugged the pumps, the LED lit up solid but only one of the pumps ran.

You can run a pump with it unplugged? Congratulations, you've got the world's first successful perpetual motion machine .
Seriously though, if only one pump runs on battery I'd double-check the FPPMk2 circuit connections; also try isolating the other pump and its PDM and then try running them from battery alone. Maybe that pump is reluctant to start up at 12V?

But with the pumps plugged in, the red LED flickered.

Did you mean with the two old PS plugged in and switched on?

 

[salty joe]

I'll be famous! What I meant by unplugged the pumps was, the PS is not plugged into 110V. I realize after re-reading how unclear that is, sorry about that weak writing.

So yes, when the power supplies are powered up and the battery backup is connected, the pumps run but the red LED flickers. When I mimic a power outage by unplugging both PS, only one pump ran but the LED was solid red.
I'll carefully check every connection and component in the circuit and re-test. Thanks a bunch!

 

[alec_t]

What output voltage are you running the old PSs at?

 

[salty joe]

I have both old PS set at about 20.6V.

 

[alec_t]

Ok. Swap that 330Ω you just added for 100Ω. That should get rid of the flicker.

 

[salty joe]

Old pumps, old PSs.
I hooked everything up to my 12V truck battery again. I took readings from the terminal points on the PDM where the pumps are connected.
With battery backup connected and the PSs plugged in and the pumps running, I got 16.9-17.4V from PDM A and 20.5-20.6 from PDM B. I got the same numbers if the pumps were switched.
With the PSs unplugged, only one pump would run. I got 12.0V from PDM A and 0.5V from PDM B. PDM A would run either pump.

Ok. Swap that 330Ω you just added for 100Ω. That should get rid of the flicker.

I swapped those resistors, but that LED still flickers when the PSs are plugged in and is solid red when the PSs are unplugged and pumps are on battery backup.
I looked at every connection and confirmed the value of all resistors, caps and checked Q1 & Q2 for the part number. I could not find anything amiss. Any idea where I should look or test?

 

[KeepItSimpleStupid]

@Joe

reversed biased Fast recovery rectifiers such as 3 A 400V or more such as these: **broken link removed**

Would go reversed biased at the power input along with the varistor. They can be used for your new jacuizzi pumps as well. I think they exist already on the 24 VDC supplies you were going to use.

As for 120 VAC suppressors. That depends on the number of GFCI's in use. They are all in parallel, so if your using 5 GFCI's consider 5 IsOBARs.

One is better than none, but the internal fuse of the ISOBAR cannot blow. If the isobar shorts internally, the fuse on the isobar blows. If the fuse doesn't bow, it still is "helping" to protect the other units, but not extremely well.

 

[salty joe]

https://www.tripplite.com/en/products/model.cfm?txtModelID=120

Hey KISS, same brand, same joules. Would this give about the same protection as **broken link removed**

 

[KeepItSimpleStupid]

The one I suggested has an RFI filter in it and I think that will be good in your setup. Spikes, it would protect. RFI-No.

We talked about this before because of controls and the motor wiring. The 24 VAC switching supplies will generate EMI/RFI, so those should get the ISOBAR.

I can't remember if the new pump power supply has a switching regulator in it.

If you have a portable AM radio in the house, you should see if you are getting any appreciable interference. Use it as a detector.

Might as well do it now and find out what the radiation sources are. You may have to shield the motor wires. You might also want to just look at the switching power supply and some minimum load like the light bulb you used.

So, a combination of efforts might be a good thing.

For surges, your selection is fine. See if you need other help with EMI/RFI. It's easier to fix now than later. Shielding reduces RFI and twisting reduces EMI.

 

[salty joe]

Thanks KISS. I read a little bit and found that Edison patented the twisting of wires and it's still useful today. I will definitely check it out with an A M radio.

 

[alec_t]

I got 12.0V from PDM A and 0.5V from PDM B.

Check D3 and D4 in the FPPMk2. Whichever was feeding PDM B seems to have gone open circuit or is not connected.
For a flicker cure, try connecting a capacitor (at least 10uF) between the R5/R6/D8 junction and ground (observing polarity). The flicker could be due to ripple on the power supplies, or switching action of the LM317. If that doesn't fix the flicker I'll have another think about the LED circuitry.

 

[salty joe]

With both old PS unplugged and everything hooked up except the battery, DMM on diode mode, D3(A) showed 256 and D4(B) showed 255.

With pumps running, the V from D3 is 16.1-16.3 and the V from D4 is 16.6-16.7.

Connecting a 10u cap across the LED and ground did not affect the flicker.

I double checked the V at both old PS and got 20.9 and 21. Somehow they both seem to have drifted up a little bit. I had them both set at very close to 20.6V a couple days ago.

Thanks for thinking about this.

 

[alec_t]

D3(A) showed 256 and D4(B) showed 255.

So those diodes are ok. Which one was connected from the battery to the PDM/pump that didn't start?

With pumps running, the V from D3 is 16.1-16.3 and the V from D4 is 16.6-16.7

A good match.

Connecting a 10u cap across the LED and ground did not affect the flicker.

I'll rethink the LED drive.

I double checked the V at both old PS and got 20.9 and 21.

Again, good.

close to 20.6V a couple days ago.

Fine. Drift could be due to temperature variation.

 

[salty joe]

so those diodes are ok. Which one was connected from the battery to the pdm/pump that didn't start?
.

d3 (a)

Tried to capitolize the d and the a. Just goes back to lower case.

 

[KeepItSimpleStupid]

Like D3a. This would be the convention.

 

[alec_t]

What is providing the Vt signals for both PDMs? Are you using the feed timer circuit?

Edit: With both PSs switched off and the pumps running off battery, check that both PDMs are getting ~12V on their +24V input, ~12V on their +12V input and ~12V on their Vt input.

Here's FPPMk3 showing the revised LED drive

 

[salty joe]

The pair of PDMs for the filter pumps have a common terminal point for both Vt connections. I ran 12V to the common Vt terminal point as well as the 12V terminal point. The 12V is coming from our controller.

The feeder timer is not built yet.

Thanks much for the diagnostic and revised LED. I picked up some work so if I’m not too wiped out I’ll get after it tonight. Saturday at the latest.

 

[alec_t]

I've just spotted a bug in the FPPMk3 in post #1996 . Here's a corrected version:
**broken link removed**

Edit: Further correction....added diodes 'D8,D9' should be labelled D9,D10 respectively.

 

[salty joe]

I think my 7812 Vreg died. I get 19.6 V on the input and less than 2V on the output. It did get pretty hot when I left a wire loose and the pumps kept running. (it was either a cathode or an anode wire from the battery backup-I can’t remember). That was a couple days ago, and the system died tonight. Plus, I must have a crossed connection because before the Vreg died, The PDMs were pulsing as if on the tide or wave mode, even though I ran 12V to the PDM’s Vt and to the PDM’s 12V. At this point, the battery backup was disconnected. I removed it to try and pinpoint the source of trouble.
The board with the dead Vreg also has the tidal timer and the first wave timer(which will not be used). I originally had a Vreg for each timer, but removed one Vreg. This is one of the very first boards I built and it wasn’t the prettiest thing on earth to begin with. After cutting into it here and there, it’s a little rough. I think the best thing is to get the parts for a new tidal timer and 12V supply. Then I can install 12V terminal points for the wave timer and battery backup and anything else that needs 12V.
I have a new wave timer that runs two pairs of pumps on another board, it seems to be rock solid.
This makes me want to scream, but better now than later. I have two questions-
Does this seem like a reasonable plan and will you guys please bear with me while I order parts and get a new tidal timer built?

 

[alec_t]

Oh dear. We'll certainly bear with you.

I think my 7812 Vreg died. I get 19.6 V on the input and less than 2V on the output. It did get pretty hot when I left a wire loose and the pumps kept running.

I have to agree....it's dead.
The 'loose wire' perhaps caused uncertainty for a PDM as to whether it should be on or off, which wouldn't have helped. With the diode failures you experienced, 24V may have somehow got back to the reg and killed it.

The PDMs were pulsing

What signal/volts did you have on the 'Fault' input of the PDMs? If that was left open circuit strange things could have happened.

Edit:

I can install 12V terminal points for the wave timer and battery backup and anything else that needs 12V.

Presumably you weren't running 12V from that failed Vreg to the two filter pump PDMs? They should be getting their 12V from the FPPMk4 instead. The two sources of 12V should not be interconnected or things could fry.