Help with Water Pump

[alec_t]

The pump only spins when the power supply is first plugged in

For how long (roughly) does it spin?
Does the FET get hot?

Any idea what the problem might be?

Possibly the 3A current-limiting isn't allowing enough grunt. Try with R3 or D5, or the link from D5 to R1/R2, temporarily disconnected to disable the limiting.

This happened for both the flick and tide.

For testing the pump drive module don't use the timing function of the wave or tide timer; just disconnect the left end of R1 from the timer and connect it to the +12V supply line directly.

 

[alec_t]

Champagne all round? (Or should we hold off on that until Joe has the pump working?). Guinness Book of Records? I see the thread has hit 1000 posts

 

[KeepItSimpleStupid]

Yea. Champagne for everyone. 1000+ posts is pretty impressive.

I like Joe's troubleshooting skills. If it doesn't work, build another. Different approach.

 

[salty joe]

"Champagne all round?"

If champagne for everyone, I'm buying and Alec gets a case!

"I like Joe's troubleshooting skills. If it doesn't work, build another. Different approach."

Yeah, but this time I am trying to figure it out instead of scrapping it. But that one driver problem is on the back burner.

U1 is soldered to board.
The suggested tests were for the 1 driver out of 6 that is not working. For the moment, I'm going to forget about the problem driver. Will the same tests help diagnose why the pump only spins when the power supply is first plugged in, even though the LED continues to toggle?

edit- Alec I missed your post 1001- I'll check it out. Thanks.

 

[KeepItSimpleStupid]

The ideas were to determine if the FET works first, I think you accidentally verified that it does, by saying it spins on power up.

The next tests, alex and I were to disable the protections so the system should work unprotected. Two different techniques of disabling. If you give up, you can always send me the driver only with a cache of spare parts.

For now, I agree, you have bigger fish to fry. Will it work?

@alex
What is going to be your suggested procedure for setting the trimmer?

Something like adjust until it trips every time and gradually increase VR1 until it just doesn't or adjust to minimize the nusence trips?

 

[salty joe]

The pump will spin only for the initial toggle. I let it go for ten minutes with the pump hooked up but not running and LEDs toggling, and the FET did not get warm at all.

"For testing the pump drive module don't use the timing function of the wave or tide timer; just disconnect the left end of R1 from the timer and connect it to the +12V supply line directly."

The pump came on and I let it run for ten minutes. The FET did not get warm at all.

Then with R3 disconected, the pump toggles along with the LED. You knew right where to look Alec - you are amazing!

 

[KeepItSimpleStupid]

With R3 out, look at the voltage across the trip test switch. Should be 12 V and drop to zero when pushed.

 

[alec_t]

The pump came on and I let it run for ten minutes. The FET did not get warm at all.
Then with R3 disconected, the pump toggles along with the LED.

That's all good news. So it looks like the current-limiting needs tweaking.

What is going to be your suggested procedure for setting the trimmer?
Something like adjust until it trips every time and gradually increase VR1 until it just doesn't

Yup, that's precisely what I had in mind. If set at the minimum it trips almost immediately; which is what seems to be happening now.
Joe, I suggest you set the 50k trimmer (I think that's the VR1 KISS is referring to) at mid-point, reconnect R3 and see what happens.

 

[KeepItSimpleStupid]

You can kinda do the bisecting method. I doubt you marked the CC direction on the schematic. In other words R and the time should increase as R is moved clockwise.

Mid point 1st
Work: Try 3/4 or 1/2 between 50% and 100%
No work: Try 1/2 between 0 and 50% or 25%
Suppose it worked between 50% and 100%
Then try 1/2(50+100) or 75%
If it still works, then try 1/2(75%+100%) = 87.5%
If it works then try (1/2(87.5+100) = 93.75%

This zeros in on the value very quickly. e.g You want to find two places close together where the motor starts and one where it doesn't.

If there are nucense trips then move it in small increments.

R below is the 50K trimmer.

You can actually measure the resistance and use that e.g. 50% of R. With the positive of your ohmmeter on R9 and power turned off, measure R. There shouldn't be any interaction measuring this way.

Once you know one of the positions, you could actually try setting the others to the same spot -10% and +10% and try from there same bisect technique only it will be faster yet.

Schematic reference: **broken link removed**

 

[salty joe]

I think the trimmers I got are junk. I installed what was supposed to be a 500K trimmer for the flick and could not adjust the flick. Turns out that the "500K" trimmer is more like 165K and it has the least resistance at both ends with max resistance in the middle. Same story with the "50K" trimmers-they max out at around 30K, also in the middle. Even if the resistance was right, it makes for a miserablely tight adjustment.

But I do have one 50K pot as well as a 500K pot so I can do that tomorrow morning. I'll check the voltage at the trip test too. In the meantime I'll chase down some good trimmers and see if I can get the problem driver squared away.

Thanks guys.

edit-Got distracted and your last post made it before mine, KISS. That sounds good. I use the "chop it in half" method to find a breaker in a hurry.

 

[alec_t]

Turns out that the "500K" trimmer is more like 165K and it has the least resistance at both ends with max resistance in the middle. Same story with the "50K" trimmers-they max out at around 30K, also in the middle.

Check the wiring. That sounds like you have the two end terminals of the trimmer shorted together.

 

[KeepItSimpleStupid]

You don;t need big trimmers. This should be a set and forget. Having a test point to easily measure the value makes sense though. That diode effectively isolates the trimmer, so using your resistance meter one polarity could have a major effect in one direction and the reading won;t be steady.

All you need is something like this: https://www.mouser.com/ProductDetai...=sGAEpiMZZMtC25l1F4XBU0s7zmsV6bK84OLncNasZks=

Those kind of pots can come in 3, 5, 10 or even 15 turns. Alec really had no idea what the range really is. You should really figure that out before searching out a 50K pot. You don't really want only the first 10% of the pot to be used. You can also add a fixed resistor in series with a smaller potentiometer. e.g. 22K+10K pot if that's what's needed. You want some adjustment. The adjustment may have to be tweaked as the bearings get more friction. Stopping and going into alarm may indicate that the motors need cleaning. If you get a better handle on the value, you may be able to use a 1 turn potentiometer. You could be in a realm where the wiper resistance makes the pots touchy.

e.g. https://www.mouser.com/ProductDetai...=sGAEpiMZZMtC25l1F4XBU0s7zmsV6bK84OLncNasZks=

 

[alec_t]

Alec really had no idea what the range really is.

True....ish. The trip time should be adjustable between ~160mS and 1100mS using the 50k trimmer with its 4k7 companion R8. The big unknown is the pump start-up time under load, especially if it gets barnacles growing on it, or whatever

 

[()blivion]

I think the trimmers I got are junk. I installed what was supposed to be a 500K trimmer for the flick and could not adjust the flick. Turns out that the "500K" trimmer is more like 165K and it has the least resistance at both ends with max resistance in the middle. Same story with the "50K" trimmers-they max out at around 30K, also in the middle. Even if the resistance was right, it makes for a miserablely tight adjustment.

Are you measuring them while they are IN the circuit? because any resistor, pot or fixed, will appear to have a reduced value if it is put in parallel with other resistances.

 

[salty joe]

Check the wiring. That sounds like you have the two end terminals of the trimmer shorted together.

Good call. I'm a little embaressed but mostly grateful you caught that mistake. Thanks. Re-did the trimmers and they show the expected values.

I re-connected R3 and slowly adjusted the 50K trimmer while the LED toggled, but the pump would not run except for the initial toggle.

 

[alec_t]

What is the approximate frequency that the LED toggles at?

 

[salty joe]

I adjusted them to about 2-3 seconds on, 2-3 seconds off.

 

[salty joe]

Don't know if it matters, but I got the same result with 5-6 seconds on, 5-6 seconds off.

 

[alec_t]

Hmm. Adjusting the trimmer shouldn't affect the off time at all. The off time as designed is fixed at ~30 secs.
For now, until we've got the trip timing sorted out, use the lamp dummy load instead of the pump. And it might be less annoying to replace the alarm buzzer with a LED (plus ~2k series resistor).
Have you got the left end of R1 connected to +12V directly (and not to the timing bit of the circuit), as suggested in post #1001?
Double-check you've got the b/c/e terminals of Q1 correctly identified and connected.(It's easy to mis-interpret a poor datasheet).

Just to make sure we're all considering the same test set-up I envision it as:-
One PDM-Mk10 is the circuit under test.
The load is a 40W lamp.
The load supply is 22V DC.
The logic supply is 12V DC from the regulator on the wave module.
Left end of R1 is not pulsed, it's at permanent 12V DC.
R3 is in circuit.
R6 does not have another resistor in parallel.
The trimmer is set mid-way.

At power-on,
Does the lamp stay lit?
Is the LED D8 lit?
What is the voltage across the trip test switch (switch not pressed)?
What is the voltage across R6?
What is the voltage at the D4/R7 junction?

 

[salty joe]

I'm sorry, I hooked the timing part of the circuit back up. My bad-sorry.

I adjusted the toggle frequency with the on/off pots from your logic circuit.

I'll get downstairs after work and try again. Thanks.