Help with Water Pump

[alec_t]

I put another 330 ohm resistor in series with R9 in the speed module. Neither pump ran and both alarm LEDs lit up.

Ok. Did you take it out again for the other tests I proposed?

The pump LED on PDM A flickered, but the pump LED on PDM B did not flicker.

Probably due to component and pump tolerances. We're on the brink of works/doesn't.

The speedcon with the added 2K2 was connected to PDM B. So I added a 2K2 to the speedcon connected to PDM A.

Those 2k2 s probably won't be needed if the Speedcon signal is routed as per test 1 in post #1610.

There was no more flicker on PDM A and the pump ran but with no speed control.

Suggests the pump current was staying just above the threshold at which the Pulse Delay Section triggers.

The pump connected to PDM B ran WITH speed control, (pump did not start reliably)

So there's a light at the end of the tunnel. More tweaking needed. Can you confirm the Speedcon connections and added/removed components you used for that setup?

Also, the pump LEDs do not turn off when the alarm LEDs turn on, but the pumps stop when the alarm LEDs turn on.

The FET is passing a few mA; enough for a LED but not start-up.

 

[salty joe]

I left the added 330Ω in series with R9 in the speed section. I did not have time for more testing.

How about if I remove the 330Ω by R9 and the two 2K2 resistors at speedcon, and then do test 1 in post #1610?

I'll double check all connections.

Thanks!

 

[alec_t]

How about if I remove the 330Ω by R9 and the two 2K2 resistors at speedcon, and then do test 1 in post #1610?

Yes, try that. And leave them out for the other tests. Try to recall too the exact set-up where you got speed control (of a sort).

 

[salty joe]

Try to recall too the exact set-up where you got speed control (of a sort).

That's easy, it's the existing arrangement. An extrra 330Ω at R9 and two 2K2 resistors, one at each speedcon terminal point in the pulse delay section.

I should have time this afternoon for the next step-thanks.

 

[salty joe]

1) Connect the Speedcon signal to the R15/D5 junction in the PDM Mk15 mod, instead of to the R1/R2 junction.
.

Alec, currently speedcon in PDM is hooked to R1/R2 and the anodes of D4/D5. I could not find R15 on PDM-Mk15.

When you get a couple minutes will you please look at this?

The two added 2K2 and the 330Ω are removed.

 

[alec_t]

With the previously-working modification of PDM-Mk15 a 2k2 resistor was added in series with D5. In my copy of the schematic I labelled it R15 and have it between D5 anode and the R1/R2 junction. Where did you connect it?

 

[salty joe]

Ohhhhh. I'm sure that 2K2 is in the right place. I was looking at my schematic and I did not add that 2K2 (R15) to the print. After work I'll double check R15 and component connections and run the test. Thanks a million!

The removed 2K2s were at speedcon.

 

[alec_t]

I'm sure that 2K2 is in the right place.

The alternative is with the 2k2 (R15) between D5 cathode and Q1 collector. If that's the way you have it then try Speedcon going to Q1 collector instead.

The removed 2K2s were at speedcon.

Fine.

 

[salty joe]

The 2K2 is between D5 cathode and Q1 collector.
I disconnected speedcon from the terminal point and taped the end. R1, R2 and the anodes of D4 and D5 are still all connected together.
I connected a wire from the collector of Q1 (pin 3) to the speedcon terminal point.
With the 50K PDM trimmer set to max, the alarm trips in 1-2 sec. whether a pump is plugged in or not. With the trimmer set to min, the alarm trips in an instant. The pump does run for that brief time. It makes no difference what position the SPDT connected to “FAST” is in, 12V, off or ground, the pump behaves the same.
The pump LED on the PDM is solid when lit. When the alarm LED lights up, the pump LED is off.
Connections between components looks good.

 

[alec_t]

All useful info.

The pump does run for that brief time.

Good. That suggests the FET oscillation has now gone away. I'll have a think and play with some sims to work out why the alarm is now doing its thing.

 

[salty joe]

Many thanks.

 

[alec_t]

What seems, from the sims, to be happening is that the Speedcon signal is now pulling the Q1 collector voltage down long enough for the alarm to trip. Try moving R15 so that it's connected between the R1/R2 junction and the D4 anode/D5 anode junction (as in post #1626); then connect Speedcon to those diode anodes instead of to Q1 collector. Apart from any spurious effects due to the wiring between the Speed module and the PDM the changes will result in the Speedcon signal and the Q1 collector signal driving the FET gate via similar paths. Hopefully the dreaded oscillation will then be absent.

 

[salty joe]

I removed the 2K2 from between D5 cathode and Q1 collector and reconnected D5 to Q1.
I installed a 2K2 between D4-D5 anodes and R1-R2 junction.
I connected speedcon to D4-D5 anodes.
The pump LED does not light up, pump only hums. The DMM showed 8.02V at the pump.

 

[alec_t]

Looks like FET oscillation again .
Leaving the PDM in that state, try test 3 in post #1610 and measure the output voltage of U1a in the Pulse Delay Section.

 

[salty joe]

OK, thanks. IDK, if I'll have time today, but wanted to mention that 8.02V at the pump was with the "FAST" SPDT in the 12V position. In the middle position (off?) 4.7V and in the signal ground position 2.84V.

 

[alec_t]

Useful voltage readings. When you say 'at the pump' do you mean from pump+ to pump-, or from pump- to ground?

 

[KeepItSimpleStupid]

"at the pump" in the past has meant at the connector for the pump. Some length of wire is permanently attached to the pump to make it submersible. My guess.

 

[salty joe]

I just put the DMM probes on the terminal point screws where pump+ and pump- are.

 

[salty joe]

I left the PDM as is. My U1a has pins 1, 2 and 3. I installed a 330Ω between pin 3 and ref. I jumped pins 1 and 3 with a 100K.
The output at pin 1 is 9.4V with "FAST" SPDT in the 12V position, 9.87V in the middle position and 10.13 in the signal ground position.
This is with the pump humming, but not running. I took these reads from the top of the board directly from pin 1. I accidently touched pins 1 and 2 together, and the pump started running!
Using the SPDT, there is a low, medium and high pump speed.
With the pump running, the output at pin 1 is 6.83V with "FAST" SPDT in the 12V position, 7.84V in the middle position and 8.6V in the signal ground position.
The pump LED did not light up at all for this test. The trip test stopped the pump and the alarm LED lit up, then 30 sec or so later, the pump started back up and the alarm LED went off.
The V readings varied and jumped around by a few hundredths of a V.
Sorry I could not get this sooner and thanks.

 

[alec_t]

I accidently touched pins 1 and 2 together, and the pump started running!

Simulation shows that shorting 1 and 2 sets U1b output high, commanding the FET to be on. You've proved that's the case.....it was enough to give that initial kick to start. I'll play further with the sim to see if I can replicate the voltage readings.

Using the SPDT, there is a low, medium and high pump speed.

Encouraging .

The pump LED did not light up at all for this test.

Can't explain that at the mo.

The trip test stopped the pump and the alarm LED lit up, then 30 sec or so later, the pump started back up and the alarm LED went off.

Good.

The V readings varied and jumped around by a few hundredths of a V.

To be expected.