Help with Water Pump

[alec_t]

I soldered a 2K2 on the speedcon wire between the pulse delay section and the PDM, but it did not seem to make any difference.

I was hoping it would cure things as per the PDM oscillation

BTW, the alarm sounds after 5-10 sec. even if no pumps are hooked up.

If the Vt inputs of the Fault Sensor aren't connected to Vt, but just left floating, stray leakage across the circuit board can cause a Fault output of the Sensor to rise enough to trip the alarm. Any unused Vt input of the Sensor should be grounded (make sure it's not already connected to 12V) and any unused Isense input of the Sensor should be connected to +12V.

 

[salty joe]

I did have two Vt and two Isense unconnected. I signal grounded VT and took 12V to Isense. There are still two fault terminals that are not connected.
With pumps plugged in, the PDM LEDs flicker then about 10 sec later the alarm sounds. One of the PDM LEDs goes off. One alarm LED comes on. When the power supply is unplugged, the alarm LED from the other pump comes on momentarily.

With no pumps plugged in, the PDM LEDs are solid and bright. The alarm sounds in 2-3 sec. The PDM LEDs remain on, even though the alarm is sounding. The alarm LEDs behave exactly as above.

Is it time for me to go over the speed and fault modules with a fine toothed comb? Would you by chance have any idea where to look first? Thanks very much.

 

[alec_t]

I signal grounded VT and took 12V to Isense.

Hopefully only for the two unused inputs

There are still two fault terminals that are not connected.

Ok.

With no pumps plugged in, the PDM LEDs are solid and bright. The alarm sounds in 2-3 sec.

So the FETs pass enough current to operate the PDM LEDS but not enough to run the pumps I wonder if there's still an oscillation problem (pity you don't have access to a 'scope ). The low current is treated as 'disconnected pump' by the Alarm section as intended.

With pumps plugged in, the PDM LEDs flicker then about 10 sec later the alarm sounds

That points to oscillation. I'll have a think and see what we can try to overcome it. Ronv (?) suggested a snubber circuit (R and C in series) across the pump. Worth a go.

Edit: Oscillation might be due to some protective feature of the pump's internal IC.

 

[ronv]

The 100 ohm & 1ufd. worked good in the sim. Right across the pump leads.

 

[salty joe]

double post-sorry

 

[salty joe]

Hopefully only for the two unused inputs

Yes, I built one fault sense module but was only using half of the FSM for two PDMs. So the two unused Vt and two unused Isense are now dealt with.
Thanks guys I really appreciate your efforts.

edit The flicker at the PDM LEDs never occurred when running only controller, two PDMs and two pumps, if that means anything.

 

[alec_t]

The flicker at the PDM LEDs never occurred when running only controller, two PDMs and two pumps, if that means anything.

The Speed module deliberately increases the 'off' time of the pump coils twice per rev, so at low revs (e.g. at start-up) there are more prominent discontinuities in the LED current, hence noticeable flicker is likely. Try temporarily connecting another 100k in parallel with R3 (100k) in the Pulse Delay Section to reduce the 'off' time.

 

[salty joe]

I added another 100K parellel with R3 to the PDM that did not get a 2K2 resistor at Isense. I could not tell any difference. Time for me to hit the rack. Thanks for kicking this around.

 

[alec_t]

the PDM that did not get a 2K2 resistor at Isense

The 2k2 should be for the Speedcon signal, not Isense.

 

[4pyros]

Sounds like its time for some pictures.

 

[salty joe]

The 2k2 should be for the Speedcon signal, not Isense.

I typed the wrong thing-sorry. The 2k2 is on the speedcon wire between the pulse delay section and the PDM. I usually run to the basement and look before responding, last night I was lazy and tired.
Should that 2K2 and the 100K be removed at this point?

I'll post a picture later today.

 

[salty joe]

The board with three dangling pots is the controller. It has two heatsinks for two 7812 Vregs. One Vreg is disconnected, I just have not removed it yet.

The red tube on the wire is to make sure the (exposed) 2K2 on Speedcon wire between pulse delay section and PDM did not short out.

The board by my glasses has two pulse delay sections and one speed set section.

The board with the pump jacks is two PDMs.

The last board that is empty in the middle has one fault sensor module, to handle four pumps, and the alarm module. I'll add another FSM later.

That little black piezo between the FSM/alarm and the controller makes some kind of major racket. Wow, it is loud!

 

[alec_t]

Should that 2K2 and the 100K be removed at this point?

I think the 2k2 should remain. The 100k can go for now.
Useful to have the pic of the setup. I see the Speedcon wiring is not exactly compact . That is one possible cause of the oscillation. It could be picking up stray RF interference and applying it to the FET gate (which has a very high impedance, making it susceptible to the interference). I suggest incorporating the 2k2 on the Speed module board and using a screened wire to connect to the PDM.

Wow, it is loud!

Temporarily disconnect until 'all systems go'?

 

[salty joe]

I suggest incorporating the 2k2 on the Speed module board and using a screened wire to connect to the PDM.

Will do. I'm going to wrap the two speed con wires (I have two PDS, two PDM and two pump) with aluminum foil. That will work as a screen, won't it? Any other wire that might need screened?

Temporarily disconnect until 'all systems go'?

Good idea, I will. Even with tape on the front as a muffler, it is piercing.
Thanks.

 

[KeepItSimpleStupid]

Joe:

What you might want to do is replace the buzzer temporarily with a LED and series resistor somewhere between 240 and 1200 ohms. Around 600 would be ideal. This assumes 10 to 50 mA for the LED.

 

[alec_t]

@Joe

That will work as a screen, won't it?

Only if the foil is grounded.
BTW, have you added the snubbers as Ronv suggested?
@KISS

replace the buzzer temporarily with a LED and series resistor

The alarm module already has one LED per pump.

 

[salty joe]

@Joe
Only if the foil is grounded.

Thanks, I ran back downstairs and grounded the foil.
I soldered the 2K2 on the pulse delay section board between the terminal point of speedcon and the wire from D1 anode . Both speedcon wires are wrapped in grounded foil. I made sure the foil did not touch the terminal points. I also confirmed that there is continuity between the foil and ground. I cannot tell any difference.


No, I have not installed snubbers.
Thanks all.

 

[ronv]

No, I have not installed snubbers.
Thanks all.

The snubber won't hurt anything so I would give it a try. Then we would know what it isn't.

 

[salty joe]

Thanks Ronv for the snubber idea. I got out my cap container and did not have a 1uf cap. Could not find one in the OEM controller, either. Before now, I don't think I've needed a 1uf cap. I'll see about scaring one up tomorrow. I have plenty of 100 ohm resistors.

I found snubber on Wikpedia and they said a snubber is hooked up in series. Did Wikpedia get it right? Does it matter what type of 1uf cap is used for the snubber?

 

[ronv]

The resistor and cap are in series, but across the pump.



I think you might have them in a bag from this order.

https://www.electro-tech-online.com/threads/help-with-water-pump.124250/#post1102437