Help with Water Pump

[salty joe]

Ronv you are as sharp as a tack! Those little poly caps should have been here by now-maybe they'll show up today. (they got ordered a few days after the other parts for the charger) I'm pretty sure everything else for your charger arrived. If they don't arrive, I'll venture out in this cold snap & get one for testing.

 

[salty joe]

Radioshack only has 1uf caps by internet sales. Not sure it's the right kind anyway. Kinda bummed the poly caps did not show up today-as soon as they arrive I'll install that snubber and post.
Thanks.

 

[KeepItSimpleStupid]

Joe:

You have a 100 nF (not 1 uf, I know),but it a 0.1 uf, in the fault module. If you have 10 in parallel you would have a 1 uf.

If you anxious, try a 100 nf and a 1K or something close resistor. It's the same time constant (0.1 uf * 1000 vs 1 uf * 100), but probably not as effective. You might have a couple of 1k2 resistors and 2 in parallel would be 1100 ohms. Close enough.

It would likely still stop any oscillations, though.

 

[salty joe]

Thanks KISS but there is not a lot of room on the PDM near the pump leads. Those poly caps should arrive any day.

 

[KeepItSimpleStupid]

Tack solder, components in midair, You know Ugly. Just a, "Does it work", kinda thing.

Anyway, it was just a thought.

 

[alec_t]

A spare 390n (as used in the PDS of the Speed module), in series with ~100Ω, would be worth a try as a snubber.

 

[KeepItSimpleStupid]

or two 390n's in parallel.

OH, snubbers can be purchased as a single component.

 

[salty joe]

I tried the snubber last night using the 1uf poly cap and a 100 ohm resistor on one PDM, but it made no difference that I could tell.
I did notice a quieter hum from the pump when the SPDT on the “FAST” line was switched.
Given the possibility that anything we try won’t work due to unknowns with the pump, I’m ready to give up on speed control. I’m also more than willing to try anything you guys come up with.
Worst case, I feel like we are in a really good place because you guys figured out drivers. I am extremely grateful for that. So even if we don’t get speed control, this project is a success.

 

[()blivion]

You guys can always fall back on controlled voltage/current based speed control. It would basically just be a crude buck converter. Nothing fancy, it's a technology that is time tested.

 

[alec_t]

Hey, don't give up on it yet Joe .
A few more things to try, one at a time:
1) Connect the Speedcon signal to the R15/D5 junction in the PDM Mk15 mod, instead of to the R1/R2 junction.
2) Put another 330Ω in series with R9 in the Speed Set Section of the speed module to raise the value of Ref.
3) Add hysteresis to U1a in the Pulse Delay Section of the speed module, by connecting the Ref input to U1a's non-inverting input via a 330Ω resistor and connecting a 100k resistor between that input and the output of U1a.

 

[KeepItSimpleStupid]

@alec

Is it possible to build a adj duty cycle 555 as a test circuit (with trigger or not) for the Speedcon signal or is it totally dependent on other circuit requirements? Just an idea?

 

[salty joe]

3) Add hysteresis to U1a in the Pulse Delay Section of the speed module, by connecting the Ref input to U1a's non-inverting input via a 330Ω resistor and connecting a 100k resistor between that input and the output of U1a.

I used pin 3 for Ref input and pin 1 for output on U1a.

Are you saying add a 330Ω resistor to pin 3 and connect the other side of the 330Ω to Ref? Then jump pins 3 and 1 with a 100K resistor?

Thanks Alec for the ideas-I'll get started in the morning.

 

[KeepItSimpleStupid]

Seems like : Ref input *330 ohm* (-) *100K* (out)

PS: I didn't look at the pin outs.

 

[alec_t]

@Joe

Are you saying add a 330Ω resistor to pin 3 and connect the other side of the 330Ω to Ref? Then jump pins 3 and 1 with a 100K resistor?

Just so.
@KISS

Is it possible to build a adj duty cycle 555 as a test circuit (with trigger or not) for the Speedcon signal

It's possible, but I'm not sure it's worthwhile (yet!). That would amount to conventional PWM which I thought we were trying to avoid?

Seems like : Ref input *330 ohm* (-) *100K* (out)

No, Ref goes (via 330R) to the non-inverting input.

 

[()blivion]

Just to note, I believe was probably the first one to fear doing PWM because it may not be a good idea. But I don't necessarily know what I'm talking about, or what the motor actually can and can't handle. Generally (as far as I know) one doesn't want to use PWM on a self contained brushless motor, as they have an internal frequency, and doing such things could cause constructive/destructive interference... In accordance with the prophecy.

I would also like to stress once again that with a low pass filter the above worries are irrelevant, as the PWM signal will be converted to smoothed DC before it ever gets to the motor. Which will effectively be the same thing as open loop buck conversion.

 

[KeepItSimpleStupid]

@KISS
Is it possible to build a adj duty cycle 555 as a test circuit (with trigger or not) for the Speedcon signal
It's possible, but I'm not sure it's worthwhile (yet!). That would amount to conventional PWM which I thought we were trying to avoid?

My thoughts were to build to nail down the values easier. I was thinking the speedcon input, but what if it was PWM direct. Might be able to gain some controller params from it. e.g. measure resistors to get freq and duty cycle Like 6 parts? Test circuit only.

 

[salty joe]

Ran into an intermittent snag yesterday-one of the PDMs was acting up. One of the legs of D2 was cracked. By the time all the components were taken apart, the problem found and everything put back together, I was out of time.

I hope to start the tests today. That is if this migraine aurora that just started goes away. It’s painless but is sort of like looking through a kaleidoscope. Sometimes they go away real quick and sometimes it gets nearly impossible to read.

 

[ronv]

Alec, Could you post the .asc file for the speed control?

Thanks!

 

[alec_t]

One .asc file coming up

 

[salty joe]

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

The pump LED on PDM A flickered, but the pump LED on PDM B did not flicker. The speedcon with the added 2K2 was connected to PDM B. So I added a 2K2 to the speedcon connected to PDM A.

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

The pump connected to PDM B ran WITH speed control, (pump did not start reliably) but now the pump LED on PDM B flickers.

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 trip test shuts everything down as designed for both PDMs.