Help with Water Pump

[salty joe]

Thanks 4pyros.
I have another pair of good PDMs, getting there.
I'll see what I can do with the problem PDM tomorrow. The alarm sounds immediately whether a pump is hooked up or not. Thanks all for the ideas to track down the problem. I'm gonna look carefully for a jumped component. I caught myself trying to do that twice today.

 

[salty joe]

The problem PDM had a 12 V wire in the wrong place. I dodged a bullet there. Anyway, I have six good PDMs. All six pumps are running in the basement right now. I'm going to let it run for a couple days. It's been running for about an hour or so and everything is cool to the touch except the power supply is just a little bit warm, not even close to precook or anything like that. The tide controller works just right, the wave controller works just right, all systems are go. I replaced the pots on the controllers with trimmers, but I need to replace the trimmer for the wave with one of those trimmers that have a real fine adjustment.
That was some super work you guys did. Thanks a million. I’ll get a picture posted after I get the boards mounted. Everything looks extra messy right now.
Everybody in the world of reefing has been talking about this new pump that just became available, so I got one. It is a very reasonably priced unit that moves an absolutely incredible amount of water. It is not even in the same league as these pumps I have been working with. There are definitely some differences between the two pumps besides how much water they move. The magnet in the motor on the new pump has twice as many poles. (With the cover off of the pump and the pump unplugged, the propeller will move one quarter turn before it is stopped by magnetism. The old pump rotates one half turn before magnetism stops it.) The new pump runs on 24 V DC just like the old one. The output on the AC/DC adapter is 24 V and 2.5 amps. The output of that adapter feeds into the controller. There are three wires coming out of the controller to the pump. I have learned that the third wire carries 5 V and has a controlling function for the pump. I know which wire is 24 V positive, which is ground, and which is 5 V. Do you think there is a ghost of a shadow of a chance that I could control this pump by cutting the 24 V wires coming from the pump and connecting them to the controllers we just built? I have learned that this pump will run well at least as low as 15 V. It doesn't bother me very much to have to splice wires back together if it doesn't work. I hate to try it if there's a high risk of frying my controllers and PDMs in the process. As far as that goes, I have a beat-up ugly PDM that does work. Do you guys think is worth trying?

If anybody is up for emergency backup, I think I have all the parts.
You guys have been absolutely amazing.

 

[alec_t]

Do you think there is a ghost of a shadow of a chance that I could control this pump by cutting the 24 V wires coming from the pump and connecting them to the controllers we just built?

Possibly, but without knowing for sure what sort of control signal is applied to the 5V wire it's difficult to say. I would guess it's for speed control, perhaps using PWM. Do you have any more technical info? A datasheet? Can you use your DMM to measure the (average) voltage on that 5V wire? To avoid cutting the wires you may be able to put a pin carefully through the wire insulation to make a measuring point.

Edit: Another guess; the 5V wire could be an output (e.g. for a speed feedback signal or a locked-rotor detector) rather than a control input.

 

[KeepItSimpleStupid]

I kinda agree.

The 5V signal could be a speed signal, a PWM drive signal, locked rotor signal
, or an over temp sensor that's "pulled-up" within the controller. It would be nice to figure that out. I doubt it's power. "pulling up" means that in the controller the 5V is actually an input and connected to a resistor to +5 V.

Why don't you post a pic of the front and back of the PCB and identity +24, GND and the 5B signal for starters?

One of us, maybe me, could actually take a look at it. Hopefully the wiring uses an easy to get plug.

 

[salty joe]

The yellow and green wire is 5V, brown 24V, and blue ground.
With the pump on hi, the green wire showed 5V, on lo the green wire showed 2.4V. On pulse modes, V jumped around from 4.4 to 0.
Brown wire showed 24.1V on hi and 24.3 on lo.
In the pic, what I think is a Vreg has L7805CV printed on it.
The IC is
12C5202AD
351-PDIP20
1230OH4F426.B

 

[alec_t]

Nice pics, Joe. Those voltages are consistent with an input control. On the other hand they are also consistent with a speed output signal.
Could you repeat the 5V line measurement, but this time with the +ve probe of the DMM connected via a cap (~100nF) to the 5V line? That would tell us if the signal is DC or pulsed (PWM).
Is the control pot wiper the same voltage as the 5V line?
Google says the 12C5202AD is a STC microcontroller, available in a variety of packages (including PDIP20). Couldn't find a datasheet for it

Edit:
I was expecting to see a beefy FET in the pics. In its absence I'm inferring the micro isn't directly switching the pump power to control speed; rather, it presumably provides a speed command to the pump and the pump has internal electronics which control the power.

 

[KeepItSimpleStupid]

I found the part, but no data: **broken link removed**

Joe: Can you tell us what the buttons do?

There MIGHT be a couple of options. Use the speed control of the OEM controller and just toggle it on and off. Force it on and let the PDM do the rest. Hopefully, I'll get a chance to look at it later.

Thanks for the pics.

 

[salty joe]

I'm guessing here-should the negative leg of the cap be attached to the 5V line, then touch the + DMM probe to the + leg of the cap?
I''l run that test and check the pot tomorrow.
I could see some electronics buried in the epoxy at the back of the pump.
The button in the middle changes mode. We have hi, lo, pulse, a rolling wave, another kinda rolling wave and random. The button on the left is for a 10 minute pump stop. There's also a way to lock the setting. Both buttons are tactile switches.

I was thinking why not feed the 5V line a steady diet of 5V? Is that what you mean by force it on?

Thanks you guys.

I was just wondering if we could lose the OEM power supply and feed the OEM cotroller with the pump termination points on the PDM? The OEM controller would be left on high for the wave pulse. The PDM would still control the wave pulse. For tide, it would be cool to have a rolling wave for 30 sec, then still for a few seconds. The OEM controller would add the rolling wave and the PDM would control the on/off times. Just wondering.
I just re-read your post KISS. That's what you are saying, isn't it? I guess I got the idea from you, but kinda took awhile to sink in. Ha ha.

 

[KeepItSimpleStupid]

I was thinking why not feed the 5V line a steady diet of 5V? Is that what you mean by force it on?

Yes, except when one says "force on", it could be force low or force high.

It's gonna take a little more studying and hopefully we can find a datasheet.

In one idea, in test form, would be put the OEM controller in the "Always on mode", that's why I asked what the controls do. I should now ask, what happens if you loose power? Do you loose the modes? I think not, because you said you can lock it. Do you have a link to the instruction manual?

What's the part in the lower right-hand corner of the top of the PCB that looks like it's soldered on the bottom layer?

The OEM controller would still have it's own 24 VDC power for now. The grounds would be connected to your power supply. The pump connections would come from the PDM. So the OEM controller would get 5V and full on from however the processor makes a full on signal.

Minimal changes to see if it works. Later, if it's possible, then tidy things up. Remove the OEM controller entirely and put a 5 V regulator in your stuff. That's the only way you get to keep all of the protections we tried to give you.

It seems safe to assume that that signal is a control and since the pump can change speeds, it's probably PWM or Pulse Width Modulated and probably 5 V is full ON. A data sheet for the processor might help.

Any processor contains a program that is permanently placed in a read/only memory The 8051 chip is really old chip, but hey could have enhanced it a lot, who knows. In the early days, you only had a few choices. Now a days, inputs can have pull-ups, pull-downs, none at all, part of a technology like I2C (inter IC Communications), Interrupts, you name it.

We think we knew what happened to your old motors. Basically thermally mismanaged.

 

[alec_t]

I'm guessing here-should the negative leg of the cap be attached to the 5V line, then touch the + DMM probe to the + leg of the cap?

A 100nF cap doesn't have a + and - leg. Use the highest value non-electrolytic cap you have (100nF, 330nF, ...)

I was thinking why not feed the 5V line a steady diet of 5V?

That's what I was hoping we could do. The wave/tide controller could then easily drive the new pump if that's what you want. However, until we know more about the 5V line it wouldn't be safe to just apply 5V to it.

It seems safe to assume that that signal is a control and since the pump can change speeds, it's probably PWM or Pulse Width Modulated and probably 5 V is full ON. A data sheet for the processor might help.

Apart from "it seems safe" I would agree with that

The 8051 chip is really old chip

What 8051? The OEM IC is a 12C5202AD STC microcontroller.

Edit: Oh. Just followed the link to the micro. So the5202 is an updated 8051.

 

[salty joe]

Do you have a link to the instruction manual?

What's the part in the lower right-hand corner of the top of the PCB that looks like it's soldered on the bottom layer?

Remove the OEM controller entirely and put a 5 V regulator in your stuff. That's the only way you get to keep all of the protections we tried to give you.

You don’t happen to read Chinese, do you? There has to be an English instruction manual. I’ll dig.
Depending on which picture, the lower right hand corner has a couple caps or the pot.
So it looks like I have to choose between the protections and a rolling wave? Ugh.

A 100nF cap doesn't have a + and - leg. Use the highest value non-electrolytic cap you have (100nF, 330nF, ...).

OK-gotcha. With a 100nf cap soldered to the 5V line, I got 0.52V in both hi and lo mode.
With the cap removed and the controller on hi mode, I got 4.98V at the 5V line, 0.52V at leg 1 of the pot and 1.56V at legs 2 & 3.
Thanks guys.

 

[alec_t]

So it looks like I have to choose between the protections and a rolling wave? Ugh.

Hopefully we can do what you suggested, to get the best of both worlds: connect the OEM 24V and ground to the PDM Pump+ and Pump- terminals, then work out what we need to do with the 5V line. However, if your new OEM supply adapter is rated at 2.5A it seems the new pump is drawing a considerably higher current than the old one, so PDMs might need modifying
From your latest measurements it looks as though the 5V line carries a DC signal; not PWM. That would make driving from the wave/tide controller easier.

 

[salty joe]

Oh, I was thinking that giving the OEM controller 24V and ground (like the OEM power supply) and the 5V wire between the OEM controller and pump would take care of itself.
That is a bummer to have to beef up the PDMs. But the difference in how much water the pumps move is like night and day.

 

[KeepItSimpleStupid]

You don’t happen to read Chinese, do you? There has to be an English instruction manual. I’ll dig.

Chinese online or scanned at 600 DPI would work. translate.google.com usually works.

I'll have to point out the component the hard way. It's a black one, like a Bidirectional LED or transistor. It has 3 leads.
Component side: The side with the IC; Solder side: the other side. This is on the "solder side"

 

[KeepItSimpleStupid]

Oh, I was thinking that giving the OEM controller 24V and ground (like the OEM power supply) and the 5V wire between the OEM controller and pump would take care of itself.
That is a bummer to have to beef up the PDMs. But the difference in how much water the pumps move is like night and day.

Kinda have to see what changes have to be made to the PDM. The FET is a possibility and the sense resistor is a possibility. The other real problem is that the PDM's basically switch ground to the pump and the pump probably needs to switch positive, so it won't be as simple.

 

[alec_t]

One further measurement: with the 100nF on the 5V line and the DMM set to AC volts what do you get?

Oh, I was thinking that giving the OEM controller 24V and ground (like the OEM power supply) and the 5V wire between the OEM controller and pump would take care of itself.

What I meant to say was that the new OEM controller could get its '24V' input and 'ground' from the PDM Pump+ and Pump- terminals respectively, and the new pump would connect to the OEM controller output unmodified. Then, as you say, the 5V line would look after itself. The tide/wave controller would then be just acting as a switched supply for the OEM controller. If each new pump comes with its own 2.5A adapter then each PDM could have an individual adapter to power the PDM, the OEM controller and the pump.

The other real problem is that the PDM's basically switch ground to the pump and the pump probably needs to switch positive, so it won't be as simple.

I don't think that problem would arise with the connections to Pump+ and Pump- as above.

 

[salty joe]

Do you mean the black tab that extends out from the board? That is where a light sensor plugs in. The idea is to slow the pump down when it's dark. It does not work well. Even if it did, I'm not sure I'd want it. With the light sensor removed, you don't get the night/darkness slowdown. I removed it from the getgo. Other than that, the only components on the solder side are 2 tactile switches, the pot and nine tiny LEDs.

 

[KeepItSimpleStupid]

@alec_t
I see, we have to provide an isolated supply for the 5V?, I'm having trouble wrapping my head around that. What am I missing. I'd figure the 5V is relative to motor- or ground and we switch ground. What a pain.

@salty
Yep, must be the light sensor. Thanks.

 

[alec_t]

@KISS
If the connections were as in post #1796 then the OEM controller looks after the 5V line; Joe would merely set the controller for permanent full speed (which I presume is one controller setting?)

I envisage the new setup as:
OEM 24V 2.5A supply >> wave (or tide) timer >> PDM >> OEM controller >> pump

 

[KeepItSimpleStupid]

@alec_t
Aaah! Your going to power the OEM controller from the PDM pump output (Motor+ and Motor-). OK, that makes a lot of sense. What we might want to do is add more capacitance on the 5V supply so that it doesn't get upset when the PDM goes into current limit. The other thing we can do is isolate the input to the 5V supply with a diode and a big cap, so it won't see the 24V supply (OEM: motor+, motor-) take a dip. Hopefully that won't cause any issues Wonder if we should current limit the control input to the motor? Just thoughts.

And your right, Joe gets the benefits of both worlds.