Help with Water Pump

To ground the TTAM, should signal ground from a PDM be used?

Click to expand...

In the tidal system the SignalGnd terminals of all four PDMs should be commoned at a single point (labelled SignalGnd in the TTAM), and the PowerGnd terminals of all four PDMs should be commoned at that same single point too. In the overall system schematic two common points are shown (SG and PG), but in practice they should be one and the same (sorry, I realise now that was confusing). The aim is to have a 'star' system of grounds, so that no large currents can flow through low-current signal lines.
Your wave system can likewise have a star ground arrangement.
I assume the filter pump system will be independent because of the different power back-up arrangement, to avoid unwanted conflicts (although a SignalGnd connection to the shared alarm section would be needed).

I have the power ground from all 8 new pumps connected to a common point. I have kept the signal ground seperate-I'll change that.

If I understand you correctly, I need to connect the four power grounds and the four signal grounds from the tidal system to a common point. Then, connect the four power grounds and the four signal grounds from the wave system to a different common point.

I almost didn't ask about grounding because I thought it was a no brainer. The last time I felt like that, things fried. I have become just a little gun shy. Good thing too.

I'll keep the filter pumps seperate from all else-that's for sure!

Thanks-tomorror I'll reconfigure the ground wires.

I have the power ground from all 8 new pumps connected to a common point.

Click to expand...

That's ok.

I have kept the signal ground seperate-I'll change that.

Click to expand...

See below.

If I understand you correctly, I need to connect the four power grounds and the four signal grounds from the tidal system to a common point. Then, connect the four power grounds and the four signal grounds from the wave system to a different common point.

Click to expand...

It really depends how you are building this. I've been thinking of the tidal system as separate from the wave system and separate from the filter system (because, historically, that's the way the designs evolved, and also because of the modular approach), in which case the tidal, wave and filter systems would have respective common points. However, if you build both the wave and tidal systems on a single board then they can share a single common point. It shouldn't be critical, and things won't fry. The main concern is to prevent high motor currents from passing through control signal ground lines.

Getting this system correctly grounded is more involved than I imagined. My approach has been all the timers, backups, PDMs etc. combine to make a single system, so their placement is not important-only the hookup is important. I hope that is sensible.

On one board I have battery backup, wave timer, filter pumps timer and the power supply backup. This board also has a row of 8 terminal points for power ground.
I have 12V for wave timer and filter pumps timer combined. I also combined signal ground for the wave timer and the filter pumps timer.
I have to go into this board and make sure the battery backup is completely isolated.

On a second board I have TTAM as well as a row of 6 terminal points connected to signal ground and another row of 5 terminal points connected to 12V. This board is still ungrounded.

I understand not wanting pump current flowing into signal ground, but don’t get how that can be achieved if the signal ground and power ground are connected.

If you could help me get this ground situation squared away, I would be most grateful. Then I can send 12V to the filter PDMs and see if their ICs are goners.
Thanks for all your help-if pics might be useful, please let me know.

@Joe

They are connected together, but only at one place. At the power supply.

Hig current grounds cause undesireable voltage drops in each of the wires for logic circuits. Motors cause a lot of ground noise. Anyting connected to a high current ground is not sensitive to voltage drops.

The low-current grounds are typically references or have very little curents.

Grounding and shielding is a difficult subject to understand.

These devices are modular enough, that te individual pieces can have their own star-like grounds. One for high current and one for low current.
The final destination is the power supply ground for both of these "star-grounds".

A bus-bar system would work for the power supply. The idea here would be to use something like a tapped copper bar where the output of the power supply and the sense leads are connected. Now you have as many places that you need to attach your 8 or so modules to.

Back to the DIN rail, DIN terminal ideas. See here: https://www.elecdirect.com/catalog/438f5907-d9f7-41d5-a973-1b8be999b7b6.aspx The pic in the middle front has about 5 screws in the middle of the terminal. This is a "bus bar". That then gives about 10 tie points for wires, probably rated for 20 to 30 Amps.

The terminal blocks are open sided and therfore require one end section per group. There is also two end sections which keep the terminals in place.

Te yellow and green striped ones are ground specific terminals.

===

Aside:
In House electrical using sub panels, you would not connect the ground and neutral in the sub-panel. Only in the main panel is the ground and neutral connected together.

This is sort of the same system your doing with each of the devices being considered as being a sub-panel.

Ground in a home system only carries fault currents and it's also a reference. Hospitals and RF transmitter sites may actually have two grounds. One for faults and one for a reference. Special orange 120 V recepticles have what's called an isolated ground terminal. The box would be connected to the fault ground and the ground terminal of te outlet would be connected to te reference ground.

The current house system would have a problem if an upstream outlet was hit by lightnig. Anything else on that leg would have the potential of getting fried. The other circuits, in theory, would not see the surge or high current spike.

KeepItSimpleStupid said:

Grounding and shielding is a difficult subject to understand.

Click to expand...

Hi KISS, you are right about grounding correctly being confusing-to me anyway.

KeepItSimpleStupid said:

A bus-bar system would work for the power supply. The idea here would be to use something like a tapped copper bar where the output of the power supply and the sense leads are connected. Now you have as many places that you need to attach your 8 or so modules to.

Click to expand...

The 8 new pumps have individual power supplies. Each 24V line goes to an individual PDM. The 8 ground wires are connected at a homemade bus-bar. (terminal points soldered together with 18g copper)

On one board I have battery backup, wave timer, filter pumps timer and the power supply backup.

Click to expand...

Now I'm confused. The battery is the power supply backup and is used only with the filter pumps.

I have 12V for wave timer and filter pumps timer combined.

Click to expand...

That's what fried things previously! The filter pumps and their PDMs and timer should get their 12V input only from the battery-backup '12V' (actually ~13.8V) output (from the LM317). The wave and tidal stuff should get 12V from the 7812 regulator.

alec_t said:

Now I'm confused. The battery is the power supply backup and is used only with the filter pumps.
That's what fried things previously! The filter pumps and their PDMs and timer should get their 12V input only from the battery-backup '12V' (actually ~13.8V) output (from the LM317). The wave and tidal stuff should get 12V from the 7812 regulator.

Click to expand...

Sorry for the confusion.
The power supply backup I referred to is where all 8 24V lines from the new power supplies go into individual anodes of beefed up, 1A diodes. The cathodes are soldered together to feed TTAM.
The battery backup is on the same board, but will be isolated electrically from all else. I really did learn my lesson there. Promise. For now, I'm trying to get everything except battery backup working.
Thanks.

@Joe

At some point your going to be dealing with shielding. Here https://www.google.com/url?sa=t&rct...=Gn2s07gVAJ8TzWasn1MwHg&bvm=bv.48705608,d.dmg is an article on that.

The quick and diry rules are that the shield gets connected at one end. The suggested end is the source of the signal. The other quick and dirty things to remember is that EMI (Electromagnetic Interference) is solved by twistieng and RFI (Radio Frequency Interference) is solved by shielding. Thus, shielded twisted pair is a popular combination.

Other quick and dirty rules are low voltage (such as Ethernet cables) should be 6" away from an AC power cable wen running parallel and they should cross in a perpendicular fashion.

In any system, you might have
1. Protective Earth Ground
2. Analog Ground
3. Digital Ground
4. High power ground
5. Reference only.

1. Carries fault currents and is used for the shield. Really connected to the Earth.
2. Analog ground - Typically a quiet ground where the current draw is small
3. Digital ground - lots of switching noise
4. High power ground - plagues the system with large voltage drops.
5. Ref only - the system definition of zero volts

All of the grounds do get connected at one point. Some of these grounds may have filtering and series resistances added intentionally.

Some PC boards have multiple ground planes. A properly designed PCB would have layes called power planes, just for power and layers for the various types of grounds. Usually these are inner layers. Most PCB's are one to two layes, but 40 layers are possible. It's also now possibe to EMBED components inside the PCB.

What we are trying to accomplish is to have the hig currents and noisy currents follow a different path than the control circuitry so they have less of a chance to interact.

Not just ground connections, but ground circuits-wow.
I am using twisted wire (solid strand copper 16g and 18g, 8'-15' long) from the PDMs to the pumps. According to the wire chart, the 18g wire is plenty beefy for this.
As far as shielding goes, how about if we get this thing up and running, then I'll find an AM radio and see how bad the interference is and go from there?

I am using twisted wire (solid strand copper 16g and 18g, 8'-15' long) from the PDMs to the pumps.

Click to expand...

Good.

As far as shielding goes, how about if we get this thing up and running, then I'll find an AM radio and see how bad the interference is and go from there?

Click to expand...

Good suggestion. If all the control stuff is in a metal enclosure that should help.

Thanks Alec.
From now on, I'll use the terms "power supply backup with beefy diodes" and "battery backup" to differentiate the two.

The battery backup is on one side of that board and is cleanly seperated from everything else. I mixed 12V and 13.8V at the terminal points.

I mixed 12V and 13.8V at the terminal points.

Click to expand...

The 12V from the 7812 and the nominal '12V' (=13.8V) from the LM317 must be kept separate.

I know, I just wanted to make the point that no wires are soldered between the battery backup and the wave timer or filter timer (turns filter pumps off for 0-15 min.)

I do have the filter timer and wave timer sharing a 12V line. Is that OK? Or should the filter timer's 12V line be disconnected from the wave timer and connected to 13.8V at the battery backup? Or, do I need to set up a dedicated 7812 Vreg for the filter timer?

Or should the filter timer's 12V line be disconnected from the wave timer and connected to 13.8V at the battery backup?

Click to expand...

Definitely. The filter timer is part of the filter system (even if, physically, it's located on a board with parts of the wave and tidal systems). Now that you've explained the board layout I think it might help if I drew up another 'how it all connects together' pic. At least we'll then be working from the same hymn sheet.
BTW, as the 13.8V is uncomfortably close to the rated 15V limit of the 4093 ICs in the filter system PDMs, for improved reliability it might be advisable to put a couple of diodes (1N4148 or 1N400x) in series between the 13.8V line and the '12V' inputs of those two PDMs, to drop the volts down to ~ 12.6. I'll try to remember to add those on the pic and re-label the battery's '12V' as 12.6 accordingly.

Did you build/are you using the Fault Sensor Module (FSM-Mk1) for either the wave or tidal system?

alec_t said:

Definitely. The filter timer is part of the filter system (even if, physically, it's located on a board with parts of the wave and tidal systems). Now that you've explained the board layout I think it might help if I drew up another 'how it all connects together' pic. At least we'll then be working from the same hymn sheet.
BTW, as the 13.8V is uncomfortably close to the rated 15V limit of the 4093 ICs in the filter system PDMs, for improved reliability it might be advisable to put a couple of diodes (1N4148 or 1N400x) in series between the 13.8V line and the '12V' inputs of those two PDMs, to drop the volts down to ~ 12.6. I'll try to remember to add those on the pic and re-label the battery's '12V' as 12.6 accordingly.

Click to expand...

Thanks Alec, I’ll get the filter timer “12V” separated from the wave timer 12V. (Neither the filter timer nor the wave timer were hooked up when I was busy cooking things by mixing 13.8V and 12V)
Very much appreciate the diode upgrade. A revised ‘how it connects together’ would be a ton of help. The power ground-signal ground connections have me at a loss. Thank you!

alec_t said:

Did you build/are you using the Fault Sensor Module (FSM-Mk1) for either the wave or tidal system?

Click to expand...

You made me re-visit FSM-Mk1 and I’m glad you did. It would be good to know if a GFI tripped!
So, yes I’ve got the parts and will build an FSM-Mk1 for the four wave pumps and another FSM-Mk1 for the four tide pumps. Could an FSM-Mk1 be added to the filter pumps without complications? If so, I’d like to.
Does the FSM-Mk1 sound the alarm in the alarm module that’s already built?

Could an FSM-Mk1 be added to the filter pumps without complications?

Click to expand...

Yes. It could be simplified, because only two of its four sections would be used.

Does the FSM-Mk1 sound the alarm in the alarm module that’s already built?

Click to expand...

Indirectly, yes. The FSM generates a 'Fault' signal which provides an input to U1d in the respective PDM, and the PDM then generates an 'Alarm' signal which goes to a respective input of the Alarm section of the TTAM. That Alarm section/module will then have 10 'Alarm' inputs (4 from the wave system, 4 from the tidal system, 2 from the filter system).

Thanks, I was not sure if hooking up the FSM-Mk1 to the filter pumps would be crazy complicated due to battery backup.

I have 10 FSM-Mk1 modules on a board. I have one terminal point for signal ground and another terminal point for 12V. Is that OK?
There are six unused pins on the third LM324. Do any of them need to be grounded?

Joe

For the LM324, you don't HAVE to ground them, but you could ground the unused terminals labeled (+) and (-), the non-inverting and inverting inputs. Definately, don't ground the ouputs.