HPA controller build

dstroy

Insanely Active Member
I'll update this thread with what I'm working on.

Still in the design process.

Right now I'm working on power stuff.

I'm going to use two power supplies (system voltage 24v), one has a battery backup and will only be used if the power goes out.

Meanwell makes these:
UHP-250-24
PSC-160B
NID-100-12
NID-65-5


Current design:
Main -

UHP-350-24

Backup -

PSC-160B

conversion
-----------------------------
24vdc -> 12vdc (NID 12v)
24vdc -> 5vdc (NID 5v)
24vdc -> 5vdc(NID 5v) -> 3v3 (linear) for sensors

24v -> 10v dimming voltage supply
24v -> 7v mega2560
24v -> 5v -> battery select (3v3) -> m4 express

on battery -- essential
1x NID 5v supply
7v mega2560
5v -> battery select board -> m4 express

on battery -- intermittent
12v NID100 (solenoids)
24v (extraction) -- depends on how big the battery I buy is.

I've got 350w main power, and 160w if it's on battery. Plenty to turn a bunch of shit on and off.

Right now I'm working on the layout for power conversion/distribution

8101

the ICs are panasonic AQW414 (photomos SSR), they are NC and short pin 13 on the Meanwell NID DC/DC converter to ground, which shuts the converter output off.





The caps and inductors are to make them comply with class B emissions

I still have to set up the trim circuits, select the correct barrier diodes so I can OR the 12v | 12v-> and 5v | 5v ->
They just need some schottky diodes.

The two main power supplies needed something different:


Anyway, that's what I'm thinking about. And how to cram all of this in the box I have.

I've got a chunk of heatsink I'm going to use for the UHP350, which I've already lapped flat.


Take it easy
 

Attachments

Capt C

Saltwater Cowboy
Way over my head, I felt dumb then I remembered I'm just uneducated in such matters. 🤣 Pulling up next to you pro and copying your notes 🤣
He does not mind dumb questions though. I have asked plenty. @dstroy i am sure you have plenty on your plate but i am sure everyone would enjoy seeing some of your oil making posts. We have a few threads started but your info would be an awesome addition.
 

dstroy

Insanely Active Member
Redundancy,

If this system goes completely offline for more than a couple of hours it would take a while for the plants connected to it to recover.

More than eight hours and there are no guarantees that the plant will survive but you could save some cuttings maybe.

Redundancy is important.

---

What exactly is essential, then? Just nutrient solution for me, about 1/4 by volume as much as normal to keep them alive, more to minimize any drought effects.

I'm choosing to feed what I normally would with the lights off.

I need to run extraction fans periodically if the power is out for a long time.

I need to sample the environment periodically, and save that somewhere. I also need to ensure that data can't be corrupted from voltage droop.

Ok, how am I going to accomplish this.

I absolutely cannot lose +12vdc or +5vdc, that's what gets the plants fed.

Therefore, I can also never lose +24vdc because that is the main input.

Using the RDN20 to tie the UHP350 and PSC160 together in parallel, will give n+1 on the +24vdc.

Using two NID100-12 and two NID65-5 will give n+1 +5v,12v.


The microcontrollers have their own supplies, powered from the main 24vdc. Small form factor switching regulators n + 1 for +5v,9v, 1A.

The m4 has its own separate battery backup, and the system behavior changes significantly if +24vdc isn't present. If the system loses +24vdc the m4 will begin to use its own battery backup to charge a supercapacitor bank that is used to actuate the solenoids the absolute minimum needed to sustain life. I tell it which areas are occupied and it doesn't try to open solenoids that aren't there, saving a bit of energy.

Still lots to think about.
 

SSGrower

Average Grower
If you power the m4 from the usb side does this not resolve the issue with the capcitor bank issue (read the feather defaults to usb when it is powered)?

If this is true, I think you can eliminate 24v from the required when power fails category. The ac infinity is 24v but you arent using the controller so 12v should run it just fine for the limited need. This way I rhink you can get away with having only a 12v backup battery (feels like they would be cheaper but idk).

Your shit runs about 3ft over my head head and this is an idea that came in the fog of am.
 

dstroy

Insanely Active Member
If you power the m4 from the usb side does this not resolve the issue with the capcitor bank issue (read the feather defaults to usb when it is powered)?

If this is true, I think you can eliminate 24v from the required when power fails category. The ac infinity is 24v but you arent using the controller so 12v should run it just fine for the limited need. This way I rhink you can get away with having only a 12v backup battery (feels like they would be cheaper but idk).

Your shit runs about 3ft over my head head and this is an idea that came in the fog of am.
The problem would still exist. In this scenario there is no input power, the main bus is down, and the only power source left is the m4 backup battery.

It's more efficient to step down from 24v to the rest of the voltages that I need. Rather than step down, and then step back up. Instead of doing that I just power the fans directly from the 24v battery, and bypass any conversion.

operation modes
1. Normal, fans powered from 24v bus
2. "Normal", main battery normal, m4 powered from main battery, fans powered from battery
3. Low, main battery low, m4 powered from its own battery, and takes over feeding but still uses main battery to do it, no extraction, no logging except critical events
4. Emergency, no main battery, m4 is trying to keep the plants alive on its battery, no extraction, no logging

I don't want to have to intervene at all. No opening tents, no touching anything. I'm busy.
 

dstroy

Insanely Active Member
So I take it there is no way (or no logical way) to use the rdn with 2 different input voltages a 24v and 12v? That way when power goes out the main bus is kicked down to 12v and anything that requires 24v just dosent run (may need diode to prevent current drain but you obviously know more about that then me)?
No, that isn't possible.

To parallel power supplies their respective output voltages need to be very close to one another.

I want the fans to run, they need 24v, I don't want powdery mildew along with my power outage. They also run a LOT more than the solenoids are open for, solenoids about 20 minutes per day, fans are on for hours (not on battery but still more than 20 minutes). Remember I would have to step that voltage up from 12v to 24v at about 80% efficiency. After having already converted 120v to 12v at about 90% efficiency, for a combined efficiency of about 70%.
 

SSGrower

Average Grower
....Remember I would have to step that voltage up from 12v to 24v at about 80% efficiency. After having already converted 120v to 12v at about 90% efficiency, for a combined efficiency of about 70%.
I dont think you would have to do this to run the ac infinty fan, i think it will just run at about 50% on 12v. But I also forget you are dealing with miltiple footprints (res is shared by 2 tents?)
Again I am just kicking out ideas with no real basis. With the voltage input range of 21-28v on the rdn20 inputs I see the flaw in that idea tho.
I am considering something like this cause I dont want the milk to spoil either.

 

dstroy

Insanely Active Member
I dont think you would have to do this to run the ac infinty fan, i think it will just run at about 50% on 12v. But I also forget you are dealing with miltiple footprints (res is shared by 2 tents?)
Again I am just kicking out ideas with no real basis. With the voltage input range of 21-28v on the rdn20 inputs I see the flaw in that idea tho.
I am considering something like this cause I dont want the milk to spoil either.

Yeah that's neat.
 

dstroy

Insanely Active Member
I think I'm going to use these to connect the smaller power supplies together.


I'm just not keen on soldering an SSOP, very small stuff. But I think this is the best solution. I couldn't justify adding the extra dissipation (heat) from power diodes, into the sealed enclosure.
 

dstroy

Insanely Active Member
I have to teach myself how to use eagle to design this pcb I need, cause I can't solder a bunch of the supporting components. I'm working on that right now.

I've got the schematic most of the way laid out for two different supply controllers with their supporting components.

I'm working on the schematic for a third type, just seeing what the total cost would be for a few different board designs.

Whatever I can make the smallest footprint with the fewest supporting will be the cheapest most likely, unless each supply input requires their own controller (two per "OR", about $5-10 each part cost).

This part is going to take a few weeks, I'm busy with other things at the moment.
 

dstroy

Insanely Active Member
From AC infinity's manufacturer:

  • 4inch 24V inline fan, the operation voltage should be 22V-28V , start voltage is 22V
  • 6Inch 24V inline fan, operation voltage from 20-28V ( start:20V, shut off 28V)
  • 8Inch 115V inline fan, operation voltage from 110V-120V ( start:110V, shut off 120V )

I only have the 4" and 6", and this matches what I tested. Probably safe to assume the values for the 8" fan are valid. Do your own testing as always.
 

SSGrower

Average Grower
From AC infinity's manufacturer:

  • 4inch 24V inline fan, the operation voltage should be 22V-28V , start voltage is 22V
  • 6Inch 24V inline fan, operation voltage from 20-28V ( start:20V, shut off 28V)
  • 8Inch 115V inline fan, operation voltage from 110V-120V ( start:110V, shut off 120V )
I only have the 4" and 6", and this matches what I tested. Probably safe to assume the values for the 8" fan are valid. Do your own testing as always.
Better than me guessing.

I havent confirmed the actual output of their power supply rated at 24v. Since you are not using their controllers are you able to "over clock" the fans by pushing 28v?
 

dstroy

Insanely Active Member
Better than me guessing.

I havent confirmed the actual output of their power supply rated at 24v. Since you are not using their controllers are you able to "over clock" the fans by pushing 28v?
Mine is 24.2v loaded.

No, pwm 4 wire fans have a pcb inside the fan that has some components on it to regulate and limit voltage. That's why it has a min turn on and a max shut off voltage. There's a regulator on that pcb and it will drop out below or above a certain voltage.

The more work that you make the regulator circuit in the fan do, the more heat it will make. The fan gets the same power regardless, probably about 30-35w to the windings, and any extra gets dissipated (turned into heat).
 

ChiefRunningPhist

Drunk on Knowledge
I'll update this thread with what I'm working on.

Still in the design process.

Right now I'm working on power stuff.

I'm going to use two power supplies (system voltage 24v), one has a battery backup and will only be used if the power goes out.

Meanwell makes these:
UHP-250-24
PSC-160B
NID-100-12
NID-65-5


Current design:
Main -

UHP-350-24

Backup -

PSC-160B

conversion
-----------------------------
24vdc -> 12vdc (NID 12v)
24vdc -> 5vdc (NID 5v)
24vdc -> 5vdc(NID 5v) -> 3v3 (linear) for sensors

24v -> 10v dimming voltage supply
24v -> 7v mega2560
24v -> 5v -> battery select (3v3) -> m4 express

on battery -- essential
1x NID 5v supply
7v mega2560
5v -> battery select board -> m4 express

on battery -- intermittent
12v NID100 (solenoids)
24v (extraction) -- depends on how big the battery I buy is.

I've got 350w main power, and 160w if it's on battery. Plenty to turn a bunch of shit on and off.

Right now I'm working on the layout for power conversion/distribution

View attachment 8101

the ICs are panasonic AQW414 (photomos SSR), they are NC and short pin 13 on the Meanwell NID DC/DC converter to ground, which shuts the converter output off.





The caps and inductors are to make them comply with class B emissions

I still have to set up the trim circuits, select the correct barrier diodes so I can OR the 12v | 12v-> and 5v | 5v ->
They just need some schottky diodes.

The two main power supplies needed something different:


Anyway, that's what I'm thinking about. And how to cram all of this in the box I have.

I've got a chunk of heatsink I'm going to use for the UHP350, which I've already lapped flat.


Take it easy
Not sure I'm comprehending but seems like you have a 24V main supply that you're splitting into 4 different lower V supplies, 5V, 7V, 10V, and 12V? And the 5V will be split into 3 parallel 5V supplies?

Sounds cool I'll be checking in. Ive been using EAGLE too.

FWIW, DC/DC buck modules are ~$0.60/ and can be 90%+.
 

dstroy

Insanely Active Member
Not sure I'm comprehending but seems like you have a 24V main supply that you're splitting into 4 different lower V supplies, 5V, 7V, 10V, and 12V? And the 5V will be split into 3 parallel 5V supplies?

Sounds cool I'll be checking in. Ive been using EAGLE too.

FWIW, DC/DC buck modules are ~$0.60/ and can be 90%+.
The main difference between the converters that I chose vs your suggestion is MTBF, switching frequency and amount of external components needed to suppress EMI. Those small high frequency converters can conduct and radiate quite a bit of high frequency noise, which needs to be suppressed so that it does not interfere with WiFi for example. The higher the switching frequency the more likely it is to occur.
 
Top