DC to DC boost converter and HBridge to drive a 5v latching solenoid

Thread Starter

Marc Piulachs

Joined Aug 13, 2019
14
Following a previous post I made about my watering hose timer reverse enginering and with the suggestions about using a step-up IC designed to boost the voltage I came up with the following design, since I decided not to use the original design and make a new implementation I have created a new post.

LoraHose-page-001.jpg

My question here is, since this is my first attempt, I would like to have some advice to check whether is anything wrong with this design? Are there any obvious flaws with it?.
  1. Is a good idea to power the HBRIDGE from the booster circuit ?
  2. The mosfet controlling the booster circuit, is a good choice?
  3. I know the led part seems wired, I haven't used any multiplexing technique nor any IC but I want to reduce the number of components of the design to improve battery life and since I will be using a MCU with lots of spare pins and the leds will be placed all on the top layer of the PCB and the mcu on the bottom it will make it easier for me to route it.
I'd also be glad to hear general opinions about the design.
 

Alec_t

Joined Sep 17, 2013
10,584
The pdf is helpful.
1) If I got my sums right, the Booster voltage exceeds, or is perilously close to, the maximum rated supply voltage of the RZ7899.
2) The AO3400 is not a logic-level type and its 30V rating is arguably too low for the application. I'd use a beefier, logic-level type.
It is never a good idea to run components near their maximum current/voltage rating if you want them to have a long life.
Is C20 (2200pF) really a 10V rated electrolytic?
Q1 should be a P-FET but the schematic uses an N-FET symbol.
 

Thread Starter

Marc Piulachs

Joined Aug 13, 2019
14
If I got my sums right, the Booster voltage exceeds, or is perilously close to, the maximum rated supply voltage of the RZ7899.
You are right, the values for the resitors were wrong , I just copy pasted the symbols . I updated the values to generate 5.1V

booster.png

2) The AO3400 is not a logic-level type and its 30V rating is arguably too low for the application. I'd use a beefier, logic-level type.
It is never a good idea to run components near their maximum current/voltage rating if you want them to have a long life.
The AO3400 should turn on at 3.3V , if I use a logic-level can I remove R38?

Is C20 (2200pF) really a 10V rated electrolytic?
Yes, it is.

Q1 should be a P-FET but the schematic uses an N-FET symbol.
The symbol was wrong, It is really a P-FET.

Thanks for your help Alec :)
 

cmartinez

Joined Jan 17, 2007
6,621
The AO3400 should turn on at 3.3V , if I use a logic-level can I remove R38?
You should remove R38 regardless. A Fet's gate behaves pretty much like a capacitor, that is because a Fet is voltage driven. Resistors at their gates should only be installed when working with higher frequencies (usually higher than 100 KHz, or when the application requires a slow turn on time), and then those resistors should be of small values. Typically between 10 and 100 ohms.
 

cmartinez

Joined Jan 17, 2007
6,621
Also, the AO3400 datasheet indicates that at a gate to source voltage of 3.3V, the drain to source resistance is about 21 mΩ, so I think you should be ok using this same nFet.

upload_2019-8-20_17-18-50.png
EDIT: I just read Alec's post recommending a beefier transistor ... I didn't know about the higher working voltage requirement.
 

iimagine

Joined Dec 20, 2010
388
The method for turning on/off the booster is inefficient, all you did there was turning on/off the chip. Power still dissipate across the rest of the circuit on the right. Consider using the DMP2035U P-Mos scheme like Q3 in your Power Supply section to switch on off VCC power to the boost converter, shutting the whole thing off

Edit: Or tie pin 2 of the MT3608 to the circuit's GND and use a beefier N-Mos to connect or cut off GND to the entire booster, not just the IC
 
Last edited:

Thread Starter

Marc Piulachs

Joined Aug 13, 2019
14
BSS84 is a P-Mos, you used the N-Mos symbol, also as for R35, you are not doing rapid switching so its value could be changed to 1-10Meg to save battery life.
The symbol was wrong also changed R35 resistor value to 1M

batteryMeter.png

The method for turning on/off the booster is inefficient, all you did there was turning on/off the chip. Power still dissipate across the rest of the circuit on the right. Consider using the DMP2035U P-Mos scheme like Q3 in your Power Supply section to switch on off VCC power to the boost converter, shutting the whole thing off

Edit: Or tie pin 2 of the MT3608 to the circuit's GND and use a beefier N-Mos to connect or cut off GND to the entire booster, not just the IC
Right, I like the second solution better.

booster.png

EDIT: I just read Alec's post recommending a beefier transistor ... I didn't know about the higher working voltage requirement.
VCC is 3xAAA batteries, AO3400 is rated at 30V, why do I need a beefier transistor?
 

Attachments

Thread Starter

Marc Piulachs

Joined Aug 13, 2019
14
Because you originally had the wrong feedback resistor values it seemed that you were trying to get a boost voltage ~27V. Now that you've corrected the values to give a 5.1V boost voltage, it is clear that the beefier FET is not needed.
Ok, It was what I thought , just wanted to be sure I was not missing something here ...the resistors had wrong values but even in that case Q2 was not affected in any way (if I'm not wrong). Q2 was switching On/Off the MT3608 IC only, the theoretical ~27V where driven by the RZ7899 IC.
 

Thread Starter

Marc Piulachs

Joined Aug 13, 2019
14
Since MT3608 has an E(nable) input, why don't you just use this pin and have no concerns about it?
The MT3608 has EN pin but it does not disable it completly , from the datasheet:

Regulator On/Off Control Input. A high input at EN turns on the converter, and a low input turns it off. When not used, connect EN to the input supply for automatic startup.

In this state, the output voltage is equal to the input voltage, therefore it will be charging the capacitor even when not needed.
 

Plamen

Joined Mar 29, 2015
98
Following a previous post I made about my watering hose timer reverse enginering and with the suggestions about using a step-up IC designed to boost the voltage I came up with the following design, since I decided not to use the original design and make a new implementation I have created a new post.

View attachment 184210

My question here is, since this is my first attempt, I would like to have some advice to check whether is anything wrong with this design? Are there any obvious flaws with it?.
  1. Is a good idea to power the HBRIDGE from the booster circuit ?
  2. The mosfet controlling the booster circuit, is a good choice?
  3. I know the led part seems wired, I haven't used any multiplexing technique nor any IC but I want to reduce the number of components of the design to improve battery life and since I will be using a MCU with lots of spare pins and the leds will be placed all on the top layer of the PCB and the mcu on the bottom it will make it easier for me to route it.
I'd also be glad to hear general opinions about the design.
Petkan:
What us the ultimate goal? Timer can be easily made using 4541. It can achieve hours.
Why boosting? What is to be controlled and what is your power source?
 
Top