PCB check, and ideas

Tizfaver
Posts: 9
Joined: Thu Jan 26, 2023 6:42 pm

PCB check, and ideas

Postby Tizfaver » Thu Feb 02, 2023 8:46 pm

I made this PCB taking inspiration from picoclick-c3t and 01Space ESP32-C3-0.42LCD. So I had bought the 01space esp32, but I only realize later that it doesn't have a Lipo battery charging module. So I wanted to create the same tab, but with the picoclick top up module.

The circuit should be fine, can you help me check? moreover if you have ideas to add you can tell me, it would be very interesting.

For now the circuit has a 0.42 display, the charging module is still to be decided (but I thought about the MCP73831/2), and also for the 3v3 converter I have more or less chosen, but it makes little difference which one. I don't know if use esp32-c3fn4 or esp32-c3fh4 (what you think is better, they are more or less the same).

Plus the idea was to keep it simple, so no rgb leds, and no leds to indicate power on (to save current). Just a display. But as I said, ideas are always welcome!

Schematic attached below the post. Even the PCB.
Attachments
bottom_.png
bottom_.png (60.47 KiB) Viewed 2566 times
front_.png
front_.png (41.61 KiB) Viewed 2566 times
Schematic_ESP32C3-LCD-LIPO_2023-02-02.png
Schematic_ESP32C3-LCD-LIPO_2023-02-02.png (126.18 KiB) Viewed 2566 times

ESP_Sprite
Posts: 8884
Joined: Thu Nov 26, 2015 4:08 am

Re: PCB check, and ideas

Postby ESP_Sprite » Fri Feb 03, 2023 1:30 am

Before I start, have you looked at the hardware design guidelines?

- Your schematic is a mess of wires. Suggest you use nets (=those labels with the same name, as a random example this uses a lot of them) to clear things up so we don't have to manually follow traces. I probably didn't catch all mistakes in the schematic because of this.
- Your LCD will not get a good reset pulse as it's reset line is connected directly to (what I presume is) the 3.3V line. Better to connect it to an IO pin or at least add a RC network in there to reset it on power-up.
- Similarly, CHIP_EN of the ESP32 only has a pullup. It needs a proper power-on reset to function. Suggest adding an 1uF capacitor to ground, combined with the existing pullup this will form a working reset RC network.
- You probably want to bring GPIO9 out to at least a testpad that you can somehow short to ground. In some cases, it may be necessary to put the C3 into download mode manually, and that would allow that.
- What LDO are you using? Are you sure it can provide 500mA plus whatever your OLED uses?
- What's up with D1 and R16? They're shorted, so they won't do anything.
- It's not good to allow the device to function off the battery while the battery is charging. Better to switch it over to the 5V from USB when that happens. This document tells you how.
- The antenna generally needs an impedance-matching pi network before being connected to the ESP. See the hardware design guidelines for more info.
- What was your plan with the circuit around D2 and D3? Soft power-off? If so, you may want to redesign it, because as is, it will simply keep EN of the LDO high as long as the battery has any voltage on it, and if it hasn't the C3 also won't be working, meaning R8,R9,D3 are superfluous.

Tizfaver
Posts: 9
Joined: Thu Jan 26, 2023 6:42 pm

Re: PCB check, and ideas

Postby Tizfaver » Fri Feb 03, 2023 6:58 pm

ESP_Sprite wrote:
Fri Feb 03, 2023 1:30 am
Before I start, have you looked at the hardware design guidelines?

- Your schematic is a mess of wires. Suggest you use nets (=those labels with the same name, as a random example this uses a lot of them) to clear things up so we don't have to manually follow traces. I probably didn't catch all mistakes in the schematic because of this.
- Your LCD will not get a good reset pulse as it's reset line is connected directly to (what I presume is) the 3.3V line. Better to connect it to an IO pin or at least add a RC network in there to reset it on power-up.
- Similarly, CHIP_EN of the ESP32 only has a pullup. It needs a proper power-on reset to function. Suggest adding an 1uF capacitor to ground, combined with the existing pullup this will form a working reset RC network.
- You probably want to bring GPIO9 out to at least a testpad that you can somehow short to ground. In some cases, it may be necessary to put the C3 into download mode manually, and that would allow that.
- What LDO are you using? Are you sure it can provide 500mA plus whatever your OLED uses?
- What's up with D1 and R16? They're shorted, so they won't do anything.
- It's not good to allow the device to function off the battery while the battery is charging. Better to switch it over to the 5V from USB when that happens. This document tells you how.
- The antenna generally needs an impedance-matching pi network before being connected to the ESP. See the hardware design guidelines for more info.
- What was your plan with the circuit around D2 and D3? Soft power-off? If so, you may want to redesign it, because as is, it will simply keep EN of the LDO high as long as the battery has any voltage on it, and if it hasn't the C3 also won't be working, meaning R8,R9,D3 are superfluous.
Ok thanks, I will remove all the wire and just use labels. Thanks for the hardware design guidelines doc esp32 (i could find it). I will come back with the all other error fixed.

Tizfaver
Posts: 9
Joined: Thu Jan 26, 2023 6:42 pm

Re: PCB check, and ideas

Postby Tizfaver » Fri Feb 03, 2023 10:30 pm

ESP_Sprite wrote:
Fri Feb 03, 2023 1:30 am
Before I start, have you looked at the hardware design guidelines?

- Your schematic is a mess of wires. Suggest you use nets (=those labels with the same name, as a random example this uses a lot of them) to clear things up so we don't have to manually follow traces. I probably didn't catch all mistakes in the schematic because of this.
- Your LCD will not get a good reset pulse as it's reset line is connected directly to (what I presume is) the 3.3V line. Better to connect it to an IO pin or at least add a RC network in there to reset it on power-up.
- Similarly, CHIP_EN of the ESP32 only has a pullup. It needs a proper power-on reset to function. Suggest adding an 1uF capacitor to ground, combined with the existing pullup this will form a working reset RC network.
- You probably want to bring GPIO9 out to at least a testpad that you can somehow short to ground. In some cases, it may be necessary to put the C3 into download mode manually, and that would allow that.
- What LDO are you using? Are you sure it can provide 500mA plus whatever your OLED uses?
- What's up with D1 and R16? They're shorted, so they won't do anything.
- It's not good to allow the device to function off the battery while the battery is charging. Better to switch it over to the 5V from USB when that happens. This document tells you how.
- The antenna generally needs an impedance-matching pi network before being connected to the ESP. See the hardware design guidelines for more info.
- What was your plan with the circuit around D2 and D3? Soft power-off? If so, you may want to redesign it, because as is, it will simply keep EN of the LDO high as long as the battery has any voltage on it, and if it hasn't the C3 also won't be working, meaning R8,R9,D3 are superfluous.
So now:
- The schematic is netlabeled.
- The lcd RES is connected to a esp32 pin, as the doc says: RES#
"Power Reset for Controller and Driver, This pin is reset signal input. When the pin is low, initialization of the chip is executed. Keep this pin pull high during normal operation".
- The LDO choosen is the Voltage Regulator MIC5504 3V3 LDO -> so can provide 500ma.
- D1 and R16 are no long needed.

Is now the CHIP_EN correct now? just added the capacitor.

I saw a lot of people connecting the annect just like and "extension" of the pin. Can't this be correct?

But I have a question, about the battery... I could add that if the USB is plugged in, it set LOW the pin EN of the Voltage Regulator MIC5504, which will not discharge the battery, right? using a transistor the invert the HIGH signal from the USB to a LOW signal?
So I removed R6, R7, R8, R9, D2, D3, C6, but now GPIO3 is not connected (can I leave it not attached?).

for the GPIO9, I haven't decided yet, becuase for the bootloader I just have this thing: https://it.aliexpress.com/item/10050049 ... pt=glo2ita
Attachments
sheet_labeled.png
sheet_labeled.png (121.89 KiB) Viewed 2451 times

ESP_Sprite
Posts: 8884
Joined: Thu Nov 26, 2015 4:08 am

Re: PCB check, and ideas

Postby ESP_Sprite » Sat Feb 04, 2023 1:03 am

Ah, that's much more readable. Just a nitpick: some people don't like 'Christmas trees' in their schematic, in other words, if you want to make it really clean, you want all your ground symbols to point down and your power rail symbols to point up. No need to change it for readability in this schematic, I'm just saying it in case you want to show off your next one.

You could indeed run a transistor circuit to disable the LDO when charging. Alternatively, the application note I linked before has a pretty simple transistor-diode circuit that feeds 5V from USB directly into the LDO when USB is connected, allowing the ESP to keep working.

I don't know what version of that LDO you're using, but https://www.microchip.com/en-us/product/MIC5504 tells me the LDO can only do 300mA? If so, that is insufficient.

On the antenna: Connecting an antenna directly to this will probably work, but will A. have diminished WiFi range and B. might not pass EMC testing if you want to make this into a commercial product.

Tizfaver
Posts: 9
Joined: Thu Jan 26, 2023 6:42 pm

Re: PCB check, and ideas

Postby Tizfaver » Sat Feb 04, 2023 4:05 pm

ESP_Sprite wrote:
Sat Feb 04, 2023 1:03 am
Ah, that's much more readable. Just a nitpick: some people don't like 'Christmas trees' in their schematic, in other words, if you want to make it really clean, you want all your ground symbols to point down and your power rail symbols to point up. No need to change it for readability in this schematic, I'm just saying it in case you want to show off your next one.
Uh, interesting.
ESP_Sprite wrote:
Sat Feb 04, 2023 1:03 am
You could indeed run a transistor circuit to disable the LDO when charging. Alternatively, the application note I linked before has a pretty simple transistor-diode circuit that feeds 5V from USB directly into the LDO when USB is connected, allowing the ESP to keep working.
I don't know what version of that LDO you're using, but https://www.microchip.com/en-us/product/MIC5504 tells me the LDO can only do 300mA? If so, that is insufficient.
I will try implement the circuit you linked. And I will change the Voltage Coverter because I think too that 300mA is not enough. But let's take this LDO as an example again, the documentation says that the input voltage for the EN pin can be the same as the IN pin. And so I don't need to do some modifications than the direct connection between the VBUS pin and the EN pin of the LDO (I mean other the transistor that invert the signal)?
The new LDO could be: SPX3819M5-L-3-3 500mA 3,3V SMD.
ESP_Sprite wrote:
Sat Feb 04, 2023 1:03 am
On the antenna: Connecting an antenna directly to this will probably work, but will A. have diminished WiFi range and B. might not pass EMC testing if you want to make this into a commercial product.
I do not really care about the WiFi range, my objective is the BLE connection with the phone (So some meters from the phone). And I think I will not make this a commercial product, just a project that people can download and order by their own.
But as I said I care more about the BLE connection. With just a direct connection how much would be the range? Internet say about 7 meter max range, is that correct? but I don't really know because it's a quite "big" ceramic antenna (about 1cm x 2mm x 1mm).

ESP_Sprite
Posts: 8884
Joined: Thu Nov 26, 2015 4:08 am

Re: PCB check, and ideas

Postby ESP_Sprite » Sun Feb 05, 2023 1:36 am

Tizfaver wrote:
Sat Feb 04, 2023 4:05 pm
Uh, interesting.
Fyi, the reason for that convention is readability: if all positive voltages point up and all grounds point down, it's trivial to see at a glance what gets pulled up and pulled down. As that's kinda important information, the easier it is to be able to see that, the better.
Tizfaver wrote:
Sat Feb 04, 2023 4:05 pm
I will try implement the circuit you linked. And I will change the Voltage Coverter because I think too that 300mA is not enough. But let's take this LDO as an example again, the documentation says that the input voltage for the EN pin can be the same as the IN pin. And so I don't need to do some modifications than the direct connection between the VBUS pin and the EN pin of the LDO (I mean other the transistor that invert the signal)?
The new LDO could be: SPX3819M5-L-3-3 500mA 3,3V SMD.
Yes, you could use an inverter on the EN pin. But given that it uses as many components as a 5V bypass, I'd go for that option.
Tizfaver wrote:
Sat Feb 04, 2023 4:05 pm
I do not really care about the WiFi range, my objective is the BLE connection with the phone (So some meters from the phone). And I think I will not make this a commercial product, just a project that people can download and order by their own.
But as I said I care more about the BLE connection. With just a direct connection how much would be the range? Internet say about 7 meter max range, is that correct? but I don't really know because it's a quite "big" ceramic antenna (about 1cm x 2mm x 1mm).
It's hard to say; it depends on the routing, PCB substrate, specific antenna and a bunch of other things. You can try without the pi network and see how far you can get. Alternatively, you can put in the footprints for the pi network but initially use a zero-ohm resistor to connect the antenna to the ESP directly: if that gives sufficient range you can keep it like that, if not you can always add the proper components later.

Tizfaver
Posts: 9
Joined: Thu Jan 26, 2023 6:42 pm

Re: PCB check, and ideas

Postby Tizfaver » Sun Feb 05, 2023 3:16 pm

ESP_Sprite wrote:
Sun Feb 05, 2023 1:36 am
Tizfaver wrote:
Sat Feb 04, 2023 4:05 pm
Uh, interesting.
Fyi, the reason for that convention is readability: if all positive voltages point up and all grounds point down, it's trivial to see at a glance what gets pulled up and pulled down. As that's kinda important information, the easier it is to be able to see that, the better.
Tizfaver wrote:
Sat Feb 04, 2023 4:05 pm
I will try implement the circuit you linked. And I will change the Voltage Coverter because I think too that 300mA is not enough. But let's take this LDO as an example again, the documentation says that the input voltage for the EN pin can be the same as the IN pin. And so I don't need to do some modifications than the direct connection between the VBUS pin and the EN pin of the LDO (I mean other the transistor that invert the signal)?
The new LDO could be: SPX3819M5-L-3-3 500mA 3,3V SMD.
Yes, you could use an inverter on the EN pin. But given that it uses as many components as a 5V bypass, I'd go for that option.
Tizfaver wrote:
Sat Feb 04, 2023 4:05 pm
I do not really care about the WiFi range, my objective is the BLE connection with the phone (So some meters from the phone). And I think I will not make this a commercial product, just a project that people can download and order by their own.
But as I said I care more about the BLE connection. With just a direct connection how much would be the range? Internet say about 7 meter max range, is that correct? but I don't really know because it's a quite "big" ceramic antenna (about 1cm x 2mm x 1mm).
It's hard to say; it depends on the routing, PCB substrate, specific antenna and a bunch of other things. You can try without the pi network and see how far you can get. Alternatively, you can put in the footprints for the pi network but initially use a zero-ohm resistor to connect the antenna to the ESP directly: if that gives sufficient range you can keep it like that, if not you can always add the proper components later.
About the antenna, I don't undersand why use 0 ohm resistor? just like a jumper? So I could try the antenna like is now?

About the inverter, I have a question, I wanted to use the NPN MMBT9014 transistor, but the datasheet (https://pdf1.alldatasheet.com/datasheet ... T9014.html) say that the Collector Current is only 100ma... do you think it will get quite hot?
This is the schematic I have done for the transistor: attached to this post.

ESP_Sprite
Posts: 8884
Joined: Thu Nov 26, 2015 4:08 am

Re: PCB check, and ideas

Postby ESP_Sprite » Mon Feb 06, 2023 2:28 am

Tizfaver wrote:
Sun Feb 05, 2023 3:16 pm
About the antenna, I don't undersand why use 0 ohm resistor? just like a jumper? So I could try the antenna like is now?
Because the pi network (of which the footprint you would be putting in) has an inductor connecting the LNA pin of the ESP to the antenna; if you put nothing there, the antenna would be disconnected. Hence a 0R link if you don't need the pi network.

Tizfaver wrote:
Sun Feb 05, 2023 3:16 pm
About the inverter, I have a question, I wanted to use the NPN MMBT9014 transistor, but the datasheet (https://pdf1.alldatasheet.com/datasheet ... T9014.html) say that the Collector Current is only 100ma... do you think it will get quite hot?
This is the schematic I have done for the transistor: attached to this post.
I don't see the schematic, but generally you can calculate that by looking up the max current through the EN pin as well as the current through any pull-up you may have in circuit. Generally, unless you have some weird value for the pullup resistor ('weird' as in <33 ohm), it should work fine.

Tizfaver
Posts: 9
Joined: Thu Jan 26, 2023 6:42 pm

Re: PCB check, and ideas

Postby Tizfaver » Mon Feb 06, 2023 2:01 pm

ESP_Sprite wrote:
Mon Feb 06, 2023 2:28 am
Because the pi network (of which the footprint you would be putting in) has an inductor connecting the LNA pin of the ESP to the antenna; if you put nothing there, the antenna would be disconnected. Hence a 0R link if you don't need the pi network.
Oh now I understand, i have done some research about the Pi network and now I understand that I need it (the 0 ohm resistor).
ESP_Sprite wrote:
Mon Feb 06, 2023 2:28 am
I don't see the schematic, but generally you can calculate that by looking up the max current through the EN pin as well as the current through any pull-up you may have in circuit. Generally, unless you have some weird value for the pullup resistor ('weird' as in <33 ohm), it should work fine.
I don't know why it didn't attached the image, but as you say now I think that should work.
Attached again the image here.
Attachments
sheet_labeled_with_npn.png
sheet_labeled_with_npn.png (198.09 KiB) Viewed 2159 times

Who is online

Users browsing this forum: No registered users and 34 guests