CP2012N - What did I do wrong??

[siconic]

Hey guys!

So I finally thought I had a CORRECT prototype, as far as I can tell. This will be my third try and protype at this, but the CP2102N is not being detected by the PC when I plug it in. Can anyone point me in the right direction? How do I troubleshoot?

Here is the design (I know the traces could be laid out a bit better, open to suggestions, but that is not my current issue, or is it?):





What am I missing? I am going for an external power, so I can power up to 1A, and there is a cap on the input that is not in the diagram.

I do kind of suspect my D+ and D- traces, but i could not figure out for the life of me how to lay them out otherwise.

[Nusa]

I don't see a signal ground connection to your USB connector.

[AndyC_772]

The bulk cap on the 5V rail, next to the voltage regulator, looks to have no GND connection either.

Doesn't your PCB software have a feature to check that a board is fully routed before generating Gerbers?

[me.rishabh12]

i designed a couple of circuit which uses cp2102 and they all work fine
i thin u should feed regin with 5v and connect vdd to ground using a cap
hope it will solve the problem
pls refer to schematic i used in my design attached below

[siconic]

The bulk cap on the 5V rail, next to the voltage regulator, looks to have no GND connection either.

Doesn't your PCB software have a feature to check that a board is fully routed before generating Gerbers?

Its connected by a ground plane, I just disabled the ratsnest, so it wouldnt show up.

Sent from my SM-G950U using Tapatalk

[siconic]

i designed a couple of circuit which uses cp2102 and they all work fine
i thin u should feed regin with 5v and connect vdd to ground using a cap
hope it will solve the problem
pls refer to schematic i used in my design attached below

Could it be that my RST is powered by 3.3v?

Sent from my SM-G950U using Tapatalk

[phil from seattle]

There is apparently a problem with USB 3 and the CP2102N.  Read this article https://www.allaboutcircuits.com/technical-articles/project-troubleshooting-solutions-from-a-problem-magnet/

[bson]

Use a continuity tester and verify that you have GND on pin 5, with 4 NC.  On the ones I use (TE Connect) pin 4 is ground and 5 is shield.

If that seems to be the issue, try bodging a bridge between pins 4 and 5.

[me.rishabh12]

No it can be connected to 3.3 v no issues
Have u install the driver for cp2102 usb to ttl drivers
Try to update the software and connect regin to 5 v
I hope it will work maybe the interal LDO in cp2102
Is not working on 3.3v
Pls refer to the datasheet they have some application reference circuit .

[phil from seattle]

I'm working on a CP2102N design. Will get my PCB back soon, hopefully.

It helps to post the entire schematic. You never know where the problem lies.

Your design uses the form from Figure 2.3 of the data sheet which has VREGIN and VDD connected.  Have you verified 3.3V? 

I do not understand the connect pin 4 to 5 comment - those are D- and D+. Surely that wasn't intended.

Your D+ and D- routing itself is ok but the proximity to your VReg is a potential problem area. If you lay it out again, I'd get more separation.  If nothing else, move the USB connector to the left to stay away from your caps. I'd bet, however, that this is not your problem.

Have you verified voltages? 5V on VBUS? 3.3V on VDD?  What do you see on the 2 suspend pins?

More a question - you seem to be using a CP2102 (non-N) in the schematic and the 28pin QFN on the board.  Have you verified the library part is the correct one for the actual chip you are using? The pin assignments look correct.

Have you looked at the CP2102N Mini Eval Kit schematic? It shows SUSPENDn being pulled to gnd via a 10K R. It also has it driving an LED to gnd so it's a bit confusing. Also, it pulls RSTb to VDD (3.3V) via a 1K R and not the 4.7K you are using. Probably not an issue but worth thinking about.

Check if you are running into the USB 3.0 problem I mentioned above.

And do report back what you find.

[Peabody]

I would just add what I've mentioned before.  You might want to get one of those $1.35 CP2102 modules from Ebay so you can confirm that all the software is working on the PC side.  You plug it into USB, and you get the dah-dah sound that something has connected to USB, and the LED on the module lights up, and then you can find the COM port in Device Manager - I'm assuming Windows here.  At least that tells you that the driver and the USB port are working.  And you can connect the module's Tx and Rx pins and do loopback testing in your favorite terminal program.

If all that works, but your circuit doesn't, that just tells you that the problem is almost certainly in your circuit.  I say almost because those modules use the non-N part.  The N is supposed to be a drop-in replacement for the original, but there may actually be differences that matter.  The USB3 issue, for example, I don't recall ever being mentioned with respect to the original CP2102.  If you're talented at reflow, you could replace the old part on the module with an N part, which could then give you a direct comparison.  So far, I haven't seen a module with the N part installed.

[phil from seattle]

@peabody, good points there.   The mini eval kit (which has the CP2102N in it) is $8 from a disti or the SiLabs store. I have one coming, should be here in a few days as well.

[bson]

I do not understand the connect pin 4 to 5 comment - those are D- and D+. Surely that wasn't intended.

I went and checked to make sure, and yeah pin 4 is shield and pin 5 is ground so the design as done is correct.

Pinout:

[bson]

You can also install Wireshark to sniff USB traffic by VID:PID.  It has an optional library you install for USB.  This lets you see if there's a setup problem.

[siconic]

@peabody, I do have a cp2102 ftdi, and it is recognized in the same port. In fact, I can use it with the FTDI pins I built into the design (so this could ALSO be used as an ftdi) and all the LED's I built in show data, and I can leave the power portion of the external FTDI disconnected and itbalso works. Just not the onboard CP2012.

Also, the only difference between the CP2102 and CP2102N pinout is the NC pins now all have functions, but can be left out (as far as I can tell) making the CP2102N capable of working on a CP2102 board, just lacking the extra features.

 Of course, I am reflowing using my stove ATM, so maybe I hurt the chip, or maybe my reflow wasnt so good.

Sent from my SM-G950U using Tapatalk

[phil from seattle]

I do not understand the connect pin 4 to 5 comment - those are D- and D+. Surely that wasn't intended.

I went and checked to make sure, and yeah pin 4 is shield and pin 5 is ground so the design as done is correct.

Ah, you were talking about the USB connector, not the CP2102N.  Though, isn't pin 4 for OTG and not the shield?

By the way, the eval kit does not connect GND and shield directly but does via a 1M resistor and a cap.

[BrianHG]

To ensure best D+/D- matching on the USB, I would eliminate the VIA and route your PCB as shown below.
Please excuse the cheap paint software touch-up job, but you should get the idea.....

[phil from seattle]

Yeah, I hate that the CP2101N swaps the D+ and D- line.  I suppose it solved a problem in the chip but it dumped the problem on us.

Are those vias really going to cause a problem? I get that avoiding it will have better signal integrity but will it degrade that much?

[Monkeh]

It doesn't swap them - your connector is upside down. Seriously, that's a bottom mount connector.

Either way, putting a via in isn't going to trouble it, it's full speed, not high speed.

[Nusa]

Yeah, I hate that the CP2101N swaps the D+ and D- line.  I suppose it solved a problem in the chip but it dumped the problem on us.

No such thing as a universal solution when you don't have a universal connector specification. USB connectors are available in a variety of mounting options and in two orientations, one of which would reverse the order of the pins on the board.

[phil from seattle]

references?

[Monkeh]

references?

For what, connector orientation?

Look at the nearest micro plug - note the side which has the USB logo. Plug it into a phone - logo faces you. Plug it into a tablet - logo faces you. Plug it into any board with that socket on the top.

[Nusa]

references?

For what, connector orientation?

Look at the nearest micro plug - note the side which has the USB logo. Plug it into a phone - logo faces you. Plug it into a tablet - logo faces you. Plug it into any board with that socket on the top.

That's an asthetic standard, not a connector specification. Which actually reinforces my point...to meet that standard regardless of the relationship of the internal board to whatever is declared the "top" of the device, you need multiple versions of the connector to exist. They're easy to find in the parts sites, if you only go looking for them. But here's a graphic to give you the idea: