LPC1768 input pin voltage is low with internal pull-up

[lpcware]

Content originally posted in LPCWare by ppescher on Wed Nov 23 04:39:01 MST 2011
I have seen this post:
http://knowledgebase.nxp.com/showthread.php?t=381

I confirm that after reset pins are configured as inputs with internal pull-up and voltage is measured at around 2.3V - 2.4V, with a VDD=3.3V (pin unloaded). This is too close to the VIH threshold 0.7*VDD = 2.31V to be really useful.

The datasheet does not specify that behavior, so I would like to know some official word from NXP about this, or someone that points me to official documentation.

As it is, the internal pull-up doesn't seem useful. Is the input voltage with internal pull-up guaranteed to be higher than VIH in all the operating range? Are the electrical characteristics reported in the datasheet correct? Because it doesn't seem so.

I definitely agree with user dpalchak about this issue: it does matter!
I opened a new thread to give it more attention.

[lpcware]

Content originally posted in LPCWare by ToBeFrank on Sat Nov 26 21:08:18 MST 2011
This is just an FYI for anyone who finds this thread while searching, this appears to be device specific. This is what the 11XX manual says:

---

Quote:

If the pull-up resistor is enabled (default), all non-I2C pins are pulled up to 2.6 V for LPC111x/101/201/301 parts and pulled up to 3.3 V for LPC11Cxx parts and LPC111x/102/202/302 (VDD = 3.3 V).

---

I have verified that the LPC11C24 does indeed pull-up to 3.3V with the internal pull-ups.

[lpcware]

Content originally posted in LPCWare by ppescher on Wed Nov 23 07:41:51 MST 2011
Well, I do understand that internal pull-ups are there to prevent unconfigured pins to float, and this is standard practice in every uC I know about. But this is the first time I come across an issue like this, that an internally pulled-up pin barely reaches the VIH. This is weird, if not crazy.

The real problem here is that this kind of behavior is not documented. I read the datasheet, page 52 figure 13, and I see that with 2.3V input voltage I should have a pull-up current of 30uA. Where does that current flow if I leave the input pin unconnected? Or should I assume it flows in the >1M Ohm load of my instruments, giving me more than 30V!? That pull-up current is way much lower, let's say 2uA, which is well under the minimum pull-up current of 15uA documented in the datasheet.

I just want official documentation about internal pull-ups which is correct, so I can design properly. I can't make assumptions based on my or other people's measures. For instance, what happens to measured values at -40°C or 85°C?

[lpcware]

Content originally posted in LPCWare by Rob65 on Wed Nov 23 07:20:44 MST 2011
IMHO the internal pull ups are just there to make sure your unused input pins are not floating. You should always use external pull ups when using input pins.

This is how I look at this from a hardware engineer point of view. The internal pull ups are just too weak to be really useful in an application - the slightest glitch on a nearby track on the PCB may result in spurious signals.
I agree that this kind of limits the use of the internal pull up/down setting.

Regards,
[INDENT]Rob
[/INDENT]

[lpcware]

Content originally posted in LPCWare by Ex-Zero on Wed Nov 23 06:51:53 MST 2011
A few unofficial notes:

#1 Older datasheets shows VIH HIGH-level input voltage: 2V :rolleyes:

#2 Measuring input voltages while input is switching high/low shows more friendly values about 1.6V and 1.2V :)

#3 If you don't trust internal pullups, listen to your hardware engineer and use external pullups :eek: