Minimal hardware set

[amirm]

I'm trying to learn ARM, and built a minimum hardware set.

I have 2 TQFP-64 to DIP adapter boards, with LPC2138 and LPC2148. Each board has LMxxxx regulator with stable 3v3 and 8 caps: 4 * 100nF and 4 * 1uF tantalum. Power connections are ok, I'm using 11.0592MHz crystal, set baud rate of FM to 9600, 4800.

For download I'm using a MAX232 powered at 5v - it accepts 3v3 input, with following connections:

DB9 pin2 (RX)  <--- MAX232 <--- LPC pin19 (P0.0/TX)

DB9 pin3 (TX)  ---> MAX232 ---> LPC pin21 (P0.1/RX, 5v tolerant)

DB9 pin4 (DTR)  ---> MAX232 ---> Multiturn Pot ---> LPC pin57 (reset)
the Multiturn Pot is 5k and set to output 3v3 when fed with 5v.

P0.14 is jumped to gorund.

I managed to read device signature with this setup a few times, but only after 4-5 attempts.
Now it is not working at all.
tryied both com ports of pc.they seem to be ok, tested with loopback.

looking RESET level with meter, it is 3.282V and pulses to 0.080V.

The MAX232 really works with 3v3 signals, I made a loopback with TX passing through Multiturn Pot to lower to 3v3 and sent back. ok.

Am I missing something?
Can P0.14 be left grounded all time, or do I have to control it with RTS?

Thanks in advance.

[amirm]

Forgot to mention: I tryied Hyperterminal at 9600, checked RESET was high, sent "?", got no response.
Even when I hold "?" down and pulse reset, there is no response.
crystal is oscilating - seen with osciloscope.

[Andy Ayre]

What do you see on the microcontroller Rx and Tx pins with your scope?
Have you checked all voltages at the microcontroller pins to make sure they are within the spec?
Have you compared your circuit to the schematic diagram of the Keil MCB2100 board, which is available for free at http://www.keil.com/mcb2100?

Andy

[amirm]

The strange thing is that after reset with P0.14 grounded, P0.0/TX stays low. I tried to pull it up to 3v3 with a 1k resistor, but got 1.1v, so the port must be really pulling down. Shouldn´t it be at 3v3 and only come down when sending data?

3v3 power source is ok.

I followed the Keil MCB2140 schematics, but did not copy it exactly. I did the circuit described in my first mail, but there are some modifications, like pull-up / pull-down resistors on JTAG port and an LM311 comparator to convert P0.0/TX from 3v3 to 5v and make MAX232 happier.
No change in results so far...

MAX232 level conversion is fine - instead of connecting RX and TX to LPC, I made a loopback:
MAX232 converts RS232 levels to 5V (RX)
Multiturn pot converts 5v to 3v3 (RX)
RX connected to TX
LM311 converts 3v3 back to 5V (TX)
MAX232 converts 5V to RS232 levels (TX)
I type on hyperterminal, get echo. Saw wavefroms on scope, RX signal was taking 200nS to go from high to low - it would take more than 2mbits/sec to make a differece.

Now that the company I work has bought a 4 trace Tektronics scope with USB, I´ll be able to look at RX, TX, RESET and P0.14.

Thanks for the reply, I´ll post results and circuit used this week, hopefully.