Problem with R-2R Ladder DAC

Dominik Przyborowski say:

Use smaller resistors and external low resistance switches driving from uC.

Click to expand...

thank,yes of course ,i have a big problem in high freq>50Khz,please help me to reduced this problem or show me how replace this circuit with other circuit.please recommend new circuit schematic for build dac circuit.
how i can calculate r2r ladder network resistor value?

Can you explain what your code is doing ? GPIO -> ODR=x ?? And how do you set the update rate in your loop for specific frequencies ?

In addition, what is your oscilloscope specs ? I cannot make out which make/ model it is.

Click to expand...

thanks
output port value increased step by step in this loop to build a ramp signal,r2r ladder convert this fragment value to analog signal,ODR is register to write GPIO port value,frequency can be increased by increase jump step in base loop.
similar technique like NCO oscillators.my osc is frequency BW is 20MHz,Micro AZ.

why after ramp signal i have a similar discharging signal??

Click to expand...

You are seeing the limited slew rate (typically 0.5 V/µs) of LM324. Use a faster OP like TL074.

FvM said:

You are seeing the limited slew rate (typically 0.5 V/µs) of LM324. Use a faster OP like TL074.

Click to expand...

i replace LM324 with TL084 but problem not solved...
TL084 Slew rate is 13V/uS but i have a discharging signal end of ramp signal...
signal before op-amp at 100Khz

op-amp signal output:

please help me

Assuming you are cycling through only 256 values (8 bits), then your update rate at 100Khz of the sawtooth implies you are outputting a 'new' data @ 100K x 256 = 25.6Mhz !! If you are cycling 10 bits, then it is even higher at >100Mhz. This is higher than your 'scopes BW and maybe causing some trouble.

Is this so ? Or are my calculations wrong ? Is your uC and code doing this ?

In addition, at this frequency, even a capacitance of 2pF will have an impedance of only 3Kohm, which is less than your r-2r ladder.

On an average breadboard implementation, chances are you have even higher parasitic capacitances.

Lastly, at the places where the ramp makes 'sharp' transitions you will see problems. These again are due to parasitic capacitances, and ALSO will be seen if your oscilloscope PROBE is not properly compensated ! What is your probes specs, and have you compensated it ?

Maybe it is time for you to post a pic of your 'real' circuit board as implemented currently ?

The "before op amp" signal looks like you are loading the node with a big capacitance, e.g. 1:1 osciloscope probe. The OP output signal isn't plausible for a high speed OP and the circuit according to post #13. What's the actual slew rate achieved in this case? It can't be seen from the photos.

kripacharya said:

Assuming you are cycling through only 256 values (8 bits), then your update rate at 100Khz of the sawtooth implies you are outputting a 'new' data @ 100K x 256 = 25.6Mhz !! If you are cycling 10 bits, then it is even higher at >100Mhz. This is higher than your 'scopes BW and maybe causing some trouble.

Is this so ? Or are my calculations wrong ? Is your uC and code doing this ?

Click to expand...

hi
thanks,your calculation have a bit wrong items,
i build 100Khz freq with 52 samples of ramp signal.
in main loop jump steps defined output freq.
in 100Khz with 52sample update rate equal ~2Mhz.
number of bit only define resolution dac.

i don't know about osc probes property specs.
probes just have a trimmer cap in circuit for tuning signal edge at higher freq.

i try to make circuit on the pcb to decries parasitic capacitance.
but i have big problem with op-amp because output freq on op-amp spend big time on discharging signal and lost sharp rarmp signal.
i use a op-amp with higher slew rate but problem not solved...

- - - Updated - - -

FvM said:

The "before op amp" signal looks like you are loading the node with a big capacitance, e.g. 1:1 osciloscope probe. The OP output signal isn't plausible for a high speed OP and the circuit according to post #13. What's the actual slew rate achieved in this case? It can't be seen from the photos.

Click to expand...

thanks
yes of course,dac output have a big rising time and falling time,after op-amp i have a sharped rising time and big falling time!!!
a big capacitance in my board.
i build pcb of circuit and try it again.
real circuit is matched by schematic.
in this picture actual slew rate is showed:

osc Time/Dive Set on 2us.
thanks for attention

Persian_Men said:

in 100Khz with 52sample update rate equal ~2Mhz.

Click to expand...

Sounds as a quite daring specification of design for assemble on a wired breadboard employing relatively standard components. Would be recommended start thinking about draw a compact PCB layout for that.

Furthermore, protoboards have inner tracks along its entire lengh which can act as real antennas, therefore becoming not a suitable environment for such tests.



+++