ADC0804 clock frequency

Hello Pablo, Your math is right and if the part is configured correctly than as you said you should expect around a 600kHz clock. I don't have any experience with this behaviour. Have seen this behavoiur on multiple parts? If not can you try a couple other parts and see if they act the same way?

Also curious when you using the scope probes you measure 360kHz but earlier in your post you measure 380kHz. How do you measure the 380kHz? Are you basing this on the digital output of the A/D?

 Were exactly are you putting the probe when you measure 360kHz?

Hi Jason,

Thanks for your answer. Yes, I have actually tried another part and got similar results. Regarding the measurements, I am measuring the clock frequency at pin 19 CLKR. At that point I get some 382 KHz. When I place the probes in the CLK IN input, that's when the frequency goes down to some 360 KHz but I assume this is due to the capacitance of the scope that adds to the Clock capacitor.

I have also observed that the threshold voltages of the internal Schmitt trigger are almost on the limit (or some times even more) of the values given in the datasheet; for example, for VT+ I measure values between 3.45 and 3.54V and for VT- I got values from 1.5 to 1.6V. If I calculate the resulting frequency taking into account these threshold voltages, I obtain something around 420 KHz...

Below are the schematics for the ADC part of my design (if you can see anything...). I feed the analog signal through a buffer and the output is fed to a D flip flop. Since I need to interface with low speed circuitry, use a divide by N counter to divide the clock from the ADC and clock the Flip Flop from here. Also you will see a simple circuit generating an initialization pulse of a few microseconds, as required for this converter to start up in free running mode.

As I mentioned before, the circuit converts just fine; no other issues than this.

Thanks.

Regards,

Pablo

Pablo, Thanks for the extra detail and schematic. I will need to find board/samples to look into this further. Please repost if you don't hear from me by early next week.

Hi Jason,

Just wondering if you found some time to dig a bit further into this issue.

Thanks in advance,

Pablo

Pablo,

I've obtained some units and started testing today. Preliminary clock frequency results match closer to your findings then the datasheet 1/1.1RC. Investigating cause of discrepency. Hope to update tomorrow.

Hi Jason,

Thanks, much appreciated for that. Yes, please let me know how you get on.

Regards.

Hello Jason,

Just wondering if you had the chance to do any further test on these ADCs?
Thanks.

Regards,

Pablo

Pablo, Apologize for the delay still working internally to determine cause of discrepency. But yes I've taken a look at a couple of ADC0804 devices and have confirmed a similar discrepency to the  flck = 1.1/RC described in Figure 8 of the datasheet.

Like I said we are investigating internally the root cause of the discrepency. Due to the holidays please expect some delays. I expect to be able to provide an update the 1st week of 2013.

Hi Jason,

Thanks for the answer. In the mean time I also sourced a couple of ADC0804 aswell but havent yet been able to test them. In any case, I have already tried two and got the same results and after your message I wouldnt expect any change. I think the reason for this discrepancy is due to the threshold voltages of the Schmitt trigger in the clock circuitry of the ADC. From the datasheet:

VT+ =  2.7V (min) / 3.1V (typ) / 3.5V (max)
VT- = 1.5V (min) / 1.8V (typ) / 2.1V (max)

Now see attached scope capture. In my circuit the positive threshold voltage is 3.48V while the negative is 1.6V. If this values are used to predict theoretically the clock frequency, a value of approx. 420 KHz is obtained, which is much close to what I am actually getting. So in my opinion, the threshold voltages of the Schmitt trigger that are at the edge of spec are the root cause of this discrepancy.

Hope you will find this information useful.

Regards,

Pablo

Pablo,

Apologize again for the delay. I needed some time to validate your hunch and it appears that you are correct. Indeed the Schmitt trigger threshold voltages are at the margin of the datasheet limits. But what I confirmed is the Schmitt trigger oscillation frequency equation (1/1.1RC) in the ADC0804 datasheet is incorrect. The actual nominal equation is more complex and depends on VCC and the threshold voltages Vt+, Vt-. For equation see App Note SNOSBC7, section 1.1 Clock Inputs (link below).  It appears that you have found this out already as well. When using that equation the frequency that I see (and that you see) fall within the nominal frequency given component tolerances and the nominal nature of this equation.

So to summarize initially you expected to see about 600kHz based on the component values and I'm assuming the 1/1.1RC equation in the datasheet. But we have found that that equation is incorrect. If we use the correct equation what is expected and what is actually seen align. Thanks for working through this with us. I will put in a request to update the datasheet with the correct equation.

http://www.ti.com/lit/ds/snosbc7/snosbc7.pdf

HI All,

Could somebody help me? Where can I find the Spice model for the ADC0804 converter? http://www.ti.com/product/adc0804-n

I can't find it in the official model list: http://www.ti.com/adc/docs/midlevel.tsp?contentId=55102

Thanks in Advance! Anthony

Antal,

Unfortunately there is no Spice model available for the ADC0804. What aspects of the device are you trying to model? What parameters were you most interested in? Perhaps we can help answer any device questions you may have.