FEDEVEL
Platform forum

QDrives , 07-29-2024, 09:08 PM

What is "mess xcel"?

Sniper2 , 07-29-2024, 09:08 PM

that 0.9-1V is still there for some reason , if it was the FB voltage then yea i vould understand somehow but it is that + something

Sniper2 , 07-29-2024, 09:08 PM

mesured - expected , or rather error

Sniper2 , 07-29-2024, 09:09 PM

the offset i wanted to also plot it since i think it can help

QDrives , 07-29-2024, 09:10 PM

So what do you measure with DAC output voltage 0V, 1V and 2.5V?

QDrives , 07-29-2024, 09:11 PM

My calculations still seem different compared to yours.

Sniper2 , 07-29-2024, 09:11 PM

maybe it is FB_v + what is droped over the DAC resistor? then again it is likesuper low current

Sniper2 , 07-29-2024, 09:13 PM

0V expected 4.56V measured 5.46
1V 3.59 measured 4.502V
2.5 2.148V measured 3.132V

Sniper2 , 07-29-2024, 09:14 PM

uploading to git xcel that i have atm

Sniper2 , 07-29-2024, 09:14 PM

here too in case u dont want to use git

QDrives , 07-29-2024, 09:15 PM

My calculations:
0V --> 5.36V
1V --> 4.36V
2.5V --> 2.86V

Sniper2 , 07-29-2024, 09:17 PM

hmm much closer but at the end it still drifts a lot

Sniper2 , 07-29-2024, 09:18 PM

ill restudi your math and see , the next evening

Sniper2 , 07-29-2024, 09:18 PM

good night*

QDrives , 07-29-2024, 09:18 PM

Could be average compared to peak of the ripple...
Good night.

Sniper2 , 07-30-2024, 02:00 PM

I can also average out like 10 100 samples

Sniper2 , 07-30-2024, 02:01 PM

Should be able to from the chip itself but I'll do it in code

QDrives , 07-30-2024, 08:20 PM

What about tolerances?

Sniper2 , 07-30-2024, 08:55 PM

standard JLC so 1% if i remember well

Sniper2 , 07-30-2024, 08:56 PM

i measured them and thy all seem to be very close to mnominal and all are a bit lower resistance so ratios are not massively changed

Sniper2 , 07-30-2024, 08:56 PM

PS did not have time to test new code or eamine your xcel due to IRL problems

Sniper2 , 07-31-2024, 07:54 PM

didnt make it in time ill return to this project in like 10-11days from now

Sniper2 , 08-02-2024, 07:29 PM

Hmm so you computed current and based on that vout

QDrives , 08-02-2024, 09:34 PM

Yes.

Sniper2 , 08-10-2024, 05:52 PM

ok i am back and redoing the math + some corrections for reference and leakage ....

Sniper2 , 08-12-2024, 07:45 PM

now it seems to work way better in terms of math

Sniper2 , 08-12-2024, 07:50 PM

at this point i think i am into tolerances and i have no idea what is the FB input current well it is pA or less BUT my current is like 0.8uA in that node

Sniper2 , 08-12-2024, 07:50 PM

then again i wander why my initial math screwed up ... maybe forgot a current direction and fb voltage at some point

QDrives , 08-12-2024, 09:17 PM

You might want to do a tolerance calculation.
If you are still outside the tolerance, then we are missing something else.

Sniper2 , 08-14-2024, 06:17 AM

It seems out of my rought math that I am in them more or less

Sniper2 , 08-14-2024, 06:18 AM

Fb input current in not documented and I did not account for fb voltage tolerance so I am sort of sure in tolerances

Sniper2 , 08-14-2024, 06:31 PM

so:
deviation from expectd is from 1.8% to 7% + 2 odd cases of 14% but that is proably typo on my part since there was no change

Sniper2 , 08-14-2024, 06:34 PM

my resistors are 1%+- AND Vfb is itself 1% +- and on top of that there is that unknown FB current needed by the IC itself

Sniper2 , 08-14-2024, 06:35 PM

and i forgot to mention that i bodged a THT for the DAC resistor and that is 5% or worse....

Sniper2 , 08-14-2024, 06:37 PM

5%

Sniper2 , 08-14-2024, 06:38 PM

ill correct the c++ code and then add code to compute on the fly deviation and % deviation

QDrives , 08-14-2024, 07:54 PM

And your DAC output is correct too?

Sniper2 , 08-14-2024, 08:06 PM

like measured?

Sniper2 , 08-14-2024, 08:07 PM

i am makeing a function for averaging readings

QDrives , 08-15-2024, 07:39 PM

Yes, did you also measure the DAC output.

Sniper2 , 08-16-2024, 04:13 AM

No, that needs to be done by hand .
The % error code is running but buggy atm since vout is wrong typo somewhere

Sniper2 , 08-16-2024, 07:09 PM

ok so new data is is basically this:
at the scale extreme error is large like 10% while in the middle it is like 1%

Sniper2 , 08-16-2024, 08:02 PM

corected code

Sniper2 , 08-16-2024, 08:03 PM

now it is way more precise and typeinf new data

Sniper2 , 08-16-2024, 08:03 PM

BUT now it is like a 0.1-0.2 V offset

Sniper2 , 08-16-2024, 08:30 PM

-6% max error

Sniper2 , 08-16-2024, 08:30 PM

error seems to grow liniarly so i think it is current related from seomething

Sniper2 , 08-16-2024, 08:31 PM

BUT good enough for now ....
ill akso need to math out the reverse fucntion + code to rectly read current

Sniper2 , 08-16-2024, 08:31 PM

@QDrives

QDrives , 08-16-2024, 11:28 PM

I would redo multiple measurements at random just to see if it is a measurement / documentation error.