#apertus IRC Channel Logs

morning se6astian!

hello!

wb anuditverma!

welcome Tanos! welcome Christofpap!

correct, the former is me and the latter is se6astian :)

thank you!

and maybe let's hear how you interpret those tasks and how you would go about solving them?

well, the microcontroller approach won't work on the FPGA

so basically it consists of a PWM controller attached via e.g. I2C or SPI and remote accessed via the main FPGA (ZYNQ)

how about T728?

correct, at different levels

Hi Bertl! I am interested for T726, T722, T734, T720. For task T726 I don’t have the required skills for the FPGA implementation, is there any chance getting it? I am asking because it is labeled as  optional .

T720 and T726 you want to talk with se6astian or alexML

Tanos: T722 you probably best chat with BAndiT1983|away when he's around

Christofpap: do you have experience with FPGAs and HDL coding? maybe any example code you wrote?

Yes, my thesis was with FPGA and VHDL coding

Christofpap: any links/URLs?

Great thank you. I have no experience with FPGAs and I don’t know if it is possible to acquire all the skills needed to complete the task. On the other hand FPGAs is within my field of interest and I plan on investing time learning the craft.

Christofpap: did you code both the RX.VHD as well as the TX.VHD?

and in the RX.VHD the inverse way

Christofpap: and why do they look so different in capitalization and indentation?

also note, that even if GSoC does not happen for motivated folks, contributions are still welcome and we will figure out a way to compensate those

Sorry, but i don't understand you

now looking at page 52 and 53 I see all lower case and indentation for port signals

so for me it looks more like it came from two different sources

Ok, i understand it was i way to understand better the code with my professor. Because i would like to do two times the same way. It was a convension with my professor

If i could help on this subject

as I said, all help is welcome, so go ahead and start contributing

you're welcome!

Hello!

Mr. Bertl?

welcome Xumerio

I we had a discussion on T732 five days ago

till that moment, temperature sensors were the only input in the algorithm

Furtheremore, as you said, the problem with fan is that it will produce noise sounds that will affect the recording

what do you think?

Ok. The noise produced by the fan is almost constant and easy to distinguish. Taking some recordings and training an algorithm can result in impresssive results

and the FPGA is very good for digital filters. Especially high order filters, because the adders work in parallel

but is that equipment close to the camera?

but it is a good idea to collect information about the fan (and record it somewhere) to compensate in post

anyway, further processing can be done

And about the Fan Application.

All users proposed to use a PID control. Which is quite easy to implement. I suppose the input depends on the sensors values, and their distance from the fan. But I am thinking at a better idea. A machine learning approach. Trained with sensors values, distances, and the appropriate fan speed (PWM vlaue)

it can provide better results than the PWM, as it can model the nonliniarities better

it deserves to be taken into account

one aspect I think is very important is that there need to be different 'modes' with different limits and target temperatures

e.g. you typically want a 'silent' operation during a shot, which doesn't really take that long (normally a few minutes)

and you want to keep a good but not annoying 'cool' during longer idle periods

in all cases we want to keep 'acceptable' temperatures in all places and warn if we go over

And all the measurements should be done in warm environments, too. It might work at 17 degrees celsius, but it won't work for 25

yup

Ok

Hi

Where can I find GSOC ideas for this organisation?

here: https://lab.apertus.org/project/view/20/

welcome rock!

off for now ... bbl

good night

nn