Scullcom Hobby Electronics #54 - DC Voltage Calibrator Part 2

[@basildouglas855]

Mr Silvio, your comment on this project is a valid point. However this project goes further than just "follow the steps and build something". It embraces a wealth of instruction in concepts and learning. This video is certainly the realm for hobbyists. If one is a professional, one would simply purchase a professional test instrument and be done with it. But I also hear what you say and it also makes good sense. So how about putting together your alternative approach using a 16 bit DAC and high end reference. To be sure, we will certainly appreciate and respect your efforts. Kind regards.

[@alanc1085]

As always, great job !! Thank you for your contributions. Your videos always inspire me.

[@IanScottJohnston]

Great work as usual! To test for noise on the output disconnect the DMM and connect your scope across the output banana sockets directly. Also, don't use the ground lead on the scope lead, fit the small clip instead.......this should make a big difference and any noise thereafter you will know is generated from within.

[@scullcomhobbyelectronics1702]

Thanks Ian for the tip, much appreciated. Regards, Louis

[@basildouglas855]

Another great follow up to this ongoing project. Brilliant!

[@tubeDude48]

Getting better!

[@coldfinger459sub0]

Awesome videos I have been binge watching your series of videos one after another for the last three days.Thank you for sharing your knowledge and educating us who want to learn your products are incredible I can’t wait to start one I just have to finish watching the rest of your videos

[@rpbale]

Great project. Can't wait for part 3. Another project that I think would be right up your alley would be a Arduino controlled variable notch filter for calibrating a low noise low frequency oscillator (10hz-100khz) for audio work.

[@scullcomhobbyelectronics1702]

Thanks for your comment and project idea. Regards, Louis

[@brettwhittingham8217]

Well done on another great video, and a useful little project for many to build up at home... I would love to see a video of your design process, and ic selection as part of the series. You have some great engineering ideas and methods, and I think your subscribers would benefit from seeing how it all happens. Keep up the good work, and have a great Christmas.

[@scullcomhobbyelectronics1702]

Thanks Brett for your comments. Interesting idea about a video on design process. This is something I may consider for a future video. I am currently busy on a couple of other projects at the moment. Regards, Louis

[@tonybell1597]

Excellent work Louis, looking forward to part /3.... Parts on order! Have a great Christmas....

[@scullcomhobbyelectronics1702]

Thanks Tony. Hoping to get some time over the Christmas holidays to design two new double sided PCB's and get them made and tested. Wish you a merry Christmas. Regards, Louis

[@hexefx]

The reason why the display update is so slow is because the way the display object is created uses a bit-banged software SPI. Even if the assigned pins are hardware SPI. I believe creating the display object with Adafruit_ILI9341 tft = Adafruit_ILI9341(TFT_CS, TFT_DC); instead of Adafruit_ILI9341 tft = Adafruit_ILI9341(TFT_CS, TFT_DC, TFT_MOSI, TFT_CLK, TFT_RST, TFT_MISO); will use the hardware SPI. However, threre is another trap. The 10k resistors in series with SPI lines will ruin it completely at moderate speeds of few MHz. A hex buffer like HC4050 would make sure the display will get a nice clean signal. I'm working now on a project using a smaller ST7735 SPI display, using a buffer to translate the levels i can easily run the display with 8MHz SPI clock, which gives it a nice speedy response. Right now it looks like there a mixture of three SPI's: hardware one used by the DAC library, software one for the touch lib and another software one using the hardware pins made by the display library. Ideally, all three could share the same HW bus and have separate CS lines. The price of using arduino libraries... ;) Anyway, another great video, really enjoyed it! Your hand drawn schematics are pure art! :)

[@scullcomhobbyelectronics1702]

Hi, Thanks for the explanation, interesting. I will have to have a further look at the libraries. Regards, Louis

[@valentinc22]

Interesting fact about the libraries. I think the ideal scenario is to have multiple hardware SPIs, not one that is daisy chained... Also, the clock of the Arduino nano is not ideal for driving this kind of color LCD. I think a higher clock MCU (like Arduino Due at 80 Mhz or Ideadstudio STM32 at 78 Mhz) would give a much faster response, even with the same libraries...

[@stevecooper1537]

Another excellent tutorial. Happy Christmas Louis. It is a long time since I used to come to your training sessions in Bradford with DER. Keep up the good work. Ng

[@scullcomhobbyelectronics1702]

Thanks Steve. Yes its a long time since I ran the technical training school in Bradford. That was in the 1970's. I still have fond memories of those days. All the best and Merry Christmas, Louis

[@pcbreflux]

Big thanks, it is always a pleasure to learn from your experience. Thanks for sharing.

[@scullcomhobbyelectronics1702]

Thanks for you comment. Regards, Louis.

[@mrjohhhnnnyyy5797]

Charge pumps may be the cause of that 10mv noise, while the spikes are from TFT. Maybe decouple the DAC from TFT by an RC filter and do the same to an opamps. -15V rail might be noisy, and that noise can creep onto the output of an opamp. With a circuit this comlex, ground layout is crucial too. Beatiful job sir! Don't take my comment negatively, I just wanna help to make this project as good as it can be.

[@scullcomhobbyelectronics1702]

Thanks Johnny, All comments are welcome. Thanks for the info. I will be having a closer look at this in part 3. Regards, Louis

[@looxylook825]

Big thanks for all your effort.

[@scullcomhobbyelectronics1702]

Thanks.

[@silvioklaic9249]

Well, this project become more complex and definitely not for hobby users. Biggest problem that I can find in hobby building is lack of precision instruments to calibrate projects. In my book if any project needs high-end calibration it is not for hobby use. So I think that better option is with 16bit DAC like AD5761ARUZ and high quality reference voltage ADR4525BRZ. With that setup basically for same expenses you can get far better precision, no need for calibration and more simplistic to build.

[@scullcomhobbyelectronics1702]

Hi Silvio, Thanks for your comments. I did look at using a 16 bit DAC for this project and may use one at a later date in another project. I wanted to use the low cost 12 Bit Dac in this project in the first instance and also introduce the hobbyist to the analog switching IC's to illustrate their uses. Regards, Louis

[@szapcsika]

When redesigning the PCB take extra care to separate analog and digital ground an power.

[@scullcomhobbyelectronics1702]

Thanks for your comment.

[@ufohunter3688]

Very nice. Perfection as usual.

[@scullcomhobbyelectronics1702]

Thanks, much appreciated. Regards, Louis

[@brainfornothing]

Merry Christmas !

[@scullcomhobbyelectronics1702]

Thanks.

[@jameshauser1507]

Very nice Louis. Very nice. Have a Merry Christmas and a Happy New Year.

[@scullcomhobbyelectronics1702]

Thanks James.

[@Flip-Flop-Rio]

Great project, my friend. Everything always done very carefully. Congratulations! Merry Christmas Michel - Rio de Janeiro - Brazil

[@scullcomhobbyelectronics1702]

Thanks Michel for you comment. Regards, Louis

[@wolfgangschweiger9557]

As always a perfect work - my next project after the load ;-) I wish you marry christmas and a happy new year. Greeting from Southern Germany, Wolfgang

[@scullcomhobbyelectronics1702]

Thanks Wolfgang. I am in the process of redesigning the DC Voltage Calibrator PCB's to double sided print and will have some boards made in China. Once I have tested them I will make some of these PCB set (2 boards) available for those who are interested. Merry Christmas, Louis

[@wolfgangschweiger9557]

Thats nice! I've made also some PCBs in China for the "Load-Project" and the result was excellent. Nice Weekend, Wolfgang

[@michaelnobibux2886]

Perhaps you could use the spare DAC to generate a correction voltage and sum that to the output. A correction table would fit in the nano's EEPROM, especially if not all 4096 steps were stored,but ,say ,only every 10th step. This would require some interpolation done in software for most steps,but this is trivial. Alternatively a separate (large enough) EEPROM chip can be used ,so that the complete correction table could fit. Great project,! Thank you so much!!!

[@birnodin]

That was a great work! I like your videos, and I like how you talk about the details!

[@scullcomhobbyelectronics1702]

Thanks for your comment. Regards, Louis

[@timothysvec8568]

Well done Sir! Well done!

[@scullcomhobbyelectronics1702]

Thanks Timothy.

[@exosdel]

Great video.. Really enjoy your projects! 👍

[@scullcomhobbyelectronics1702]

Thanks.

[@tommyallehamn]

Awesome 👍 Merry Christmas Louis 🎄

[@scullcomhobbyelectronics1702]

Thanks Tommy.

[@WaldemarHeimann]

You have very good content! Thanks for doing this!

[@scullcomhobbyelectronics1702]

Thanks for your comment. Regards, Louis

[@esnam6557]

This is a great improvement for less than 1V. setting accuracy and assembling everything, thank you very much. Looking forward for the 3rd part for reducing the noise task. I am wondering if you can also consider a project for making a "cap leakage tester" please (e.g. a solid state version of Heatkit IT-11 ). This is highly demanded and very useful test gear to have.

[@scullcomhobbyelectronics1702]

Thanks for your comment. With regards a capacitor leakage tester, these usually require high voltages of up to 500V or more and so could be dangerous for the hobbyist who normally work at low voltages. So I am not sure about such a project. Regards, Louis

[@esnam6557]

Many thanks for your response, I fully understand your concerns about the high voltage dangers in a KZbin project. However, there are many electronics hobbyists who restoring tube radios in KZbin and dealing with high voltages. I am planning to make a leak tester and have seen a couple of KZbin videos but prefer your approach in projects so wondering if I can ask you technical questions during the design and build process. Wishing you Merry Christmas and Happy New Year.

[@scullcomhobbyelectronics1702]

I wish you well with your project. Feel free to ask any technical questions, I will help if I can. Regards, Louis

[@isidoromaich7226]

As always great job, and thank you

[@scullcomhobbyelectronics1702]

Your welcome. thanks. Regards, Louis

[@johncoops6897]

5:44 - how can that resistor array give +/- 1 mV of trimming? It can decrease the output by attenuation (voltage divider), but I cannot see how it can possibly increase the voltage. What am I missing?

[@scullcomhobbyelectronics1702]

Hi John, I have found that the tolerances on these reference IC's are such that most are slightly high of the 1.024V. So the trim circuit brings it down slightly. Regards, Louis

[@eddierodriguez6644]

Great job as always love your projects , and yes please make available the new PCB for purchase thanks again for your time much appreciated and keep up the videos.

[@scullcomhobbyelectronics1702]

Thanks Eddie, I will be getting a few sample PCB's made by a company in China over the next week or so and will test them. If all goes well I hope to have a few ready for Part 3 of this project. Regards, Louis

[@thecombatengineer7069]

Louis, are you using EasyEda? I noticed someone drafted ‘Pt.1’ plans on the site but after I commented about the TFT package, they deleted the thread. I also found that we CAN upload gerber files (KICAD, EAGLE, etc) directly to their board maker and will be uploading that video shortly.

[@scullcomhobbyelectronics1702]

Hi Combat, I am using KiCad. I have now redesigned the PCB's and have created some gerber files. I will be using PCBWAY in China to make some prototype boards to test. If all goes to plan I will try and make some available. Regards, Louis

[@OneBiOzZ]

i was going to comment about the cables and noise before you ran the scope test, i would shield, add a ferrite and an output cap at the terminals , along with a ferrite over the arduino replacement cable and maybe some caps on the switch input of the switches but that might be pushing it a bit

[@scullcomhobbyelectronics1702]

Thanks Alyx for your comments. We will have a further look in Part 3. Regards, Louis

[@roelandriemens]

Nice project again. I have only one question. Is the wiring arangement at 8.39 correct? It looks like 100k and 1k resistors exchanged places.

[@ceptimus]

I was about to post that - but you beat me to it! :)

[@scullcomhobbyelectronics1702]

Hi Roeland, On the layout I made a silly mistake on the whiteboard and did not notice it, sorry about that. The 100K resistor goes to the slider of the trimmer and the 1K resistor goes to pin 6 of the IC and the top of the trimmer. I have corrected the layout download on the link I gave below my video so that is now correct. Below also the link to the download zip file with the corrected layout and schematic: www.scullcom.com/DC_Cal_1_024V_option.zip Regards, Louis

[@irawarnaca8133]

Great project, thank you! I have a question. On the schematic at 7:11 you show the output of the trimmer going to the 100K resistor, but just after that on the whiteboard at 7:40 you show the output going to the 1K resistor. Which way did you end up doing it?

[@scullcomhobbyelectronics1702]

Hi Ira, Thanks for pointing this out. The schematic is correct. On the layout I made a silly mistake on the whiteboard and did not notice it, sorry about that. The 100K resistor goes to the slider of the trimmer and the 1K resistor goes to pin 6 of the IC and the top of the trimmer. I have corrected the layout download on the link I gave below my video so that is now correct. Below is also the link to the download zip file with the corrected layout and schematic: www.scullcom.com/DC_Cal_1_024V_option.zip Regards, Louis

[@STRATOS13PAO]

i think you probably already know this but if you hold the screen back light button(hi res) on the fluke for 1 second you can get higher resolution, it would be useful for all those voltage tests. get video and project btw, thank you for the content.

[@scullcomhobbyelectronics1702]

Thanks Stratos. Yes I did know but simply forgot to press the back light button long enough. Hopefully, the Keithley meter readings helped. Thanks for bring it to my attention. Regards, Louis

[@fluiddynamicist]

Well done! Why don’t you use higher accuracy (ref V 1.024V) up to 4V?Your op amps are able to multiply by 4!

[@pirateman1966]

Use of trim-pots with their high temperature coefficients ( ±100 ppm/°C) are not needed. Use of calibration values in the EEPROM is the way to go. The initial accuracy is not that important and can be compensated in software. The overall stability over time and temperature, is more important. The display is too slow as is. Using it in parallel mode, instead of I2C or SPI, would've dramatically improve its speed. XMEGA family with its 2x12bit DAC (not ATMEGA), and plenty of I/O lines (12+ for display), would've been a better choice for the MCU and eliminated the need for an external DAC. Nice project and effort on your part though.

[@scullcomhobbyelectronics1702]

Thanks for your comments and suggestions. I will be having a further look at calibration in Part 3 so there may be some changes to come. Regards, Louis

[@pirateman1966]

I'll look forward to it.

[@qsptecnologia]

Thanks again. How can I calculate the N digits in a specification of 0,005% + N digits? In data sheet of multimeters always have this notations.

[@scullcomhobbyelectronics1702]

Hi, The N digits specified in multimeter specification with regards its ranges accuracy is usually calculated during the testing of mutimeter after manufacture, when batch testing is undertaken. The N digits is usually referred to on datasheets as number of counts ±. This error factor is usually caused by ADC errors, noise and offsets errors, which will vary from meter to meter due to component tolerances and manufacturing processes. For example if we measure a voltage of say 2.000V on a particular range of the a multimeter where it quotes a accuracy for that range of ±(0.05% + 3). This means the actual reading could show a difference of ±0.05% of 2.000V meaning it could read between 1.999V and 2.001V, in addition to that we have other factors mentioned above and they will add additional counts (or digits) to the reading. In the case illustrated here it was quoted as 3 counts, so the actual reading could range between 1.999V - 3 least significant digits and 2.001V + 3 least significant digits. So in this case the reading could be in the range of 1.996V to 2.004V. Hope this helps. Regards, Louis

[@qsptecnologia]

Thanks so much. I appreciate your channel. It's very good to learn. I am from Brazil and I have a channel too, but in portuguese. If you want to know, just search qsptecnologia on youtube. Thanks a lot.

[@scullcomhobbyelectronics1702]

Thanks for the info. I will have a look at your channel. Regards, Louis

[@CliveChamberlain946]

On ebay dot com, I didn't find anything with that search string. Using "Plastic Electronics Instrument Case w/ handle" gave me good results though. I sent you a PM regarding EEVblog video sharing opportunity during January - would you consider, if you still have enough video footage on hand, to feature a wrap-up (part #8 summary) on the DC Load?

[@jlegen]

Cliff Matthews i got this case from Banggood as ‚plastic instrument case‘; it‘s only about 9 bucks...

[@scullcomhobbyelectronics1702]

Hi Cliff, Just seen your private message and have answered it. Regards, Louis

[@First2ner]

on ebay I find one. I searched "7.8 x6.9 x2.7 plastic" and it finds one for 15 and half $

[@CliveChamberlain946]

Thanks! What I ordered was 200x175x70mm @ $11.85 CAD ( www.banggood.com/Electronic-Plastic-Shell-Cartridge-Handle-Project-Case-Desk-Instrument-200x175x70mm-p-1035473.html )

[@eddierodriguez6644]

here is a link to the case...www.banggood.com/Electronic-Plastic-Shell-Cartridge-Handle-Project-Case-Desk-Instrument-200x175x70mm-p-1035473.html?rmmds=search&cur_warehouse=CN

[@kennethnilsson6976]

Nice project. Perhaps use 1x oscilloscope probes instead of 10x as the screen of your scope shows. That would probably increase the noise measurement accuracy though the increased capacitance could perhaps influence the measurement as well.

[@scullcomhobbyelectronics1702]

Thanks Kenneth. Regards, Louis

[@jonka1]

As far as noise is concerned your Fluke and its leads are a large pick up source of ambient noise. Try disconnecting it.

[@Gooberslot]

Is there any way to calibrate this without the need for high-end test gear? Also, what happened to your website?

[@scullcomhobbyelectronics1702]

Hi, I will be looking at calibration in more details in the next part of this project. With regards my website, I had to take it down temporary due to the fact it had been attacked by a hacker and some of the data had been corrupted. Hopefully, I will find time in January to sort it out. Regards, Louis

[@bodzio_stawski187]

What is the difference between DC voltage calibrator and... the power supply?:) Why is it better to use calibrator instead of lab power supply?

[@ydonl]

A voltage reference is a device or circuit that's intended to produce an exact, precise, stable voltage that can be used for comparison with other voltages. It can't produce much current, so there is very little power available. In fact, if the load current changes, the output voltage will probably change to some degree, and that would defeat the purpose of the device. A calibrator uses a reference to produce other voltages. The usual technique is to divide the calibrator's output voltage, and compare it to an accurate voltage reference, and that will determine wither the output voltage is too high or too low, using the accuracy of the fixed voltage reference. A power supply is intended to not only produce a specific voltage, but also with enough current that the resulting power can be used in a useful circuit of some kind. Most power supplies are less accurate than most calibrators. However, it's possible to build a poor calibrator, or a very precise power supply, where the voltage in the power supply might be more accurate than the calibrator.

[@TheFunkman]

my display works but the touch screen needs calibration. I have to push in blank areas to select values and the set button is not working at all. Edit: I inverted the display (option 3 in line 69) and that problem is solved. My touch screen is designed upside down from the one in your project.

[@thecombatengineer7069]

Since all the links are to your website, none are working. Did anyone download the files before his site got attacked and can put it on GitHub?

[@scullcomhobbyelectronics1702]

Hi, All the download links I give under my videos for the projects are on my www.scullcom.com website which is not affected and works OK. I also use this website to display my youngest sons comics and artwork he draws and writes. So if you go to the home page that is what you will see. I simply use some free web space on that site for my project downloads. The website that was attacked was my other website www.scullcom.uk. My eldest son is in the process of cleaning that website and hopefully will recover all the web pages. All been well it should be back up and running in the New Year. Regards, Louis

[@First2ner]

why didn't you use meter in 4.5 digit mode, you would get one more digit, wouldn't you?

[@scullcomhobbyelectronics1702]

Yes agree I should have set the Fluke 87V to its Hi Res Mode, which would have given the extra digits. I simply forgot to do it at the time, sorry. Having said that I did use the 6.5 digit Keithley also to show the output voltages, so the results were visible. Regards, Louis

[@fexmate]

I was thinkig about the same, he just had to hold the backlight button for 1 sec, and it would be better some cases I think.

[@qwaqwa1960]

Re LCD speed, there seem to be a number of "optimized" libraries out there if you go looking. Here's one relevant video: kzbin.info/www/bejne/aF6Vep2Qf6iBgsU

[@scullcomhobbyelectronics1702]

Thanks.

[@qwaqwa1960]

Also a number mentioned on the Teensy site: www.pjrc.com/store/display_ili9341_touch.html

[@SeanBZA]

Would suggest some screening in the case, along with some screening on the output lines, and perhaps a screen sheet between display and the analogue side. Should get a lot of the noise down with the output, and probably a 100n film ( not ceramic) cap across the output terminals.

[@scullcomhobbyelectronics1702]

Thanks Sean for your comment. I will be having a further look at this in Part 3. Regards, Louis

[@MrNedelcuBogdan]

Did you see my message?

[@scullcomhobbyelectronics1702]

Yes. Been busy over Christmas. Regards, Louis

[@MrNedelcuBogdan]

Scullcom Hobby Electronics Do you think the regulated high voltage power supply it is a great project?

[@MrNedelcuBogdan]

Do you think there might be someone interested?

[@scullcomhobbyelectronics1702]

I am sure there will be someone out there with an interest. But i do not think it is for the hobbyist. Regards, Louis