Author Topic: DIY SCPI programmable dual channel bench PSU 0-50V/3A (now EEZ H24005)  (Read 285660 times)

0 Members and 2 Guests are viewing this topic.

Offline rikkitikkitavi

  • Contributor
  • Posts: 22
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #125 on: June 20, 2015, 08:32:23 am »
OK, then I'll revise my thinking on painting heatsinks.   

By the way, that's a great link to the PDF of the guys Master Thesis on the analysis of the heatsink, thanks for that.  It's great to see a detailed analysis of heatsinks, and I'll read it to update myself.  My thinking on the matter seems like it might be out of date.
Compare emissivity of natural aluminium with the highlited areas.
My comment was based on engineering , i e thermal conduction thrpugh a thin paint layer (do the math) , convective heat transfer (similar) and radiative (very different) between natural and painted aluminium.

Doing the numbers you see that there is some difference for natural convection, but with forced convection radiation is neglible due to low face surface to colder surfaces and convective heat transfee increases rapidly with air velocity due to the stagnant film layer thins.
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Mosfet pre-regulator PCB r2B5
« Reply #126 on: June 20, 2015, 06:38:04 pm »
Mosfet pre-regulator PCB is more or less completed. It now include the following sections:
  • Power mosfet pre-regulator
  • Bias SMPS pre-regulator with LDO post-regulation
  • Sync isolation
  • V/F converter for temperature sensor (NTC)
Items 2, 3 and 4 is reusable for other pre-regulators (like SMPS buck or buck-boost). Last two items are optional: sync isolation can be used to synchronize switching frequency of all SMPS controllers. In dual channel version with SMPS pre-regulators their number can be up to 5 (2 high power, 3 small power).
V/F converter is used since there is enough room to place it on PCB and couple of unused pins were available on I/O expander (MCP23S08) located on the post-regulator PCB. It is not precise (~100Hz/oC in 25-70oC range) like dedicated temperature sensor devices but since it should be used for measuring temperature on main heatsink it is more then enough for monitoring if alarming temperature is reached or not.
In this version is fixed Power good signal issue what is important for proper functioning of post-regulator during power up and power down phase.

Schematic for r2B5 version is in the attachment. In new post that follow I'll attach Eagle files, Gerbers and BOM.
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Mosfet pre-regulator PCB r2B5 layout, Gerbers and BOM...
« Reply #127 on: June 20, 2015, 06:41:31 pm »
Here is all files required for creating above mentioned mosfet pre-regulator. PCB is maybe a strange or unnecessary mix of THT and SMD components. If this project comes to group buy phase I can redesign everything to use SMD to the maximum extent.

Don't know if custom library is also required for viewing Eagle files.
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Assembled Post-regulator PCB...
« Reply #128 on: June 25, 2015, 05:55:38 pm »
A small update about post-regulator. It's now assembled as it is shown on following pictures. Initial testing is made with new current shunt monitor section with 40V/V gain and OP27 op amp.



 

Offline timofonic

  • Frequent Contributor
  • **
  • Posts: 904
  • Country: es
  • Eternal Wannabe Geek
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #129 on: June 25, 2015, 07:44:06 pm »
It's getting amazingly nice!  Congratulations!  ;)
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #130 on: June 30, 2015, 07:16:11 am »
I made some progress when inductive load is connected to the PSU. In the post #100 some kind of oscillation was visible. In the new CC loop gain is lowered to 40V/V and OP27 is used instead of LTC2057. With clamping diode added in parallel to the inductor I got the following response:







« Last Edit: June 30, 2015, 07:18:06 am by prasimix »
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
PSU in enclosure with provisional front panel
« Reply #131 on: July 12, 2015, 02:13:01 am »
I almost succeed with mounting all required components into enclosure. Said almost, since as you will see the front panel has to be defined. Also some other mechanical issues have to be resolved. At the beginning I was planned to make a first prototype that is only manually controlled. I also had a chance to order a sample front and back panel for free from shop that belongs to brother's friend. The front panel is prepared for mounting multi-turn pots and cheap (and unusable :(, see post #85) eBay grade DMMs, binding posts for power and sense signals and protective earth:



I adopted it to carry the MCU, keypad and BP (Binding posts) PCBs. Now it looks like this:







Please note that this funny look is a provisional and a new front panel has to be designed. In the process I'll probably make a fusion of three mentioned PCBs. More about BP PCB will follow in the next post, here I'd like to present some mechanical obstacles.

First, AUX PCB has to be mounted on the removable top cover because it ends up that there is no enough space to mount everything inside even when MCU, keypad and BP PCBs are mounted on the front panel. This possibly don't looks nice but it's functional and require a little bit longer AC main cables and +5VDC for MCU and BP. The later is also provisional (twisted white-blue) and will be replaced with shorter and shielded one. Actually everything around front panel is provisional.

Second, "double decker" approach has obvious advantage in saving space but from other side if something went wrong on bottom channel, upper channel PCBs have to be removed first. Also even if everything is ok, and you'd like to concurrently monitor signals on both channel you are in trouble.
One can say that even sharing single main heatsink is not a good idea and makes channels chained to each other. I thing that is of secondary importance in comparison with the previously mentioned fact.

Transformer: few years ago I ordered a custom made toroidal which is declared as 2x150VA (2x48VAC). I know that is too low for continuous planned capacity of 2x150W. But, I expected that it can deliver without hassle at least 1.5A per channel or 150W in total. Some preliminary testing told me that this is not a case. Ok, it's summer time and my room temperature is 29oC but with such kind of load it can go easily to ~80oC. Don't looks nice :confused:. What are you thing?

Now we comes to the most problematic part: front panel. First please note that USB cable (see bottom left corner) is connected directly to the MCU PCB regardless the fact that small "breakout" PCB for Ethernet and USB socket was made:



It's not mounted even provisionally since I had no idea where to put it: on the front panel or maybe on the back panel. Both are pretty populated and that is one more indication that this enclosure is too small. But, the same manufacturer have another slightly bigger (and not too big) that could serve the purpose better.
If USB socket (Ethernet will be covered in some of the following topics) can be mounted on the front panel it will be nice to put it on the MCU. Also CC/CV LEDs could be placed on the MCU and at the end the Keypad which is  currently a separate PCB could be merged with MCU. But, here we came to the problem. The MCU PCB is "piggybacked" on the LCD and that adds ~10mm depths which is too big for chosen keypad's tactile switches. That means that "Front panel mainboard" should have a hole for LCD and some simple and not too expensive mean of connecting LCD. Currently I have no clear idea how to do that and especially what kind of connection to use. Possibly my description is also to confusing that someone understand what I wants to point out.
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Binding posts PCB...
« Reply #132 on: July 12, 2015, 02:15:56 am »
A few words about BP PCB. Main intention is to provide simple and secure handling of Sense inputs and making serial and parallel connections without additional external cables. The former will double voltage (up to 100V!) and the later will do the same with current (up to 6A BUT, with proper transformer you can expect up to 10A since everything can be easily scaled up to 5A per channel!). It also includes output protection (combination of TVS, VAR and SAR).
And some extra: since all mentioned cases (remote sensing and channel interconnection) are not ordinary cases, an additional LED indication (MCU programmed) is provided on top of what will be displayed on the LCD screen. An 16-channel LED driver SPI device is used (TLC5925) to drive LED's, small relays for remote sensing and power relays for PSU's power outputs.
The following scenarios are covered with current BP PCB version:



LEDs over sense binding posts are yellow and LEDs over power posts are green or bicolor and can switch to red to indicate special wiring: serial or parallel. To avoid MCU programming error on BP PCB is "hardcoded" protection of concurrent activation of relay in chage for serial and parallel connection which could cause a major disaster.
And also few pictures ...






 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Hello world and simple channel setup...
« Reply #133 on: July 12, 2015, 02:17:34 am »
Some progress is also made on MCU programming side. I had some tough time understanding SPI modes and particularly with many presumptions how to handle MISO signal (data sent from peripheral device to MCU) :). Another story was communicating with LCD but after all I can say that first code is done which provides basic functionality using MCU control.
When code will be in good shape (decently debugged and tested) I'll make it public on Bitbucket. Maybe some sequences will be also recorded with camcorder and provide to public.
Right now I have welcome screen, self-test (check SPI devices availability) and setting of voltage, current and their limits.
Welcome screen (info about model, firmware, number of channels, range, pre-regulator type, ADC and DAC resolution):



If self-test passed screen will looks like this:



Here is how self-test screen will be in case that one channel is faulty (a SPI cable is removed from Ch#2):



Green LED's that belongs to Ch#2 (left side) should not be active. Not really a bug because proper BP PCB handling is wait to be added.

Three values for each channel will be presented most of the time in two rows. Depending of detected mode of operation (available on I/O Expander inputs) in the first row with bigger numbers will be shown measured output voltage (CV mode) or current (CC mode). Second row displays measured and max. value of current (CV mode) or voltage (CC mode). Here is an example when both channel is in CV mode (first with load, second without load connected):



If Ch#1 enters CC mode voltage value will be replaced with output current:



If i.e. second channel is not defined or did not passed initial self-test, screen will looks like this:



Please also note color coded keys in the keypad's first column: they are used as "Function" keys and match right side menu (four position). The rest of the keypad is standard phone pad (0-9, * and #). Idea is to reach any menu options with minimum keys pressed. Also important function like "fine" and "coarse" output adjustment (increasing and decreasing) will be assigned to "hotkeys" on phone pad.

So far all SPI functionality is tested, namely: ADC, DAC, I/O expander (on post-regulator PCB), LED driver (on BP PCB), RTC, external EEPROM, LCD, Ethernet (on MCU PCB). And now one sad fact: Ethernet is out of game with current MCU if Arduino sketches is used. It's library (EtherCard) in combination with other already used (LCD and Arduino itself), requires more memory than ATmega32u4 has :(. That calls for another MCU PCB version and next time I'll not go below ATmega2560. But with current set of ideas what could be added it's quite possible that even 256K will not be enough :).
Anyway, idea is to put set of nice functionality in 32K and release MCU PCB with existing MCU. I hope that software feature list will be ready for release soon.
« Last Edit: July 12, 2015, 05:10:06 am by prasimix »
 

Offline Liv

  • Regular Contributor
  • *
  • Posts: 133
  • Country: by
Re: PSU in enclosure with provisional front panel
« Reply #134 on: July 12, 2015, 05:18:40 pm »
At the beginning I was planned to make a first prototype that is only manually controlled.

In the 21st century, power supply with manual control are not wanted. No need to spend time to support this version.
PSU PSL-3604 Pulse gen. PG-872 Freq. cnt. FC-510
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Arduino shield as PSU front panel
« Reply #135 on: July 30, 2015, 03:46:26 am »
I'd like to present here the current results of thinking what to do with PSU front panel and all bits and pieces that is planned to be mounted on it, namely MCU PCB with LCD 128x64 display, 4x4 keypad and what I call BP PCB which is presented in post #133. Starting with MCU programming shows also that current MCU (Atmega32u4) will not be sufficient to support all SPI devices and that even if i.e. Ethernet is skipped that adding some advancing functionality could become questionable.
First I decide to merge all mentioned PCBs into single one. Also thanks to suggestion from one member I decided to put in the core of the new PCB a 3.2" color touch-screen TFT display. That automatically remove from the scene keypad, provides better presentation in color and four times more pixels and even comes with connector for the SD-card. Chosen display can be found on many places and price is very competitive in comparison with monochrome especially if you take into account that keypad is not necessary any more, and that SD-card support is included (I bought mine here). It looks like this:



... and it's accessible directly via i.e. UTFT and UTouch libraries. Such type of displays usually use 16-bit parallel bus, few more wires for touch screen and standard SPI for SD-card (SD card operate in SPI mode). So many lines require more capable MCU (more I/O pins and more memory for sw library) but as I already said current MCU was not right choice and adding TFT/touch-screen/SD-card create strong demand for migrating to something else where at least two scenarios becames possible: adding a new MCU to the new PCB or maybe create front panel PCB as an Arduino shield for models with more capable MCUs such as Mega2560 (still 8-bit but with many I/O's and 256K program flash memory) or Due (32-bit, 512K memory). Idea to create "MCU-independant" solution sound very attractive but, there was a "small detail": supply voltage is different for mentioned boards, and of course later which works with 3.3V cannot tolerate 5V on its I/O pins. Supporting both of them will require some level translation. That is accomplished using SN74LVCC3245 in the following setup: all peripherals are set to 3.3V, main supply is 5V and if Mega2560 is used 5V<->3.3V translation will be done. In case of Due, same level shifters will be used but with 3.3V on both side ("A" and "B") and on-board 1A LDO will step down 5V to 3.3V. Arduino Due require up to 800mA, therefore a new AUX PS is designed which I'll present in the next post.
The new PCB, what is as I mentioned before an Arduino shield now has the following functionality:

  • BP PCB (sheet 1/5) is almost identical with previous one, but channels binding posts and LEDs are now positioned one on top of another (vertically). Such orientation provide more space for TFT which is oriented as "portrait" (240x320) because it that case it's possible to put Arduino connectors over TFT area (but on the opposite side) and in that way USB connector is on the side, not on the top which simplify connection while staying within the limits of the chosen enclosure (115mm height).
  • On the left ("analog") side I also added, because there is a lot of space now, two V/F converters (sheet 5/5) that can be used for interfacing with battery's NTC that "intelligent" or MCU-driven PSU can include in algorithm for battery charging and to some extend discharging since post-regulator could enter 2nd quadrant and drain up to 300mA from the "load".
  • Interface to the Arduino which will be directly "piggybacked" to this "shield" can be found on sheet 2/5. All connector's pins location was actually mirrored (and rotated) in Eagle files to ensure proper connection without any intervention on Arduino board side (what is of course default).
  • Previous MCU board functionality is now shown on sheet 3/5 and 4/5. That include digital isolators which are mandatory since two channels will be controlled from single source. Digital isolators transfer SPI, interrupt/data ready, reset and chip select for DAC, ADC and 8-bit I/O expander. On the sheet 4/5 you can see used SPI peripherals: external EEPROM, RTC, Ethernet. There is also buzzer driver and one analog input for reading NTC (this one is separate from before mentioned channel dependent for battery charging). At the end there is also 40-pin TFT/touch-screen/SD-card.











When all mentioned is put together on the one place it looks like this (no auto-router function is used):



... or without ground planes displayed:



The front panel should looks like this (main switch is copied from the previous design, it can be much more smaller now, more about that in the next post):



So, at the end such solution can be used to "computerize" any PSU with SPI devices added on the power supply PCB side. All your comments and suggestion are welcome. I'll check all in the coming days and if nothing horribly wrong is found I can go with ordering PCB (together with new AUX PS).
« Last Edit: July 30, 2015, 04:17:19 pm by prasimix »
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Pinouts for Arduino PSU shield...
« Reply #136 on: July 30, 2015, 03:47:39 am »
Please find below an overview of all Arduino pins used for this "shield" (that possibly could simplify understanding schematics).

EDIT 2015-08-05: ETH_SELECT signal also require level shift


« Last Edit: August 05, 2015, 07:04:52 pm by prasimix »
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
New AUX PS with soft-start/stand-by
« Reply #137 on: July 30, 2015, 07:59:05 pm »
With introducing possibility to works with ARM MCU that can be found on Arduino Due board, requirement for power is changed. Existing AUX power supply is designed to deliver up to 300mA with LM(2)5574 controller that cannot deliver more then 500mA. New controller is selected, LM(2)5575 from the same family that can deliver up to 1.5A.
(version with prefix 2 is designed for lower max. Vin). PCB layout is redesigned in a way to insure better power inductor cooling and possibly better EMI performance.
In parallel with redesigning SMPS part of this PCB I start thinking about adding some sort of soft-start capability since toroidal transformer and huge input capacitor that is deployed in mosfet pre-regulator cause huge inrush current when main power is applied. I started with simple delayed relay which bypass power NTC that is connected in serial with transformer. Thanks to discussion with some people I ended up with replacing relay with two triac. Here is how the new AUX PS PCB schematic looks like:



Soft-start that is shown in upper right section. Solution is compromise between two contradicting requirement that require transformer and input/bulk capacitor. Optimally former has to be switched when input voltage is in its peak, and later when it is in zero. But there is only one power up and decision when to start is based on what is less stressful for the circuit. Since speed of current increase for transformer "pre-magnetization" is slower then one for charging up capacitor two triacs is deployed: one with ZCD (zero-crossing detection) optotriac and another with "non-ZCD". Power up sequence is controlled with MCU, at the beginning both triac are turned off.
Then one with ZCD is used to apply power resistor and after some period od time (up to 1 sec) another one with non-ZCD is fired to bypass power resistor. Now the first one can be turned off. Small heatsink is added to "direct" (second) triac to help dissipate couple of watts that is expected during operation.
Such solution also allows PSU stand-by mode and high current AC main voltage switch is no longer required.

New PCB layout is shown below.



« Last Edit: July 30, 2015, 08:03:13 pm by prasimix »
 

Offline lunemec

  • Contributor
  • Posts: 7
  • Country: cz
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #138 on: August 04, 2015, 03:10:28 pm »
Hello parasimix,

these suplies are awesome. You did a superb job. How much did it cost to build the last prototype (in CZK/eur). I want to build some supply and your design looks the most complete here on the forums. I have 19V 2.5A adapter lying around, is it possible to change input voltage? Also I want to build cheaper version, I think analog version with multiturn pots without lcd would be more than enough. Any suggestions on modifying this scheme for that purpose?

Thank you,
Lukas
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #139 on: August 04, 2015, 06:31:53 pm »
Hi Lukas, you can found pre-regulator details in post #128 with BOM for TME and Farnell. For lower Vin you need to change voltage divider R19/R21 (Sheet 2/2) that define UVLO of LM5574. Current values define cut-off at ~43V and you need to decrease it. Let say with R19=18K and R21 unchanged you'll move that point down to ~18V.

Basic solution (without MCU) needs also post-regulator. I'm working on final PCB layout and it will be published soon together with BOM. In case of post-regulator you need to decrease gain of voltage error amp (four resistors) to be within 2.5V for full scale. Similar thing could be doe for current error amp, but with 2.5A its pretty close to what I have now (3.15A for full scale). Again changing four resistors and possibly current shunt resistor.

Important: if you'd like to proceed with basic solution take into account issue presented in post #85. I gave up trying to find a reason why such DMM generate such problem in CC mode. You have to find somehow what's going on in case that you are going to use such DMMs.

Please also note that I'd like to run a GB for PCBs in coming months if I received enough inquiry.
 

Offline lunemec

  • Contributor
  • Posts: 7
  • Country: cz
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #140 on: August 04, 2015, 06:37:22 pm »
parasimix, I'd be interrested in the pcb's but I need to know the price estimate it would cost to build this supply (let's say without the transformer) ... Because if it would be something like 5000 CZK/200 eur, that would be way too much for DYI project ...

Thanks,
Lukas
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #141 on: August 04, 2015, 06:52:50 pm »
If you check pre-regulator BOM you'll see that is around ~25EUR without PCB. Soon we'll know what's about post-regulator.
 

Offline Rolo

  • Regular Contributor
  • *
  • Posts: 204
  • Country: nl
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #142 on: August 05, 2015, 05:28:31 pm »
Just a quick thumbs up for this project. I like reading it and learn from the way you design. Can't wait to see the display lite up !
Just curious, did you already do some tests with the new LCD ? 
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #143 on: August 05, 2015, 06:26:09 pm »
Just a quick thumbs up for this project. I like reading it and learn from the way you design. Can't wait to see the display lite up !
Just curious, did you already do some tests with the new LCD ?

Working on it ;). That will require some time. I'm trying to use UTFT and UTouch libraries. Arduino demo sketches works "out of box" (I'm currently using Mega2560 with level translation shield) but first disappointment was that only monospaced fonts are supported and there is no many fonts available. I need pretty big fonts (over 48pts) for displaying V and A values. Therefore the first step was to find some free truetype and convert/prepare it that can be used for this project. That is already done and proportional fonts is now supported. Nice thing with new display is the colors, so it will be possible to highlight some information using different color. More news will follow soon...
 

Offline Rolo

  • Regular Contributor
  • *
  • Posts: 204
  • Country: nl
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #144 on: August 05, 2015, 11:37:01 pm »
I agree, most Fonts are simple upscaled versions of a small font, resulting in blocked characters. Not wat you want on a fancy TFT. One reason is that they struggle to get it to fit in the flash memory of an Arduino Uno, once you make the step to controllers with more flash (as you did) it possible to load fonts with more details.
I did some experiments with a 2.2 inch TFT and found this article very informative. He uses Adafruit libs as base but made an extension with nice proportional fonts, even some big numbers only and 7 segment fonts.
Maybe this is of any help to you.
http://www.instructables.com/id/Arduino-TFT-display-and-font-library/

 
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #145 on: August 07, 2015, 09:34:42 pm »
Thanks Rolo for that link. Issue with fonts is solved in a way that selected TrueType font is converted into BDF and then further handled with some additional code.
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
IEEE 488.2 specification
« Reply #146 on: September 05, 2015, 04:11:17 pm »
While waiting for new batch of PCBs that include new MCU board designed as Arduino shield and accompanying aux power supply I'm working on specification for interfacing MCU with the rest of the world using USB (or serial over USB) and Ethernet. I found very simple (text based) and pretty straightforward SCPI that is used for communication with test equipment. SCPI specification is free but IEEE 488.2 this has to be supported partially is not. IEEE asks $313 for non-members or "only" $250 for members. I'd not ready to spent such amount of money for it and wondering if anyone knows if some good reference or book exists that cover basic IEEE 488.2 without violating owner's copyright.

Thanks.
 

Offline TerraHertz

  • Super Contributor
  • ***
  • Posts: 3544
  • Country: au
  • Why shouldn't we question everything?
    • It's not really a Blog
Re: IEEE 488.2 specification
« Reply #147 on: September 05, 2015, 04:35:09 pm »
IEEE asks $313 for non-members or "only" $250 for members. I'd not ready to spent such amount of money for it and wondering if anyone knows if some good reference or book exists that cover basic IEEE 488.2 without violating owner's copyright.

http://www.mksa.deit.univpm.it/biblioteca/sala_tecnica/scaffale_strumenti/Strumenti/ieee488/
  std488dot1 1987.pdf   09-Mar-2004 10:16    3.3M
  std488dot1 2003.pdf   09-Mar-2004 10:10    1.8M
  std488dot2 1987.pdf   09-Mar-2004 10:20     10M
  std488dot2 1992.pdf   09-Mar-2004 10:21    1.9M

It should be the law that nothing can be a standard in any way for anyone, unless the full details are freely available.
Copyright was supposed to be so individual authors could profit from their creativity (for a limited time), not so corporations could screw us all forever.
« Last Edit: September 05, 2015, 04:37:59 pm by TerraHertz »
Collecting old scopes, logic analyzers, and unfinished projects. http://everist.org
 

Online prasimix

  • Supporter
  • ****
  • Posts: 1218
  • Country: hr
    • EEZ
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #148 on: September 05, 2015, 04:46:12 pm »
Wow! That was quick and it's the Full Monty! Massive thanks for that TerraHertz. That definitely will simplify the rest of the journey :).
 

Offline technix

  • Super Contributor
  • ***
  • Posts: 3087
  • Country: cn
  • From Shanghai With Love
    • My Untitled Blog
Re: DIY programmable dual channel bench PSU 0-50V/3A
« Reply #149 on: September 15, 2015, 11:41:48 pm »
My M130v1 power supply plan no longer have that complex human interface at all. The user interface is now a simple 16x2 HD44780 LCD, 4 buttons and a USB connection - the device is controlled almost exclusively over USB from a computer. Since at least for me I always have at least one computer next to my workbench, building the complexity as a computer program instead of in hardware makes more sense to me. It is just still too expensive for me to use Bluetooth Low Energy or I may even start rolling test gear using iPhones as control interfaces.

Also I almost never used level shifting of any sort, but I use lots of PCF8574 and MCP23017 I2C GPIO expanders because a 3.3V I2C bus will be able to have both 5V and 3.3V devices on it running at the same time. This also allows me to use smaller MCUs like the ATmega328P which is also a lot cheaper. I did use two microcontrollers but still.
 


Share me

Digg  Facebook  SlashDot  Delicious  Technorati  Twitter  Google  Yahoo
Smf