EEVblog #232 – Lab Power Supply Design Part 5Posted on January 6th, 2012 25 comments
Dave goes through the final base schematic for his Open Hardware lab power supply design.
Schematic is here:
Great video, one of my favourites (and I’ve watched them from episode no. 1). Nice to see Open Source Hardware logo on Altium schematics!
Hi Dave, I missed the shot by asking this on the wrong blog post the last time, sorry.
- Will we be able to purchase a kit from you, when ?
- (Excuse the silly question as I don’t know if there is a standard for this) The Serial port on the Arduino pro mini is arranged as TX0/RX1/VCC/GND, would it be possible to change design to use it (if not, I’ll just make my own adapter, no big deal) ? It would enable us to use directly the available usb converter to programm the AtMega168.
Altough I already knew you would not like the idea, I still have to mention it.
You could get 12 Bit precision from a 10 Bit ADC by oversampling. The average from 16 samples should be equal to a 12 Bit conversion.
There is an app note from Atmel about this topic:
Are you still considering >7V input voltage DC, as stated in a previous video? I can’t see it in the schematic.
Are you also still sticking to the initial specs (see part 1), 0->6Vmax, 0->1A?
I think that a back light will be good if you put it in
Great video, as usual, Dave!
I really worry about that heat sink, though. That’s quite a trap for a kit builder! I would worry far more about that than a few SMD devices. It’s interesting how worried people are about SMD soldering. I hear often that, once people give it a shot, they prefer SMD soldering to through hole. I know I do!
Anyway, looking forward to the kit!
I would also love to buy this as a kit. Or even buy a bare board and have a Mouser BOM to click on like some other open hardware projects do. It’d be awesome to have this on my bench, particularly the microcurrent readout.
A perfect 8 MHz clock will not give you a perfect RS-232 baud rate, since none of the baud rates are factors of 8,000,000. I tend to use resonators designed for baud rates, like 7.37 MHz (115200 * 64) or 18.43 MHz (115200 * 160)
Failing that, one trick I like (when the hardware supports it) is to use a slightly faster baud rate, but transmit two stop bits while expecting only one in return. That way, the next character always shows up a little later than the receiver expects.
great video as always!
Are you aware that with the Atmega8/168/328 you can’t use the RESET pin as GPIO without losing ISP programming capability? (You can still use high voltage programming after you have disabled the reset pin. But this can’t be done using the ISP connector as it is a bit parallel protocol.)
So I’d suggest to move LED1 to another pin or not populate the ISP connector for the kit as noone would be able to use it with pre-programmed chips anyways..
Awesome video !
I’d love to see more “how to design” type of videos.
From about the middle of the video I was hoping you haven’t done the pcb layout yet. But unfortunatly you alredy have. I had hoped you could do the layout on video.
Any chance you could make a video about the pcb layout and the proccess of designing it ?
Great Video(s) as always.
I was in parallel finalising my own power supply based on your initial videos, I considered higher resolution (bits) than a micro can provide but unsure that I need that precision on a bench top supply for hobby use, anyway we’ll see. Maybe now I’ll just get the kit.
Would be great to see the industrail design aspect ie. fitting the LCD & encoder/buttons to the front of an enclosure with the heat sink at the back for example.
Keep up the great work!
I was considering building a lab power supply of my own a while back now but couldn’t really get the design optimized for accuracy and reliability.
Elektor has published a digital control power supply back in 2001. It served as a good reverence, but I didn’t quite like the topology they used.
Don’t you think that adding a Programmable Gain Amplifier (PGA) would be better for measuring voltages & currents in the lower range instead of 2 separate circuits?
What happended to DaveCad?
I guess that format has a huge impact on the tooling costs from the pcb fab…
Like [hpux735] I also prefer soldering smd above through hole as long as the pitch is not to small and there is a pcb to fit it on. And there are plenty microcontrollers around with 12Bit ADC and DAC. And with some more IO line’s would make it easier to use a cheaper and more hacker friendly (availability) HD22780 compatible LCD.
Some extra free IO’s would also make it easier for hobbyists to add some unforseen functionality. For Example I would personally like to add an RS-485 Interface instead of RS232 but that needs an extra enable pin.
Did you ever think of making some kind of contest for who can write the best software for this baby?
Your average DSPIC has a 12 bit ADC and a 2 16 bit PWMs… was arduino a requirement?
The 30F2011 also showed in as one of the cheapest in your digikey search
I specifically did not name any brand of microcontroller in my previous post because it turns far to easy into a “my micro is better than yours” contest.
Please let it stop here and respect Dave’s choice of using all through hole components. His design seems to be almost finished and he already found a box to fit everything in.
Besides that the costs of the IC’s alone are fairly insignificant for small batches of pcb’s.
Take the uCurrent for example. It’s basically a $2 opamp in a box but there are lots of different reasons why it adds up to a $50 sales price. Dave has even shed some light on this subject in several of his previous blogs.
And if that’s not enough, using through hole for the micro makes it really easy to use a socket and an adapter print with the (smd?) controller of your own choice. As a matter of fact I might just do that to get some spare IO for my own mods.
just had an idea, I realy like the ability to measure small currents accurately but low side shunts give all kinds of problems. For example when you’re designing a circuit and you want to use separate power supply’s for a 5V and a 3.3V section with a common ground.
Have you thought of using 2 max4080′s in parallel. The 2nd one can have a shunt resistor of 100 ohm or even 1k. It doesn’t matter very much if there is a voltage drop over that shunt. The 0R1 shunt can be disabled easily with an P-channel fet or an PNP transistor driven into saturation. Or you can also set the shunt in series and short the 100Ohm shunt with a fet if it’s not needed…
Just my 2 pennies, it’s your design of course.
nice advance with that project. one idea: can we get the software to act as a battery charger circuit on demand? theres all in: voltage and current sense, we might add an 10k NTC input on one of the ADC…..
and software that is adaptable when we get a new battery chemistry… a winner!
Excellent video series. Will you be addressing the design of a dual tracking power supply with + and – floating outputs?
awful piece of information, I had come to know about your blog from my friend vimal, mumbai,i have read atleast 13 posts of yours by now, and let me tell you, your blog gives the best and the most interesting information. This is just the kind of information that i had been looking for, i’m already your rss reader now and i would regularly watch out for the new posts, once again hats off to you! Thanks a million once again
Nice project. I like it. I think it is very helpful.
no, Digikey didn’t get it wrong. You didn’t understand the way they sort. Number of channels = number of multiplexed inputs. Number of converters = number of converters working at the same time. So by that, the Microchip one you are looking for is sorted correct as four channels/one converter. Just like for example with all Atmega µCs, a number of channels that share one converter.
Leave a reply