Markus Lab Powersupply Project

Go To Last Post
38 posts / 0 new
Author
Message
#1
  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've been putting together a design for a microcontroller based lab powersupply for me. I want it to be flexible and easy to tinker with the software. Precision is nice, but not a main concern. I think that this may be interesting for others too and am contemplating providing PCBs/kits/units at a later stage.

Microcontroller based Lab Power-Supply
Inspired by the Dave's EEVBlog power-supply series, see www.eevblog.com
Basic Premises
Modular construction: You can build and test individual modules separately, independent of others working or even being present. Also, if you don't need all parts/gadgets you can just build the parts you need and leave the rest off. For example you can forego the switching pre-regulator or the ucurrent uamp measurement circuit.
The construction is mixed, mostly SMD with some trough-hole. Some parts are through-hole, especially if variants are used. The micro and the op-amp are SMD, because they are not available in through-hole packages. I don't now how big of a draw-back the SMD construction is if this gets distributed as kit one day.
The power supply does not connect directly to mains power, but uses a standard power supply as input. This allows us to skip all mains safety related questions as those are cared for by the upstream mains power supply. The power supply input is 10V to 20V. There are plenty of laptop power supplies who supply DC power in this range. There may be limitations, like 30V-3Amps only being available on 20V power. This mainly to keep the switching preregulator reasonable.
Modules:
- Linear power regulator, takes two input voltages (vset/iset) and translates it into 0-40V/0-4A output with 10mV/1mA resolution. This can me modified, for example to 0-10V/0-1A with 2.5mV resolution. We use the LT3080 (1 amp) or LT3083 (3 amps), for the regulation. These regulators allow to go down to zero volt/amp. The target precision is 1%.
- Switching pre-regulator, takes 10-20V input and translates this into 2-32V. It follows the output voltage, reducing the drop-out voltage of the linear regulator to 2V. This reduces cooling requirements for the linear regulator a lot. It also allows to use any available DC power from a 12V battery to a laptop power supply and get up to 30V output.
- Microcontroller based. The micro (AVR atxmega32A3U) has an USB interface, drives a graphical LCD display and reads rotary switches. It generates the 0-1V ouput (via DAC) used as regulator input. If desired the microcontroller can be replaced with two potentiometers and a 1V reference. The USB interface does not provide galvanic isolation !
- High precision reference. If a better precision than the built-in 0.5% reference of the microcontroller is desired a separate 1V reference can be added. In this case the relevant resistors in the circuit must be of high precision too.
- uCurrent small current measurement device. This is a small shunt with amplifier to measure small currents (uAmps).
- Bluetooth interface. There is a slot to add a simple bluetooth serial interface module. This allows to interface the power supply with a computer wirelessly (galvanic isolation).
- Simple Volmeter. A spare ADC channel is wired to provide a simple 0-40V voltmeter (10mV resolution). Signal ground hardwired to power supply ground. This is not for precision, ut for convenience.
Specifications:
Output:
- 0-30 V
- 0-3 Amps (LT3083), 0-1Amps (LT3080)
- 0-30V in 2.5mV steps (12 bit)
- 0-3A in .025mA steps (12 bit)
Input:
- 10-20V
Hybrid regulation:
- Switching regulator stage -> low power dissipation
- Linear regulator -> low noise
- LT3083 or LT3080
Microcontroller controlled (xmega32A4U)
Computer interface (USB, Bluetooth)
Graphical LCD (128x65)
Optional components/modules:
- uCurrent
- Bluetooth
- 1 Amp version by using a LT3080
- Switching preregulator
User Interface
- Graphical LCD display
- Rotary knob to regulate voltage
(One or two rotary knobs ?)
- 'Menu' and 'Cancel' button
- Beeper

Open Issues and Caveats
Protection against user error on the output side:
- Input voltage connected when power supply is off
- Reverse polarity input voltage
- High input voltage
Protection against user error on the input side, reverse polarity protection
USB interface not isolated
USB ground not equal to ouput ground due to ucurrent device in the return path

Markus

Attachment(s): 

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think it's a great idea to make a nice power supply. Getting it price competitive with those being made in China will be tricky, but that is something that can be worried about later!

Overall your schematic looks good! I haven't had time to digest a couple parts (I'm reading through it while I munch on breakfast) but here are my suggestions:

Give it an AC input. A non-isolated power supply is not super useful, nor is one that needs another power supply to run it (yes I get that there are cheap AC to 12V supplies out there, but still...)

There are USB isolators out there. At least add in an optional isolation circuit for those that need it. Otherwise this is pretty much guaranteed to cause issues at least once.

You might want to add a pot on the I and V sensing to allow for calibration.

Bluetooth doesn't seem like a useful feature, but that's just me (I just struggle to imagine how I'd use that). At least bring out the serial lines to an FTDI compatible connector (though that brings up isolation again - which I guess would be the only reason I see bluetooth being useful).

There is not any ESD protection anywhere. You need it on the USB data lines, buttons, output terminals, etc.

Why the two different current sensors? It looks like the one will be measuring output current + LT3080 quiescent current (but with lots of noise, due to its close proximity to the flyback), while the other will only be measuring output current. I think the other one will also contribute an output voltage error at higher currents.

Keep us updated with your progress!

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

nleahcim wrote:
I think it's a great idea to make a nice power supply. Getting it price competitive with those being made in China will be tricky, but that is something that can be worried about later!

Overall your schematic looks good! I haven't had time to digest a couple parts (I'm reading through it while I munch on breakfast) but here are my suggestions:


I'm not really worried about competition right now. I have a day-job to feed me and my family, so I'm not looking to make money on this. But I want to prevent the BOM price to grow too much though, some pricey stuff can be optional.

nleahcim wrote:
Give it an AC input. A non-isolated power supply is not super useful, nor is one that needs another power supply to run it (yes I get that there are cheap AC to 12V supplies out there, but still...)

The main problem is that then I have to worry about certification, compliance and other such things. This can be done but is expensive. Essentially I have decided to outsource this problem to our Chinese friends with their $15 laptop power supplies :-) Maybe later, but only after all of the exiting parts start to fly.

nleahcim wrote:
There are USB isolators out there. At least add in an optional isolation circuit for those that need it. Otherwise this is pretty much guaranteed to cause issues at least once.

Yes, an optional USB isolator is a good idea. For now my thought was that the Bluetooth module fulfills this need. The USB interface would be used mostly for firmware updates via boot loader.

nleahcim wrote:
You might want to add a pot on the I and V sensing to allow for calibration.

My intent is to use good precision resistors (<=1%) to alleviate the problem. But basically I can add pots, who wants them uses them, who doesn't just uses a jumper instead.

nleahcim wrote:
Bluetooth doesn't seem like a useful feature, but that's just me (I just struggle to imagine how I'd use that). At least bring out the serial lines to an FTDI compatible connector (though that brings up isolation again - which I guess would be the only reason I see bluetooth being useful).

I found that all recent PCs do have built-in Bluetooth and that this can be used like a serial port. This gives an easy, insulated interface to a PC.

nleahcim wrote:
There is not any ESD protection anywhere. You need it on the USB data lines, buttons, output terminals, etc.

Yes, you are right, protection is not well though out yet. The buttons will be on the PCB, do I need ESD protection then ?

How do I choose the ESD suppressors, for example for USB ?
I have had a look things like 'XMEGA - USB hardware design' form Atmel or 'USB 2.0 circuit protection' from Littlefuse. But they don't say the exactly the same so I have information, but am looking for real-world wisdom.

nleahcim wrote:
Why the two different current sensors? It looks like the one will be measuring output current + LT3080 quiescent current (but with lots of noise, due to its close proximity to the flyback), while the other will only be measuring output current. I think the other one will also contribute an output voltage error at higher currents.

They serve a different purpose. The high-side current sensor is part of the (analog) current limiter. The low-side current sensor is amplified by the precision-opamp. Its purpose is to measure (very small) currents (uAmps) while keeping a small burden voltage. It will be bypassed by default.
I'm aware that there will remain a small low-side resistance, but the entire thing is optional. So you can just not include it if you hate the downsides.

In the end, many of our circuits and devices are very low power and measuring it is interesting and not always easy (burden voltage of many multimeters).

nleahcim wrote:
Keep us updated with your progress!

I will don't worry.

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Its a nice idea to mix linear and switching topologies to minimize power losses while the linear accuracy is kept.The linear regulator takes in the input Vin=VOut+0.7V so the losses are neglible.
What seems a little strange to me is why use a step up converter and what C26 1uf does in the circuit.
If a buck mode was used 30v had to be applied in the input and output always would be equal or less than this voltage.
Now the input is in range 10-20v so a buck-boost mode converter must used,SEPIC topology can be made connected C26 not in series but in the output and the mosfet drain.

1V reference divided with 4096 steps multiplied x30 gain for 30v output gives 0.00732v per step.

Thats what i am thinking at a glance.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

geoelec, you are asking good questions. I'm afraid I don't have perfect answers for all of them, that's why I need the discussion here to gain insights and fix problems.

The way how I couple the switcher with the LT3080 using Q22 comes straight out of the LT3080 data sheet. There they use a buck converter, but the idea is the same here. I'm fairly confident that the basic idea is sound. It may be that 0.7V is not enough, then I'll add a diode or two in the base line of Q22. We'll see in practice.

I need a SEPIC because the input and output voltage ranges overlap. Buck or boost alone will not do as I can have 12V in and 20V out and 12V in and 5V out. The circuit comes directly out of the TI webench tool. The tool does not allow for variable output, but I used the worst-case (10V in, 30V/3Amp out) for the component sizing.
You are right about C26 being funny, I misplaced it, it should be between the transformer output and the drain of Q21.

You are right that 4096 steps over 30V give 0.00733V per step. But I would like to have some headroom, so I'll use 4096 steps over 40.96V. I prefer to deal with 1/4 steps over 1/3 steps in software too.

Thanks for the feedback.

Attachment(s): 

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I've discussed this in the meantime also on the EEVblog forum (http://www.eevblog.com/forum/projects-designs-and-technical-stuff/general-purpose-power-supply-design-7488/). Was pointed out some bugs and improvements and have reworked the supply for the internal voltages (3.3V, backlight, etc). This in preparation to potentially beeing able to operate it from USB power. This with a severely limited power budget, of course.

I've ordered the parts for a prototype ($100 at Mouser) and a set of PCB's (Seeed). I may have some spare PCBs to give away, if someone is interested.

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Nice to see the progress you have made, Markus.

I am in the process of enclosures. Which is a lot of work ;)

Cheers

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, enclosures and packaging it all nicely are an entire chapter. For now I think I'll box it into a plastic enclosure from Hammond (RM2015L).

I've come up with the following mock-up. Not really happy with it yet. But I'll have to have an enclosure in my hand, but they are out of stock everywhere...

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

This is my mock-up

Size matters. I prefer large knobs ;)

Attachment(s): 

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

If you want a really nice front panel, you should try frontpanelexpress.eu.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

My powersupply is just for home-use and low cost. Nevertheless, you can be proud of yourself, Jeroen: many moons ago we discussed, here on AVRfreaks, the design of a lab supply. Initially I considered to control the outputvoltage with a DAC, like Markus does. You made a remark about what could go wrong if the controller would get upset by f.i. an EMI-pulse: so you recommended a more rugged control for output voltage: a potentiometer. I took your advice to heart and changed the design accordingly.

Thanks for that !

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Yes, I like large knobs too, but also a large display, the bigger the better. But in real life there have to be compromises.

I find there is an imbalance between the big knobs and the small display on your panel.

I had a look at frontpanelexpress, the front panels are nice. They seem mostly geared towards rackmount equipment/sizes. Prices start at almost $100, that is about the same as the rest of the BOM.

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
I find there is an imbalance between the big knobs and the small display on your panel.
You're so right, Markus. It's all compromises here as well. I compensated for the small display by using a large font.
I have had a few of these old-style mobile phone displays for 6~7 years now (when I started with AVR's), and only used one so far. This was a project in which I could use all of them. Although small, they're pretty good readable. The backlight is a pain: it takes 60mA from an isolated floating supply. Need to find a fix for that. It is and will remain a Project ....

The Dual Lab Supply is better proportioned IMO.
Same PCB's but different assembled.

It's nice to share idea's in a co-development a simular project.

Cheers

Nard

Attachment(s): 

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I like the dual supply, good display-size button size balance :-).

You are very right about the backlight, mine wants 45mA at 3V. A real pain to get from a 12-20V input. In the end I decided to simply size a resistor for 12V and do the rest with PWM in software. It's one more thing in software, but the xmega I'm using has a smorgasboard of timers, etc. so it a set-up and forget type of thing.

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Long ago I had these panels made:

So you can use FrontPanel Express for smaller panels :)

I'd use a big graphics VFD as display, unfortunately they are very expensive. Though RS has a 12 digit 14 segment module for a silly low price.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

That is a nice front panel. What was the cost ?

I've decided to use a 128x65 graphical LCD. It is similar in price as a same size character LCD ($14) but gives me a lot of flexibility in what I can display. The biggest limitation will be the available flash size (32k).

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I think it was around 25 euro or so; it was a batch of 25. The cost is (or was) directly proportional to the number of operations, so every extra character added cost :)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

In the short term I'll concentrate on getting the electronics and firmware working. The we'll see again about the details for the box for the device.

One related problem to solve is with the display. It is a chip-on-glass display with no specific provisions four mounting it. In the prototype I'll probably just hot-glue it to the front panel from behind. Will look ugly, but only on the inside :-).

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

No project is complete without some good load of hotglue. At least, that is for my projects :)
F.i., to lock-tite the M3 nuts in the wood. Put a silicone sleeve over the bolt-thread that should stay unglued, embed the nut in hotglue, let it cool down and pull-off the silicone sleeve. Re-usable.

Very nice display, Jeroen.

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Is that COG LCD one from Displaytech by any chance?

My experience in the past with a Displaytech 128x64 COG was not that good, unfortunately. I started a thread on that a couple of years ago.

If the display has pins, I'd mount on a PCB that has at least two holes. Depending on the case construction you either place threaded studs on the backside of the frontpanel or use a separate frame for it. Like having two frontpanels, one internal where you mount buttons, pots, jacks and possibly the display on, and the visible one. This will give a clean look without any visible mounting hardware.

Nard, there is no display in the picture I posted, so which display is nice? :)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

No I've fallen for the COG displays from Newhaven. The specific one I'm using is the NHD-C12865AR-FSW-GBW. It has a flex-cable and I provide a mating connector on my board. The display area is quite large ca 50x30mm with not much border. There is also a bigger version, ca 10mm bigger, if I have the space to upgrade.

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
There is also a bigger version, ca 10mm bigger, if I have the space to upgrade.

ShowOff :lol:

Quote:
Nard, there is no display in the picture I posted, so which display is nice? :)
Nice frontpanel I meant of course.
And in the meantime I found a flaw in the sketch-up of mine: I was about to start milling. Now it's too late ( 23:10h )to do so, ... too much noise and likely too much me swearing when the millbit runs hot in the perspex (lexan, very tough) That was a narrow escape.

Sleep well guys

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Not completely finished (some varnish on the wood, calibration and a fiber), but this is how my PSU Queck turned out. Quack will follow soon. Quick will take a bit longer ;)

Attachment(s): 

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Hi Nard,

Looks great. What is the "Cool, Cooking, Hot" connector?

Cheers,

Ross

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Status of temperature. A 3mm fibre comes in there.
Original plan was this text:

We're cool
We're cooling
We're freaking HOT

But that took too much space :)

There is active cooling (fan) but under normal use it will not come in. If Queck is forced into Linear Mode, and the user decides to pull 5A @ 5V, dissipation will be around 100W and then we need quite some airflow to cool the linear regulator. When used in Pre-Switched mode, dissipation will be far less.

There are three cooling profiles in the menu: Quiet, Normal and Cool. Up to the user.

OffTopic: My youngest is moving out of the house, so I'll have other things to do in the coming week(s) ;)

Cheers

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

:lol:

My old eyes thought it said "cooking" ... I guess that is an alternative to "freaking Hot".

Best of luck with the moving. Do you get to change the locks? :lol:

Cheers,

Ross

Ross McKenzie ValuSoft Melbourne Australia

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Do you get to change the locks?
My house, ....... or his ? :lol:
This was the first day and I feel as if it's been a week already. How is that possible ?

Locks will be re-renewed. Which will be quite expensive as all cylinders share one key now. If we want to keep it like that (the one key for all) I guess we're looking at a few hundred euro's. I wonder what they will do ....

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Why would want to change the locks? I still have the key to my parents house, and I moved out more than a decade ago.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Quote:
Why would want to change the locks?
We were just kidding, Jeroen. Both my kids have keys to my house. But moving into a new home of which you can't be sure that you have all keys: then you'd better change all locks.

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

I see :) And of course to make sure the previous owners/tenants can enter the house a bit too easily ;)

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Markus,

What's the theory of operation for the LM334 (current source) on the output pin of the power supply?

What's the expected power dissipation of the LT3080/83 at the highest allowed output current and worst case voltage across the LT3080? What type of heatsink are you using to dissipate this heat?

Linear Regulators are not always the best low-pass filters. Check out the Ripple Rejection graphs on page 7 of the LT3080 data sheet. You'll probably want to put some power filtering between the switcher output and the linear input.

BTW, what IS the switcher frequency? (It's a good idea to annotate your schematic with key design parameters like this one - e.g. switching frequency, roll-off and gain for op-amp circuits, power dissipation for regulators and other power devices, etc.)

My two biggest pet peeves with power supplies:

A. Not enough output terminals or binding posts, especially DC Common. Power supplies don't play in a vacuum, they usually connect to lots of other equipment in a lab set-up. Why not 2 or 3 positive binding post and 2 or 3 Common binding posts? Also a good idea to orient the binding posts along a vertical edge of the case rather than the bottom edge as most are (for bench-top use), and make sure the wire hole in the binding post stud is oriented horizontally. (This may affect panel design if you are using D-holes for the binding posts.)

2. Lately, I have seen a number of commercial lab power supplies with red terminals for both plus and minus outputs and a black terminal for chassis/earth ground. What planet are they living on? Stick with the tried-&-true conventions:
RED -Positive
BLACK - Negative/Common
GREEN - Chassis/Earth Ground

In EE-Land we've been doing it that way since the Egyptians and Romans used that color scheme on their power supplies. It's what everybody is expecting, so less chance of a disaster in the heat of battle.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Was out on the back of some horses with my family, good fun, but keeps you away from AVRFreaks :-).

I could not wait, so I took some basic tools with me to spend some time with the prototype. That's how it looks right now:

@Nard: How did you create your front panels. They look mighty good !

@Chuck: The LT3083 does need a minimum amount of load current for stable operation. The LM334 is there to consume a minimal amount (1mA), to keep the LT3083 happy, even if there is nothing connected to the output. I know this will stop working correctly at low voltages (<1V), I'll have to measure the real effects on the prototype.

If the switching converter is in place, then the voltage across the LT3083 is around 0.7V (essentially the base-emitter voltage of Q31). The diode D31, shown in the schematic will probably be a 0R resistor instead. This comes straight out of the LT3083 datasheet, so I think it will work out.

The switcher will run around 400kHz. I've got the wrong transformer and have to order it with the next round of things, so the switcher has to wait for now. Enough other stuff to get running first. I do have a separate document, where I note design decisions, calculations, etc. Not everything can go onto the schematic.

My PCB is quite small and the box I'm aiming for is not much better, so I may not have physical space for more than the three binding posts I'm planning for. I will use black for negative and red for positive, for sure !

For now working with the xmega has been a pleasure, the naming conventions make sense and most things did work right out of the box (except if you compile for 32k and have only 16k :$).

Now I'm working on the LCD, I found it easier to roll my own, than to use ASF. As ASF does not have fonts anyway, I'll try to get the dogm library working. Or is there a better option ?

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Both casings are made of plywood. The frontpanel of the dummyload is a printout on photopaper, covered with transparant foil. Glued that to the plywood.
For the powersupply: the top-half is lexan (perspex), smoky version, one side is matted and is what we look at. The cutting for the display was a tough thing to do: not a rectangular shape, but curved. The controller is fitted to this panel and connects with the (isolated) PSU-board with just two small flatcables: power and comm. The top knob sits on a subframe, so it suffies to remove the knob and then the complete upper half of the front can be taken out. The length of the flatcables allows for that.
The bottom half is the same principle as used on the dummyload.

I use mostly scrap material for these projects. Low cost is a necessaty for me. Nevertheless, I try to stick to conventions like red for +, black for - and green or green/yellow for safety-ground. For the dual lab supply the + is yellow and the - is blue. Not such a bad choice, considering I can use it as a " + gnd - " supply. And there is a second set of output terminals: the loudspeaker connector, left of the 4mm banana terminals.

I wrote my own routines for the (surplus) displays, including fonts. What I really like about using a graphic display is that I can mix fonts. And it allowed me to add a bargraph at the bottom of the display to show the temperature of the heatsinks.

The 15V @ 5A max is underspec'd. It delivers 7A easily @ 0 ~13V and above 15V linear derating to 2A @ 20V. Pulling more current results in a 100Hz ripple on the output. But the SW checks for that and shows it on the display.

It's a very forgiving supply and that's what I went for.

Two things that bit me: the backlight for the displays. Solved that by taking out the original BL and replaced it with 4 modern leds, running of the floating 10V for the controller.
The second thing was the power disspation of the mains transformer. In the dual lab supply I use a 250VA toroid transformer which consumes only 6W in idle mode. The powerunits in the single lab supplies, 150VA, conventional EI-core, dissipate 13W in idle. I assumed it would be far less, but as we all know: assumption is the mother of all FU's. So some extra ventilation was required.

Looking forward to your results, Markus.

cheers

Nard

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Markus,

So with 0.7 Volts across the 3083 what is the maximum power dissipation you need to accomodate? Also important in a lab supply is the maximum power dissipation encountered when the output is shorted for an indeterminant time period. How does the 3083 contend with a dead short on its output? (I.e. Foldback or constant current?).

With 400 KHz coming into the 3083, you may see a lot of this fed thru to the output of the 3083. The 400 KHz component on the input may also screw-up the 3083 in some way. I'd be looking forward to adding some power-level filtering components between the switcher and the 3083 as a first-cut approach. Once the whole thing is built and running properly, I'd try removing, or reducing, the filtering to see if there are detrimental effects under all possible operating conditions of the supply.

Does your micro design allow you to "modulate" the output voltage in any way? For example: at 50 or 60 Hz, or between two set limits at a certain rate, moderate-term dropouts and spikes, etc. I would find this feature useful for testing standalone devices that would normally run on a battery, automotive power supply, UPS,etc.

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Plons wrote:
Locks will be re-renewed. Which will be quite expensive as all cylinders share one key now. If we want to keep it like that (the one key for all) I guess we're looking at a few hundred euro's. I wonder what they will do ....

How many locks do you have?! Last I checked it took me all of about 5 minutes to re-key a (very common over here) Weiser lock with a master key. No locksmith would charge you more than about 5$ per lock + price of the keys if you bring them to him.

Unless they are all Abloy or whatever other coded cylinders...

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

Chuck-Rowst wrote:
So with 0.7 Volts across the 3083 what is the maximum power dissipation you need to accomodate? Also important in a lab supply is the maximum power dissipation encountered when the output is shorted for an indeterminant time period. How does the 3083 contend with a dead short on its output? (I.e. Foldback or constant current?).

The LT3083 has a internal current limiter to 3.3 Amps. So the maximum power dissipation would be 0.7V x 3.3 Amps = 2.3W. The LT3083 is a through-hole version right now in a TO-something package, ready to be attached to a beefy heatsink. Essentially I did want to be able to work without the switching preregulator, but then I will need a worst case of about 100W of power dissipation.

Quote:
With 400 KHz coming into the 3083, you may see a lot of this fed thru to the output of the 3083. The 400 KHz component on the input may also screw-up the 3083 in some way. I'd be looking forward to adding some power-level filtering components between the switcher and the 3083 as a first-cut approach. Once the whole thing is built and running properly, I'd try removing, or reducing, the filtering to see if there are detrimental effects under all possible operating conditions of the supply.

That is one aspect I need to measure the entire thing in real life. Yes, this is a distinct potential problem area, we'll see.
Quote:
Does your micro design allow you to "modulate" the output voltage in any way? For example: at 50 or 60 Hz, or between two set limits at a certain rate, moderate-term dropouts and spikes, etc. I would find this feature useful for testing standalone devices that would normally run on a battery, automotive power supply, UPS,etc.

The DAC of the micro can provide up to 1M samples/sec. Its output goes with minimal filtering to the set pin of the LT3083. Yes, you can generate interesting output modulation, if so desired. It may be that you would like to reduce the size of the two 100n capacitors in the signal path (C45, C46), in this case.

Markus

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

hugoboss wrote:
Plons wrote:
Locks will be re-renewed. Which will be quite expensive as all cylinders share one key now. If we want to keep it like that (the one key for all) I guess we're looking at a few hundred euro's. I wonder what they will do ....

How many locks do you have?! Last I checked it took me all of about 5 minutes to re-key a (very common over here) Weiser lock with a master key. No locksmith would charge you more than about 5$ per lock + price of the keys if you bring them to him.

Unless they are all Abloy or whatever other coded cylinders...

The cylinders we use here in the Netherlands cannot be re-key'd: fixed "code" only.

A GIF is worth a thousend words   They are called Rosa, Sylvia, Tricia, and Ulyana. You can find them https://www.linuxmint.com/

Dragon broken ? http://aplomb.nl/TechStuff/Dragon/Dragon.html for how-to-fix tips

  • 1
  • 2
  • 3
  • 4
  • 5
Total votes: 0

very nice
thank you