An (almost) linear power supply for my music PC
- Thread starter thatguy
- Start date
Thad E Ginathom
Well-Known Member
I'm not good at these sort of numbers and technical specs, but it seems that there are some very good ones in this review for the Antec CP-850 switched power supply.
Reading it made me very happy because
--- I like the author's sense of humour.
--- I have one of these power supplies.
But it's huge: so non-standard huge it only fits a limited number of Antec's own cases, and needless to say, they are big cases.
Reading it made me very happy because
--- I like the author's sense of humour.
--- I have one of these power supplies.
But it's huge: so non-standard huge it only fits a limited number of Antec's own cases, and needless to say, they are big cases.
My plan is to run my PC off solar power too, so that, at least in the day, it will be immune to power cuts, but preliminary investigations suggest something like Rs.40,000 to supply 160W. If you could do it for less to a DC-only 12v supply, that would be neat
Last edited:
I am sure it is possible (trickle or float charging as Borg has mentioned), but won't you lose the main advantage of running from the battery. (I don't know much about power supplies but) I think it would be hard to completely isolate your PC from the main power supply.
My uninformed opinion
picoPSU is (in all probability) an SMPS. If you choose the one with 32 volt input, all four of the ATX buses will be switched outputs. The advantage of the 12v model is that the +12v bus is directly powered by whatever source you use. Any component which uses +12v will get a clean supply of power. I am quite certain the Asus Xonar ST uses the +12v as its main power source.
It is a desi product. Vartech, bought from S P road.
Last edited:
Here is what I propose. Lets design a power supply that can work with any input voltage from 15 to 24. It regulates the input down to 12 volts and feeds into a pico PSU.
One can either use a battery or the rectified and filtered output from a transformer as the input.
josephjohnt, santhosh and cmsajith: Would it possible for you guys to estimate your power consumption. The highest rated pico PSU is 160 watts. I would be wary of using it for more than 30-40 watts if I want to run fanless.
Once we build up some skills in doing power supplies, we can attempt a fully linear ATX PSU.
One can either use a battery or the rectified and filtered output from a transformer as the input.
josephjohnt, santhosh and cmsajith: Would it possible for you guys to estimate your power consumption. The highest rated pico PSU is 160 watts. I would be wary of using it for more than 30-40 watts if I want to run fanless.
Once we build up some skills in doing power supplies, we can attempt a fully linear ATX PSU.
to be precise i would say we need to "float it" not trickle
The battery need to be connected to a power source with voltage regulated at 13.8V and load connected to battery(load, battery and powersource parallel) and source need enough amps to supply both the load and battery charging current so that the battery is not discharged and you get a constant voltage all the time.
of course to isolate from grid you need to use some other power source like a solar panel
the battery will absorb the ripple and the advantage is not really lost. the output will be rock solid with the enormous capacitance of a battery.also keep in mind that if you just use the battery without a power source it cannot hold the voltage at constant level and it will taper off
here is a schematic of 13.8v 20A powersupply with short circuit protection i got from net, you can increase capacity based on your requirement
edit: it is aso possible to skip the smps portion and make linear regulators off this for the +/-5v v and -12v,+12v can be made just off this itself with some series diodes to drop off the 1.8v
Then you will get a totally linear supply
Attachments
Last edited:
Thanks borg for the info and the schematic.
I was hoping that we could run it on pure power even while the charging device is not charging (solar for example) or not connected. Is this true even if the battery's capacity (amperage) is high?
santhosh
if you want you can switch off the charger while listening especially with high capacity battery. we can add a battery charge monitoring and indication circuit. i assumed people dont want too big battery and also donot want to bothered about the dropping battery voltage while discharging, hence the idea of keeping it afloat
if you want you can switch off the charger while listening especially with high capacity battery. we can add a battery charge monitoring and indication circuit. i assumed people dont want too big battery and also donot want to bothered about the dropping battery voltage while discharging, hence the idea of keeping it afloat
Last edited:
@Borg, Though I'm aware of some of the facts, you have explained it wonderfully well.
@Santhosh, AFAIK, it will taper irrespective the battery capacity; only the tapering will be less steep. For a given time period, the voltage drop would be lower.
@Santhosh, AFAIK, it will taper irrespective the battery capacity; only the tapering will be less steep. For a given time period, the voltage drop would be lower.
Time now for some action.
Checked my closets and it looks like I already had most of what I needed (leftovers from an abandoned amp project). One more visit to SP road should clear the road. Will buy the toroidal transformer after I decide on the rail voltage.
After reading this post I decided to add a 5 volt regulated output so that the sound card can by fully powered by the linear supply.
Progress will be slow, but will update the thread as I make prototypes and test them.
Cheers
Checked my closets and it looks like I already had most of what I needed (leftovers from an abandoned amp project). One more visit to SP road should clear the road. Will buy the toroidal transformer after I decide on the rail voltage.
After reading this post I decided to add a 5 volt regulated output so that the sound card can by fully powered by the linear supply.
Progress will be slow, but will update the thread as I make prototypes and test them.
Cheers
Last edited:
There is a part of me (the Audiophile part) which wants to go the full way and make a complete ATX power supply.
But ...
Traditionally, an ATX power supply has had 5 voltage levels, +12, +5, +3.3, -5 and -12. The -5 was deprecated (just checked my PSU, no white wire) but the latest ATX spec still has the -12v.
Since I want the power supply to work from a simple 15-30 volt power supply, -12 volt poses a problem. I could request a toroidal with two secondaries, but I will lose the ability to run from a battery.
After a little search on Google, I realized that most motherboard do not use the -12 lines. Here is an article on ATX supplies written in 2011 which mentioned a 10 year old motherboard as an example of one that uses the -12 volt line.
Voltage Rails : Power Supply 101: A Reference Of Specifications
The following video gives some more hope. A guy tries to run his PC without the white and the blue wires.
ATX Power Supplies - Negative Voltages - YouTube
Another PSU will have its wires cut tomorrow
But ...
Traditionally, an ATX power supply has had 5 voltage levels, +12, +5, +3.3, -5 and -12. The -5 was deprecated (just checked my PSU, no white wire) but the latest ATX spec still has the -12v.
Since I want the power supply to work from a simple 15-30 volt power supply, -12 volt poses a problem. I could request a toroidal with two secondaries, but I will lose the ability to run from a battery.
After a little search on Google, I realized that most motherboard do not use the -12 lines. Here is an article on ATX supplies written in 2011 which mentioned a 10 year old motherboard as an example of one that uses the -12 volt line.
Voltage Rails : Power Supply 101: A Reference Of Specifications
The following video gives some more hope. A guy tries to run his PC without the white and the blue wires.
ATX Power Supplies - Negative Voltages - YouTube
Another PSU will have its wires cut tomorrow
Last edited:
I rigged up my first circuit today. Just an assortment of components I picked up from S P road. Although it has been almost two decades since I did any electronics, Zener diode and the emitter follower are amongst the few things that I still remember
Here is my first attempt:
The zener has a breakdown voltage of 12 volts. 2N6059 is a NPN Darlington pair.
I wasn't very happy with the regulation, but I shouldn't have expected anything better from something this crude.
Here is the circuit driving a four ohm load, with the input voltage at 14.9. The multimeter reading (11.83) is the output voltage.
When I pushed the input to 29.9 volts, the output rose too (12.57 volts). And so did the temperature of the power transistor. The transistor and the heat sink became too hot to touch. Lesson learned: It would be stupid to keep the input higher than 15-16 volts.
The sausage like things on the plate are one/two ohm high wattage resistor. These get quite hot with 3 amp of current.
Here is the no load voltage with an input of 15. Still under control. Almost.
Too late to do anything more today, but I intend to take some more readings with the circuit before abandoning it and building the next one.
Here is my first attempt:
The zener has a breakdown voltage of 12 volts. 2N6059 is a NPN Darlington pair.
I wasn't very happy with the regulation, but I shouldn't have expected anything better from something this crude.
Here is the circuit driving a four ohm load, with the input voltage at 14.9. The multimeter reading (11.83) is the output voltage.
When I pushed the input to 29.9 volts, the output rose too (12.57 volts). And so did the temperature of the power transistor. The transistor and the heat sink became too hot to touch. Lesson learned: It would be stupid to keep the input higher than 15-16 volts.
The sausage like things on the plate are one/two ohm high wattage resistor. These get quite hot with 3 amp of current.
Here is the no load voltage with an input of 15. Still under control. Almost.
Too late to do anything more today, but I intend to take some more readings with the circuit before abandoning it and building the next one.
Last edited:
@thatguy. Congratulations on some groundsup DIY experimentation. I remember reading up on a similar requirement being discussed in DIYaudio. Let em search for the thread, If I find it I will post the link,
cheers
cheers
josephjohnt
Active Member
Another PSU will have its wires cut tomorrow
Did you get to sacrifice a PSU and find out if the negative voltages are required?
Not yet. Got an iBall PSU on Saturday but didn't have the time for the experiment. Hopefully next weekend.
why zener and transistor, best would be regulator chips for this kind of app. for large drops, use two different voltage levels e.g. to drop from 18 to 12, drop 18 to 15, and then drop 15 to 12, dont know how much power chips can handle though.
I just wanted to play around with a few circuit before building the real one
There are regulators with large current capacity available, but I believe they are expensive (haven't checked though). I bought a few 7812s, which apparently can do 1 amp, but have no faith the ability of to-220 to dissipate much heat. I need 5 amps and would prefer to have the current passing through tranistors in to-3 packaging.
Last edited:
Hmm ... on a quick scan of the spec, the LM340-12 doesn't look very different from 78xx. The short circuit current is 1.5A. The only thing different for the TO-3 package seems to the operating temperature ratings.
This is the spec I was looking at:
http://people.ece.cornell.edu/land/courses/ece4760/FinalProjects/s2007/aw259_bkr24/LM340T5.pdf
Last edited:
it was written 5A here Linear Series - Voltage Regulators
now i checked the national datasheet and you are right
now i checked the national datasheet and you are right
Just wanted a closure on the last circuit. Here is the Output and VB (voltage at the base), plotted against the input voltage.
The circuit does regulate the output but I wouldn't use it to run my computer. If I needed a power supply for a 12 volt fan, this would be good enough. The output also seemed to vary with the temperature, though I did not investigate it much.
The next circuit (which should be ready tomorrow) will be more interesting
The circuit does regulate the output but I wouldn't use it to run my computer. If I needed a power supply for a 12 volt fan, this would be good enough. The output also seemed to vary with the temperature, though I did not investigate it much.
The next circuit (which should be ready tomorrow) will be more interesting
Last edited:
Purchase the Audiolab 6000A Integrated Amplifier at a special offer price.
