Where is the information on the ATX v2 board?
Thanks
Trev
Sorry, that is still locked away in the workshop while we toil away learning KiCAD :) I'll see what we can post up today.
Ian: Are you going to be a permanent Eagle defector? :-) Or are you just paying around with KiCAD?
Okay, was confused. I thought the whole design process was open?
Trev
In general our process is open, but there are two limiting factors. One is time to get stuff to the wiki. Now that Arakis is back I think we'll get caught up. Also, I respect it when our team members are learning something new and don't have the same penchant for failing-while-learning in public that I do :)
Thanks for the reply, Ian. It is very true that designing something new can make you look a complete fool if done in public. lol.
Keep up the good work, DP is excellent.
Thanks
Trev
Look for this and the v1.3 update in the forum tomorrow and blog on Thursday (?). :)
One other thought about changes to V2+ --- should the ATX connector be 24pin? The PSU I scrounged had a 24 pin. With the Free PCB I hacked together it worked OK with 4 pins hanging over the edge, but would it be a better design to have a 24 pin connector? The top 20 pins on the 24 pin are the same, so it doesn't really matter which you have. I don't know if the 20 pin connector that you are using will accept a 24 pin plug?
I would like to suggest replacing the push button with a slide switch making it much easier to determine if the supply will be providing power when the ATX supply is turned on.
The ATX PSU wont turn on until you press the button, the led below the button will light if you have power at the outlet or power cord and the led s below the banana plugs will light up when there's power to the plugs.
Will this revision have bigger polyfuses? And how much current v1.2 can handle if I change (or bypass =/) the actual polyfuses?
Regards
We're going to stick with this value but maybe add a footprint for a through hole fuse. For safety sake, I wouldn't recommend removing the polyfuses.
Are you going to be using the same sick of beige case as the V1.x design?
Or are you guys moving stuff around enough that it won't work?
And would putting the resettable fuses in parallel give you 2.5A max (for 2) instead of just 1.25? If so, for folks that want it, why not just put 4 footprints in places for the fuses and they can add more as needed?
[quote author="kenyee"]Are you going to be using the same sick of beige case as the V1.x design?
Or are you guys moving stuff around enough that it won't work?[/quote]
Here's a link to the post that talks about the fact that the board in v2 is now bigger. You can download the files and check for yourself and see that the board is now 176 x 111 mm that's a lot bigger than in Eagle.
http://http://dangerousprototypes.com/forum/viewtopic.php?f=19&t=4530#p44247
[quote author="kenyee"]Are you going to be using the same sick of beige case as the V1.x design?
Or are you guys moving stuff around enough that it won't work?
And would putting the resettable fuses in parallel give you 2.5A max (for 2) instead of just 1.25? If so, for folks that want it, why not just put 4 footprints in places for the fuses and they can add more as needed?[/quote]
V2 is a bigger board, but currently it was only a proof of concept thing, there are no plans to produce it in the upcoming future...
[quote author="kenyee"]
And would putting the resettable fuses in parallel give you 2.5A max (for 2) instead of just 1.25? If so, for folks that want it, why not just put 4 footprints in places for the fuses and they can add more as needed?[/quote]
That's not a good idea, the results would be unpredictable at best, as there's no gaurantee that they will pass equal amounts of current at any given time. They are, by there very nature, a Positive Temperature Coefficient device, so when the current increases in one so it will heat up and the resistance will increase, bringing the current down again, so the current will have more of a tendency to balance out. However the effect is not very linear and in places it's actually very sharp, so it would be all over the place with two (or more) and overall just not recommended. There is no reason you can't fit a single higher rated Polyswitch/PTC thermistor or whatever you want to call it. For that matter, there's no real reason you couldn't fit a fuse, other than perhaps space and maybe footprint, but plenty of small PCB fuses that could be adapted.
[quote author="Sleepwalker3"][quote author="kenyee"]
And would putting the resettable fuses in parallel give you 2.5A max (for 2) instead of just 1.25? If so, for folks that want it, why not just put 4 footprints in places for the fuses and they can add more as needed?[/quote]
That's not a good idea, the results would be unpredictable at best, as there's no gaurantee that they will pass equal amounts of current at any given time. They are, by there very nature, a Positive Temperature Coefficient device, so when the current increases in one so it will heat up and the resistance will increase, bringing the current down again, so the current will have more of a tendency to balance out. However the effect is not very linear and in places it's actually very sharp, so it would be all over the place with two (or more) and overall just not recommended. There is no reason you can't fit a single higher rated Polyswitch/PTC thermistor or whatever you want to call it. For that matter, there's no real reason you couldn't fit a fuse, other than perhaps space and maybe footprint, but plenty of small PCB fuses that could be adapted.[/quote]
I think one of the big pro's of this design is the resettable part of it :)
Yep, I agree. There's really no reason you couldn't fit higher rated polyswitches, providing of course the supply is capable of pumping out more than that rating and of course that the board traces can handle it. That was one of the BIG problems with the Sparkfun design - pathetically small traces.
I did similar to this about 10 years ago with a PS2 supply, bolted an offcut of G-Rail (DIN rail) onto the lid, fitted some DIN Rail terminals, cut the plugs off and wired up to the terminals. That worked fine for years until eventually I overloaded the thing pulling 20+ Amps from the 5V rail and it curled up its toes (even though it was rated at 22A on that rail), but then I needed a decent current like that and polyswitches don't go that high (not that I've seen anyway).
The v2 should be fitted to a SoB standard, but we made that board mostly to try out KiCAD so consider it just a brainstorm PCB and not an actual candidate for production.
I think the parallel fuses would also have a voltage difference between them that could cause problems. Sort of like how it isn't possible to put two 5volt regulators in a circuit for twice the current, they have minor voltage differences.
It seems like one of the manufacturers of polyfuses doesn't see any problems with paralleling polyfuses. Page 5 from http://www.te.com/content/dam/te/global ... entals.pdf (http://www.te.com/content/dam/te/global/english/products/Circuit-Protection/knowledge-center/documents/circuit-protection-psw-fundamentals.pdf) states the following:
Devices in Parallel
When two identical PolySwitch devices are placed in parallel, the
hold current of the devices willincrease and the combined
resistance should be half the resistance of one of the devices.
The magnitude of the hold current increase is dependent on the
configuration of the devices and the consequent impact on the
power dissipation. If the power dissipation doubles, the hold
current will roughly double as well. If the power dissipation
increases by less than a factor of two, then the hold current for
the two devices will be less than twice that of a single
component. Two examples illustrate this:
1. Two devices are placed in parallel and are soldered to individual
individual traces that are thermally isolated from each other
(this can be done by placing the traces far away from each
other). By doing this, the power dissipation will be double that
of a single part. The resistance will decrease by half and the
hold current will double.
2. Two devices are placed in parallel and are soldered within
close proximity, perhaps on a single trace. In this case,
depending on the trace width, the power dissipation ranges
from that of a single device to double that of a single device. If
the power dissipation is the same as a single device, then the
hold current will increase by roughly 40%. If the power dissipation
is somewhere in between, then the hold current can be
approximated using the following equation:
Really interesting, I stand corrected :)
I still stand by what I said earlier, there is a lot of 'Should', 'if', 'roughly','depends' and other non-specific terms there - in short 'it depends' on many factors. As I said, the results will be all over the place and to me that isn't a good idea, I like to have a much better idea of what's going to happen when it comes to protection - just look at the MOV thread and the posts on the blog to see why!
There's no reason why you can't fit a bigger capacity unit if you want it to pass more and if you need it 'big' for one thing and smaller for another, then you could add a switch. So yes it can be done, but the results won't be accurately predictable - that's what I suggested and that's what the app note seems to be saying too - at least the way I read it.
Hey guys !
Is there still no plan to go in production for the v2 ?
Sincerely,
Jegeva