Tact switch and mosfet transistor questions

[Visus]

Visus, what project are you planning to work on if not a box mod?

TI might have some mosfets suitable for the pcb. I think they'd have to be reflow soldered since the tabs are all underneath the case.

I have some plastic cases I posted a while back in parts,  been procrastinating filling them.
08100's with 18650 finally
I have been loyal to 14500's and still are but cannot beat the runtime of 18650's using a subtank so I may as well build it.

I also want to build a mcu mod  and a box mod for my son and for that I would use your board..

Yeah really really tiny,  I came up with a really cool method of soldering  by tinning the wires positioning on top of the pads and sorta wave soldering to them makes a really solid connection and can use 18ga.  drain pretinned that area is 1mm 18 ga is ~1.5mm using tape to help mask ..  It also acts as a rudimentary heatsink..  Really tiny use of space so its a win win for aa box mod.. Ya wouldn't even know its there lol..  I got really good at soldering 0402 items and
 I actually like that style better than leads style for ghetto wire no board..   I am testing/trying,  it may not be doable without a board and reflow but why not try lol..  They have other 3x3mm and 5mm MOSFETs I also will try..

[Visus]

Lol I was answering Davids question,  sorta answered your first part lol.
I squint really good lol with a 100 watt fluro lighting it up..
Want a good 2s fet from ti n chan  but also wanna play with the tiny stuff rfol..
pls recommend one thats definitely doable wired ..


Ego
Really just want to filter the noise out, Zander gave it  a go and could not get his fet to open..   I know some ego boards are cv --no worries there,  but if the one I get is pwm, I figured a low pass filter, lil knowledge from google search was easy enough but no-- you say I need to ~know what ~frq it is, so how to filter the signal to open the fet without knowing,  that would be a better question...

[CraigHB]

Craig would IRL3103STRL N Chan 30V 64Amps work? Thanks Craig You Rock!

Thanks and you're welcome.  David4500 already answered your question, but any MOSFET in an SO-8 or compatible leadless package will fit on that layout. 

In looking for something with a curve "a little more to the left",  narrow your parmetric search to parts that have a maximal gate-source voltage tolerance of 12V over the more typical 20V.  That's the way it works with MOSFETs, the lower the maximal V_GS the further to the left the curve is.  Though the tradeoff there is tolerance to abuse so it's better to use a 12V tolerant part than an 8V tolerant part, the 12V part is a little tougher.

[fsors]

Thanks and you're welcome.  David4500 already answered your question, but any MOSFET in an SO-8 or compatible leadless package will fit on that layout. 

In looking for something with a curve "a little more to the left",  narrow your parmetric search to parts that have a maximal gate-source voltage tolerance of 12V over the more typical 20V.  That's the way it works with MOSFETs, the lower the maximal V_GS the further to the left the curve is.  Though the tradeoff there is tolerance to abuse so it's better to use a 12V tolerant part than an 8V tolerant part, the 12V part is a little tougher.

Thanks Craig! Any recommendations on a specific mosfet from you for this board would be highly valuable to me and very much appreciated. Another question, is do I need to use 3 mosfets on the board? Thanks in advance Craig! You Rock!     

[CraigHB]

Lol I was answering Davids question,  sorta answered your first part lol.I squint really good lol with a 100 watt fluro lighting it up..Want a good 2s fet from ti n chan  but also wanna play with the tiny stuff rfol..pls recommend one thats definitely doable wired ..

TI does make some really nice N-channel MOSFETs.  Mostly they're the 5x6mm leadless packages, but there's a number of smaller ones too.  Weird though, they have few P-channel ones.  They do make one really nice PMOS in a 3x3mm package, but that's the only noteworthy one they make.  In any case, the small stuff is do-able by hand, I use parts like that all the time.  Though it's taken me a lot of soldering practice and skill to get there.  Preferably you can use a hot air station or even frying pan or toater oven reflow.  Mainly the trick is being able to see what you're doing.  I use a 10x stereo microscope, but there's options there as well.

Ego
Really just want to filter the noise out, Zander gave it  a go and could not get his fet to open..   I know some ego boards are cv --no worries there,  but if the one I get is pwm, I figured a low pass filter, lil knowledge from google search was easy enough but no-- you say I need to ~know what ~frq it is, so how to filter the signal to open the fet without knowing,  that would be a better question...

I know there's a schematic posted for the eGo board around here somewhere, but I don't feel like taking the time to look for it.  An ideal time constant to allow 33Hz and block higher frequencies would be 3mS.  If this on a gate drive you can't put too much resistance inline or you'll slow down the switching speed of the MOSFET too much causing it to overheat.  In the extreme like 10K Ohms the MOSFET won't even turn on at all. 

With the limitation on resistance you'll end up with a large capacitance to make your filter ideal since the bandpass frequency is so low.  What you'll have to do is use the highest resistance you can get away with and the largest capacitance you can comfortably put on the PCB.  Probably like 100 Ohms and 10uF which gives you a time constant of 1mS.  That's pretty close to the ideal 3mS.  That should filter most of the electrical noise you may run into.  100 Ohms might still be too high, probably 50 Ohms would be better dropping the time constant to 500uS.  It will still catch a lot of the noise since it tends to be higher frequency stuff anyway.

[Visus]

Thanks and you're welcome.  David4500 already answered your question, but any MOSFET in an SO-8 or compatible leadless package will fit on that layout. 

In looking for something with a curve "a little more to the left",  narrow your parmetric search to parts that have a maximal gate-source voltage tolerance of 12V over the more typical 20V.  That's the way it works with MOSFETs, the lower the maximal V_GS the further to the left the curve is.  Though the tradeoff there is tolerance to abuse so it's better to use a 12V tolerant part than an 8V tolerant part, the 12V part is a little tougher.

Woot answered my question too
I was unsure of my choices because of the 12v 20v etc I can read the bar graphs but still shaky legs lol..
Its why mostly everybody just goes with 3034 so many variables assumed when selecting a fet..  But its a good fet...

I cancelled my ego purchase lol..  Not even gonna mess around with it until I build the box mod an I can use my sons  ego board, also check it to know what its doing. 

Thanks Craig Never mind the ego part I will ask ya later Im more than sure when then...

[Visus]

We keep posting at the same times lol

Thank ya very much, theres actually a spot open on the ego board for a resistor and a capacitor on the output so it would be an easy task too..   

[CraigHB]

Oh, I have to add a correction, the time constant is simply RC not 1/RC, spaced out on that, the step response for a low pass filter is Vin times one minus the natural log of -t/RC so I thought 1/RC for some reason.  It's actually still meaningful, it's just the time constant in terms of frequency not period.

[CraigHB]

Thanks Craig! Any recommendations on a specific mosfet from you for this board would be highly valuable to me and very much appreciated. Another question, is do I need to use 3 mosfets on the board?

Well, here's a couple NMOS with a low threshold and low on-state resistance;

http://www.ti.com/lit/ds/symlink/csd17312q5.pdf

http://www.vishay.com/docs/64815/sir404dp.pdf

Those are leadless SO-8 packages.  They can actually be soldered by hand, I do them all the time, just use a lot of flux under the pads to make sure you get good flow and plenty of heat on the iron, 650F will do with 63/37 solder.  Otherwise you can reflow them with a hot air pencil or even in a frying pan or toaster oven.

There may be some regular S0-8 ones worthy of looking at, but I usually don't bother with them because all the best ones are coming out in those leadless packages.

You could probably run just one if you wanted, but the idea is to lower on-state resistance.  There's really nothing to be gained by running fewer MOSFETs except cost of parts, but just two will still halve the on-state resistance and be plenty low enough for all but the highest outputs.

[fsors]

Well, here's a couple NMOS with a low threshold and low on-state resistance;

http://www.ti.com/lit/ds/symlink/csd17312q5.pdf

http://www.vishay.com/docs/64815/sir404dp.pdf

Those are leadless SO-8 packages.  They can actually be soldered by hand, I do them all the time, just use a lot of flux under the pads to make sure you get good flow and plenty of heat on the iron, 650F will do with 63/37 solder.  Otherwise you can reflow them with a hot air pencil or even in a frying pan or toaster oven.

There may be some regular S0-8 ones worthy of looking at, but I usually don't bother with them because all the best ones are coming out in those leadless packages.

You could probably run just one if you wanted, but the idea is to lower on-state resistance.  There's really nothing to be gained by running fewer MOSFETs except cost of parts, but just two will still halve the on-state resistance and be plenty low enough for all but the highest outputs.

Thanks Craig! I have booked those for tomorrow. The ti looks particularly interesting. Thank You for sharing your knowledge and kindness! 

[CraigHB]

I actually like the Vishay one better.  Though the on-state resistance is a little higher, it has 12V tolerance on Vgs and it has a very left curve that is well into the flats at 3V.  Also, the Vishay package has tits on the leads making it easier to solder.  The TI package is ground flat on the lead sides.

[fsors]

I actually like the Vishay one better.  Though the on-state resistance is a little higher, it has 12V tolerance on Vgs and it has a very left curve that is well into the flats at 3V.  Also, the Vishay package has tits on the leads making it easier to solder.  The TI package is ground flat on the lead sides.

Well that makes me like it best then too! What resistor would it take? Should I use 3 of these fets? Thanks Craig! 

[CraigHB]

The short answer is a single resistor between 4.7k and 47k should be fine.  I'd probably use a single 10k pull-down resistor on those three MOSFETs.

Here's the long answer.  The pull down resistance is not particularly critical, nor it is critical to pull each gate down individually with three separate resistors.  It's fine to tie the gates together and pull them down with one resistor. 

One consideration for pull down (or up) resistors and MOSFETs is the gate-source leakage.  Typically it's in the nano-Amps.  If you go high enough, and I mean really high like 10meg, the leakage can cause a voltage drop high enough to partially turn on the MOSFET.

Another consderation is noise sensitivity.  When resistances are high enough, voltages can be picked up from currents running on the board and even from EMI radiated by the power grid or radio.  For things sensitive to only a volt or two it can cause problems.  Again this only for really high resistances over several mega-Ohms. 

Finally there's the turn off and turn on times for the MOSFET.  When pulling the gate up, resistance is only limited by the source (the battery) and current paths which is quite low, but when the gate is pulled down, charge stored in the transistor's gate-source junction has to dissipate through the pull-down resistor.  For high resistances and/or MOSFETs with high gate charge, that can cause the MOSFET to shut down slowly generating unnecessary heat and power loss.  Again this is going to be for higher resistances, probably above 100k, but it also depends on gate charge.  With the three MOSFETs in parallel you are tripling the gate charge.

You may want to limit the current that flows by using a higher resistance, but typically for a device that draws as much power as an atomizer, there's not much concern over power consumption in peripheral circuitry.  Using a milli-Amp to drive the MOSFET is not a concern.  Though for some battery powered devices other than e-cigs it can be.

[david4500]

Thanks for all the wisdom Craig.

That Vishay mosfet looks interesting. Think I'll give it a try with my next parts order.

[david4500]

No on-board fuses this time.

https://oshpark.com/shared_projects/PqyAD9aT

[fsors]

The short answer is a single resistor between 4.7k and 47k should be fine.  I'd probably use a single 10k pull-down resistor on those three MOSFETs.

Here's the long answer.  The pull down resistance is not particularly critical, nor it is critical to pull each gate down individually with three separate resistors.  It's fine to tie the gates together and pull them down with one resistor. 

One consideration for pull down (or up) resistors and MOSFETs is the gate-source leakage.  Typically it's in the nano-Amps.  If you go high enough, and I mean really high like 10meg, the leakage can cause a voltage drop high enough to partially turn on the MOSFET.

Another consderation is noise sensitivity.  When resistances are high enough, voltages can be picked up from currents running on the board and even from EMI radiated by the power grid or radio.  For things sensitive to only a volt or two it can cause problems.  Again this only for really high resistances over several mega-Ohms. 

Finally there's the turn off and turn on times for the MOSFET.  When pulling the gate up, resistance is only limited by the source (the battery) and current paths which is quite low, but when the gate is pulled down, charge stored in the transistor's gate-source junction has to dissipate through the pull-down resistor.  For high resistances and/or MOSFETs with high gate charge, that can cause the MOSFET to shut down slowly generating unnecessary heat and power loss.  Again this is going to be for higher resistances, probably above 100k, but it also depends on gate charge.  With the three MOSFETs in parallel you are tripling the gate charge.

You may want to limit the current that flows by using a higher resistance, but typically for a device that draws as much power as an atomizer, there's not much concern over power consumption in peripheral circuitry.  Using a milli-Amp to drive the MOSFET is not a concern.  Though for some battery powered devices other than e-cigs it can be.

Thanks Craig You are a asset to the community and are a Master with a wealth of knowledge! So if I tied them I can pull them down with a 15K that's cool! I will go that way. Do I have to use fuses on this board?........I probably will use the fuses anyway since there are spots for them but just want to know if it's by-passable?   

[fsors]

No on-board fuses this time.

https://oshpark.com/shared_projects/PqyAD9aT

david4500 those look cool too and I might have to order some after the first set gets done which I just got an email about. They say they will be back from the printer around March 13th. Look forward to building with them. 

[CraigHB]

Do I have to use fuses on this board?

It's never a bad idea to use fuses, but they do introduce additional power loss.  For parallel removable cells it's rather sketchy to run without some kind of reverse polarity or charge mismatch protection.  For anything powered by a Li-Ion battery there really has to be some kind of protection from excessive currents.  They're just too hazardous otherwise.

I don't use fuses myself, but cells are permanently installed and an MCU monitors current and temperature so the system knows if there's a problem.  Even then it would not be a bad idea to add fuses as a backup, I don't because I'm covered and I don't want to introduce the power loss.

[dc99]

Quick question, is the heavy solder enough?

[david4500]

Do I have to use fuses on this board?........I probably will use the fuses anyway since there are spots for them but just want to know if it's by-passable?   

Yeah I would use them since there is room on-board. Use the new fuse free board if you wish to not to use them or have them off-board.

But if the fuses were to be omitted, run a wire from 510+ or battery+ and connect to "510+" on the board.

If the on-off switch was going to be omitted as well, connect that wire to "SW1-" instead of "510+".

[Visus]

Craig I found something of interest and wanna know why you two EE's agree to disagree,  are mosfets that much better now, more closely matched?  The post is from 2011 and thats huge past days  for todays electronics, that is mindboggling cool..

Heres your quote:

Here's the long answer.  The pull down resistance is not particularly critical, nor it is critical to pull each gate down individually with three separate resistors.  It's fine to tie the gates together and pull them down with one resistor. 

Heres the post of interest of this..  It also confirms your post about MOSFET choices further to the left on the graph as heat rises RDson grows so ya..

MOSFET RDS(on) has a positive temperature coefficient - the warmer it gets, the higher the RDS(on) gets. If you connect two MOSFETs in parallel with similar characteristics (i.e. the same part number from the same manufacturer), drive them identically, and don't have huge asymmetry in your PCB layout, the MOSFETs will indeed share current quite nicely. Always make sure each MOSFET has an independent resistor in series with each gate (never parallel gates without resistors) as gates tied directly together can weirdly interact with each other - even a few ohms is better than nothing.

http://electronics.stackexchange.com/questions/16882/dissipating-1w-on-a-to-220-without-heatsink

I would tell guy my instructor said its not critical but I have read where you also said the same in a post but that was for pwm circuit .... So ya why agree to disagree..

[fsors]

Yeah I would use them since there is room on-board. Use the new fuse free board if you wish to not to use them or have them off-board.

But if the fuses were to be omitted, run a wire from 510+ or battery+ and connect to "510+" on the board.

If the on-off switch was going to be omitted as well, connect that wire to "SW1-" instead of "510+".

Thanks david4500! I like these kind of options! 

[CraigHB]

Craig I found something of interest and wanna know why you two EE's agree to disagree,

You can model the gate source connection with a capacitor and resistor in series.  For matched parts, gate charge is similar enough that gates will charge equally and MOSFETs will turn on equally.  In that case it's two of the same resistors and two of the same capacitors in parallel.  Currents will equally flow between them.  For non-matched parts, tying gates together could result in one part carrying the bulk of the load during on-off switching.  That can be a problem in terms of unequal loading and asymmetric heating.

Furthermore, a user on-off switch is not the typical application for a Power MOSFET.  They're actually used for all kinds of different switching applications and that can have a big impact on design considerations and part selection.  For higher speed stuff like PWM or even converter switches, it's not uncommon to parallel MOSFETs to obtain higher current ratings.  Motor controllers are a good example.  High output ones can have tens of MOSFETs running in parallel.  Equal loading of the gate-source junction is critical.  Even variations between parts coming off the same line can be an issue.  However, for a user on-off switch, this does not become much of a consideration since the speeds are pretty slow, unless you plan to cycle the switch as fast as humanly possible.

Hope that addresses your confusion.

[Visus]

Thanks,

Wasn't confused,  I woulda argued with guy in class per se'  but needed my argument.   

Past few days I've crammed enough in my head to explode on some breadboards lol

[CraigHB]

Well, eh, question then.  Hopefully that made plain sense to you, basically the lower the switching frequency the less of a concern there is.  A user on-off switch is about as low a frequency as you can get.  And I can relate to exploding head syndrome, get it sometimes myself.  Problem is the more you know the more you realize you still need to learn and then when you want to try some new circuit you try to understand it all in one sitting.  I call it burning brain cells.  Programming actually gets me worse than circuit design. 

[Visus]

Well, eh, question then.  Hopefully that made plain sense to you, basically the lower the switching frequency the less of a concern there is.  A user on-off switch is about as low a frequency as you can get.  And I can relate to exploding head syndrome, get it sometimes myself.  Problem is the more you know the more you realize you still need to learn and then when you want to try some new circuit you try to understand it all in one sitting.  I call it burning brain cells.  Programming actually gets me worse than circuit design.

Lol ya its awesome learn mosfets, controllers, computer, and control everything lol...
They are producing new items that change the way it was done/did daily.  I love it, efficiency is awesome.  Example: If Tesla home batteries are as amazing as posted many things are going to change..  But whoa the future incoming faster than the economy can handle so ya,  you engineers rolling hot in the brain cells..

[david4500]

In a spacious 1590B

[doobedoobedo]

Very tidy.

[Visus]

Wow that looks good
love ur switch choices the led in em is baller

[dc99]

http://www.mouser.com/ProductDetail/STMicroelectronics/STD100N10F7/?qs=sGAEpiMZZMvMXeU%252bpcMAVNQG84RiK2Um
Any reason this wont work? Its available in almost every package
Im still looking for the elusive p-chn that is easy to work with. http://www.mouser.com/ProductDetail/Infineon-Technologies/IPD042P03L3-G/?qs=sGAEpiMZZMvplms98TlKY8VID4ASP18oAUQnIRM03p0%3d

[david4500]

For the nfet, the rds on is too high. You'll want around 2 milliohms. Vgs th is also slightly high, 4.5v. You want it to be less than 2.5-3v. If for any the smd pcbs, of the packages shown in the data sheet they are all too large.

I think Craig has posted some well suited pfets in the past if you look through his posts.

Got the quad board today:



When soldering the drain tabs on the back, I didn't realize the solder was flowing though the vias to the other side.

[fsors]

Very cool david,  will the solder flow through mess with anything? 

[fsors]

david do you think I can get one of these boards in this size package?

[CraigHB]

I think Craig has posted some well suited pfets in the past if you look through his posts.

I'll save you the trouble of searching, this one is the best PMOS I've come across for a single cell or parallel cells;

http://www.vishay.com/docs/62860/si7157dp.pdf

I've mentioned that part a few times already on the forum here.  I use this part myself and I've measured 2 mOhms at nominal voltage for a single battery.  This one has lowest on-state resistance and lowest gate-source threshold of any PMOS I've come across.  Granted you might be able to find a PMOS with lower on-state resistance and higher gate-source threshold, but it's doubtful.

Not that it matters for a user on-off switch, but FYI, notice the incredibly high gate charge on that PMOS.  You really couldn't use it for much else besides a user on-off switch, or something that runs at a very low frequency like a 33Hz PWM driver.  The gate charge on that one is like a 1uF capacitor which is a huge amount of capacitance for a MOSFET gate.  It has a very large gate junction, probably covers the entire channel.

[david4500]

Very cool david,  will the solder flow through mess with anything? 

Something to be aware of... just be little less generous with the solder than I was. I kept feeding solder not realizing it was going through to the other side. I can clean up the blobs with some solder wick if I wanted.

david do you think I can get one of these boards in this size package?

TriParaMos in a Hammond 1590G:


TriParaMos-NF (no fuses)
In a 1590G:



1590P:


1590A:

[fsors]

Awesome david Many mounting options really look forward to getting the boards. 

[CraigHB]

I can clean up the blobs with some solder wick if I wanted.

I use the hell out of that solder wick.  I wouldn't be able to do any PCB work without it.  I usually keep a couple rolls of the #1, #2, & #3 sizes on hand.  It's a crisis if I run out.

[dc99]

Question, could I use the boards David designed for the SI7157DP? I worked on a board for two days and decided Im not smart enough to design boards.
I was going to post a picture of how far I got and ask for help but

[CraigHB]

I think all the boards he's posted switch the low side requiring an NMOS.  The SI7157DP is a PMOS so you need a board designed for a high side switch.

What software were you using to design a PCB that you had too much trouble with?  There's other free PCB design programs that are easier to use, KiCad is one, gEDA is another. 

Eagle is one of the hardest programs to learn to use, it does things in a way that's not similar to other computer programs.  The reason is that it started out as a DOS command line program.  They never totally redesigned the program for Windows, only added buttons and menus in the graphical interface to automate the scripts the old DOS program used.  You can actually still use all the original DOS commands in the command line box.

[dc99]

ExpressPCB, My biggest problem is that I really don't understand what is required and Im not ashamed to admit it. The board design stuff is driving me nuts. Not like theres anything out there that's really a lot of help. Its not that I cant put the parts in the design, its more of how big a pad to use and such. If there anything that will help Im up for learning. I just cant figure out how because everything Ive seen either assumes your an engineer or a genius

[Visus]

Right there with ya dc99 I loaded up PCB for linux played with it for a few hours and watched few vids on yt.
im linux savvy to advanced level but whoa that thing is stupid hard...

It takes 20 operations just to do simple task such as use the library and make your own files for layers,  it is a rudimentary pcb designer thats made for the people who made it.  lol

Any suggestions for linux Craig?

Checking the other two compatibility now..

Woot Kicad good to go for fedora ..  going to give it a whirl..
Its called electronic lab for linux but its kicad with add ons

Anyone else with linux its in yum dir
# yum groupinstall 'Electronic Lab'
It supposedly has everything in library (userloaded) to design and simulate anything up to date 2015 in electronics to micro-nano particle research

Exactly what I was looking for an extensive library

Thanks Craig

[CraigHB]

Welcome, gEDA was originally written for Linux, but now has Windows ports for some of the components.  KiCad is highly regarded. I believe it runs natively on Linux though there is a full Windows version as well. 

@dc99, most PCB software provides a WYSIWYG interface meaning you can put up a "snap to" grid that reflects actual measurements.  I use an inexpensive digital micrometer I got off eBay and it's surprisingly accurate.  I'll measure my parts then base my PCB drawings directly off those measurements.  It's something I do quite often. 

For example, you can measure the thickness of a wire or component lead to determine the size hole you need.  I usually make the pad wall equal to the hole size so the overall pad diameter is three times the hole diameter.  You'll want to add a little to the hole diameter for production tolerance.

Generally it's better to use parts drawn yourself than rely on canned libraries.  They can have inaccuracies or fail to meet needs for the type of soldering you do.

If you're using ExpressPCB software it's going to be somewhat limited, though it's very easy to use.  You'll do better if you invest the time to learn a general PCB design tool i.e. Eagle, KiCad, gEDA.  ExpressPCB is a proprietary program not able to output the standard Gerber and Excellon plot files used by most PCB fabs.  It's also limited in output options for printing templates and making your own PCBs.

[CraigHB]

Oh, @Visus, if you're talking about circuit simulators, there are some free PCB design packages that include them.  One is Design Spark from RS Electronics (I believe they own Farnell) and another is MultiSIM Blue from Mouser.  I haven't tried the simulators on those so I don't know how well they work.  Personally, I use OrCAD PSpice which is not a free program, it's a professional package and expensive, I managed to get a license for it from an engineer buddy.  I also use LTSpice which is very good for what it's designed to do, mainly analog stuff.  That one is free.

In terms of circuit simulators native to Linux, I'd have to look them up.  I have used Linux in the past, but typically I run everything on Windows.  Linux gets too time consuming to configure, though it does work a lot better than Windows once you get it set up. 

It is handy to have everything in one suite, but for example I don't care for OrCAD's PCB design software, but the simulator is good.  Sometimes it's better to pick and choose components than use a single suite.

[Visus]

Woot already drew my 1st schematic input symbols and  library components,  footprinted, and populated a pcb, moving along nicely.  My brain is tired now lol..  Its an awesome program really well done, quite easy  for a noob compared to the other program.

Doing a course from contextual electronics on kicad called, "getting to blinky" using a 7555 to blink a led. 

I had no idea it would draw the connections and vias even insert proper package dimensions on its own after the schematic was made and/or added library symbol parts that usually have a similar footprint in the cvpcb program.  I was doing it arse backwards just wanted to define layer, draw pads,  and connect the lines..  I found it impossible to do lol..

Awesome I figure in a few weeks ill have plans for synthetic food made from excrement -- no really someone in Japan actually designed a food machine that makes synthetic meat from excrement..  It's not that far fetched if you think about what fertlizes plants and then we eat the plants, the cows, pig, goat,  sheep, etc. eat the plants ewwww...lol A lot fish only eat other fish lol..  Hmn Jesus was on to something there, but stones in glass house,  he intelligently called us a bunch of neanderthal nasty eating crap mofo's in a way...     He was/is wayyy too smart lol..

Woot Thanks again,  Love this stuff..

[dc99]

Downloaded KiCad. Seems way easier. Im still a little dumb at it but I think I can figure this one out. Thanks Craig.

[david4500]

Hey dc, if kicad doesn't work out for you check out Diptrace. I downloaded multiple pcb programs and was clueless. Wasn't until diptrace I was able to jump right in and start making boards. The only thing I had to look up in the tutorial was how to make a custom pattern/footprint.

[Visus]

Dc your gonna love it,  I already have a knack for it.  Just watch and follow along with a yt tutorial on it and it's easy.

This guy takes ya along to build a blinking led and when ya get done with that.
https://www.youtube.com/user/contextualelectronic

Go to Nicholas next,  Nicholas will simplify what Contextual made ya do from scratch.
It makes it click in the head fast.
https://www.youtube.com/channel/UCctMiEpCvXzipAYyB-RjmYg

For really cool 1st time review use from an engineers point of view.  Dave from EEVblog reviews the program,
he actually had a tough time with it and then boom he had it down in basically 3 hours...  Powerful package was his impression and that was from 2012 they have refined it, so it's even better today.
https://www.youtube.com/watch?v=bg0sEjD7R6M
-----------------------------------------------------------------------------------------------
Craig the electronic suite actually comes with maybe a hacked similar program called gnuspice and three other simulators,  it was a 4gb download, it was huge almost 7gb total install, took forever.  It loaded many programs along with Kicad..   I probably will never use most of them;  even chemistry and petri dish electrical simulations..  That area of building for me is a good ways down the road, if  lol...

Today I did learn I could just draw a simple pcb and no need for schematic if I did not want to use one, woot..  Had to learn to use the editors properly and I have those semi- down now...

[CraigHB]

Yes there's two ways to go about it.  You can start with a schematic and have the program "rubber band" a board for you.  Then you can manually draw out the traces or use the autorouter.  That aspect of a PCB program is one that comes under a lot of criticism.  Some autorouters can do a good job for you and others fall down on the job.  The quality of a PCB suite's autorouter is usually a big point in terms of how well software scores in a review.

The other way to go is to start with a blank board and add your own components and your own connections.  I tend to jump right to the board for most stuff, though for large designs, doing it the automated way can save a lot of time.  It just depends on what you feel more comfortable with.  Either way works.  In the professional world it's typically done in an automated fashion because of the big reduction in the time it takes to lay out a board.  For hobby stuff it's your time and up to you how you use it.

[CraigHB]

Craig the electronic suite actually comes with maybe a hacked similar program called gnuspice and three other simulators,  it was a 4gb download, it was huge almost 7gb total install, took forever.  It loaded many programs along with Kicad..   I probably will never use most of them;  even chemistry and petri dish electrical simulations..  That area of building for me is a good ways down the road, if  lol...

Yeah that's the thing with Linux, you can probably find ten different programs for any one thing you're trying to do.  Though the one issue wth open source stuff is there's no real assurance of any quality, the software can be the best thing since sliced bread or a total POS.  You sometimes have to kiss a few frogs, I've had that happen before with open source stuff.  In any case, if you find a good circuit simulator for Linux, then great.  For most of the stuff I do as a hobbyist, LTSpice is a very good simulator and it's free so I just use that.  For things LTSpice can't do, I use OrCAD PSpice which is not that often.

[dc99]

Is the TPS3700DDCR suitable for a low voltage shut-off on an unregulated box? I remember a wiring diagram somewhere on here but I cant find it.