Re: PCB making software, What do you use?
|
Re: PCB making software, What do you use?
I happen to use EAGLE, because I had (and still do) a hobby license.? That practice ended starting at 8.0 when Autodesk bought the program and the program bought the farm.
Since I detest the idea of "rent a software", I have never upgraded.? Autodesk's suggestion is "go do the subscription program anyway".
However, their latest versions only do something on the order of 3in by 4in boards at the max, with a few other limitations that say "NO".
You can parse the eagle board file and extract enough information to make an OpenSCAD model of the board, using the preview in OpenSCAD to visualize the board.? You could also print a model of the board (and import it for OpenSCAD designs) for board fit.
I've considered using Kicad, but the lack of an autorouter tends to be a bit much.? I'll stick with EAGLE since it's paid for. I've switched board houses from SeeedStudio to JLCPCB because of their price on 4X4 4 layer boards.
It's really what you want to use, and if the tools match your needs.
Harvey
toggle quoted message
Show quoted text
On 11/21/2022 11:02 AM, Paul Amaranth via groups.io wrote: For historical reasons, I use the gEDA suite and it works
pretty well. I just made up a VHF amplifier using it
and JLCPCB. Using the slower shipping option, it was something
around $10 for 5 boards that got delivered in a week. $8
of that was shipping. That's a bargain in my book.
The gEDA suite has strong community support, but KiCAD is
a bit more mainstream with institutional sponsors. Either
is highly capable and free.
There's a learning curve with all of these cad programs and
the biggest issue is having the physical footprints available
for whatever parts you are going to use. You are going to
need to be able to make your own footprints in order to
make a reasonable board. And, be sure you have the correct
footprints before you route the board.
Pick a good professional level tool and you only have to
learn it once. Pick a hobby level tool that ties you to
a single vendor and you're at their mercy. Pick a cloud
based service and cry when it disappears.
Paul
On Mon, Nov 21, 2022 at 07:20:05AM -0800, Mikek wrote:
I have made pcbs with software but it has been many years since I last used a program.
I want something that will do surface mount as well as through hole. Fast learning curve is a big plus!
Also Free!
I don't have anything very complex, 2, 3, or 4, transistors maybe later I'll try some SMD ICs.
What do you recommend.
BTW, there used to be programs where you could get 3 pcbs for $xx.00,
but I'm not finding any hits for those.
Mikek
!DSPAM:637b972935241052014296!
|
Re: PCB making software, What do you use?
Hi, ? I use DipTrace. ? They have different purchase options including free. ? I have used lots of Schematic Capture/PCB Layout and Simulation packages over the years and I purchased a full functionality, unlimited pins, Non-profit (Hobby), seat in June 2021 for $375 US. I have created lots of PCBs with it. ? Like most of the other free or low-cost packages out there, it is an easy to learn integrated package with Schematic, PCB Layout, Library Editor, and good Tutorials. It has a large library and includes an interactive 3D modeling of the board with components on it. The 3D Tool can be quite helpful to make sure everything is correct. ? Like all of the other tools out there, each one has its strengths and weaknesses. I will just give my impressions of the tool. I have no affiliation with DipTrace, just a customer. ? ? Schematic Capture Tool The Schematic Capture tool is okay. You have to know what parts you want to use for your design including discrete components like Resistors, Capacitors, Connectors, etc. It expects you to know something about the parts you want to use at the Schematic design. You can¡¯t just add a 10k 1/4W 1206 SMD Resistor or a 0.01uF 50V X7R 0805 Capacitor. You have to select a Resistor, then the package, then the Value. You can go back later and change the characteristics later - I just think it is a little slower and annoying to do that. Be careful, just changing values, you can get bit sometimes. ? The Component Library is arranged by component type (Resistor) > component package (Axial) > Vendor Part Number. You have to look at the Datasheet to select or make sure you have what you want. Which as I mentioned above is a little slow and annoying. ? As warped as I am, I would prefer to put down a 10uF 50V 5x10mm Low ESR Radial Capacitor this way¡ Capacitor > Value > Type (Low ESR) > Package > Part Number. Even assigning the package later just before or during PCB Layout would be better for me. *** This order may be totally different from what someone else would do. *** ? When you go to the Properties for the Schematic Symbol or Component, it provides a link to their Datasheet database and a Mouser Part Number ¨C if available, (it used to be Digikey ¨C I prefer Mouser), for each component. That helps some. ? ? PCB Layout Tool The PCB Layout Tool is okay. It transitions from the Schematic Capture Tool seamlessly is my experience with it. The Auto-Placement Tool will just stack all components to one side of the board ¨C not very useful at all. I don¡¯t use it. ? It would be nice to have intelligent placement rooms. I just do that outside the PCB border and then move them into the PCB area with the Net lines then place the blocks per Signal Integrity/Power Integrity desires. This is a very common design process. Interconnectix (ICX), was the first tool that I used that was a Signal Integrity Based Router that really worked. Then Mentor bought them out. ? The Auto-Router is about the same as many others. It will allow you to assign some properties to the nets, allows a DO-File, you can group Differential Pairs, etc. I gave it a cursory check and it seem fast enough, although I hand-route my boards. ? The version I purchased allows unlimited layers. The highest layer count I ever worked with Mentor Expedition was 30 Layers before PCB Thickness and Via aspect ratio became a serious problem. I doubt that I will go over twelve for any hobby PCBs I¡¯ll ever do with DipTrace or any tool. ? The Planes Processor works well with DipTrace. It can be poured and removed and seems fast enough. You can assign different Pad sizes and Pad/Route keep-outs easily. Of course, you need to process the planes any time a component is placed once it has already been placed, then a final one before Gerber Translation. I say this because moving components is an easy thing to do and you just need to make sure the planes provide the proper return path for traces. ? ? Library Editor Tool The Library Tool is a little different. The best way I found to use it was to use an existing Symbol/Cell/Footprint, modify it and save it into your own directory within the tool. I had to make a bunch of 74AS Parts once and that was the easiest for me. I did the same thing with an Arduino board too. ? ? The Gerber Import/Export Tool is okay. The Layout Database to Gerber Tool is a little different. You need to follow the Tutorial/on-line content to get it right. ? The Silkscreen Tool works okay, although, I have had problems with all of the text not making it into the Gerbers. You just have to check to make sure it all gets there. ? ? The 3D Tool is okay. As mentioned above the 3D Tool works well. The Print function was a little different but easy to use. ? ? Other Updates are free and they update it often ¨C which is good. No silly ¡°Maintenance Fee¡± like the Big Boy tools. ? It has good Import / Export functionality. It can import / export just about any smaller layout package products (Altium, OrCAD, Eagle, PADS PCB, P-CAD, KiCAD, Netlists), you would want except the bigger Mentor, Cadence or Zuken products. ? The Tool also helps you order PCBs from a couple of vendors. It gives you Bay Area Circuits and PCBWay for an instant quote and placing the order. I just use PCBWay. I haven¡¯t used it through DipTrace. PCBWay has only had a problem with the Drill Table once ¨C but it lets you know you need to fix it before it lets you order PCBs. ? The Support Staff seems good and the Tutorials/Videos are well written and quick to the point. ? ? Anyway, I like DipTrace and I was glad I decided to purchase it. ? Ross Wellington ?
toggle quoted message
Show quoted text
From: [email protected] <[email protected]> On Behalf Of Donald H Locker via groups.io
Sent: Monday, November 21, 2022 8:30 AM
To: [email protected]
Subject: Re: [Test Equipment Design & Construction] PCB making software, What do you use?? KiCAD. Free, open-source. Incredibly capable; great support crew; moderate learning curve. Diptrace. I've not used it in years, but it was fairly easy way back when. I don't know cost; used to be free. Have a look at for some other packages to consider. HTH,
Donald. On 11/21/22 10:20, Mikek wrote: I have made pcbs with software but it has been many years since I last used a program.
?I want something that will do surface mount as well as through hole. Fast learning curve is a big plus!
?Also Free!
I don't have anything very complex, 2, 3, or 4, transistors maybe later I'll try some SMD ICs.
?What do you recommend.
BTW, there used to be programs where you could get 3 pcbs for $xx.00,
?but I'm not finding any hits for those.
?????????????????????????? Mikek
|
Re: PCB making software, What do you use?
For historical reasons, I use the gEDA suite and it works
pretty well. I just made up a VHF amplifier using it
and JLCPCB. Using the slower shipping option, it was something
around $10 for 5 boards that got delivered in a week. $8
of that was shipping. That's a bargain in my book.
The gEDA suite has strong community support, but KiCAD is
a bit more mainstream with institutional sponsors. Either
is highly capable and free.
There's a learning curve with all of these cad programs and
the biggest issue is having the physical footprints available
for whatever parts you are going to use. You are going to
need to be able to make your own footprints in order to
make a reasonable board. And, be sure you have the correct
footprints before you route the board.
Pick a good professional level tool and you only have to
learn it once. Pick a hobby level tool that ties you to
a single vendor and you're at their mercy. Pick a cloud
based service and cry when it disappears.
Paul
toggle quoted message
Show quoted text
On Mon, Nov 21, 2022 at 07:20:05AM -0800, Mikek wrote: I have made pcbs with software but it has been many years since I last used a program.
I want something that will do surface mount as well as through hole. Fast learning curve is a big plus!
Also Free!
I don't have anything very complex, 2, 3, or 4, transistors maybe later I'll try some SMD ICs.
What do you recommend.
BTW, there used to be programs where you could get 3 pcbs for $xx.00,
but I'm not finding any hits for those.
Mikek
!DSPAM:637b972935241052014296!
--
Paul Amaranth, GCIH | Manchester MI, USA
Aurora Group of Michigan, LLC | Security, Systems & Software
paul@... | Unix/Linux - We don't do windows
|
On Mon, Nov 21, 2022 at 07:38 AM, Alan wrote:
I did say PILES, piles are produced when things ( wires) are layered on top of others, that is pile winding. Pies are baked pastry encased food. My callsign ( G8LCO) tells that I am in England so I speak English.
?Ok, I'm sure you speak the proper English, we will just have to agree we use different words for the same thing. Instead of piles we would say layers. But having multiple sections on the same form, we call those pies. Plural of pi, (I think? :-) If I search pi wound inductor with google I get a lot of pi wound inductors, searching pile wound inductors gets, me just a bunch of conventionally wound inductors. ??????????? Its all fun! ???????? Mikek
|
Mikek
I did say PILES, piles are produced when things ( wires) are
layered on top of others, that is pile winding. Pies are baked
pastry encased food. My callsign ( G8LCO) tells that I am in
England so I speak English. The Piles in your picture are wave
wound, that normally means cotton covered wire. I have used an
automatic winding machine to wave wind inductors years back, an
Uncle had a coil winding business when I was a kid. Cotton covered
wire layers more readily than enamel, it is also a slightly
thicker coating than enamel so has less self capacitance. Cotton
covering is hard to find now, large wound parts are very much out
of fashion and there are very few machines that can wind an
insulator ( cotton or silk) onto thin wires. Such machines can be
tempremental to run, with labour costs and demand being what it is
covered wire is now hard to find. A few years back the last wire
coating winder in the UK fell out of use because the lady that ran
the machine retired. My Wife remembers that happening? because she
used to sell double silk covered silver wire in her business.
Your line that self capacitance is always there is absolutely?
correct BUT it needs to be factored into any LC calculation. That
is why the multiple frequency method is used, the resonance curve
is offset by the self capacitance, plotting the curve back to
Cext=0 leaves the value of the self capacitance.
The self capacitance is distributed over the turns depending on
the construction the distribution may vary. Pile winding allows
self resonances to occur at much higher frequencies,? if an
inductor is used as a Radio Frequency Choke ( RFC) then pile
winding increases usable bandwidth. Dividing the inductance into
several parts means that the L and C for each pile are also
divided so the resonance frequencies are higher. Ideally each pile
has a different number of turns so the resonances are staggered
but that significantly adds to the cost of pile wave winding. The
issue of RFC resonance has largely dissapeared with the
replacement of? thermionic valves ( Tubes) with much lower
impedance semiconductors but it was a significant problem in RF
power amplifiers.
On 20/11/2022 17:02, Mikek wrote:
toggle quoted message
Show quoted text
On Sun, Nov 20, 2022 at 05:30 AM, Alan wrote:
The real issue of Q with additional resistors
is the self capacitance of the inductor. If you look
at the HP manual there is a nice graph showing the effect of
self capacitance on indicated Q.
That is all true, but once you coil is built, the self capacitance
will always be there, all you can do is build to minimize it.
Don't you pretty much have calculate with indicated values when
designing.
?About half way down this page is a program that calculates self
capacitance from measurements you enter. You can do it with two
data points, but this will allow at least 15 and give you
inductance also. Oh, looks like coping the program name makes a
link, the page has other useful programs, worth a look.
It is a very real issue, the self capacitance can
easily be around 30pF for a multi-turn close wound coil. So if
you have 30pF of external tuning capacitance your resonance is
at 0.707 of the true resonance frequency AND the inductance that
you get from the resonant frequency? and the 30pF external
tuning is a factor of 2 out !? Even with a fully meshed tuning
capacitor the self capacitance is very significant with many
inductors that have many turns of wire. It gets worse when the
wire is close wound into many layers and becomes more random
when scramble winding is used. Wave winding helps reduce
internal C as does multiple piles.
? I think you meant to say pies, instead of piles.
???????????????????????? Mikek
?
--
Thanks, Mikek
|
Re: PCB making software, What do you use?
To add to this, is OK as an online site for PCB design if you want a cloud solution/to collaborate. I prefer KiCAD though as I need to integrate it with FreeCAD for the case design.
You can get PCB's made for very little money at PCBWAY.com or JLCPCB.com. The latter has a really nice collection of parts you can use and very cheap surface-mount assembly costs. Kind regards, Marcus
On Mon, 21 Nov 2022 at 15:29, Donald H Locker via <dhlocker= [email protected]> wrote:
KiCAD. Free, open-source. Incredibly capable; great support crew;
moderate learning curve.
Diptrace. I've not used it in years, but it was fairly easy way
back when. I don't know cost; used to be free.
Have a look at
for some other packages to consider.
HTH,
Donald.
On 11/21/22 10:20, Mikek wrote:
I have made pcbs with software but it has been many years since I
last used a program.
?I want something that will do surface mount as well as through
hole. Fast learning curve is a big plus!
?Also Free!
I don't have anything very complex, 2, 3, or 4, transistors maybe
later I'll try some SMD ICs.
?What do you recommend.
BTW, there used to be programs where you could get 3 pcbs for
$xx.00,
?but I'm not finding any hits for those.
?????????????????????????? Mikek
|
Re: PCB making software, What do you use?
KiCAD. Free, open-source. Incredibly capable; great support crew;
moderate learning curve.
Diptrace. I've not used it in years, but it was fairly easy way
back when. I don't know cost; used to be free.
Have a look at
for some other packages to consider.
HTH,
Donald.
On 11/21/22 10:20, Mikek wrote:
toggle quoted message
Show quoted text
I have made pcbs with software but it has been many years since I
last used a program.
?I want something that will do surface mount as well as through
hole. Fast learning curve is a big plus!
?Also Free!
I don't have anything very complex, 2, 3, or 4, transistors maybe
later I'll try some SMD ICs.
?What do you recommend.
BTW, there used to be programs where you could get 3 pcbs for
$xx.00,
?but I'm not finding any hits for those.
?????????????????????????? Mikek
|
PCB making software, What do you use?
I have made pcbs with software but it has been many years since I last used a program.
?I want something that will do surface mount as well as through hole. Fast learning curve is a big plus!
?Also Free!
I don't have anything very complex, 2, 3, or 4, transistors maybe later I'll try some SMD ICs.
?What do you recommend.
BTW, there used to be programs where you could get 3 pcbs for $xx.00,
?but I'm not finding any hits for those.
?????????????????????????? Mikek
|
On Sun, Nov 20, 2022 at 06:14 PM, John KN5L wrote:
Value for parallel resistance was computed a few emails back:
On 11/20/22 12:32 PM, John KN5L wrote:
Xc = 1/(2*Pi*1.7e6*0.4e-12) = 234kohm
Rp = 234k x 45 = 10.53Mohm
?Getting to late for me to think, but how can you calculate the Q if you don't consider all of the resistance (11M¦¸ and CpRp of 9.484M¦¸) across the capacitor? Probably just two ways to do it, and I'm just not following your way. ??????????????????????????????????????????????????? Thanks, Mikek
|
On 11/20/22 7:39 PM, Mikek wrote: I think you should add the loss resistor across
?C1 in your SimSmith drawing and simulation. Attached. C1 with capacitance and Q (0.4pF and 45) is replaced with C1 capacitance (0.4pF) and R2 parallel resistance (10.53Mohm). Value for parallel resistance was computed a few emails back: On 11/20/22 12:32 PM, John KN5L wrote: Xc = 1/(2*Pi*1.7e6*0.4e-12) = 234kohm
Rp = 234k x 45 = 10.53Mohm Notice Q-Meter trace L.V.M*50=634.57 is close to previous email SimSmith attachment where L.V.M*50=634.62 John KN5L
|
On Sun, Nov 20, 2022 at 04:29 PM, John KN5L wrote:
9,848,000¦¸ is Rp for the capacitive component of the added 11M resistor.
It was much easier for me to understand by viewing the SimSmith model
(attached again), where the capacitive component is isolated with C1
with C=CpC and Q=CpQ, with values of 400fF and 45.
One can discuss if the Q should be for the capacitive component or for
the resistor in total.
Both of our answers are equivalent. 45 for the capacitive component and
22 for the total resistor.
Maybe I'm biased, after spending 10 days trying to figure out why the Q meter reading was low. And the answer is because the self capacitance of the resistor is lossy and that loss is equivalent to, 9,848,000¦¸. That does not show up in your SimSmith, I think you should add the loss resistor across ?C1 in your SimSmith drawing and simulation. The way it is doesn't solve for what (I) was looking for. Maybe it doesn't matter, but you are not finding the reason that the 11M¦¸ reads low without it. That's my 1.816 cents. With the inflation we had, 2 cents isn't worth what it used to be. :-) ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ? ?? Mikek ????????????????????????????????????????
|
I¡¯m reading your observations about high Q measurements and calculations . This is very interesting. High Q is vey dif¨ªcilt to measure. Q-meters was designed as a practical instrument to design inductors mainly and also capacitors . On inductors Q are very important but also self resonance frequency. In radio practice when the components gets connected with tubes or transistors que Q goes down very much . For this reason I believe that HP q-meter specs are more than enough for practical proposes. 10% for Q cose to 1000 it¡¯s very good !!
toggle quoted message
Show quoted text
On 20 Nov 2022, at 19:29, John KN5L <john@...> wrote:
?On 11/20/22 5:49 PM, Mikek wrote:
So, what I think we have is loss resistance (from the self capacitance)
of 9,848,000¦¸ 9,848,000¦¸ is Rp for the capacitive component of the added 11M resistor.
It was much easier for me to understand by viewing the SimSmith model
(attached again), where the capacitive component is isolated with C1
with C=CpC and Q=CpQ, with values of 400fF and 45.
One can discuss if the Q should be for the capacitive component or for
the resistor in total.
Both of our answers are equivalent. 45 for the capacitive component and
22 for the total resistor.
John KN5L
|
On 11/20/22 5:49 PM, Mikek wrote: So, what I think we have is loss resistance (from the self capacitance)
of 9,848,000¦¸ 9,848,000¦¸ is Rp for the capacitive component of the added 11M resistor. It was much easier for me to understand by viewing the SimSmith model (attached again), where the capacitive component is isolated with C1 with C=CpC and Q=CpQ, with values of 400fF and 45. One can discuss if the Q should be for the capacitive component or for the resistor in total. Both of our answers are equivalent. 45 for the capacitive component and 22 for the total resistor. John KN5L
|
On Sun, Nov 20, 2022 at 02:45 PM, John KN5L wrote:
Are we talking about the Q of the resistors self capacitance?
Yes, Per Steve's measurement, 11Mohm resistor has parallel capacitance
of 0.4pF with Q=45.
John KN5L
? OK, I see it a little different. The parallel capacitance has its own loss resistance associated with it. ?I think you ignored that? 11M¦¸ = R1, ? Rmeasured or Rm= 5,190,000? Rcp = Loss resistance of parallel self capacitance, it's the unknown. Rt = R1 + R2, rearranged it is R2 = Rt - R1, Changing the names? 1/Rcp = 1/R1 - 1/Rm = 1/5.19M¦¸ - 1/11M¦¸ = 1/Rcp or 9,848,000¦¸. So, what I think we have is loss resistance (from the self capacitance) of 9,848,000¦¸ in parallel with 11,000,000¦¸ which equals, 5,190,000¦¸. Leaving us with a Self Capacitance Q of 22.17. What do you think? ?????????? Mikek ? -- Thanks, Mikek
|
On 11/20/22 1:20 PM, Mikek wrote: John KN5L wrote: Yesterday's post using SimSmith to solve for 11Mohm
resistor capacitance
and Q by match Steve's Q measurements. Results were 0.4pF with Q=45.
Are we talking about the Q of the resistors self capacitance? Yes, Per Steve's measurement, 11Mohm resistor has parallel capacitance of 0.4pF with Q=45. John KN5L
|
Thanks for all that Jacques.
--
Thanks, Mikek
|
Mikek,
?? ??? ??? ??? ? It is my contention that on the Q meter that 0.2pf is resonated out, because after adding the resistor I need to retune to resonance.
?? ??? ??? ??? ? I'm using an 0805 smd resistor with 1" leads of #32 wire.
??? ??? ??? ??? After adding that resistor to the resonate circuit, I need to adjust the tuning cap 0.12pf to get back to resonance.
?? ??? ??? ??? ? I appreciate any enlightenment you can give me. Mikek
Sure, the capacitance of the resistor is in parallel with the Q meter tuning cap.
??? ??? ??? ??? P.S. You mentioned 'high impedance' amplifiers, I don't recall if I said anything here about testing the input of high input impedance amps,
?? ??? ??? ??? ? but that is what I want to do. Put the amp input across a resonant circuit, see haw much I need to readjust the capacitance to get back to resonance and
??? ??? ??? ??? see how much the Q dropped. Then do the math,? Rp = Q1 x Q2? /? 2 x pi x f x C x Delta Q.
???? ??? ??? ??? Do you see this as a reliable method to find the input impedance of a "high input impedance' amplifier?
Correct !? As long as nothing saturates.? Just monitor the output with a scope.
You may test the amplifier with the capacitive divider that will be used in the final circuit.? This is where you want a high Rp.
Note that with a minimum freq of 1 MHz, Q=1000 and Cmin. =20 pF, the corresponding Rp is 8 Mohms.
So if the amplifier Rp is 80 Mohms, the Q error is 10 %,? and 1% with 800 Mohms.
As the frequency increases, the source resistance will start to give appreciable error.? See my attached doc.
Jacques
VE2AZX
|
On Sun, Nov 20, 2022 at 11:13 AM, John KN5L wrote:
How are you calculating the Q of the capacitor?
??????????????????????????????????????????????? Mikek
John KN5L wrote: Yesterday's post using SimSmith to solve for 11Mohm resistor capacitance
and Q by match Steve's Q measurements. Results were 0.4pF with Q=45.
Are we talking about the Q of the resistors self capacitance? ?Jacques Q was about 9 on his 10M¦¸ resistor according to the graph. ?Jacques, was your resistor a 1/4W through hole or smd? ? -- Thanks, Mikek
|
Using SimSmith, as shown in yesterday's post. Adjusted CpQ such that
Q-meter model measured Q=634.62 is a close match to Steve's measured Q=635
John KN5L
toggle quoted message
Show quoted text
On 11/20/22 12:59 PM, Mikek wrote: On Sun, Nov 20, 2022 at 10:32 AM, John KN5L wrote:
Yesterday's post using SimSmith to solve for 11Mohm resistor capacitance
and Q by match Steve's Q measurements. Results were 0.4pF with Q=45.
How are you calculating the Q of the capacitor?
??????????????????????????????????????????????? Mikek
|
Leon
Patricio A. Greco
