Hi marce,
Yups - the dimensions are correct. I am doing a basically two-layer digital design where I need a good attenuation from signal-to-ground layer. Thus the thin PCB ... Any idea as to where I may find this?
Cheers,
Jesper
P.S.: Maybe I should add that I only need small amounts like e.g. 40*50 cms.
Yups - the dimensions are correct. I am doing a basically two-layer digital design where I need a good attenuation from signal-to-ground layer. Thus the thin PCB ... Any idea as to where I may find this?
Cheers,
Jesper
P.S.: Maybe I should add that I only need small amounts like e.g. 40*50 cms.
Last edited:
Are you talking controlled impedance, if so are you using DDR memory or gigabit Ethernet, or RF if not them it will not be so critical, the problem is that the laminate thickness you are thinking of is too thin for assembly... post a circuit or more details would help......
Hi again marce ...
Thanks for asking considerate questions ;-)
Well, the point is that I'm doing my own PCBs because I like the "artistic freedom" in doing so and also don't have the finances to make multiple 4 layer trial boards. I already have some 0.2 mm boards (and yes, they are flexible) but the copper layer is a bit too thick for soldering and unsoldering components in a way that the components may not be damaged.
So I'd like to find a bit of similar PCB with just a little thinner copper thickness.
What I'd like to keep low with the 0.2 - 0.5 mm board is the crosstalk coefficient.
There are many uses so no particular board to post right now. Anticipating circuits ;-)
Cheers,
Jesper
Thanks for asking considerate questions ;-)
Well, the point is that I'm doing my own PCBs because I like the "artistic freedom" in doing so and also don't have the finances to make multiple 4 layer trial boards. I already have some 0.2 mm boards (and yes, they are flexible) but the copper layer is a bit too thick for soldering and unsoldering components in a way that the components may not be damaged.
So I'd like to find a bit of similar PCB with just a little thinner copper thickness.
What I'd like to keep low with the 0.2 - 0.5 mm board is the crosstalk coefficient.
There are many uses so no particular board to post right now. Anticipating circuits ;-)
Cheers,
Jesper
Last edited:
Crosstalk coefficient, what are you reading...
Use a decent thickness laminate and don't worry about such things, I doubt you will be doing a circuit that is going to be any different if you used a decent thickness laminate!
Seriously I don't know where your got this little nugget to be worried over, my recommendation would be to use a standard laminate and not worry about crosstalk coefficients... I would also NOT use such a thin laminate, the stress on solder joints etc will be more problematic thanwhat you are worrying about. Normally in the real world we use a certain thickness of laminate between signals and a ground plane for impedance control, not crosstalk coefficients, if we are worried about crosstalk between parallel conductors then the best option is to space the conductors apart.
You do seem to be going at this PCB thing the wrong way and in a way I have never heard anybody use before, there are more problems to worry about than the crosstalk coefficient and solving it using what are effectively laminates designed for multi-layer boards.
Keep things simple you are more likely to end up with a working assembly rather than a dodgy assembly with numerous problems... So firstly use a sensible laminate that can carry an electronic assembly...
Also 2 layers, unless you are really good and the circuit is really simple then you will have routes on two layers so you wont have a ground plane but a ground copper pour, if any signals cross a split in the return path you will have more problems to worry about than CC!
Jeez when we do things like this we have to use a 3D field solver and signal integrity software to highlight any problems so we only do it on really critical designs where the speeds and circuitry are at a level where signal integrity could be degraded, so for what you are doing I would step back and look at the basics and more important design criteria.
Use a decent thickness laminate and don't worry about such things, I doubt you will be doing a circuit that is going to be any different if you used a decent thickness laminate!
Seriously I don't know where your got this little nugget to be worried over, my recommendation would be to use a standard laminate and not worry about crosstalk coefficients... I would also NOT use such a thin laminate, the stress on solder joints etc will be more problematic thanwhat you are worrying about. Normally in the real world we use a certain thickness of laminate between signals and a ground plane for impedance control, not crosstalk coefficients, if we are worried about crosstalk between parallel conductors then the best option is to space the conductors apart.
You do seem to be going at this PCB thing the wrong way and in a way I have never heard anybody use before, there are more problems to worry about than the crosstalk coefficient and solving it using what are effectively laminates designed for multi-layer boards.
Keep things simple you are more likely to end up with a working assembly rather than a dodgy assembly with numerous problems... So firstly use a sensible laminate that can carry an electronic assembly...
Also 2 layers, unless you are really good and the circuit is really simple then you will have routes on two layers so you wont have a ground plane but a ground copper pour, if any signals cross a split in the return path you will have more problems to worry about than CC!
Jeez when we do things like this we have to use a 3D field solver and signal integrity software to highlight any problems so we only do it on really critical designs where the speeds and circuitry are at a level where signal integrity could be degraded, so for what you are doing I would step back and look at the basics and more important design criteria.
Hi marce,
Thanks for your comments, however, as you may see from my replies below I do think I have taken into account your comments in my design thinking, and so I remain interested in such a laminate:
As I wrote previously I prefer to do my own PCBs for various reasons and the base in this may be a two layer board with modifications. I can see the point in your comments, however, I do think that I am reasonably taking them into account with this approach - and may still gain something from using a thin PCB.
So I remain interested in a tip on where to find such a thin PCB laminate (in not too large quantities).
Cheers,
Jesper
Thanks for your comments, however, as you may see from my replies below I do think I have taken into account your comments in my design thinking, and so I remain interested in such a laminate:
Well various literature (provided mostly by you, actually ;-) ) and simulating using Saturn PCB Design simulator.
I would agree in case I did not intend to reinforce the laminate. But I do and my guess is that it will be as mechanically "sound" as a thicker laminate, yet still with the advantages of the thin laminate.
Well, there will be (a) ground plane(s) without any splits - at least not where it may be critical for return paths etc.
As I wrote previously I prefer to do my own PCBs for various reasons and the base in this may be a two layer board with modifications. I can see the point in your comments, however, I do think that I am reasonably taking them into account with this approach - and may still gain something from using a thin PCB.
So I remain interested in a tip on where to find such a thin PCB laminate (in not too large quantities).
Cheers,
Jesper
My comment on what are you reading was because you seem to have got the wrong end of the stick regarding signal integrity and what is important...
I will say again the laminate is to thin and will create more problems than it is likely to solve for DIY I would not go below 1mm, you are not going to gain anything... I do this every day for a living and would never consider using such a thin laminate without some extra support, most likely bonding to another piece of FR4.
What do you think you will gain by using such a thin laminate....
I will say again the laminate is to thin and will create more problems than it is likely to solve for DIY I would not go below 1mm, you are not going to gain anything... I do this every day for a living and would never consider using such a thin laminate without some extra support, most likely bonding to another piece of FR4.
What do you think you will gain by using such a thin laminate....
0.2mm laminate 0.3mm trace 55R characteristic impedance
0.5mm laminate 0.3mm trace 87R characteristic impedance to get around 50 requires a 0.8mm trace.
Cross-talk will only be an issue if you have lots of traces running closely together for a distance, DDR memory interface routed 0.1/0.1mm track space work so for simple DIY jobs cross-talk isn't a big concern, also you'll have the room to do the best thing to avoid cross-talk, spread the traces out...
If you insist on using a totally unsuitable laminate then you are going to have to bond it to something more solid when its an assembly and to assembly it...
Again if it is a digital circuit (or analogue) and you are after the best results (you worry about -130dB in other threads) then why compromise a layout, go for a 4 layer design. Some circuitry would be beneficial in giving advice as at the moment this is all general and not focused on real design...
0.5mm laminate 0.3mm trace 87R characteristic impedance to get around 50 requires a 0.8mm trace.
Cross-talk will only be an issue if you have lots of traces running closely together for a distance, DDR memory interface routed 0.1/0.1mm track space work so for simple DIY jobs cross-talk isn't a big concern, also you'll have the room to do the best thing to avoid cross-talk, spread the traces out...
If you insist on using a totally unsuitable laminate then you are going to have to bond it to something more solid when its an assembly and to assembly it...
Again if it is a digital circuit (or analogue) and you are after the best results (you worry about -130dB in other threads) then why compromise a layout, go for a 4 layer design. Some circuitry would be beneficial in giving advice as at the moment this is all general and not focused on real design...
@peranders: Thanks for the tip on printline - they are even in Denmark ... However, I'm a bit wary of asking for samples unless I genuinely am considering becoming a customer and in this case I'm not sure I will ... But I'll keep it in mind - thanks for suggesting ;-)
@marce: As you may well know I have a great deal of respect for the competences & knowledge you possess. And, as a matter of fact, I've read much (if not most) of the links you've provided over time - and I unhesitantly thank you for those as I have learned much.
In this case, however, I'm not so much asking for comments or suggestions on what I'm doing as to just suggestions on where to find such a laminate - which IMHO I still reckon is quite useful for my purposes (which is not only a clock circuitry).
I hope this comes through with the positive intention that I sent it with.
Cheers,
Jesper
@marce: As you may well know I have a great deal of respect for the competences & knowledge you possess. And, as a matter of fact, I've read much (if not most) of the links you've provided over time - and I unhesitantly thank you for those as I have learned much.
In this case, however, I'm not so much asking for comments or suggestions on what I'm doing as to just suggestions on where to find such a laminate - which IMHO I still reckon is quite useful for my purposes (which is not only a clock circuitry).
I hope this comes through with the positive intention that I sent it with.
Cheers,
Jesper
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.