Maker Pro
Maker Pro

Through hole capacitors

C

colin

Jan 1, 1970
0
Hi,
I had been trying to decouple the power supply of a prototype circuit with
200mhz+ fast pulses and precision timers, with a smd capacitor on a double
sided pcb, but found there was still measurable noise on the gnd side of the
capacitor due to stray inductance. the back side of the pcb is undisturbed
ground plane with smaler gnd planes on the top side conected with vias. I
didnt realy have room to route a better gnd conection for another capacitor
and rather than put the capacitor on the back and disturb the gnd plane I
had an inspiration and drilled a 1 mm hole and found a 10nf smd capacitor
fitted through the hole perfectly, the end caps were flush on either side of
the pcb and cld be soldered ok and the chips power pin put on top of the
buried capacitor, therefore taking up very little space.

I had one go short circuit some time later (and it took a while to find as I
had forgoten it was even there!) but then I find smd components can get
upset quite quickly when reworking a prototype many times, the capacitor
cheked out ok when I removed it, although it seemed low in value but
increased when pressure applied, i find other capacitors can also fail
simlarlarly varying capacitance with a smal flex of the pcb after being re
soldered a few times.

A small array of through hole capacitors at various points solved the noise
problem.
I remember old tv tuners used to have square block ceramic wich went through
a slot in the pcb,
I was wondering if this is still considered ...

Colin =^.^=
 
F

Fred Bloggs

Jan 1, 1970
0
I had been trying to decouple the power supply of a prototype circuit with
200mhz+ fast pulses and precision timers, with a smd capacitor on a double
sided pcb, but found there was still measurable noise on the gnd side of the
capacitor due to stray inductance. the back side of the pcb is undisturbed
ground plane with smaler gnd planes on the top side conected with vias. I
didnt realy have room to route a better gnd conection for another capacitor
and rather than put the capacitor on the back and disturb the gnd plane I
had an inspiration and drilled a 1 mm hole and found a 10nf smd capacitor
fitted through the hole perfectly, the end caps were flush on either side of
the pcb and cld be soldered ok and the chips power pin put on top of the
buried capacitor, therefore taking up very little space.

I had one go short circuit some time later (and it took a while to find as I
had forgoten it was even there!) but then I find smd components can get
upset quite quickly when reworking a prototype many times, the capacitor
cheked out ok when I removed it, although it seemed low in value but
increased when pressure applied, i find other capacitors can also fail
simlarlarly varying capacitance with a smal flex of the pcb after being re
soldered a few times.

A small array of through hole capacitors at various points solved the noise
problem.
I remember old tv tuners used to have square block ceramic wich went through
a slot in the pcb,
I was wondering if this is still considered ...

Colin =^.^=

They are actually working on integrating the capacitor right into the
PCB dielectric during board manufacture.
 
P

Pooh Bear

Jan 1, 1970
0
colin said:
Hi,
I had been trying to decouple the power supply of a prototype circuit with
200mhz+ fast pulses and precision timers, with a smd capacitor on a double
sided pcb,

Just the one ! ?

That's your problem !

Graham
 
L

Leon

Jan 1, 1970
0
Fred said:
They are actually working on integrating the capacitor right into the
PCB dielectric during board manufacture.

The Pulsonix PCB software I use has that feature (Embedded Component
Technology). Rs, Cs and inductors may be placed on the inner layers.
I've never used it, though. :cool:

Leon
 
L

Leon

Jan 1, 1970
0
colin said:
Hi,
I had been trying to decouple the power supply of a prototype circuit with
200mhz+ fast pulses and precision timers, with a smd capacitor on a double
sided pcb, but found there was still measurable noise on the gnd side of the
capacitor due to stray inductance. the back side of the pcb is undisturbed
ground plane with smaler gnd planes on the top side conected with vias. I
didnt realy have room to route a better gnd conection for another capacitor
and rather than put the capacitor on the back and disturb the gnd plane I
had an inspiration and drilled a 1 mm hole and found a 10nf smd capacitor
fitted through the hole perfectly, the end caps were flush on either side of
the pcb and cld be soldered ok and the chips power pin put on top of the
buried capacitor, therefore taking up very little space.

I've heard of that technique being used on PCBs working at microwave
frequencies.

Leon
 
C

colin

Jan 1, 1970
0
Leon said:
The Pulsonix PCB software I use has that feature (Embedded Component
Technology). Rs, Cs and inductors may be placed on the inner layers.
I've never used it, though. :cool:

Leon

Interesting, yes I supose the next logical step would be to have ceramic
instead of fr4 for one of the layers beteewn the power planes :)

The manafacturing guys would probably have kittens if they had to put
capacitors inbetween layers, and what about reparing them ?

I was wondering if there are any big difference in stresses puting them
through a hole like this.

we already have pcb printable resistors maybe printable capacitors are next
?

Colin =^.^=
 
C

colin

Jan 1, 1970
0
Pooh Bear said:
Just the one ! ?

That's your problem !

Graham

Well I already had one or two but that wasnt enough so was trying to put one
more in but kinda ran out of room, adding a leaded capacitor directly accros
the supply rails didnt cut it either, eventualy had 2x 10nf 1x .1uf and 1x
1uf in one spot as close as possible (all but the 1uf buried), this cleaned
up the supply rail nicely for the 100ns 1-3v ramp generator and comparator
(right next to a 74vhc flip flop but with seperatly fed supplies)

I made a square wedge wich makes the 1mm hole have corners to acept the 0603
size wich is 1.1mm corner to corner, maybe they should make round capacitors
just for this ? althought ive come acros round resistors and diodes,
soldering is a little bit tricky.

Colin =^.^=
 
K

Ken Smith

Jan 1, 1970
0
Fred Bloggs said:
They are actually working on integrating the capacitor right into the
PCB dielectric during board manufacture.

You can already get high K PCB material so to some degree the capacitors
have already been made into the material.

The some years back, Russians made circuit boards with some of the
discrete parts internal to the PCB. It was done very like blind vias are
done but capacitors and resistors were placed part way through the
process. The result was a PCB with chips on both sides and many of the
other parts hidden inside.
 
K

Klaus Kragelund

Jan 1, 1970
0
Ken Smith skrev:
You can already get high K PCB material so to some degree the capacitors
have already been made into the material.

The some years back, Russians made circuit boards with some of the
discrete parts internal to the PCB. It was done very like blind vias are
done but capacitors and resistors were placed part way through the
process. The result was a PCB with chips on both sides and many of the
other parts hidden inside.

Have anyone tried this in practice? And is this technique viable for
cost sensitive products? And lastly any PCB production companies
offering this?

Thanks

Klaus
 
K

Ken Smith

Jan 1, 1970
0
Ken Smith skrev:


Have anyone tried this in practice?

Well the russians space folks did do they count? They were trying to sell
the technology to others.
And is this technique viable for
cost sensitive products?

NASA thought it was too costly.
 
J

Joseph2k

Jan 1, 1970
0
Ken said:
Well the russians space folks did do they count? They were trying to sell
the technology to others.


NASA thought it was too costly.
Sounds more like something the NSA would do.
 
Top