Terminator3
Advanced Member level 3
newbie: Measureing Er constant without good equipment
1. My first question is about measuring relative dielectric constant Er of copper clad laminate.
I read about ring resonators and other microstrip methods. But what if we do not have Network Analyser or any Spectrometer able to deal with high frequencies? Is using a RCL-meter is the only alternative? I just made RCL-measurement of raw 85x45mm PCB material:
calculating area
S=0.085*0.045
constant in formula
E0=8.85*10^-12
E0 = 8.85e-12
thickness 0.5mm
D=0.5*10^-3
D = 500e-6
RCL-meter gives measurement of 0.22nF
C=0.22*10^-9
C = 220e-12
Final computation:
Er=(C*D)/(E0*S)
Er = 3.249510727
is it right way of calculation? Er is 3.24 Farad/Meter?
And this method can be inaccurate for high frequencies. So what we can do if there is no handy Network analyser/spectrometer?
For example, we have frequency source of 5GHz. I come up with some stupod method: etch 20 bandpass filters for this frequency, then connect each filter and measure amplitude (simple RLC-voltage with some detector element), make chart, choose filter with maximum voltage amplitude. Or maybe we can come up with some clever microstrip power dividers and mixers, then etch them and analyse IF output?
Method#2: Etch copper on both sides of CCL, so no any copper exists. Take 5GHz antenna, measure signal power using RLC-meter with hf "detector". Take etched dielectric board and put in in front of 5GHz antenna, measure signal power again. do some calculations. I think it can be possible, we know air constant, then we know how it changed after inserting plate, right?
2. My second question is about HJ-FET and Bipolar transistors.
How engineer decide wich type of transistor can be used? I mean he knows the price, noise and power characteristics already. So here we have two transistors: cheap HJ-FET and cheap Bipolar with almost the same characteristics. What will he choose and why?
As an example: **broken link removed**
we can put BFP405 bipolar with common emmitter, +v to collector pin and we get RF from collector microstrip. (emitter pin to Gnd)
ALSO we can put some HJ-FET, +v to Drain pin and again we get RF from Drain microstrip. (source pin to Gnd)
If price, power and noise is the same, maybe there are some other pros and cons?
3. And the third question is. Why LDPE is not used as dielectric material for microwave CCL?
I just read http://members.tm.net/lapointe/Plastics.htm table and see that:
LDPE is 2.2 at 1GHz
PTFE is 2.0 at 1GHz
So i see that LDPE can be put as dielectric. Why not? Is it aging effects, problem with water absorption(?), can't resist high temperatures?
Good points: easy adhesion, cheap(?).
Thanks!
1. My first question is about measuring relative dielectric constant Er of copper clad laminate.
I read about ring resonators and other microstrip methods. But what if we do not have Network Analyser or any Spectrometer able to deal with high frequencies? Is using a RCL-meter is the only alternative? I just made RCL-measurement of raw 85x45mm PCB material:
calculating area
S=0.085*0.045
constant in formula
E0=8.85*10^-12
E0 = 8.85e-12
thickness 0.5mm
D=0.5*10^-3
D = 500e-6
RCL-meter gives measurement of 0.22nF
C=0.22*10^-9
C = 220e-12
Final computation:
Er=(C*D)/(E0*S)
Er = 3.249510727
is it right way of calculation? Er is 3.24 Farad/Meter?
And this method can be inaccurate for high frequencies. So what we can do if there is no handy Network analyser/spectrometer?
For example, we have frequency source of 5GHz. I come up with some stupod method: etch 20 bandpass filters for this frequency, then connect each filter and measure amplitude (simple RLC-voltage with some detector element), make chart, choose filter with maximum voltage amplitude. Or maybe we can come up with some clever microstrip power dividers and mixers, then etch them and analyse IF output?
Method#2: Etch copper on both sides of CCL, so no any copper exists. Take 5GHz antenna, measure signal power using RLC-meter with hf "detector". Take etched dielectric board and put in in front of 5GHz antenna, measure signal power again. do some calculations. I think it can be possible, we know air constant, then we know how it changed after inserting plate, right?
2. My second question is about HJ-FET and Bipolar transistors.
How engineer decide wich type of transistor can be used? I mean he knows the price, noise and power characteristics already. So here we have two transistors: cheap HJ-FET and cheap Bipolar with almost the same characteristics. What will he choose and why?
As an example: **broken link removed**
we can put BFP405 bipolar with common emmitter, +v to collector pin and we get RF from collector microstrip. (emitter pin to Gnd)
ALSO we can put some HJ-FET, +v to Drain pin and again we get RF from Drain microstrip. (source pin to Gnd)
If price, power and noise is the same, maybe there are some other pros and cons?
3. And the third question is. Why LDPE is not used as dielectric material for microwave CCL?
I just read http://members.tm.net/lapointe/Plastics.htm table and see that:
LDPE is 2.2 at 1GHz
PTFE is 2.0 at 1GHz
So i see that LDPE can be put as dielectric. Why not? Is it aging effects, problem with water absorption(?), can't resist high temperatures?
Good points: easy adhesion, cheap(?).
Thanks!
Last edited: