Call Search
     

New to Ham Radio?
My Profile

Community
Articles
Forums
News
Reviews
Friends Remembered
Strays
Survey Question

Operating
Contesting
DX Cluster Spots
Propagation

Resources
Calendar
Classifieds
Ham Exams
Ham Links
List Archives
News Articles
Product Reviews
QSL Managers

Site Info
eHam Help (FAQ)
Support the site
The eHam Team
Advertising Info
Vision Statement
About eHam.net

donate to eham
   Home   Help Search  
Pages: [1] 2 Next   Go Down
  Print  
Author Topic: Calculating the value of a bypass/decoupling capacitor  (Read 1172 times)
KE8HAG
Member

Posts: 62




Ignore
« on: April 08, 2018, 07:15:29 AM »

Hypothetically. Not having a schematic, is there a general rule of thumb to approximate the value of a decoupling cap to ground in a IF circuit, that has no identification and one of the leads is broken off flush to the body?

I'm far from being an engineer in electronics and have wondered if there is there is a simple way (rule of thumb) to approximately figure an unknown value decoupling cap, in say a 455 kc IF circuit. In the past I have guesstamated and the results seemed to of work ok but have no idea of what conflicting frequencies I've eliminated. I'm sure during the designing of this circuit there were frequencies being generated that needed to be suppressed or gotten rid of to improve the quality.

Nobody likes to guess, but sometime ya have too.

I'm interested in your opinion.

Logged

Shoot for the Moon !! Even if you miss...you'll still land in the Stars.
WB4SPT
Member

Posts: 618




Ignore
« Reply #1 on: April 08, 2018, 07:19:57 AM »

Well,  if you are trying to cause a near short at .5MHz, a 0.1uF should be close enough at under 5 Ohms. 
Logged
KE8HAG
Member

Posts: 62




Ignore
« Reply #2 on: April 08, 2018, 09:35:43 AM »

Well,  if you are trying to cause a near short at .5MHz, a 0.1uF should be close enough at under 5 Ohms. 

In the past, I was using 0.02 and 0.05 uf. Guess I was over doing it then.

Thanks
Logged

Shoot for the Moon !! Even if you miss...you'll still land in the Stars.
WA3SKN
Member

Posts: 6642




Ignore
« Reply #3 on: April 08, 2018, 10:45:25 AM »

Usually .1 uf, .01 uf, .001 uf disc caps were used for bypassing, depending on circuit impedance and frequencies involved.
They were good to around 270-280 MHz.
73s.

-Mike.
Logged
WB6BYU
Member

Posts: 17476




Ignore
« Reply #4 on: April 08, 2018, 11:58:38 AM »

Depends on the impedance of the circuit.

With high impedances in tube circuits, 100 ohms of reactance is close enough to a dead
short that lower value of bypass capacitors can be used.

In a solid state circuit where the impedance may be 50 ohms (or lower), then a
proportionally larger capacitor (smaller reactance) is needed.

At 455 kc, .001 uF or 1000uuF is about 350 ohms or so.  (The same applies for 455 kHz
and 1nF, of course, for modern readers.)

So .01 uF / 10nF would  be 35 ohms and 0.1 uF / 100nF would be 3.5 ohms.

.001 uF was generally a good value in tube circuits for HF, but may be marginal at IF,
and I'd probably grab .01 uF instead.  For a solid state circuit I'd use 0.1 uF due  to
the lower circuit impedances.  Of course, it depends on the specific location in the circuit.



[edited to add]   Note that you can have too large of a bypass capacitor.

The capacitance plus the internal inductance and the lead length create a resonant circuit.
The bypass function is most effective at self-resonance, and the old ARRL Handbook had
tables of values and approximate lead lengths to use to optimize performance on various bands,
particular VHF and high HF.

This can be important at higher frequencies:  we found that a 12pF 0402 SMD chip capacitor was
about self resonant in the 5 - 6 GHz range, and any larger value didn't work any better.
« Last Edit: April 08, 2018, 12:03:52 PM by WB6BYU » Logged
KE8HAG
Member

Posts: 62




Ignore
« Reply #5 on: April 08, 2018, 03:00:55 PM »

Depends on the impedance of the circuit.

With high impedances in tube circuits, 100 ohms of reactance is close enough to a dead
short that lower value of bypass capacitors can be used.

In a solid state circuit where the impedance may be 50 ohms (or lower), then a
proportionally larger capacitor (smaller reactance) is needed.

At 455 kc, .001 uF or 1000uuF is about 350 ohms or so.  (The same applies for 455 kHz
and 1nF, of course, for modern readers.)

So .01 uF / 10nF would  be 35 ohms and 0.1 uF / 100nF would be 3.5 ohms.

.001 uF was generally a good value in tube circuits for HF, but may be marginal at IF,
and I'd probably grab .01 uF instead.  For a solid state circuit I'd use 0.1 uF due  to
the lower circuit impedances.  Of course, it depends on the specific location in the circuit.



[edited to add]   Note that you can have too large of a bypass capacitor.

The capacitance plus the internal inductance and the lead length create a resonant circuit.
The bypass function is most effective at self-resonance, and the old ARRL Handbook had
tables of values and approximate lead lengths to use to optimize performance on various bands,
particular VHF and high HF.

This can be important at higher frequencies:  we found that a 12pF 0402 SMD chip capacitor was
about self resonant in the 5 - 6 GHz range, and any larger value didn't work any better.

Lot of good info there. I don't do to much work with the newer stuff. SMD stuff is to small for my eyesight. I have a hard time seeing the component let alone reading whats on it.  Grin
I do a good bit of tube radios, old stuff. I posed the question because sometimes I'm working in the blind and thought there might be a rule of thumb I could use. Electronics is a pass time hobby for me. I'm still learning electronics and know just enough to get myself in trouble. I understood most of what you said but have never set down to learn the math to calculate impedance or reactance for resonance.

I appreciate the time you took for me to explain the differences and precautions.

Thank You
Al
Logged

Shoot for the Moon !! Even if you miss...you'll still land in the Stars.
W6EM
Member

Posts: 1725




Ignore
« Reply #6 on: April 08, 2018, 05:21:07 PM »

And, of course, don't forget the DC voltage rating either.  Important because most of today's leaded components are in the neighborhood of 25V to 100V, and if working with tube circuits, you will need a higher voltage rating if decoupling in the plate circuits.  600V or higher are getting hard to find these days.....and pricey.

73.

Lee
W6EM
Logged
KE8HAG
Member

Posts: 62




Ignore
« Reply #7 on: April 08, 2018, 07:59:45 PM »

And, of course, don't forget the DC voltage rating either.  Important because most of today's leaded components are in the neighborhood of 25V to 100V, and if working with tube circuits, you will need a higher voltage rating if decoupling in the plate circuits.  600V or higher are getting hard to find these days.....and pricey.

73.

Lee
W6EM

Thanks Lee. I try to stay very conscious of that. On the older equipment, especially the electrolytic caps, when I replace those I bump them up a little on the voltage because in the day most of that equipment was designed for 117 volts ac and today's voltage is a bit higher.  I like to keep at least 35% above the dc working voltage.

David @ Just Radios prices aren't to bad. His inventory isn't to bad either.  The only cap I have a hard time finding is the chassis mount can multi cap.

Again Thanks for the reminder.

Al
Logged

Shoot for the Moon !! Even if you miss...you'll still land in the Stars.
G3RZP
Member

Posts: 163




Ignore
« Reply #8 on: April 10, 2018, 02:35:24 AM »

A very rough 'rule of thumb' is that normal - 20 or 22 or 24 AWG leads look like about 20nH/ inch. So a 0.1 mFd with 0.5 inch leads is self resonant about 3.55 MHz. Actually, it will be a bit lower because of the capacitor's self inductance, and above that frequency, it will have an increasing impedance. I normally figure on 0.5 to 1.0 microfarads for about 20 to 150kHz, 0.1 mFd  from 100kHz to 2 MHz, 0.01 mFd  for 1 to 10 MHz and 0.001 mFd from about 10 to about 30 MHz. But it's not critical.

What can be critical in a production environment is choosing a capacitor to be self resonant because of its internal inductance. Then the manufacturer changes the design or purchasing decides that a 0.1microfarad 350 volt capacitor is just that and ABC company supplier is 10 cents cheaper per hundred off than XYZ supplier......

Then the fun starts: the original guy who perpetrated the problem has either left or been promoted out of the way and it is all now the fault of the poor so-and-so in engineering who has to sort out the mess.....

Cynical? Me?
Logged
KE8HAG
Member

Posts: 62




Ignore
« Reply #9 on: April 10, 2018, 04:25:18 AM »

A very rough 'rule of thumb' is that normal - 20 or 22 or 24 AWG leads look like about 20nH/ inch. So a 0.1 mFd with 0.5 inch leads is self resonant about 3.55 MHz. Actually, it will be a bit lower because of the capacitor's self inductance, and above that frequency, it will have an increasing impedance. I normally figure on 0.5 to 1.0 microfarads for about 20 to 150kHz, 0.1 mFd  from 100kHz to 2 MHz, 0.01 mFd  for 1 to 10 MHz and 0.001 mFd from about 10 to about 30 MHz. But it's not critical.

What can be critical in a production environment is choosing a capacitor to be self resonant because of its internal inductance. Then the manufacturer changes the design or purchasing decides that a 0.1microfarad 350 volt capacitor is just that and ABC company supplier is 10 cents cheaper per hundred off than XYZ supplier......

Then the fun starts: the original guy who perpetrated the problem has either left or been promoted out of the way and it is all now the fault of the poor so-and-so in engineering who has to sort out the mess.....

Cynical? Me?


"Then the fun starts: the original guy who perpetrated the problem has either left or been promoted out of the way and it is all now the fault of the poor so-and-so in engineering who has to sort out the mess....."  Smiley Smiley

That's kinda the rule of thumb I was looking for. For me, it's food for thought when working in the blind.

Thanks much.

And it's OK to be Cynical. I'm that way sometimes.

Al
Logged

Shoot for the Moon !! Even if you miss...you'll still land in the Stars.
WB4SPT
Member

Posts: 618




Ignore
« Reply #10 on: April 10, 2018, 05:39:32 AM »

I suppose we should take a stab at completing this SRF exercise for at least VHF and UHF.  In the leaded world: 

330pf with 1/8" leads work good at 150MHz+/-.    33pf at 450MHz.   10pf at 800MHz. 
Logged
G3RZP
Member

Posts: 163




Ignore
« Reply #11 on: April 10, 2018, 06:21:16 AM »

In the SMT world, you still have to think about the track lengths.....and maybe the bond wire lengths in an IC.
Logged
WA3SKN
Member

Posts: 6642




Ignore
« Reply #12 on: April 11, 2018, 03:42:30 PM »

No need for math if you have an ARRL Handbook. (Just about any year).
There is a inductive and capacitive vs freq nomograph that covers 100 Hz to 100 MHz.  You just check the freq and the reactance in ohms  shows for both inductors and capacitors.  More important, you can just shift the zeros to cover other frequencies.
Simple and works!
73s.

-Mike.

I keep an enlarged version in the shack on the wall... handy!



Logged
G3RZP
Member

Posts: 163




Ignore
« Reply #13 on: April 12, 2018, 04:38:19 AM »

However, one must not forget the stray inductance of leads and tracks with regard to series resonant frequencies.
Logged
KM1H
Member

Posts: 3503




Ignore
« Reply #14 on: April 14, 2018, 09:45:23 AM »

In the late 20's the bypasses were mostly .1 uF for HF and IF's. That gradually went to .05 and then .02 with a few steps in between.

The reason was partially cost plus the engineers eventually found out what was adequate and not the best via theory. Consumer radio production was very competitive and every penny counted. You also had inductance with the large paper caps and lead dress plus cap placement were necessary to keep the stage from taking off. Sometimes the reverse was true when a disc cap increased stage gain and a low value carbon comp or film resistor right at the control grid pin calmed it down.

In extreme cases a 100 Ohm right at the screen pin and bypassed at both ends were needed; this was more common after a tube swap to something with a much higher transconductance and lower noise resistance. It could really wake up older radios on the higher frequencies.

Carl
Logged
Pages: [1] 2 Next   Go Up
  Print  
 
Jump to:  

Powered by MySQL Powered by PHP Powered by SMF 1.1.11 | SMF © 2006-2009, Simple Machines LLC Valid XHTML 1.0! Valid CSS!