Here are some bode plots of a Fairchild LM317 under various conditions:
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
An externally hosted image should be here but it was not working when we last tested it.
As I know, Bode plots are straight lines approximating the real curve of a circuit. Yours appear to be real graphs. Or simulated, but not Bode drawings. I believe. :-D
Bode plots are the real (curvy) thing. The straight line approximations are very useful as they are usually good enough to spot problems.
No, they are straight lines !
Bode plot - Wikipedia, the free encyclopedia
They represent straight approximation of a function, not the function itself. Sometimes, the asymptotic straight line to the function.
Wikipedia disagrees. It says the Bode plot can be approximated by straight lines, not that the Bode plot is straight lines. The Bode plot is the full curvy plot, as I said.Wikipedia said:
Schematic -- nothing exceptional or creative
An externally hosted image should be here but it was not working when we last tested it.
it does not look exceptional.
What is does show is that much thought needs to be implemented to get good test jigs. Then these good test jigs can reveal lots that we do not yet know about our circuits.
eg. Hamon divider
I only recently discovered that this existed. It is so simple to implement and yet is capable of far greater accuracy than I can ever achieve by ANY other method. And that is not an exaggeration. I can get +-20ppm with a few hours work on a two stage Hamon Divider giving 1:10:100 ratios that I just cannot get otherwise.
Mr Hamon is brilliant.
Your schematic shows that you too can think and devise tests.
What is does show is that much thought needs to be implemented to get good test jigs. Then these good test jigs can reveal lots that we do not yet know about our circuits.
eg. Hamon divider
I only recently discovered that this existed. It is so simple to implement and yet is capable of far greater accuracy than I can ever achieve by ANY other method. And that is not an exaggeration. I can get +-20ppm with a few hours work on a two stage Hamon Divider giving 1:10:100 ratios that I just cannot get otherwise.
Mr Hamon is brilliant.
Your schematic shows that you too can think and devise tests.
DIYAUDIO member Conrad has a Hamon divider tutorial on his site:
How to Build a Hamon Resistor Divider Network
The schematic of the bode tester isn't original to me.
How to Build a Hamon Resistor Divider Network
The schematic of the bode tester isn't original to me.
But I will still give you the credit for recognising a test circuit that can produce results that would have the rest of us wondering:
How do we do THAT?
How do we do THAT?
That's excellent stuff! IMO, the 317 is remarkably capable if used properly. There was an old paper/datasheet that showed the noise peaking if the output cap esr was too low. It's on line somewhere, though I can't locate it right now. No doubt it was part of your inspiration!
One thing the test setup here shows is how useful transformers can be, solving isolation and common mode problems that would otherwise be difficult.
One thing the test setup here shows is how useful transformers can be, solving isolation and common mode problems that would otherwise be difficult.
+1, very interesting.
The paper Corad refers to is 'Understanding and reducing noise voltage on 3-terminal regulators' by Erroll Dietz, National Semiconductor. Appears as an appendix in Bob Pease' excellent book on' Troubleshooting analog circuits'
This is the graph:
The paper Corad refers to is 'Understanding and reducing noise voltage on 3-terminal regulators' by Erroll Dietz, National Semiconductor. Appears as an appendix in Bob Pease' excellent book on' Troubleshooting analog circuits'
This is the graph:
Could we extrapolate from that graph and guess that 47uF to 220uF may be even better at attenuating that noise peak?
Alternatively, would adding a low ohm resistor in series with the cap bring down the peak?
Which would be more effective?
100uF with the intrinsic esr of a standard commercial electrolytic, or a 22uF with 0r1 added to damp the peak.
What do you think?
Would Jack's test rig show this? I am trying to interpret the last pic in the series (47uF+1r0). What is it showing us?
Alternatively, would adding a low ohm resistor in series with the cap bring down the peak?
Which would be more effective?
100uF with the intrinsic esr of a standard commercial electrolytic, or a 22uF with 0r1 added to damp the peak.
What do you think?
Would Jack's test rig show this? I am trying to interpret the last pic in the series (47uF+1r0). What is it showing us?
- Status
- Not open for further replies.