An in-depth practical visualisation of how bypass capacitors work at both high and low frequencies.
Bulk decoupling capacitors vs bypass capacitors.
Capacitor placement and types are tested and the results examined.
How package inductance can have a large effect.
Loop area and what is means, it's impact on EMC emissions, and how currents flow in ground planes is demonstrated
Links:
http://web.mst.edu/~jfan/slides/Archambeault1.pdf
Use that HCMOS oscillator to drive all the gates of a 74F00 in parallel, and use one output to drive the transmission line. 74F was the fastest standard TTL, and had massive power current needs. They were quickly replaced with S and then LS, as they were almost as fast but more importantly did not run ar so high a current.LSTTL in many cases was happy with little decoupling, but F series would make a very nice oscillator.
Did make a stack of 3 54S00 chips to make a clock driver, using all 12 gates in parallel, soldered the decoupling capacitors directly on the pins top and bottom of the stack along with a 10uf 25v wet tantalum package there. It ran hot, but the clock edges were really fast, you got to love the CERDIP for being able to handle heat well.
74F was the fastest standard TTL, and had massive power current needs. They were quickly replaced with S and then LS, as they were almost as fast but more importantly did not run ar so high a current.LSTTL in many cases was happy with little decoupling, but F series would make a very nice oscillator.
You are writing about the 74H or 74S series, the initial fast TTL designs that were replaced by 74LS.
The 74F and 74AS was later designs introduced in 1985.
Great stuff, Mr. jones! Even when I have gotten an intellectual grip on these things I find it really handy having the practical demo to hang it on. I have a very visual memory and I find the demos help with remembering the maths involved. When they are purely abstract I have to keep looking stuff up, but once it's related to a physical setup it seems to stick.
Dave, is there a reason why you didn't look at showing how a tantalum cap can offer both bulk and local decoupling ?
We often see them used on longer supply rails to clean up both HF and LF.
That was neat. I never really considered much how the physical layout of something can make such a difference until I started watching your videos, and this particular one really is cool as you can easily visualize the difference.
About visualizing the path of electrons, I wonder if it could be done using a higher voltage, like mains, and then you could probe different parts of the board with a multimeter. Could use something less conductive than copper as well, maybe lead. Would make for a fun experiment.
Or solder a bunch of small LEDs and they should glow at different intensities.
Well done and explained!
Just wonder why did you use that switching psu instead dp832 behind it, there is a specific reason? I guess to have on the screen of the scope some ripple that the linear one does not have, is this right?
I was wondering if you could take that test jig and do another experiment: take a second scope and place a probe with its ground lead attached inside of the "loop area" It would be cool to see how much noise the ground wire picks up.
Dave, is there a reason why you didn't look at showing how a tantalum cap can offer both bulk and local decoupling ?
Because this wasn't supposed to be a video on bypass capacitor types, the video was already longer than I wanted.
Well done and explained!
Just wonder why did you use that switching psu instead dp832 behind it, there is a specific reason? I guess to have on the screen of the scope some ripple that the linear one does not have, is this right?
Because it was sitting there.