Re: Questions
Hi
Couldn't resist
1) Yes, he suggests that. Termination solves all misunderstood problems, except groundbounce ofcourse.
2) Happy to hear that you prefer HC logic over AC
Desoldering is a nighmare, but you can cut all pins, remove the package and desolder pin by pin. Use a socket for future experiments and make sure the power supply is LOW noise
succes
Dr.H said:
Hi
Couldn't resist
1) Yes, he suggests that. Termination solves all misunderstood problems, except groundbounce ofcourse.
2) Happy to hear that you prefer HC logic over AC
Desoldering is a nighmare, but you can cut all pins, remove the package and desolder pin by pin. Use a socket for future experiments and make sure the power supply is LOW noise
succes
In case anyone hasn't seen this.........
Guido made reference to this article in another post. I do advise reading it.
Goes double if you insist on using logic that is too fast for the application.
http://www.tentlabs.com/Info/Articles/Supply_decoupling.pdf
Jocko
Guido made reference to this article in another post. I do advise reading it.
Goes double if you insist on using logic that is too fast for the application.
http://www.tentlabs.com/Info/Articles/Supply_decoupling.pdf
Jocko
Dear all!
I am terribly sorry - or rather ashamed! [again, given that it has happened already once..]
My only excuses are these phrases:
So, what happened is this: being a not serious person, when I was reaching for the old jig I did go by memory, and supposed that the on-board connector goes for the flop Q out. So I just connected the scope probe to pin 14 to monitor the supply jiggles [being this my main point..], and has taken the diagrams.
Well, that connector in the reality was going to the CLK pin! and terminating that pin really has the miracolous effect of - eliminating the output signal...[by going below threshold], and thus eliminating the switching transients.. Ok, just leaving a small glitch so as to create a greater confusion..
So, if you don't mind, and still have patience for me, I would put up again some repeated tests. By the way, all this inflicts only those tests done "terminated". The others remain fully valid - though channel 1 is not the output, just the clock.
In the previous setup I also was not really happy about the long ground lead of the probe - so now I used a nicer probe with a direct ground lead [P6243 - 1pF/1Mohm/1G]. Notwithstanding, I still would not dare to claim that what I see is purely a power glitch.
So, here it is: an F74, CH1 is Qout, connected to the scope through a short piece RG174 through LEMO connectors. CH2 is thel scope probe between pin 7 -14, now short leads.
I am terribly sorry - or rather ashamed! [again, given that it has happened already once..]
My only excuses are these phrases:
And also this:
So, what happened is this: being a not serious person, when I was reaching for the old jig I did go by memory, and supposed that the on-board connector goes for the flop Q out. So I just connected the scope probe to pin 14 to monitor the supply jiggles [being this my main point..], and has taken the diagrams.
Well, that connector in the reality was going to the CLK pin! and terminating that pin really has the miracolous effect of - eliminating the output signal...[by going below threshold], and thus eliminating the switching transients.. Ok, just leaving a small glitch so as to create a greater confusion..
So, if you don't mind, and still have patience for me, I would put up again some repeated tests. By the way, all this inflicts only those tests done "terminated". The others remain fully valid - though channel 1 is not the output, just the clock.
In the previous setup I also was not really happy about the long ground lead of the probe - so now I used a nicer probe with a direct ground lead [P6243 - 1pF/1Mohm/1G]. Notwithstanding, I still would not dare to claim that what I see is purely a power glitch.
So, here it is: an F74, CH1 is Qout, connected to the scope through a short piece RG174 through LEMO connectors. CH2 is thel scope probe between pin 7 -14, now short leads.
Attachments
And the same, terminated. These two pic are really a mess, I have tried to position trace 2 [CH2] in the middle, so CH1 extends symmetrically up/down.
What I see here, is the contrary of those previously said [and faithful to good old logic..] The AC logic creates high current glitches at crossing over, which wreaks havoc here with my so-so bypassing structure. [Do not forget, this was my intention] These current spikes are getting worse by the termination..
Now, it is definitely possible to use a better bypass setup, and have much less ringing - but, in the same time, look out for the F logic test - much smoother in the same setup. In fact, F logic in a proper bypass situation have shown diminishingly small glitches.
In the future I would like to see also other logic families.
Thanks again, and forgive me!
George
What I see here, is the contrary of those previously said [and faithful to good old logic..] The AC logic creates high current glitches at crossing over, which wreaks havoc here with my so-so bypassing structure. [Do not forget, this was my intention] These current spikes are getting worse by the termination..
Now, it is definitely possible to use a better bypass setup, and have much less ringing - but, in the same time, look out for the F logic test - much smoother in the same setup. In fact, F logic in a proper bypass situation have shown diminishingly small glitches.
In the future I would like to see also other logic families.
Thanks again, and forgive me!
George
Attachments
- Status
- Not open for further replies.