I don't understand what asking for AC coupling has to do with not taking classes?
...
That's where Ron H's suggestion of ac coupling and DC offset come in, and if it is a solution that he suggests, then i guess it's not wrong asking for a little more clarification.

AC coupling, biasing, and DC coupling are concepts given in every introductory course on electronics, these methods are used in all types of circuits, from simple transistor amplifiers to sample and hold ADCs. That is why I asked why you seem to have never heard of it.

An example is Question 5 in the AAC E-Book (Book located at top of page). The biasing of transistor, with the AC coupling capacitor, are part of the basic 1 transistor (or 1 tube) amplifier. This circuit is given in all intro to circuits classes, do you recognize it? Ron H is suggesting you use a similar method to bias your signal into the operating range of your BBD.

 

AC coupling, biasing, and DC coupling are concepts given in every introductory course on electronics, these methods are used in all types of circuits, from simple transistor amplifiers to sample and hold ADCs. That is why I asked why you seem to have never heard of it.

An example is Question 5 in the AAC E-Book (Book located at top of page). The biasing of transistor, with the AC coupling capacitor, are part of the basic 1 transistor (or 1 tube) amplifier. This circuit is given in all intro to circuits classes, do you recognize it? Ron H is suggesting you use a similar method to bias your signal into the operating range of your BBD.

Hi,

Yes i recognize it.

Maybe i worded my questions differently, but hey, English is my Third language, not first or second.


Anyway, what i meant was, why do i need to use AC coupling to bias my BBD?

My current understanding is that as the signal fluctuates between +ve and -ve values, the +ve ones might go fine through the BBD, but when the -ve part comes, i get spikes, like in the simulation results i posted earlier.

I want to know what the issue is here, the circuit, the signal or the Cadence software itself.

Cadence does give glitches, and tries to give results which it shouldn't. So i need to confirm if i am wrong or Cadence.

Thanks.

 

What Ron H is proposing is essentially the same thing as biasing a BJT transistor so the signal is in the operating area.

This "moves" the entire signal from -ve to +ve to be entirely +ve, the signal then goes through the delay, and at the output, another capacitor removes the bias applied so the delay could function, leaving an AC -ve to +ve signal.

If you tried to amplify the -ve to +ve signal, the BJT transistor would only amplify the +0.6V to +ve portion, since any lower voltages would turn the transistor off entirely. This is the problem you are seeing with your current attempt of a BBD.

Think of a sinewave alternating from +1V to -1V, if you feed that through a capacitor, which is connected to a voltage divider, the -1V to +1V AC signal then becomes a 0 to 2V signal, this can be amplified, delayed, or any other action, then the output goes through another capacitor to remove the +1V "DC Bias" from the signal, resulting in a -1V to +1V Sinewave.

Since semiconductors are generally polarized (only work with positive voltages or only negative voltages) this issue comes up frequently.

There are more elaborate and accurate methods, such as an op amp to add the bias, but for simple signals, capacitor coupling suffices.

 

What Ron H is proposing is essentially the same thing as biasing a BJT transistor so the signal is in the operating area.

This "moves" the entire signal from -ve to +ve to be entirely +ve, the signal then goes through the delay, and at the output, another capacitor removes the bias applied so the delay could function, leaving an AC -ve to +ve signal.

If you tried to amplify the -ve to +ve signal, the BJT transistor would only amplify the +0.6V to +ve portion, since any lower voltages would turn the transistor off entirely. This is the problem you are seeing with your current attempt of a BBD.

Think of a sinewave alternating from +1V to -1V, if you feed that through a capacitor, which is connected to a voltage divider, the -1V to +1V AC signal then becomes a 0 to 2V signal, this can be amplified, delayed, or any other action, then the output goes through another capacitor to remove the +1V "DC Bias" from the signal, resulting in a -1V to +1V Sinewave.

Since semiconductors are generally polarized (only work with positive voltages or only negative voltages) this issue comes up frequently.

There are more elaborate and accurate methods, such as an op amp to add the bias, but for simple signals, capacitor coupling suffices.

Except he needs gain on the front end, but I have no idea how much.

 

Anyway, what i meant was, why do i need to use AC coupling to bias my BBD?

I have told him the reason several times. I think that my suggestions go in one eye and out the other. He is waiting for a schematic that he can build.

It ain't gonna happen, at least not from me. I have explained how to do this until I'm blue in the face.

 

What Ron H is proposing is essentially the same thing as biasing a BJT transistor so the signal is in the operating area.

This "moves" the entire signal from -ve to +ve to be entirely +ve, the signal then goes through the delay, and at the output, another capacitor removes the bias applied so the delay could function, leaving an AC -ve to +ve signal.

If you tried to amplify the -ve to +ve signal, the BJT transistor would only amplify the +0.6V to +ve portion, since any lower voltages would turn the transistor off entirely. This is the problem you are seeing with your current attempt of a BBD.

Think of a sinewave alternating from +1V to -1V, if you feed that through a capacitor, which is connected to a voltage divider, the -1V to +1V AC signal then becomes a 0 to 2V signal, this can be amplified, delayed, or any other action, then the output goes through another capacitor to remove the +1V "DC Bias" from the signal, resulting in a -1V to +1V Sinewave.

Since semiconductors are generally polarized (only work with positive voltages or only negative voltages) this issue comes up frequently.

There are more elaborate and accurate methods, such as an op amp to add the bias, but for simple signals, capacitor coupling suffices.

Thanks for the explanation.

This explains why i only see clocks showing for the -ve part of the signal in my CMOS design.

I will try the biasing then.

 

Thanks for the explanation.

This explains why i only see clocks showing for the -ve part of the signal in my CMOS design.

I will try the biasing then.

If you didn't understand what Ron H was suggesting you try, it was due to the fact that Ron H assumed you already knew that information from earlier classes.

If you did not understand the purpose or concept of what he was telling you, you SHOULD have stated that you did not understand the concept/purpose of the change.

Instead, you've been simply asking for Ron H to do all the work for you, then, rather than thanking him with the "Thanks" button, you have either argued, or asked for a "Working Circuit", rather than asking for "How will this work".

 

If you didn't understand what Ron H was suggesting you try, it was due to the fact that Ron H assumed you already knew that information from earlier classes.

If you did not understand the purpose or concept of what he was telling you, you SHOULD have stated that you did not understand the concept/purpose of the change.

Instead, you've been simply asking for Ron H to do all the work for you, then, rather than thanking him with the "Thanks" button, you have either argued, or asked for a "Working Circuit", rather than asking for "How will this work".

My Posts 96,98 and 100 on page 10 highlight the fact that i did ask for clarification where needed.

I am not looking for a circuit. I am looking for the direction to go to. After all, the concept of BBD itself i looked on Google and IEEEXplore pages, not here.

Anyway, i am sorry for any misunderstandings.

 

The likelyhood of finding an analog BBD on silicon from google is very slight.

However, the suggestions Ron H has been supplying to you can be found Easily by searching Google for them, if you didn't fully understand his intent. You keep asking for him to do more work. I, personally, would have bailed on this thread a long time ago. Ron, however, is a great guy and is really trying to help you understand. From my read, however, you seem to think "AAC (and who knows how many other forums you posted this same question on) owe you a full explanation of how to perform this task". The only work you've done is take screenshots and simulate, then come back and say it didn't work, without putting thought into why. This is a sad state of affairs.

I don't mean to come off all grouchy old man here, BUT: The specific project of the class you are in has NOTHING to do with learning how to create an Analog Delay in Silicon with fab restrictions!

The class is attempting to tie together all of the concepts, math, algorithms, and rules you (should have) learned in previous classes, and apply them in silicon design to prove you have a complete understanding of the course work.

I hope you didn't sell your intro to electricity/electronics books as soon as the semester was over, at which point you purged everything you learned as "I'll never be working with THAT!, I'm Majoring in xxxxxx"! It's silly that I should even need to be exposed to that headache of stuff.

If you do fall into the the latter category, then, for nothing more than not degrading having an EE/CS degree further, PLEASE realize that fact and change majors!

<RANT>

You WILL NOT be taught everything you need to know to succeed in EE/CS in college. You WILL be taught theories, common methods, algorithms, and a lot of math that will allow you to refine your work to a single subset of a subset in CS or EE. If you are unable to take two abstract concepts, and realize that sticking them together may be a partial solution to your problem, you need a lot of course re-work and tutoring by a professional (Luckily, you have many here who will assist for free, but our patience isn't that extensive, and you need to be an active participant.)

I've stated the following MANY times on this forum, and it will hold true well into the future: Most of what I am organizing design teams to create did Not Even EXIST when I was going to college. If you choose EE/CS as a career, College is the easy part. You need to Love it, like an an artist likes slapping paint on canvas. If not, you will not be motivated to learn the latest designs and standards which are continually being updated and published by think tanks. This includes multiple programming languages, analog circuits (Which I suck at, BTW), Digital logic, mixed digital and RF, all on one board with components, sometimes on the same piece of silicon. It's not a field where a month can be spent to try "something different". This project would be a DSP project, designed, tested, and done in a few days. Trying to make it analog and perform as well as digital is... backwards, especially when you are theoretically being paid megabucks to think about it. Businesses turning out product simply cannot afford to have a half engineer/half R&D guy spending a ton of time on something just to be different, especially if the design parameters do not call for it.

You need to do some serious inspection of your goals, what you like to do, what you love to do, and what you don't like doing. There's nothing worse than going to a career everyday in a field you can barely keep up with and are continually worried about new technologies leaving you without a job. I will take a job I Love for less than half the salary of a similar job where I am miserable. Money isn't everything.
</RANT>

Do not take this as a personal attack. Look at it as "advice from self 30 years into the future".

 

Well said, TOG!
I posted a short rant a couple of hours ago, but deleted it within about a minute.
One statement that I deleted I will repost, though:
I am totally burned out on this thread.

 

I am totally burned out on this thread.

You have the patience of Job. I burned out on the thread 3/4 of the way through it, and I was only READING it!

 

The likelyhood of finding an analog BBD on silicon from google is very slight.

However, the suggestions Ron H has been supplying to you can be found Easily by searching Google for them, if you didn't fully understand his intent. You keep asking for him to do more work. I, personally, would have bailed on this thread a long time ago. Ron, however, is a great guy and is really trying to help you understand. From my read, however, you seem to think "AAC (and who knows how many other forums you posted this same question on) owe you a full explanation of how to perform this task". The only work you've done is take screenshots and simulate, then come back and say it didn't work, without putting thought into why. This is a sad state of affairs.

I don't mean to come off all grouchy old man here, BUT: The specific project of the class you are in has NOTHING to do with learning how to create an Analog Delay in Silicon with fab restrictions!

The class is attempting to tie together all of the concepts, math, algorithms, and rules you (should have) learned in previous classes, and apply them in silicon design to prove you have a complete understanding of the course work.

I hope you didn't sell your intro to electricity/electronics books as soon as the semester was over, at which point you purged everything you learned as "I'll never be working with THAT!, I'm Majoring in xxxxxx"! It's silly that I should even need to be exposed to that headache of stuff.

If you do fall into the the latter category, then, for nothing more than not degrading having an EE/CS degree further, PLEASE realize that fact and change majors!

<RANT>

You WILL NOT be taught everything you need to know to succeed in EE/CS in college. You WILL be taught theories, common methods, algorithms, and a lot of math that will allow you to refine your work to a single subset of a subset in CS or EE. If you are unable to take two abstract concepts, and realize that sticking them together may be a partial solution to your problem, you need a lot of course re-work and tutoring by a professional (Luckily, you have many here who will assist for free, but our patience isn't that extensive, and you need to be an active participant.)

I've stated the following MANY times on this forum, and it will hold true well into the future: Most of what I am organizing design teams to create did Not Even EXIST when I was going to college. If you choose EE/CS as a career, College is the easy part. You need to Love it, like an an artist likes slapping paint on canvas. If not, you will not be motivated to learn the latest designs and standards which are continually being updated and published by think tanks. This includes multiple programming languages, analog circuits (Which I suck at, BTW), Digital logic, mixed digital and RF, all on one board with components, sometimes on the same piece of silicon. It's not a field where a month can be spent to try "something different". This project would be a DSP project, designed, tested, and done in a few days. Trying to make it analog and perform as well as digital is... backwards, especially when you are theoretically being paid megabucks to think about it. Businesses turning out product simply cannot afford to have a half engineer/half R&D guy spending a ton of time on something just to be different, especially if the design parameters do not call for it.

You need to do some serious inspection of your goals, what you like to do, what you love to do, and what you don't like doing. There's nothing worse than going to a career everyday in a field you can barely keep up with and are continually worried about new technologies leaving you without a job. I will take a job I Love for less than half the salary of a similar job where I am miserable. Money isn't everything.
</RANT>

Do not take this as a personal attack. Look at it as "advice from self 30 years into the future".

Thanks for the mentoring lesson, but i am passionate about my studies and goals in life.

I think we got off on the wrong wavelengths. I was looking for the right direction to be pointed to, that's all.

I have repeatedly said this, and would say again, i appreciate the efforts by you and Ron H to help in this regard.

I have also clarified that i know DSP is a much simpler and quicker approach, however i am designing my circuit as part of bigger system on chip. Everything is supposed to be analog to minimize power consumption, design and construction complexity, because this is for an implant to be used for many years.


I know that typing long responses on the forum might be tedious for frequent posters like you guys, but hey, it helps newbies like us in understanding. When you explained the biasing concept, it made it crystal clear for me and brought the picture of I-V characteristics of a BJT in my mind, so i could go back and dig up more info.

That's all i want. Being pointed in the right direction.

I don't take things personally.....definitely not on the internet.

So here is a question:

I really didn't understand what you meant by not being able to find BBD lines in silicon on Google. Can you please clarify that?

 

I really didn't understand what you meant by not being able to find BBD lines in silicon on Google. Can you please clarify that?

We get many homework questions here. Most people just want the answer handed to them.

Sometimes, after a few questions, they will figure out the right words to talk to google with to find the answer to their problem. I was simply stating that the odds of finding a BBD circuit completed and tested published somewhere on the internet is very slight.

 

Ron,
It surprises me that it took you this long to burn out. 130 posts!
"You're a better man than I am, Gunga Din!" (R. Kipling)
Ken

 

Ron,
It surprises me that it took you this long to burn out. 130 posts!
"You're a better man than I am, Gunga Din!" (R. Kipling)
Ken

Yeah. I kept thinking, "If I just answer this question, he can take it from here."

"Do you like Kipling?"
"I don't know. I've never Kippled."

 

Yeah. I kept thinking, "If I just answer this question, he can take it from here."

Sounds like a marriage falling apart. "If we can just get over this particular fight, we'll be fine...".

"Do you like Kipling?"
"I don't know. I've never Kippled."

I know that was a joke, but try it more often, they're almost all available at the Library of Congress online.
Gunga Din

 

We get many homework questions here. Most people just want the answer handed to them.

Sometimes, after a few questions, they will figure out the right words to talk to google with to find the answer to their problem. I was simply stating that the odds of finding a BBD circuit completed and tested published somewhere on the internet is very slight.

Well, i did find some stuff on Google and IEEExplore pages.

Then i looked through the datasheets of the chips that were built on the BBD concept. The datasheets show what is contained in the chip. The only difference seemed to be the Fab process they used and what i can use now. But when i implemented the design, i didn't get the results.

I guess i will post my simulation results for further discussion. The delay that the chips offered was something from 1.5ms to 52ms!

I am looking for 1ms, meaning not so long of the BBD line and hence not so much distortion.

Plus i have yet to do the layout as well, to see if this all can fit physically on the die or not.

 

Yeah. I kept thinking, "If I just answer this question, he can take it from here."

"Do you like Kipling?"
"I don't know. I've never Kippled."

I didn't mean to make you angry. I apologize for that.

You've been great help so far!

 

I didn't mean to make you angry. I apologize for that.

You've been great help so far!

Not anger, as much as frustration.

 

Not anger, as much as frustration.

Okay okay, from now on i will post my schematics when i have them, and clearer questions.