Continue to Site

Welcome to EDAboard.com

Welcome to our site! EDAboard.com is an international Electronics Discussion Forum focused on EDA software, circuits, schematics, books, theory, papers, asic, pld, 8051, DSP, Network, RF, Analog Design, PCB, Service Manuals... and a whole lot more! To participate you need to register. Registration is free. Click here to register now.

Has anyone used any so called Analog Synthesis tools?

Status
Not open for further replies.

hdmi

Junior Member level 3
Joined
Feb 1, 2008
Messages
29
Helped
1
Reputation
2
Reaction score
0
Trophy points
1,286
Activity points
1,549
I wonder if anyone has any experience using any kind of so called analog synthesis tool and what is your opinion on the usefullness of these tools from scale 1 to 10, with 10 being extremely useful.
What I mean is tools like neockt from Cadence, or Circuit Explorer from Synopsys (acquired from Analog Design Automation couple years ago),

https://www.cadence.com/products/custom_ic/neocircuit/index.aspx
**broken link removed**

What these tools claim they can do are given any circuit topology, and the spec (GBW, noise margin, PSRR, etc.), they will call SPICE (either external or internal)and their optimization engine hundred of thousand times, finally come up with an optimized ckr which is supposed to meet the spec.
 

Yes, you could say that I have used such tools, since I was the architect for analog synthesis at Antrim Design Systems. The tools you mentioned both perform circuit optimization, and they can be a very useful tool for designers. Like all tools, you need to understand how they work and not rely on just pushing a button.

What questions do you have?

Mike D.
**broken link removed**

p.s. I should disclose that I work at Synopsys, but I am not directly asspciated with Circuit Explorer. (Though I can get expert answers from that team for you).
 

I think the approach of push button synthesis tool is very limited. You often stick with a mismatch of the synthesis topology with the circuit requirement. Take an example!

If you need an differential opamp used as integrator in an active RC filter the common mode input voltage is nearly fixed. Second the loop stability have to test including the integrator and the filter.

Does the topology take advantage of the limited input common-mode range? Is the stability reflecting the filter feedback or embedding network?

This is a simple circuit where most synthesis assumption fail! And that for the most stressed topology domain. Furtheron and to my opinion the designer lacks learning from the synthesis result. So to prepared for the untested situations. Because you could only verify would you expect. All the circuits which fails after fabbing where not verified for the case happen in silicon or lacking modeling quality.

I think it is better to focus on analog design automation to improve designers productivity.
 

The_World_is_Analog said:
Yes, you could say that I have used such tools, since I was the architect for analog synthesis at Antrim Design Systems. The tools you mentioned both perform circuit optimization, and they can be a very useful tool for designers. Like all tools, you need to understand how they work and not rely on just pushing a button.

What questions do you have?

Mike D.
**broken link removed**

p.s. I should disclose that I work at Synopsys, but I am not directly asspciated with Circuit Explorer. (Though I can get expert answers from that team for you).

Mike:

My question or my curiosity is that conceptually, this type of tools sounds great, however, we all know NeoCircuit first from CMU folks (NeoLinear) has been around for more than 10 years, and in reality, 8 out 10 analog designers today never heard this kind of tool, if it is truly so useful, why isn't it more widespreading in so many years, nore does Cadence or Synopsys seem to push hard to sell this kind of tool into the user. Also, Barcelona is a total disaster, although the approach is differnt. So can you shed some light on what issues in specific that prevent it from widespreading? Set up issues? Hard to use? Can you be a little bit more specific when you say "Like all tools, you need to understand how they work and not rely on just pushing a button.", before a user push the button, what does (s)he have to do in order to use the tool successfully, and how difficult is it?

Thanks,
 

Hi,

First I should say that I make a distinction between tools that describe themselves as analog synthesis solutions, and circuit optimizers - which is what NeoCircuit and CircuitExplorer are. I don't have the time or space to write all my thoughts here on the topic of analog synthesis, but you can find some of my recent thoughts on my blog at http://synopsysoc.org/analoginsights/

I see optimization as a necessary beginning step toward more automated analog synthesis. That was also the starting point I used when I was designijng our system at Antrim. Early optimizers were not good enough for production use. A second issue that slowed their adoption was that by its very nature, analog optimization requires a lot of simulation. It is only in recent years that parallel SPICE and Fast-SPICE have been powerful enough to be used as engines for optimization. Prior to that, optimizaers could only be used effectively at companies that had unlimited seats of their in-house simulator.

So, I think this is a good time for designers to look at optimization tools to speed up the design process. When I would talk to engineers about my work at Antrim, I often challeneged them to tell me how they used simulation to do design. If you look at it, it is generally a lot of manual optimization. That is unecessary, but it takes time to prove it to designers who are well schooled on SPICE, but not opimizers.

-Mike
 

Mike:

When I say Synthesis Tool, I already mean [Analog Synthesis = SPICE + Optimizer], as if the circuit simulator is a built-in engine or is external to the tool (like NeoCircuit or Circuit Explorer), it is a different story and there are pros and cons to both approaches. I don't want to get into Parallel SPICE, fast SPICE, etc. either, it's probably off the topic. But execuse my ignorance, I don't know Atrium's approach much, since you have experiences on both Circuit Explorer and Atrium's tool, how do you compare both? Strength and weakness on both?

thanks,

Added after 30 seconds:

Mike:

When I say Synthesis Tool, I already mean [Analog Synthesis = SPICE + Optimizer], as if the circuit simulator is a built-in engine or is external to the tool (like NeoCircuit or Circuit Explorer), it is a different story and there are pros and cons to both approaches. I don't want to get into Parallel SPICE, fast SPICE, etc. either, it's probably off the topic. But execuse my ignorance, I don't know Antrim's approach much, since you have experiences on both Circuit Explorer and Antrim's tool, how do you compare both? Strength and weakness on both?

thanks,

Added after 3 minutes:

Mike:

When I say Synthesis Tool, I already mean [Analog Synthesis = SPICE + Optimizer], as if the circuit simulator is a built-in engine or is external to the tool (like NeoCircuit or Circuit Explorer), it is a different story and there are pros and cons to both approaches. I don't want to get into Parallel SPICE, fast SPICE, etc. either, it's probably off the topic. But execuse my ignorance, I don't know Antrim's approach much, since you have experiences on both Circuit Explorer and Antrim's tool, how do you compare both? Strength and weakness on both?

thanks,
 

Status
Not open for further replies.

Similar threads

Part and Inventory Search

Welcome to EDABoard.com

Sponsor

Back
Top