Figured this is as good a place as any to express an opinion.
Seems we are seeing the rampant spread of simulatoritis around here. Folks are spending all their time pouring over simulations and not actually DIYing as much. A bad case of simulatoritis leads to a cold and neglected soldering iron!
Don't let this happen to you! The cure is simple; grab some parts, assemble a circuit and apply power. Don't worry about a little blue smoke, it's part of the gig...
Seems we are seeing the rampant spread of simulatoritis around here. Folks are spending all their time pouring over simulations and not actually DIYing as much. A bad case of simulatoritis leads to a cold and neglected soldering iron!
Don't let this happen to you! The cure is simple; grab some parts, assemble a circuit and apply power. Don't worry about a little blue smoke, it's part of the gig...
Simulation is a wonderful thing to show you what doesn't work. That can save a lot of time and aggravation. It's true that a circuit that sims well may not live up to its promise, so at some point you need to build the circuit and test it.
Hm... good point!
It does nowever seem to have very much the same symptoms as beuracratitis vulgaris - it completely paralyzes all practical action , - BV however also severely affects logic and reason.......
It does nowever seem to have very much the same symptoms as beuracratitis vulgaris - it completely paralyzes all practical action , - BV however also severely affects logic and reason.......
I'm not against the use of simulators, I use LTspice regularly. For me it a virtual playground where I can do stupid things and get away with it. It also allows for 'proof of concept', which is nice for checking wether or not an idea is valid or not.
What I see is this fixation or simulated results and people passing up building something because design X isn't good, so the simulator says. My point is that some designs that sound really good are passed over due to what a simulator said. Things that should be built and appreciated are being left by the wayside solely due to virtual results from a piece of software. Very sad in my humble opinion.
What I see is this fixation or simulated results and people passing up building something because design X isn't good, so the simulator says. My point is that some designs that sound really good are passed over due to what a simulator said. Things that should be built and appreciated are being left by the wayside solely due to virtual results from a piece of software. Very sad in my humble opinion.
Well, some people like the theoretical side, and some simply enjoy building circuits. Some like both. Plenty of room for everybody.
I wholeheartedly agree that there is room enough for everybody. Different strokes for different folks, so it is said.
The issue is that the simulator, though very useful, has its caveats. It appears to be commonly accepted that the simulated output of a theoretical design can be an order of magnitude, or even substantially more, off from reality. We are at the mercy of the models we use, whether or not we can include realistic parasitics into the simulation and then wether or not the user has set appropriate parameters by which to run the simulation. Any of which singly or in combination may keep someone from creating something. This makes it tough to make the theory close to reality.
To me, the essence of DIY is the act of creation. Mental masturbation is fine during the conceptual phase and it is even beneficial to enable the exploration of various avenues a design could take. Ultimately the proof is in the pudding - you have to build it and via whatever means one chooses or has available, evaluate it.
I certainly agree you have to build it before you know if it works well. Generally though, the sims are optimistic, sometimes wildly so. I've seldom built something that simed badly, but turned out to work well on the bench.
Hi,
Sims are just a tool, can be used well, can be used badly, I like TinaTi.
Will tell you some things, won't tell you others, which mainly for hifi
is the schematic doesn't usually accurately reflect good build and
layout practise, and even if did, say star earthing, the sim doesn't
care unless you start adding nebulous components into the rails.
I'm interested in what sims badly but actually works well ..... and why .....
rgds, sreten.
Sims are just a tool, can be used well, can be used badly, I like TinaTi.
Will tell you some things, won't tell you others, which mainly for hifi
is the schematic doesn't usually accurately reflect good build and
layout practise, and even if did, say star earthing, the sim doesn't
care unless you start adding nebulous components into the rails.
I'm interested in what sims badly but actually works well ..... and why .....
rgds, sreten.
Anything that takes advantage of unmodelled properties, or anything that is exceedingly sensitive to numerical errors. Regarding the first class, you could think of transitron-like pentode circuits, or bipolar transistor circuits based on avalanche multiplication. I haven't heard of anyone using these for audio, though. From the second class, crystal oscillators come to mind, whether they start up in a transient analysis depends on the numerical integration method.
I was working on an LLC SMPS and got a bit stuck.
So I got into ltSpice and that simulated my circuit very well.
I was able to change component values and see what effect it had.
I now have a working SMPS thanks to ltSpice.
So I got into ltSpice and that simulated my circuit very well.
I was able to change component values and see what effect it had.
I now have a working SMPS thanks to ltSpice.
LTSpice (Switcher Cad) was designed specifically to simulate Linear Tech SMPS designs and does that quite well, fortunately it does analog simulations quite well in general.
These days everything I design and build gets simulated first. I usually have a very strong idea of what it is I am going to build ahead of simulation and just use Spice to refine the finer points of the circuit design, and to avoid obvious mistakes.. It has proven effective in reducing iterations as it gets me close and a little additional time on the bench to refine it is generally all I need to do. It's a big time saver for me, I build a lot of stuff. I generally enjoy the end result the most of all until the bug bites again and I have to build something else. 😀
These days everything I design and build gets simulated first. I usually have a very strong idea of what it is I am going to build ahead of simulation and just use Spice to refine the finer points of the circuit design, and to avoid obvious mistakes.. It has proven effective in reducing iterations as it gets me close and a little additional time on the bench to refine it is generally all I need to do. It's a big time saver for me, I build a lot of stuff. I generally enjoy the end result the most of all until the bug bites again and I have to build something else. 😀
actually, simulation tendency appeared more hot a few years ago, and its like more actually being built these days
just look at the many active GB threads
not sure there have ever been this many before
but may of little comfort 😀
but I also notice a strange thing
its like some things never changes, and keeps popping up again and again, with no end
just look at the many active GB threads
not sure there have ever been this many before
but may of little comfort 😀
but I also notice a strange thing
its like some things never changes, and keeps popping up again and again, with no end
A useful analogy might be that of simulators used by airlines; it is better that mistakes be made and corrected before it "hurts" some one or something?
Abs
Abs
As sreten pointed out earlier in the thread it's just a tool, a good one imo, and there have always been the armchair rail fans, err blabbers rather than builders, err non-participant speculators, umm I rather seem to be putting my foot in it.. 😛 😀
It is nice to see all of the group buy threads, and I hope a large number of people here are actually deriving the pleasure that comes from knowing they built it themselves. Shame not to experience that sense of accomplishment.. I love music which is also why I enjoy building my own audio gear - in some sense it gets me closer to the music I enjoy via a process that itself is enjoyable to me. And I can't play anything to save my life.. 😀 😀
It is nice to see all of the group buy threads, and I hope a large number of people here are actually deriving the pleasure that comes from knowing they built it themselves. Shame not to experience that sense of accomplishment.. I love music which is also why I enjoy building my own audio gear - in some sense it gets me closer to the music I enjoy via a process that itself is enjoyable to me. And I can't play anything to save my life.. 😀 😀
Yes it does help in that regard imo, still nothing better than caution and care in building and checking it twice or thrice.. 😀 My most costly mistakes (rare fortunately) have always been those of execution rather than design. I've never smoked anything more costly than a cap or resistor however... (Some of which aren't particularly cheap either, but sure beats a power or output transformer.)
Yes. That is because sims use ideal components, but we don't.fpitas said:
In audio, almost nothing. In RF, this may happen sometimes when a bad design happens to work OK because of a lucky combination of parasitic components. It would still be a bad design and probably unreliable, and might stop working after a repair a few years later, but may still work well enough to fool/satisfy a DIYer for a whiloe.sreten said:
Physicists sometimes do very strange things with transistors and diodes, using exotic phenomena that aren't covered by any compact transistor or diode model. For example, using 1 A rated transistors to drive 40 A current pulses into a transmission line by using current-mode second breakdown, see R. J. Baker, "High voltage pulse generation using current mode second breakdown in a bipolar junction transistor", Review of Scientific Instruments, April 1991:
http://cmosedu.com/jbaker/papers/1991/RSI621991.pdf
In "Time domain operation of the TRAPATT diode for picosecond‐kilovolt pulse generation", the same author uses some obscure breakdown phenomenon in an ordinary rectifier diode to make subnanosecond pulses. Of course none of this has anything to do with audio.
http://cmosedu.com/jbaker/papers/1991/RSI621991.pdf
In "Time domain operation of the TRAPATT diode for picosecond‐kilovolt pulse generation", the same author uses some obscure breakdown phenomenon in an ordinary rectifier diode to make subnanosecond pulses. Of course none of this has anything to do with audio.
Quite a few of the schematics posted could have used a good dose of simulation.
Not the truth, but they can be very educational. Magic smoke is expensive. They can give hints to problems we can't easily measure. I have no way to measure more than the first harmonic of 20K. I can simulate it. I can put a bad load on an amp and probe around hoping to not find oscillations, or I can simulate the stability and may be close.
I have found them handy to see how different typologies behave and compared that to my selection of amps, noting what is different in the sim and what does my wife hear. Yes, we found correlations.
Besides, we need a good drafting tool anyway. Why not push a button and let the drafting tool do a simulation? Too easy not to use.
Not the truth, but they can be very educational. Magic smoke is expensive. They can give hints to problems we can't easily measure. I have no way to measure more than the first harmonic of 20K. I can simulate it. I can put a bad load on an amp and probe around hoping to not find oscillations, or I can simulate the stability and may be close.
I have found them handy to see how different typologies behave and compared that to my selection of amps, noting what is different in the sim and what does my wife hear. Yes, we found correlations.
Besides, we need a good drafting tool anyway. Why not push a button and let the drafting tool do a simulation? Too easy not to use.
- Status
- Not open for further replies.