LTSpice I/O Device modeling

As I mentioned in a Lab thread, I'm trying to get up to speed with LTSpice to prototype some stuff virtually to reduce the times I have to disconnect the smoke detector near my workbench. :roll:

In addition to the circuit I'm prototyping, How crazy does one usually need to get in modeling the characteristics of the signal source and destination? Like for example, if I'm designing something to be fed from a typical (or specific) passive instrument pickup, do I need to include inductance and such of the pickup; and so on?

Tnx

--Bob

[quote author="stickjam"]As I mentioned in a Lab thread, I'm trying to get up to speed with LTSpice to prototype some stuff virtually to reduce the times I have to disconnect the smoke detector near my workbench. :roll:

In addition to the circuit I'm prototyping, How crazy does one usually need to get in modeling the characteristics of the signal source and destination? Like for example, if I'm designing something to be fed from a typical (or specific) passive instrument pickup, do I need to include inductance and such of the pickup; and so on?

Tnx

--Bob[/quote]

Depends on how accurate you want your sim to be :grin:

If the source impedance is going to be significantly loaded by the amplifer etc. then the inductive component could be quite important.

But you can have some idea/guess as to the source Z and throw in some values around the vicinity just to see how sensitive things are.

From a noise standpoint the inductive component of a source will make the current noise at the amp input important at frequencies where the inductive reactance gets significant compared to the resistance. An example: moving magnet phono cartridges.

> passive instrument pickup, do I need to include inductance and such of the pickup

Stuff where there is a Standard, or a Custom, you can sometimes ignore the reality of the source.

If you build a guitar amp input that looks an awful lot like 470K plus an open grid (say 100pFd), and drive it with a low-Z audio oscillator, you can compare your results to existing guitar amps tested with a low-Z source.

If you want to know what REALLY happens on stage, you better plug in a bunch of milliHenries, an on-guitar volume pot, maybe a tone pot and cap, and the 100pFd-1,000pFd of the guitar cord. And you better take account of the fact that the pickup is sorta velocity sensitive, the string is neither constant-amplitude nor constant-velocity, you have six strings all different and each string will be fretted to all different lengths....

Actually you don't design a guitar amp, you build something and play it, tweak, repeat daily for 5 to 50 years. Since others have been down this road, you can try approximating the input Z and overall gain/response of an existing design measured from a handy low-Z source, and let the pickup impedances take care of themselves. You know a guitar amp will "work" with hi-Z input, about 10mV-100mV input sensitivity for max output, a mild treble rise and optional midrange dip.

If you are going to try low-Z loading of guitar pickups, you better collect data on a LOT of pickups and onboard vol/tone circuits and figure your EQ to get the tone back to the Fender/Gibson mainstream reference.

If you have any feedback to the input device, you need to check for stability with odd reactances at the input. The canonical phono preamp is 47K||300pFd and if you do that, you can let cartridge impedance be the pickup designer's problem. But there was a large class of 2-transistor phono preamps where the feedback into the first emitter mucked-up the actual input impedance, in ways that were not obvious with low-Z drive but gave significant response errors with real pickups. OTOH there are feedback values that will motorboat badly with zero-Z drive yet be stable with the few-K resistance of most (not all) phono pickups.

Or in short: it depends. Use your brain and expect to be wrong.

Posted: Thu Oct 20, 2005 11:36 pm by PRR:

If you build a guitar amp input that looks an awful lot like 470K plus an open grid (say 100pFd), and drive it with a low-Z audio oscillator, you can compare your results to existing guitar amps tested with a low-Z source.

If you want to know what REALLY happens on stage, you better plug in a bunch of milliHenries, an on-guitar volume pot, maybe a tone pot and cap, and the 100pFd-1,000pFd of the guitar cord. And you better take account of the fact that the pickup is sorta velocity sensitive, the string is neither constant-amplitude nor constant-velocity, you have six strings all different and each string will be fretted to all different lengths

Click to expand...

Thanks PRR. It's 66% worse than that, cause I've got 10 strings! :wink: Acutally I suppose it's easier because I'm designing for a specific instrument only--a Chapman Stick with a Villex passive Block. There is a volume pot, a tone switch and another to select pickup configuration (bridge/neck*/inphase/outphase.)
*Ha! It's all neck! :grin:

I know I'll need to actually build something and play through it to know for sure, but thinking the LTSpice route is a more convenient route to get something close without having to deal with stray capacitances and intermittent behavior on a prototype block or soldering up a perfboard kludge. Besides, I'm starting to dabble with tubes after a 30-year absence from my repertoire, so my prototyping methods would obviously need to adapt. Of course if I was stupid enough not to care about safety, a proto-block with DIP-to-octal adapters might look amusing! :roll:

Here's my idea, and feel free to make suggestions or even shoot holes in it if it won't work: I'll play the Stick dry, directly into my DAW interface to create WAV files of various playing styles at various pickup/eq settings. then I'd set up a circuit in LTSpice like this...

• pickup/cable model
  ---> voltage source using a WAV file
  ------> circuit under design
  ---------> appropriate output load

Then I'd use Transient Analysis and the .WAVE command at the output node to create WAV files which represent what the thing should theoretically sound like.

Have any of you folks tried something like this? How well might/would/does this work? I'm assuming one would put the input voltage source there because the WAV file I'm using represents what the signal looks like at that point in the circuit. Would putting the pickup/cable model there be effective as expected, or would it simulate an unrealistic situation since the signal really starts at the pickup coil as opposed to being injected midstream?

Thanks. -- Bob

> create WAV files which represent what the thing should theoretically sound like.

Well, yes, you could do that.

Though: most soundcard/DAW inputs are medium-Z and line-level, while it appears your axe is very high-Z and low-level. Since it is not really a problem to give it a hi-Z input, I'd at least do that. And maybe some gain too, to overcome DAW noise.

And then, looking at their suggested preamp (the modded RANE), you are nearly done. Their preamp adds some switching and gain control. It has a steep adjustable low-cut to knock down thuds. It has a rather ordinary graphic EQ, which may be fun, but is more easily bought than built, or you might just try a simple 2 or 3 band tone control. Then some level controls and output mixing.

I think you could nail it up on a wooden board and have a test rig that worked good. If you could live without any EQ, two stages of 6SL7 volt-amp and a 6SN7 cathode follower (to drive solid-state inputs) would about do one channel. Two 6SL7, one per pickup, and a cathode follower might do for the dual-channel axe. Two cathode followers can make a fine low-cut filter. A 6SN7 and a James tone control, or 6SN7 and a Baxandall, would give some tonal freedom. Stages like this don't need simulation: wire 47K-270K plate to B+, 1Meg grid to ground, and cathode resistor 1K to 3K to set the plate about halfway up the B+. It can be done without hardly any smoke at all.

Thanks PRR.

I was Googling about tube breadboarding--actually finding surprisingly little. But I found enough to jog long-buried memories as a preteen, building a shortwave receiver on a scrap 2x6 studded with brass tacks. I now recall that was the contraption that made my grandfather remark that "one of these days, you're gonna have a spark fly ten feet out of your a$$!" :grin:

Anyway, I grabbed a chunk of plywood and started to throw together a prototype warmth/drive stage in an attempt to give my all-solid-state Hammond X5 some balls. See this thread on that project--I'd still like more feedback on the various stuff I'm thinking about doing ot it. (B3 Jim? :sam:*) It ended up a bit noisy with the old used "mystery brand" 12AX7 I had on hand. I've gotta make a trip to the only "boatanchor heaven" left in town for some NOS tubes and a real transformer to replace the two Radio Shack jobs that I kludged back-to-back to get a respectable B+. Still, the aroma of decades of grime and fingerprints being baked onto that solitary vacuum envelope was like incense, time-transporting me back to 1969!

:?: Before I go tube shopping, any suggestions on what tube types and brands might be the best choices for low noise yet overdrive in a most pleasingly "Hammondy" way? (I'm not stuck on the 12AX7--just that and a couple 6V6's is all I had in the boneyard at that time.)

Regarding the Stick preamp, I have a design in mind... I acquired a schematic from someone to diagnose/repair a 5ad0w5ky Bass Preamp. Unfortunately, I was specifically asked not to pass the schematic on to anyone else--sorry. (I'm sure it's reverse-engineerable, although I didn't want to--it's SMD and my eyesight isn't what it used to be. But you know, if there's a component value you just can't read... :wink: ) It is JFET-based so I thought I'd try my hand at adapting it to tubes. (Ironically, I've heard that that the sand-state version was based on an existing tube design!) I'll probably pick up a pile of 6SL7s and 6SN7's since a phono pre is also on my DIY to-do list. :?: What brands of these have you found to be the nicest and cleanest?

Thanks

--Bob

*one of these days I'm actually going to have to get out to meet and buy ya' a Founder's! :wink: :sam:

> pick up a pile of 6SL7s and 6SN7's

Think 12SL7 12SN7: they are still cheaper and mostly unused 1950s stock.

> make a trip to the only "boatanchor heaven" left in town

Patronize your local shops while you can. But in most towns, the "local tube store" is http://TheTubeStore.com or http://TubesAndMore.com or http://www.triodeelectronics.com