EMRFD Message Archive 5700
Message Date From Subject 5700 2011-01-09 13:46:28 Rob Steenburgh Distorted output waveform from Wilderness Sierra Greetings:
Some folks over at the Wilderness Sierra Yahoo group suggested I post my
I've attached a photo of the oscilloscope trace of my Sierra output. I am
trying to understand what the distortion of the waveform is telling me.
The photo was taken when the Sierra was operating at 7MHz.
Power output was approximately 10W with the onboard DL QRP-PA2008 amplifier.
The output waveform was similarly distorted before I added the amplifier, and
seems to be present on all bands. Maximum drive for the QRP-PA2008 is 10mW.
I'm driving it with about 7mW.
I've found if I reduce the output to 1W or less, I get a smooth sine wave.
On the air, the tone sounds good and I've never gotten any reports of poor
quality. Similarly, I haven't experienced, or had any reports of RFI.
Any thoughts? Perhaps spurious harmonics superimposed on the original signal?
[Non-text portions of this message have been removed]
5702 2011-01-09 15:10:44 Tim Re: Distorted output waveform from Wilderness Sierra Attachment didn't make it through but a quick google turned up this pic (it might be it?) on Flickr:
One thought is, what sort of LPF does the QRP-PA2008 have? I know nothing about the QRP-PA2008... I would venture to guess based on the waveform that it's a transformer-coupled output push pull stage. The waveform you show is quite nice for a small push pull stage with no LPF but have sizable 3rd harmonic. Add on a simple LPF and it's a beauty.
You won't hear anything at all wrong with the note
5703 2011-01-09 16:02:19 rasteenb Re: Distorted output waveform from Wilderness Sierra Thank you for the reply, Tim.
You were absolutely right; the PA uses a Mitsubishi RD06HHF as the
driver and a pair of RD16HFFs as a push-pull PA. The PA uses the
band-pass filter on each of the Sierra's band modules.
The photo you found is an earlier version. It was taken before I
re-tuned the band-pass filter, yesterday, peaking the output as
described in the construction manual. Before peaking, I was getting
about 7W out. The current waveform picture is now posted at
Perhaps I need to un-peak the filter?
Cheers & 73,Rob
5704 2011-01-09 18:00:42 Tim Re: Distorted output waveform from Wilderness Sierra That's not bad either, pretty typical for a push-pull output when operating in class C. Now in addition to 21 MHz you're seeing 35 MHz and 49 MHz :-).
Just put it through a low pass filter and it'll be beautiful. I found a PA2008 schematic at http://www.qrpproject.de/qrppa2008.html - you need a low pass filter, that's all.
The Norcal sierra has a PA output lowpass filter (C47/L5/C48/L6/C49
5705 2011-01-09 18:34:43 rasteenb Re: Distorted output waveform from Wilderness Sierra I run the output of the PA2008 through the low pass filter on the Sierra's band module, so the LPF is after the PA.
I did some more research and found out later models of the Sierra included a ferrite bead on the base lead of Q6, the driver. This was added for parasitic suppression. My Sierra was one of the first models produced in the mid 90s and didn't have this ferrite bead.
So, I ordered a couple.
I'm also going to have another look at the PMO bandpass filter.
Tim, thanks again for your help with this. As I get the bead in place and tweak the PMO filter caps, I'll keep you posted on the results.
5706 2011-01-10 06:31:16 Tim Re: Distorted output waveform from Wilderness Sierra Is is possible you're measuring the waveform looking into the low pass and not at the output of the low pass, at the dummy load? Or that there's a lifted ground at the scope probe or at the low pass filter module?
Look in the EMRFD chapter on measurements for the "right way" of hooking the scope up to the dummy load and the many "wrong ways".
I'm really struggling trying to come up with a way to make a waveform like yours, after the low pass. Before the low pass is easy. After... is hard unless there's a lifted ground or something making the LP be completely ineffective. I can't imagine a missing ferrite bead in the pre-driver doing that. This is looking more and more like a missing ground
5707 2011-01-10 14:20:57 rasteenb Re: Distorted output waveform from Wilderness Sierra You can see the schematic of circuit I'm using to do the measurements
It combines a design for an RF probe and an RF sampler in one box. I
had to pitch the wire wound resistors. Found out even if they say
"non-inductive", they can still be inducing.
I just connect a dummy load to the part Antenna/Meter jack and put the
switch in the "scope" position.
Interestingly, when I leave the dpdt switch in the "meter" position, I
get a nice clean sine wave output on the oscilloscope, but it's only
5708 2011-01-10 15:29:07 Tim Re: Distorted output waveform from Wilderness Sierra The 5pF series capacitor and back to back diodes don't seem appropriate for measuring 10W of RF (never mind that 50 ohm 50 w resistor whatever that was there for). The 5pF is effectively a high-pass and will make it look like there's a lot more harmonic content than there actually is and above a certain level (That you may be reaching) the diodes will be clipping.
What's wrong with EMRFD figure 7.15, or 7.16, the "right ways" to hook up a scope?
5709 2011-01-10 15:59:43 kb1gmx Re: Distorted output waveform from Wilderness Sierra Your scope output will distort any signal that is greater than
5710 2011-01-10 19:01:23 rasteenb Re: Distorted output waveform from Wilderness Sierra I doubt there's anything wrong with EMRFD figure 7.15, or 7.16, the
"right ways" to hook up a scope but I don't have EMRFD in hand yet, so I
have no idea what the figures look like. I checked our local library
today, but they didn't have a copy, either. Don't keep me in suspense!
In any case, the sampler design comes from Phil, K6LS. His drawing says
it's good for about 200W <http://www.qsl.net/k6ls/rfsampler.html> .
The 50W, 50 Ohm resistor acted as part of both the sampler and the RF
probe, but being wire-wound, was not useful. I switched it out for four
2W non-inductive types. When in the "probe" position, the resistors are
an 8W dummy load.
5711 2011-01-10 19:25:07 davidpnewkirk Re: Distorted output waveform from Wilderness Sierra 5712 2011-01-10 20:39:36 William Carver Re: Distorted output waveform from Wilderness Sierra I started to write MY opinion on what a 5 pF divider into a scope probe
with diodes back to back would act like. To whit, the 5 pF forms a
divider with the probe capacitance which is conceptually flat with
frequency if it wasn't for the 10 megohm DC resistance of the probe.
That resistance, in parallel with, say, 20 pF of capacitance at that
node, forms a highpass with a corner frequency under 1 KHz. But there is
NO lowpass function.
Then instead of just opinion, or even a SPICE model, I thought I'd hook
it up to some test equipment and report the measurements. With an 8640B
into a 50 ohm terminator, and a 5pF (zero lead length) to two
back-to-back 1N4148 diodes, monitored with an active high impedance
probe into an 8560E spectrum analyzer, here's what I measure:
With 1 MHz RF, at 10 MILLIwatts, there's visible appearance of harmonics
that were not there before. The second harmonic is still in the noise,
third harmonic is down 50 dB.
At 100 MILLIwatts, 0.1 watt, the second harmonic is -40 dB, third
harmonic is -16 dB, fifth -30 dB. The relative strength of even and odd
harmonics is exactly what you'd expect of back-to-back diodes. This is a
LOT of harmonic energy: you'd have absolutely no idea what the
transmitter itself was putting out unless it was HORRIBLE. And it only
gets worse: with more power the diodes will be producing nice square
waves no matter what the transmitter looks like.
Switching to 2MHz, then 4 MHz, still 10 mW, amplitudes didn't change one
bit....neither absolute, nor relative to each other. And switching to
500 KHz, the lowest thing the 8640B will generate, also produced the
same result. As I expected: the highpass nature of the network only is
seen down at audio. We're way above that and only see the capacitive
voltage division of the 5 pF and the probe capacitance.
This was into a 50 ohm load. Tuning an inductive resistor to be
resistive at the fundamental does NOT make it resistive at the
harmonics. In fact the tuning capacitor, on the high side of resonance,
acts as a lowpass function. Until resonance of the tuning capacitor and
its leads came into play....then at some point all bets are off on what
impedance the harmonics see. And you need to be testing at a few
milliwatts to believe it with a wideband 50 ohm load.
That published circuit works fine for counters: they don't need a sine
wave, they work better on square waves in fact. So that circuit is great
for a counter. But for any measurement where you care about the shape of
the waveform, or harmonic content those diodes are deadly to the
If you remove the diodes, you remove the nonlinearity. You're left with
a capacitive divider with 5 pF and capacitance of the scope. Now IF YOU
HAVE A 50 OHM LOAD, you can view the output with more confidence.
An approach to an HF dummy load at modest power levels is one of the
many resistors in TO220 and TO247 transistor packages. They are used on
900 MHz cell phone transmitters, Collins uses them as terminators in
their antenna tuners. Examples: Mouser sells Caddock which IN 50 OHM
VALUE, has "50 ohm performance to 500 MHz". Certainly good for HF. 15
watts MP915-50, sells for $2.98. Thirty watt MP930-50 sells for $3.72,
MP9100 in a TO247 case sells for $9.40. Those are all 50 ohms, 1%
tolerance. They need to be bolted to a heatsink to achieve full rated
power for more than a second. But if you're starting out in homebrewing
and don't have a dummy load they are a good investment.
Bill - W7AAZ
5713 2011-01-10 20:52:56 rasteenb Re: Distorted output waveform from Wilderness Sierra Thanks Dave!
I did, in fact, have a 10x probe for my scope and it easily handled the output voltage (<25V). Using just the 10x probe connected to the dummy load, the output was a beautiful sinusoid.
So as Tim and Allison noted, it looks like the distortion was an artifact of the device I was using to couple the oscilloscope to the transmitter.
So now I'm curious about the K6LS sampler. Perhaps it was designed to examine modulated signals.
As far as the correct instrumentati
5714 2011-01-10 21:04:56 William Carver Re: Distorted output waveform from Wilderness Sierra Rob, good deal on getting the right appearance with the scope probe
straight into the rig. At some point RF can fry a scope probe, but not
with a Wilderness Radio. When you go QRO you could go to a capacitive
divider so the scope probe isn't exposed to hundreds of volts.
The diodes will clip the tops off amplitude modulated signals, too,
remove the positive modulation as you go up toward 0.1 watt. I think he
just got a little overenthusiastic in his list of what it would do.
BTW, that last Caddock resistor for about $9 I mentioned, I didn't say
that it is rated at 100 watts. And Mouser has no minimum order so even
if you get nothing else you only pay actual shipping cost.
5715 2011-01-10 22:14:42 rasteenb Re: Distorted output waveform from Wilderness Sierra Thanks for doing that experiment, Bill, and sharing the results.As I
slowly increase my arsenal of test equipment, and with any luck my
ability to use it correctly, I'll have to try it.I was intrigued by the
thought of modeling it in SPICE or LTSPICE. I may have to get my hands
on that software.
Thanks for the tip on the resistors, too. I was working away from home
for 18 months and was experimenting with my own. Now that I'm back
home, it dawned on me that I do have a 300W dummy load. I still might
pick up a couple of the resistors you mentioned, though. They'd make a
more portable QRP dummy load when strapped to an adequate heat sink.
I'm really glad I came over to this group to post my question; I learned
a lot and the topic seemed to generate some good discussion.And, I'm
happy about the origin of the distorti
5716 2011-01-11 06:54:01 William Carver Re: Distorted output waveform from Wilderness Sierra What would really be great would be the ability to have multiple guys
get on here all at once, and hash out details in real time. I recall
that some of the best ideas used to occur over a burger at lunchtime.
Little flaws in one's understanding (mine, too) are quickly corrected in
a group of peers. But we have to live with the connectivity that
internet provides. And that's substantial. It just takes a while, and
not everyone joins in.....takes too much time to participate in
The main thing to do is to not believe anything you read on internet. As
Al Gore called it, "The Misniformation Highway". Didn't he? Anyway,
think everything through YOURSELF and make sure it all makes sense
before you believe it!
SPICE/LTSPICE is a fantastic tool. You learn tons from it.
5717 2011-01-11 08:37:31 Tamás Fábián Re: Distorted output waveform from Wilderness Sierra Here is an "over-a-burger" kind of probably-wrong-but-hey-what idea: :)
The waveform distortion occures when power is high. Every time Rob raised
the power, the wave got worse. I suspect that the LPF's inductors are
saturated or heated up to a point where core parameters change, resulting in
this kind of distortion.
On 11 January 2011 15:53, William Carver <firstname.lastname@example.org> wrote:
> What would really be great would be the ability to have multiple guys
> get on here all at once, and hash out details in real time. I recall
> that some of the best ideas used to occur over a burger at lunchtime.
> Little flaws in one's understanding (mine, too) are quickly corrected in
> a group of peers. But we have to live with the connectivity that
> internet provides. And that's substantial. It just takes a while, and
> not everyone joins in.....takes too much time to participate in
> The main thing to do is to not believe anything you read on internet. As
> Al Gore called it, "The Misniformation Highway". Didn't he? Anyway,
> think everything through YOURSELF and make sure it all makes sense
> before you believe it!
> SPICE/LTSPICE is a fantastic tool. You learn tons from it.
[Non-text portions of this message have been removed]
5718 2011-01-11 09:11:33 William Carver Re: Distorted output waveform from Wilderness Sierra This morning I was looking at the current ratings of Triad
choke/inductors for LPF of a class D audio amp-as-modulator. They said
the inductance at "rated current" could be 50% of the indicated value.
So that comment about distortion COULD be right on the mark.
I've noticed that when something isn't working right, instead of my mind
being open to far-out possibilities I focus on what I THINK is going on.
Getting someone to spit out a different idea, even if it's not THE
answer, kicks me out of the rut and is often very helpful.
On Tue, 2011-01-11 at 17:37 +0100, Tamás Fábián wrote:
> Here is an "over-a-burger" kind of probably-wrong-but-hey-what idea: :)
> The waveform distortion occures when power is high. Every time Rob raised
> the power, the wave got worse. I suspect that the LPF's inductors are
> saturated or heated up to a point where core parameters change, resulting in
> this kind of distortion.
> On 11 January 2011 15:53, William Carver <email@example.com> wrote:
> > What would really be great would be the ability to have multiple guys
> > get on here all at once, and hash out details in real time. I recall
> > that some of the best ideas used to occur over a burger at lunchtime.
> > Little flaws in one's understanding (mine, too) are quickly corrected in
> > a group of peers. But we have to live with the connectivity that
> > internet provides. And that's substantial. It just takes a while, and
> > not everyone joins in.....takes too much time to participate in
> > everything.
> > The main thing to do is to not believe anything you read on internet. As
> > Al Gore called it, "The Misniformation Highway". Didn't he? Anyway,
> > think everything through YOURSELF and make sure it all makes sense
> > before you believe it!
> > SPICE/LTSPICE is a fantastic tool. You learn tons from it.
> > Bill
> > >
> [Non-text portions of this message have been removed]
> Yahoo! Groups Links
5719 2011-01-11 12:42:15 Robert Cerreto Re: Distorted output waveform from Wilderness Sierra Bill,
"Real Time" discussions sure would be nice. I would give anything to have "real time" and/or "close by" friends that can share the experimenter part of the hobby. But, alas, for me, there is nobody around this neck of the woods that is interested in this part of the hobby. So, I am thankful and content for groups such as this one. Besides, this form of communications provides much more diversity as well. The spread and sharing of talents/education /experience cannot be matched any other way.
I would surely like to meet all who have contributed here in this group and just cook them all a good Italian meal before/after sharing time at the lab bench!!!!
Yes, you must do a a lot of "filtering" with Cyber learning/discussing....then again it gives this guy one more reason to get into the lab. Yes. SPICE is a great tool. From Ohm's Law to the more complex analysis, it sure gives much more dimension to the hobby of experimenting. SPICE is a good start to developing an idea and a good finish to the final analysis. Still, I will still never give up smelling the solder...and... maybe some of the "other" smoke from a failed idea or venture.
Thanks for your sharing Bill.
5720 2011-01-12 06:10:08 nm0s_qrp Re: Distorted output waveform from Wilderness Sierra I doubt that the distortion seen is from saturating magnetics. The losses associated with this would be readily apparent with very high losses in the core - you would smell it overheating.
Another possibility - could you be clipping your scope front end? Do you have a 1X or 10X probe?
73 Dave NM0S
5721 2011-01-12 07:29:22 davidpnewkirk Re: Distorted output waveform from Wilderness Sierra You're welcome. I have modified the "too too" comment I made on your Flickr photo page--I'm tapering down on the Prednisone, now, thanks; bronchitis flareup; much better now :-D--to say, merely, for subsequent tuners-in:
"Just connecting your scope to the ungrounded end of the dummy load via a 10X probe will give you the waveform measurement you're looking for. (The leg that includes the antiparallel diodes, 5-pF capacitor and big 50-ohm resistor is unnecessary, and the diodes would distort the signal.) For the best possible waveform, and certainly if you decide to hook the setup to a spectrum analyzer to see spectral components that span a dynamic range of many tens of decibels, I'd disconnect the RF-to-dc detector leg to remove all possibility that it will add distortion in the form of harmonics. (For the waveform measurement with the scope, its presence probably doesn't make a difference.) Finally, be sure to account for dc voltage division between that 4.7-meg resistor and the DMM's input resistance."
I do hope you'll have the fun of viewing the output on a spectrum analyzer. As others have mentioned, low-pass filtering is needed to tame the output spectrum for FCC compliance, but being able to see the spectrum with and without filtering is great fuel for the intuition.
Dave, amateur radio W9VES
5722 2011-01-12 12:25:16 Gian Re: Distorted output waveform from Wilderness Sierra Hi Robert,
this is a very interesting group....
5723 2011-01-12 22:55:03 davidpnewkirk Re: Distorted output waveform from Wilderness Sierra 5724 2011-01-13 08:29:40 William Carver Re: Distorted output waveform from Wilderness Sierra I didn't mean that tuning the reactance out of a power resistor was
"wrong", just that it wasn't flat for purposes of measuring harmonic
output. There are lots of places little tweeks, intelligently applied,
can enhance the frequency response/bandwidth of circuits like wideband
amplifiers, SWR bridges, etc.
Dave, I'm sure the Ansoft model for the diode isn't amiss! Unthinkable!
A diode detector with series coupling capacitor and shunt diode to
ground charges the capacitor with a DC voltage that can be measured at
the end of the resistor. 'Course the resistor and 10 megaohm scope probe
resistances form a voltage divider, so the scope reading will be about
2/3 of the voltage the capacitor is charged to. For DC measurements,
where you kinda like to not have ANY rf floating around, the usual
practice would be to put a small bypass capacitor from the voltmeter end
of the 4.7 megohm to ground.
Without a bypass there's RF on the series combination of resistor and
scope, as you observed. The 4.7 megohm resistor's parallel capacitance
(maybe 1/4 pF?) will form a voltage divider with the 10X probe
capacitance (18 pF?), so the scope would see perhaps 1/72 ?? of the RF.
There will be a transition from this 1/72 at rf to 2/3 at DC, at a
corner frequency determined by 4.7 megohms and 10 megohms in parallel
(3.2 megohms, and 5 pF in parallel with 18 pF (23 pf), which is about
When modeling you can get as complicated as you want since the software
is doing all the work. As you go finer and finer, modeling lead
inductance on the resistor, coax connector as a short transmission line,
you will reach a point where you find a physicist down there doing some
modeling, too. He will recognize you and nod his head wisely and you can
wave "Hi Wes".
5726 2011-01-13 12:15:08 davidpnewkirk Re: Distorted output waveform from Wilderness Sierra 5727 2011-01-13 13:46:22 William Carver Re: Distorted output waveform from Wilderness Sierra I use a very old version of PSPICE, typing in node-to-node components
and signal descriptors on a DOS machine. It works fine but that DOS
machine may die before I do. I'd not downloaded LTSPICE because this is
a Ubuntu machine, thanks for the heads up that LTSPICE is
WINE-compatible, Dave. That motivates me to get out of a rut.
5729 2011-01-14 06:07:53 Tim LTSPICE LTSPICE (I use it on Ubuntu Linux under Wine as well, it just works) is good stuff. For those who don't know: http://www.linear.com/designtools/software/ltspice.jsp
If you know how to write netlists in PSPICE this will be useful in LTSPICE because really, under the hood, are all the netlists etc. that you already know.
LTSPICE has very good tools for power MOSFET models "built in", better than any other versi
5730 2011-01-14 07:04:13 William Carver Re: LTSPICE Tim, tnx for the comment about MOSFET modelling: a jab in the posterier
since I'm in the process of making a "VIKING III" class E AM rig for
5732 2011-01-14 09:05:51 Tim Re: LTSPICE On the subject of AM modulation of class E (or, probably class C or any class) solid state finals, the Jan/Feb 2001 QEX has an article called "A keyed power supply for class E amplifiers" also 5733 2011-01-14 09:46:49 William Carver Re: LTSPICE > On the subject of AM modulation of class E (or, probably class C orSo many interesting things to do, never enough time! This started as a
> any class) solid state finals, the Jan/Feb 2001 QEX has an article
> called "A keyed power supply for class E amplifiers" also on the web
> at the author's website
> where the authors used a MC34025 class chip to make a switching power
> supply (100kHz switching range) to generate a specifically controlled
> envelope for CW keying.
> It seems to me if you want to do a high efficiency AM modulator that
> this approach is perhaps the slickest and most efficient way to do it.
> Even slicker than class D or class H, because the modulator output
> filtering magnetics will likely be way smaller.
> LTSPICE has a very good library of switching regulators and examples
> and I ought to give it a shot myself :-)
> Tim N3QE
weekend monoband TX with a donated 6146B. It hasn't turned into a
monster, but it certain is far from the initial concept of a "quickie".
That 6146B is back in K7SU's junkbox and I'm a month into a solid state
An offline switcher is "now", and it would be interesting to be able to
tuck the completed TX under my arm, although I don't know why that
matters. For class E AM you need to swing the supply from near zero from
DC to a 3 KHz rate, a bit different than the CW supply described in the
QEX article. I have a heavy-but-perfect transformer, multiple 25A
bridges and a very large new computer grade cap. Not high tech, but a
sturdy, reliable power supply essentially already exists. Time and cost
outlay are both zero.
Jim Tonne, who designed AM modulators for Continental Electronics 100's
of KW AM transmitters, has a canned 4-pole filter design for a PWM
modulator, W4ZCB had some large pot cores and Litz wire that looks like
wire rope. Voila, like magic Jim's lowpass filter inductors are
realized. My junkbox yielded 1 uF and 2.2 uF polypropylene capacitors
with super low ESR and high current capability. A junkbox IRF9540 MOSFET
completes the power portion of the modulator. Both AM and CW shaping
will be done by shaping the AC + DC to the PWN modulator. A few days of
time pondering all this, undouubtedly a few more WINE-ing LTSPICE and
verifying/refining penscil sketches. Cash outlay: zero.
(I'm not really "cheap", it's just an amusing aspect of the project) I
did purchase an Analog Devices SSB2166 to play with for audio
processing/limiting so can't clip the power modulators output and
further infuriate the few SSB ops that wake up early Sunday morning to
find an AM group near THEIR frequency.
The modulator is interesting, but no big deal technically. I'll limit
the seriously-different aspect of this to the RF deck itself. Time
outlay significant, but future potential for makes it worth doing.
5734 2011-01-14 10:39:25 KF4BQ Re: Distorted output waveform from Wilderness Sierra > This opportunity to correct error was more than interesting enough to spark modeling spot-frequency-equivalent R, L, and C potentiometers in Ansoft SerenadeSV 8.5. The modeling schematic I built is atHi Dave,
> and the graph of all three responses, treated as forward gain, is at
Your term:1E6 in the models is not working as you think. Didn't the -42dB response of the simple resistor divider bother you?? (i.e. should be -6dB).
73, Mike Collins KF4BQ
5735 2011-01-14 15:07:45 davidpnewkirk Re: Distorted output waveform from Wilderness Sierra 5736 2011-01-14 16:14:47 William Carver Re: LTSPICE Indeed LTSPICE runs under WINE just fine, Tim.
Too well, in fact. I missed a weekly sked at 9:30 AM, I missed W4ZCB at
2 PM. Luckily I put out the garbage at 7:30 AM so I'm not in hot water
with the XYL. However it's 5PM and I'm starting to get hungry. Didn't
look up at lunchtime.
By the end of the day I'm getting results faster than if I was on the
DOS machine using a 20 year old version of PSPICE.
What's the story on editing device model parameters? I can look at 'em,
but not change them. Do I have to find the model file and create models
for new devices manually? There are SPICE models for many active
devices, is they are means to add them to LTSPICE?
5737 2011-01-15 02:39:37 Nick Re: LTSPICE Hello William,
I started using LTSpice 3 months ago and went through this phase also! I
have also installed it under debian linux and it works fine but most of
my examples were ran on the windows version.There are a lot of
tutorials available for LTSpice if you make a google search.I have never
used another SPICE before in order to compare, but I think that you
can add your models the same way and if you have any previous models you
can add them without too much effort.
for example, if you look into your installation directory (on windows it
is under "Program Files\LTC\LTSpiceIV", on linux i think the "Program
files" folder is located in the ".wine" sub-directory of your home
directory) you can find a folder named "lib" (do not confuse it with
the Ubuntu's system lib folder). under the lib folder there is a folder
named "cmp". there you can find the standard models for jfet, bjt,
diode etc (the file extension clearly refers to the file content--bjt
for BJT). if you open the files with a simple text editor you will
recognize imeediately the declarations for the device models. Additional
models are located under the "sub" folders (do not let the .sub, .mod
and .lib file extensions to fool you, there are plain text files) with
their corresponding schematic symbols located under the "sym" folder. It
is very easy to make you own component.
on the following site
i found (for my taste) the two most usefull tutorials for LTSpice.
PART 1 http://184.108.40.206/~krausg/LTSwitcherCAD/SwitcherCAD-Tutorial_English/pdf-File/LTspice_4_e2.pdf
PART 2 http://220.127.116.11/~krausg/LTSwitcherCAD/SwitcherCAD-Tutorial_English/pdf-File/RF_English_01_final.pdf
Part 1, is a series of examples where all the features of LTSpice are
explored (also how to make your own models from scratch is described) in
details. Part 2 describes the design and simulation of a full 137Mhz
weather satelite converter.
on the same site you can also find a big collection of the author's spice models.
I also usually google the device name followed by "spice model" and 99%
of the times i get the model from the manufacturer pages.so far, both
NXP and infineon provide spice models on their web wites.
What I have not tried yet, and so I cannot tell for sure, is whether
existing libraries from other spice software (e.g. orcad) can be
imported just by copy-paste. at this location http://homepages.which.net/~paul.hills/Circuits/Spice/ModelIndex.html%c2%a0 there is a huge reference of spice models
which links to libraries of other spice software. I have extracted some
models from there but I have not tried to grab the whole file and saveit
under the "sub folder . I think that in general, I have to save the
file, add an ".include file.lib" directive into the schematic and use
the exact device name from the library in the schematic.This is
something that I still investigate.
I hope the above help a bit to start with LTSpice. It helped me a lot !
5738 2011-01-15 08:12:35 William Carver Re: LTSPICE Thanks for all the info. I'm printing, and have saved, your email. I've
been able to use models of devices on my very old PSPICE without a
I've spent 30 minutes looking around this machine, and used "Search",
and I can't find any .WINE, .wine, or any reference to LTSPICE. Suddenly
I feel like I'm using Windows where things are hidden and/or stashed
places. It is not in my home directory. Hmmm. Maybe Ubuntu has a hiding
ability like Windows.
5739 2011-01-15 09:45:08 William Carver Re: LTSPICE I found the LTSPICE files. They are ON THE DESKTOP. I may have to
download it again to put it in the file system.
Bill - W7AAZ
5740 2011-01-15 10:24:52 Nick Re: LTSPICE Bill,
I checked the installation on my Linux machine just not to propagate false info.
I am running Debian Linux with Gnome, so my file explorer in Nautilus. If you use KDE it may be different ( I think it uses Konqueror)...but the idea must be similar. Under the menu "View" select "show hidden files", This will immediately display a number of files starting with a dot (e.g .vlc .wine) which are linux hidden files . Assuming you are in your home directory (e.g /home/bill/) enabling the "show hidden files" will display the ".wine" folder. in there you will find (usually) folder "drive_c" (which is the equivalent of windows drive C) and in there there is hopefully the folder "Program Files" . the rest is easy.
Alternative, if you have a shortcut of the LTSpice executable on the desktop , by right clicking and selecting properties, you will probably find a tab that contains the command that wine uses to launch the program. On my system, the tab is named "Launcher" and the command is
"env WINEPREFIX="/home/nikos/.wine" wine C:\\Program\ Files\\LTC\\LTspiceIV\\scad3.exe" where "/home/nikos/.wine" is the location of my wine installation. You should find something similar on your system.
Otherwise, by starting a terminal window in your home directory, no matter if it is visible or not, you still can do "cd .wine/drive_c/" and navigate in command line if you feel comfortable.
During this weekend, I had on my plans to investigate more on the issue of re-using libraries from a demo version of orcad that I have somewhere. I will let you know if i have any success.
Besides this, the other thing I need to discover is how to re-use existing blocks in more than one design without re-drawing them.
5741 2011-01-15 10:43:19 William Carver Re: LTSPICE Yes, I found the "env......etc...." but when I look there, there's no
WINE subdirectory. As I said in a later mail, I found LTSPICE is
installed on the desktop. Dunno why the "env...." command doesn't
reflect that, but it, and the model files, are all on the desktop. IE,
on the desktop is a file "LIB" and "EXAMPLES", and in the "LIB" director
are the "CMP", "SUB" and "SYM" directories. I looked in one and see the
part numbers in filenames.
Tnx for the help, I'll figure details out later, can't fool with this
any more right now. I might have to do another download with me head
screwed on this time because I'm sure I simply can't make an "LTSPICE"
folder and move these things to it.
5742 2011-01-15 10:51:58 Harold Smith Re: LTSPICE Bill,
Are you sure you have wine already installed? My Fedora computer didn't.
It wasn't difficult to locate and install. Just Google it.
de KE6TI, Harold
5743 2011-01-15 11:02:19 Nick Re: LTSPICE I really dont know why it is installed on the desktop.Normally it should go somewhere else!
maybe installing and re-installing (so window-ish!!!!) it...i dont know.in my case, I just downloaded the .exe file from Linear Technologies web site and double clicked it. wine took over and the installation proceeded as if it was windows.
the last thing that comes to mind is to check your wine installation/configuration by running "winecfg".
just two more links that may be of help for wine
5744 2011-01-15 11:06:39 William Carver Re: LTSPICE Yes, I installed WINE and I have LTSPICE running. I just couldn't find
the LTSPICE itself to play with it's libraries. As it turns out, they
are all on my DESKTOP. I didn't install it to my "file system". It runs
fine, just looks a little funny to see the LTSPICE file folders on my
When I downloaded the default installation was /PROGRAM FILES, etc. The
usual Windows BS. I thought "I don't HAVE such a directory" so told it
to put it on the desktop. Dumb, but trial-and-error seems to hold an
I don't have time right now but I'll probably have to download and
install it again, this time get it into the file system.
Bill - W7AAZ
On Sat, 2011-01-15 at 13:51 -0500, Harold Smith wrote:
> Are you sure you have wine already installed? My Fedora computer didn't.
> It wasn't difficult to locate and install. Just Google it.
> de KE6TI, Harold
5745 2011-01-15 12:42:24 William Carver Re: LTSPICE In "Applications", right above "add/remove software" is a new picture
labeled "WINE". Appeared since I installed WINE yesterday. Under it is
C: and LTSPICE.
I ran "Applications: add/remove software", but it doesn't find LTSPICE
on its list. It's probably not seen since it's under WINE. Perhaps WINE
has its own add/remove function. It certainly has ADD, hi. ADDing
LTSPICE with a copy already on the machine could have pitfalls. There is
a reason for the REMOVE function. I'll ignore the issue of WHERE the
LTSPICE is sitting for the moment. Other things to do, want to USE it,
not play with the damned computer.
5746 2011-01-16 03:35:49 davidpnewkirk Re: LTSPICE 5747 2011-01-16 04:23:41 davidpnewkirk Re: LTSPICE 5748 2011-01-16 04:40:40 davidpnewkirk Re: LTSPICE 5751 2011-01-16 20:33:18 ae5ew Re: LTSPICE I miss my Amiga when I could name drives/media.