"Standard parts" for rf amps?

[Roy Lewallen]

. . .
IMHO if the results from spice differ from the
real results then spice is wrong.
. . .

If the results from SPICE differ from the real results, then your model
is inadequate.

Roy Lewallen, W7EL

 

[Henry Kiefer]

Spice is as good as the input is!
Spice is not a Maxwell field equations solver!

You can predict within 5% without Spice?? How?

- Henry

 

[Jim Thompson]

Spice is as good as the input is!
Spice is not a Maxwell field equations solver!

You can predict within 5% without Spice?? How?

- Henry

[snip]

Some of us have done better than 5% prediction BC (before CAD)... like
for close to 20 years before I ever saw a computer... and 25 years
before simulation software.

...Jim Thompson

 

[Henry Kiefer]

Sure Jim. I understand you. That was the time where time was cheaper. Once I
programmed a Gauss linear equations system solver on my Casio FX-602P -
having 512 bytes and even left one byte (for extensions ;-) It was with
minimal input help system (Showing the indices to input) and the matrix was
of dynamical order. (Sorry for my bad english)

I can buy resistors and capacitors with 5%, sometimes with 1%. If you build
a rc oscillator with 1% component values, you get an error typical 1,5%. And
then you add the active components with horrible semiconductor strayung
values. Of course, it is possible with feedback structures to linearize and
stabilize such systems.
So, where is the difference between measurement of the component values and
calculating by hand the circuit OR setting the measured values in the Spice
component dialog boxes? The difference is the time needed and the
possibility to make errors.

5% is a real good value for an analog system.

Hey Jim - Why you don't answer my second private message? :-(

- Henry

Spice is as good as the input is!
Spice is not a Maxwell field equations solver!

You can predict within 5% without Spice?? How?

- Henry

[snip]

Some of us have done better than 5% prediction BC (before CAD)... like
for close to 20 years before I ever saw a computer... and 25 years
before simulation software.

...Jim Thompson

 

[Saandy , 4Z5KS]

The NE85633 bipolar transistor from California Eastern Laboratories
(CEL) ! very cheap quite stable and with a good noise figure. very well
documented by the manufacturer!
try to fish a development kit of SMD coils from Murata, sometimes they
give them away for free.
I personally prefer to go with air wound inductors, they're better and
quite small at these frequencies.
Saandy 4Z5KS

 

[Henry Kiefer]

Thanks Saandy! I knew CEL but forget them ;-)

Seems the NE85633 is really cheap and good performance. The package is
reasoable for hobbyists.
Do you already converted the parameters to spice and is it possible to
provide the file to me? Thank you!

Should be a good oscillator and all-purpose amp...


Sure, plain wire is better than the manufactured coils if possible to apply.

cu -
Henry

 

[Henry Kiefer]

Sorry. I found the needed file at CEL website by searching for part number.
- Henry

 

[Winfield Hill]

Henry Kiefer wrote...

I can buy resistors and capacitors with 5%, sometimes with 1%.
If you build a RC oscillator with 1% component values, you get
an error typical 1,5%.

I have a nice line of 2% caps in stock, but where do you
get the elusive (affordable) 1% capacitors?

 

[Rich Grise]

Henry Kiefer wrote...

I have a nice line of 2% caps in stock, but where do you get the elusive
(affordable) 1% capacitors?

You have undergrads select them.

Cheers!
Rich

 

[Henry Kiefer]

For example http://www.wima.de/fkp2.htm
But I think there are other manufacturers too.

- Henry

 

[Henry Kiefer]

OR two capacitors in parallel. The support cap has a nominal value of 1/100
of the bigger cap. So you can tune the bigger.

- Henry

 

[Hal Murray]

I have a nice line of 2% caps in stock, but where do you

get the elusive (affordable) 1% capacitors?

Digikey stocks a handful of Panasonic ECQ-P(Z),
POLYPROPYLENE, Radial, 0.001 to 0.47 uF.

Not a great selection, but probably good enough for
many timing applications if you do the fine tuning with
the size of the R.

 

[Winfield Hill]

Hal Murray wrote...

Digikey stocks a handful of Panasonic ECQ-P(Z),
POLYPROPYLENE, Radial, 0.001 to 0.47 uF.

That's good to hear of, thanks! The 2% parts I have in stock
are the same series, but cheaper 2% parts. I didn't realize
DigiKey carried so many 1% film capacitor values in stock.

 

[Henry Kiefer]

For timing purposes I suggest using a ready-to-use clock generator in tiny
SOT-23 or alike. For example from www.linear.com and www.maxim-ic.com
They have a accuracy well in the 2% range and can be tuned with a little
resistor 1%. Maxim sells a 32768Hz variant - crystal controlled.
They're all digital ouput!

Building a 32KHz oscillator with a clock crystal and having sinus output is
really simple with CMOS buffers.
You can mix the signal with RF getting an IF somewhere. Look at the app note
at www.cypress.com how to build a time-code receiver using that concept -
just of the cystral and cypress mixed-signal controller.

- Henry

 

[Joel Kolstad]

Henry,

Building a 32KHz oscillator with a clock crystal and having sinus output is
really simple with CMOS buffers.
You can mix the signal with RF getting an IF somewhere. Look at the app note
at www.cypress.com how to build a time-code receiver using that concept -
just of the cystral and cypress mixed-signal controller.

I don't suppose you have the name and/or number of that application note? I
did a little searching without success to find it.

---Joel Kolstad

 

[Henry Kiefer]

Here it is Joel:
http://www.spectrum-benelux.com/docs/AppNote radioclock.pdf
(Seems that it is not on the cypress site)

If you realize something with the PSoC, then I would like to hear from you.

- Henry

 

[hayseed]

(Snippage)

Building a 32KHz oscillator with a clock crystal and having sinus output is
really simple with CMOS buffers.

Like snot???

(more snippage)
Sorry. I couldn't resist it.

Merry Christmas to all!!!

-Bob

 

[Don Lancaster]

Like snot???

(more snippage)
Sorry. I couldn't resist it.

Merry Christmas to all!!!

-Bob

Turns out that if you really want micropower sinewave oscillation at
32kHz, you are far better off using bipolar transistors than CMOS gates.



--
Many thanks,

Don Lancaster voice phone: (928)428-4073
Synergetics 3860 West First Street Box 809 Thatcher, AZ 85552
rss: http://www.tinaja.com/whtnu.xml email: [email protected]

Please visit my GURU's LAIR web site at http://www.tinaja.com