ce
pHz
----------> out
han
low
do
ose
e 5
NXP says that it shouldn't be used for new designs. The nice thing about the BFR92 is that it's multi-sourced, cheap and widely available.
-- Bill Sloman, Sydney
I am going to use the 10MHz output of the Rubbitin frequency reference to make 50.000000000 MHz for my FPGA board. The plot is a bit complicated, and is planned like this:
25 MHz crystal >- FPGA board Spartan2 -> x2 DLL (delay locked loop that is a build in thing in the Spartan2 FPGA)-> 50 MHz -> divide by 5 in FPGA -> 10 MHz
-> xor with 10MHz from Rubbitinn unit -> low pass -> 25MHz xtal VCO. But now how to tune the 25 MHz xtal?
+5 | | [ ] 680 25M | ----|[]|--- | | 100k |--------------------> out | ---===-----| c 100k | |/ -----===--------------| BC548B control- |\/ e voltage | /// Programmed the verilog, made a small crystal module that fits the normal FPGA crystal oscillator module socket (see picture link below). Soldered some experimental loop filter...Divided the 25MHz in the FPGA by 5 Divided the 10 MHz from the rubbitin frequency reference by 2. xored both 5 MHz, and used that to control the 25 MHz oscillator.
one enty per second:
25,000,543 Hz
Pity about the BFG25A/X--its beta holds up extremely well down to very low collector currents. It's an old favourite of mine. I suppose I'll have to switch to the BFG403W, or perhaps the BFT25A.
The BFR92 is okay, but has twice the C_CB, and gets weirdly slow at even mildly low collector currents.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 USA +1 845 480 2058 hobbs at electrooptical dot net http://electrooptical.net
Statek used to have low cost lower Q 60 kHz crystals.
On some crystals, you could further lower their Q by breaking the hermetic seal for air damping.
-- 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: don@tinaja.com Please visit my GURU's LAIR web site at http://www.tinaja.com
Who needs a crystal for WWVB...
--
| James E.Thompson, CTO | mens |
| Analog Innovations, Inc. | et |
| Analog/Mixed-Signal ASIC's and Discrete Systems | manus |
| Phoenix, Arizona 85048 Skype: Contacts Only | |
| Voice:(480)460-2350 Fax: Available upon request | Brass Rat |
| E-mail Icon at http://www.analog-innovations.com | 1962 |
I love to cook with wine. Sometimes I even put it in the food.
"Jim Thompson" wrote in message news: snipped-for-privacy@4ax.com...
A new format is due out for WWVB that includes PM coding. Will that circuit still work?
See:
..and lowering the frequency by use of a lead (graphite) pencil - scribble away!
Have you ever opened a tuning fork crystal can? That's the long cylinder style used in watches. Or the molded SMT packages?
Some 60 KHz crystals are in the 60-70 cent range at Digikey.
Why bother, when you can get crystals a couple hundred Hz either side of 60 KHZ?
Lower Q so you can pull them?
How far do you want to pull them? They are available at +/- a half percent of the center frequency. How narrow do you want to make it? If only a couple cycles, just use a 60 KHz crystal in a series resonant filter to take advantage of the Q factor. If anyone is running an old NTSC security monitor or analog TV on a converter that loses it's video input, harmonics from the horizontal oscillator can drift right through
60 KHz. I could pick that up a quarter mile, from my nearest neighbor who would go to sleep with a TV on. When the station went off the air for the night, it would drift & wipe out WWVB. It would saturate the preamp in my loop antenna which was tuned to 60 KHz, and had another filter between the preamp stages. The antenna put out a 8V P-P sinewave at the decoder end of the coax. The TV pushed a nasty 15V P-P signal.
I use my portable foxhole digger for them..
That BFG403 has an astonishingly low 20fF Cbc! Nice part...
-- Thanks, Fred.
loop
25MHz 5] 680
----------------> out
/ s sry
hm do
. aso
nd
ays,
tty
I'll
And Digikey has got it ex-stock for $A.0.50 in small quantities.
The base-emitter junctions is even more vulnerable than the BFR92's - Vebo of only 1V - and the data sheet doesn't give any data about operation at collector currents below 1mA, but it's still a distinctly appealing part. An Ft of 17GHz is a whole lot better than 5GHz, and both are plenty good enough for a 4MHz oscillator.
n-- Bill Sloman, Sydney
Quite frankly, using a hot-rod like that for a low-frequency oscillator is asking for trouble. One must use layout rules and components that prevent parasitic modes up to where the transistor runs out of gain.
I've seen many a young buck try to make circuits treating few-hundred MHz signals with GHz rated components. They are invariably puzzled about strange circuit behaviour and completely unaware that their gadget screams at the top of its little voice at 5GHz or so.
Jeroen Belleman
ocked loop
MHz
ss -> 25MHz
+5 | | [ ] 680 |-------------------> out
///
ess
very
kohm do
so.
hing
ye I'll
33R in series with the base, and close up against it, certainly tames the BFR92. My guess is that it would work with the BFG403 too.The original proposal was to get a 20MHz oscillator to run at 20uA, which rather constrains the collector currents that might be used - neither the BFR92 nor the BFG403 would have GHz bandwidth with a drain current of 20uA.
-- Bill Sloman, Sydney
It isn't a hotrod at low collector currents. I've used the BFG25A/X for years and years. At 1 uA of collector current it's about a 20 MHz transistor, which is pretty good going, but it won't oscillate.
I do use SiGe:C transistors in bootstraps a fair amount, though. You just put an 0402 bead in the base lead, and they become pretty docile. Their base regions have graded composition, so their saturation behaviour is excellent and their Early voltage is effectively infinite. That means that you can run a cascode stage at high voltage gain. That plus a PNP wraparound (aka a series-shunt pair) makes a brilliant, very quiet bootstrap. SPICE says that the voltage gain is about 0.9998 or a bit higher, including vaguely realistic strays. When you build it, of course, it's nearly always limited by the residual capacitance to ground that doesn't get bootstrapped.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 USA +1 845 480 2058 hobbs at electrooptical dot net http://electrooptical.net
With a multilayer PCB it's easy to bootstrap that too... Taming it for a high impedance source like the probe you spoke about a while ago would be real difficult though (if not impossible).
-- Thanks, Fred.
The input amp on this one is a BF862. The BFP640s are the cascode stage and the bootstrap. They're really good for that, because their Early voltages are so high and their beta and R_BB' values are to die for. An
0402 bead in the base (one that has highish impedance up to ~10 GHz) makes them very well behaved, if you don't hang long traces on them.The bootstrap loop bandwidth is only about 700 MHz, so it isn't that bad.
Cheers
Phil Hobbs
-- Dr Philip C D Hobbs Principal Consultant ElectroOptical Innovations LLC Optics, Electro-optics, Photonics, Analog Electronics 160 North State Road #203 Briarcliff Manor NY 10510 hobbs at electrooptical dot net http://electrooptical.net
ElectronDepot website is not affiliated with any of the manufacturers or service providers discussed here. All logos and trade names are the property of their respective owners.