Register a SA Forums Account here!
JOINING THE SA FORUMS WILL REMOVE THIS BIG AD, THE ANNOYING UNDERLINED ADS, AND STUPID INTERSTITIAL ADS!!!

You can: log in, read the tech support FAQ, or request your lost password. This dumb message (and those ads) will appear on every screen until you register! Get rid of this crap by registering your own SA Forums Account and joining roughly 150,000 Goons, for the one-time price of $9.95! We charge money because it costs us money per month for bills, and since we don't believe in showing ads to our users, we try to make the money back through forum registrations.
 
  • Locked thread
longview
Dec 25, 2006

heh.
it's a beautiful thing



replaced a ton of capacitors and now it works well enough to self calibrate. i've ordered more replacements and now intend to fully recap this thing once they get here.

here's the brochure, if anyone's interested:
http://www.avionteq.com/Document/Wavetek-4015-Specification-Sheet.PDF

Adbot
ADBOT LOVES YOU

atomicthumbs
Dec 26, 2010


We're in the business of extending man's senses.

longview posted:

it's a beautiful thing



replaced a ton of capacitors and now it works well enough to self calibrate. i've ordered more replacements and now intend to fully recap this thing once they get here.

here's the brochure, if anyone's interested:
http://www.avionteq.com/Document/Wavetek-4015-Specification-Sheet.PDF

dat Finlux EL display :eyepop:

Captain Foo
May 11, 2004

we vibin'
we slidin'
we breathin'
we dyin'

That thing seems super loving cool and I barely know what it's for

longview
Dec 25, 2006

heh.
right now i'm just playing with it in duplex mode, receiving broadcast FM, running it out on the demod port and running it back to the receiver tester ext. mod port to broadcast on 70cm

this weekend i'll probably make the required jig cable to calibrate my two icom handhelds, this rig is basically all the instruments required to calibrate those in one. might give the IC-775 and 706 a try later since this can do HF and 50W CW on the input

sure sign there are a lot of marginal caps: it reports faults when it's cold, leave it on for 10 minutes and it's fine once the caps warm up a bit and the ESR goes down. been running for 3 hours now and still fine.

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
so past week i picked up a 33 crystal mixpack of random HF freqs, most good and usable and interesting. Couple of 75m's, lots of 8 12 16 20 24mhz type stuff, etc

Real fun tho is the HC-49u pack i picked up from a guy that....sells crystal packs

for 15 bucks i got crystals for:

quote:

3535 3550 3560 3579.545
7030 7040 7050 7055 7070 7110 7114 7122
9000 10106 10116 10118 10120 10140
14050 14060 21050 21060

so i neeeeeEEEEEd to get my code fluent again and build a qrp piece of poo poo transmitter.

e: oh and I'm still scheduled to pick up that Tek 2235 for $130 this weekend, keeping in touch with that dude.

Jonny 290 fucked around with this message at 09:41 on Sep 20, 2016

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
load of chinashit shipped woooo come on counterfeit transistors

oh i ordered some ICL7660's too b/c i saw them in a Sony repair video, neat little negative voltage power converter. figured might be nice if i ever need a few volts of - to bias something or w/e

spankmeister
Jun 15, 2008






I like the TLE2426 for rail splitting a single supply into +-

e: image attachment didn't work

longview
Dec 25, 2006

heh.
both those look like useful ICs, should probably keep some around

i usually just use a spare opamp for virtual grounds and use a series resistor + bypass cap everywhere it's used (LM8261 is pretty ok).

ordered some ICs yesterday myself, 20 pcs 74HC4046 VCO+phase detector, 20 pcs 74HCT9046 (upgraded 4046), 100 75HC4051 8:1 mux and 200 BB833 1-10 pF varactor diodes

the bb833 is a good choice for a lot of circuits even up to GHzes and it's cheap enough that if i need more capacitance range i can just put in a few more. the tuning range is effectively 1-10V as well, and a decent delta from 0-5V.

i got a box with these yesterday: http://www.ebay.com/itm/Lot-of-400-New-14-4MHz-TCXO-Ham-Marker-REF-Microwave-/222251085155?hash=item33bf326d63:g:YscAAOSwmfhX2qp2

the BB833 can be directly installed to make a VCXO with a nice fairly linear response (35 Hz/V positive slope)

atomicthumbs
Dec 26, 2010


We're in the business of extending man's senses.

longview posted:

i usually just use a spare opamp for virtual grounds and use a series resistor + bypass cap everywhere it's used (LM8261 is pretty ok).

man i still don't understand electronics

longview
Dec 25, 2006

heh.
the stabilock now works around half the time, it seems to be temperature and/or mechanically sensitive as well.

so today i checked the IF filter tuning and the FM discriminator setting, IF tuning was more or less ok but i did manage to improve the out-of-band rejection a bit while maintaining the symmetry.
they put a buffered test point on the board which made probing very easy, i tried my noise source with a band pass filter as a broad band source, but ended up using the RF generator set to generate an overmodulated AM signal. that made a nice fairly wide signal to adjust off of.

fm discriminator was easier to adjust; since the FM circuitry in this rig is DC coupled the tank is adjusted with a continuous tone input, reading DC voltage off the demod output BNC. tuning is just set a frequency in that gives 450 kHz, tune to 0V+-50mV, then go 10 kHz up or down and adjust gain for 1V+-50mV, easy.

didn't actually work of course but it was nice to get it adjusted.

so there's still things left to check, but i need data:



that's the entire self test sequence for the instrument recorded on my scope and plotted in Kst (the only software i could find that will accept a 12 million line CSV file)

the readout is from the demod output; in the test sequence the muxes are set up to buffer out everything the computer board ADC is reading, so this should be a good place to look since the only things i'm missing is the IF frequency counter and the various data-bus lines.

i believe the failing test is around the 600k point near the middle, now that i know how to plot and what format to save data in i can wait until it passes a test and get a few data sets of good and not good test sequences. with those it should be possible to spot exactly what is different and narrow it down to a test sequence number which will tell me what it's trying to do.

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp

longview posted:

200 BB833 1-10 pF varactor diodes

oh for fucks sake just a 4001 :P they'll give you 10pf!


actually i got some varactors on the way too, complete other end of the spectrum though



time for a little WHAT THE gently caress ARE THESE NERDS GOING ON ABOUT interlude!

--

What is a varactor? Take a diode. Wire it up backwards. Put voltage across it. Its capacitance varies in rough proportion to the voltage.

We discovered this effect, and then developed special-purpose "varactor diodes" that have a: more capacitance and b: a smoother curve.

These are in MANY electronic devices and radios you own. Since it's a variable capacitance part, it can be used, in many cases, to replace the bulky and fragile variable capacitor in many radios. All you need to do to tune then is provide a variable DC voltage on the varactor, which can be done with a simple potentiometer. It's still a semiconductor though, so you can't use this hack for anything where "power" is involved - it's strictly a small signal/receiver technique.

--
END INTERLUDE

QuantumPotato
Feb 3, 2005

Fallen Rib
Things I have learned working my local repeater:

* Jimmy Carter was a bad president.
* PROSTATES
* Trucking is boring.


I think I need to get into HF sooner than later. I did cop a Upconverter for my RTLSDR and have been poking around what I can find with my stupid little antenna. Waiting on a cheap china order with parts to build a real antenna.

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
holy poo poo, they talk about a president besides Obama?

Truly amazed.

lol

longview
Dec 25, 2006

heh.

Jonny 290 posted:

oh for fucks sake just a 4001 :P they'll give you 10pf!

actually the 4001 is my go-to PIN diode

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
i read somewhere that LEDs can make for pretty fair ones too, i saw those 10 for a buck though and said what the hell

One of the projects is to make an electronically tuned preselector front end for the new rtl-sdr stick, so if we want to start ruminating on that, i will not object

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
my initial thoughts are a tank circuit on the front end into a FET buffer, then a tunable bandpass filter maybe? and a gain amp after that. but i am welcome to any other notions

Storysmith
Dec 31, 2006

longview posted:

so there's still things left to check, but i need data:



that's the entire self test sequence for the instrument recorded on my scope and plotted in Kst (the only software i could find that will accept a 12 million line CSV file)

i believe the failing test is around the 600k point near the middle, now that i know how to plot and what format to save data in i can wait until it passes a test and get a few data sets of good and not good test sequences. with those it should be possible to spot exactly what is different and narrow it down to a test sequence number which will tell me what it's trying to do.

if your data is time,number and you have gnuradio already you can pretty easily massage it and get a gnuradio flowgraph that reads the failing and passing and correlates them and plots the difference. may be useful if your plotting software doesn't make that easy enough.

also that thing is a beast and I'm so jealous please keep posting about it

longview
Dec 25, 2006

heh.

Jonny 290 posted:

my initial thoughts are a tank circuit on the front end into a FET buffer, then a tunable bandpass filter maybe? and a gain amp after that. but i am welcome to any other notions

i actually have a sony lp-1 which is kind of what you're suggesting, here's the antenna board schematic. control input is a single analog voltage + on/off control



it's broad band enough that this design just switches in different fixed voltages for each band

obviously this will look a bit different if the input is meant to be 50 ohm

with a 50 ohm input you could fine tune the tank with varicaps, and use pin switches to select different inductors.

i'd also suggest a pin limiter on the front end instead of clipping diodes if possible; HSMP-4820 is pretty cheap and designed specifically for limiters. avago has a decent app note on the design of limiters using that diode, and another with the HSMP-3814 on high linearity pin attenuators.

i used a bf998 as the buffer in my mini-whip PCB, it seems to work pretty well

probably best not to put gain in the selector except to compensate for losses, SDR radios struggle enough with overload as it is

Storysmith posted:

if your data is time,number and you have gnuradio already you can pretty easily massage it and get a gnuradio flowgraph that reads the failing and passing and correlates them and plots the difference. may be useful if your plotting software doesn't make that easy enough.

also that thing is a beast and I'm so jealous please keep posting about it

thing about gnuradio is installing it on windows is super annoying so i'd have to install fedora/centos as my desktop os to use it properly.

i think the alignment should be fairly possible to do in kst, it looks like it might be able to do some of that by itself, and there's a lot of good reference points in the data to manually align it by

longview
Dec 25, 2006

heh.


so i got a few traces of a good test, this stuff takes a while since the scope takes literally 40 minutes to write a 12M point file to a USB drive in CSV format. it can do a binary format but no software i've found can actually read files that big.

the tests are not entirely deterministic so aligning is kind of a problem since it would have to time-stretch, and some tests seem to be skipped or shortened in the failed trace.

this is part of the demodulator test and by far the most obvious difference i've found. i'm not sure exactly where in the demod test sequence this is, but in general the failed trace has a lot more noise during idle. this also holds true in the other two data sets (total two failed, two good).
i'm pretty sure the test is running some FM testing in the picture, immediately to the left off screen there's a fairly long idle section, in the two bad traces there's a lot of -0.1V spikes around 2-3 samples apart. the two good traces are slightly different in density but both are significantly lower than the two bad ones.

since this was captured at 200 ksps it's not unlikely that the data is aliased. note also that the good signal doesn't look amazingly great during the test either so it's obviously not perfect.

current assumption: assuming it's testing the FM discriminator, this is likely to be noise originating in the local oscillator, most likely what i'm seeing is the reference frequency for whatever PLL is having trouble maintaining lock. most likely cause is obviously the remaining electrolytics in there; i replaced all the 10 and 2.2µf caps with ceramics but there's several 47 and 100µf caps remaining.

so i'll replace those when they get here; then i'll probably go over the alignment procedure for the synth. it's a little complicated and requires some desoldering of components to inject test signals but the actual tests look to be manageable. i'll also need a 0.1" header extender cable and possibly some MCX patch cables to run the unit with that board out.

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
Badass. anybody else I'd say would be in the weeds, but i know youll figure this out :D

Thanks for posting that sony schem. Definitely very similar to what i'm looking at.

do i want to do a 50 ohm in, do an impedance transformation to a tank circuit (i've heard about 'loading' tanks so I'm thinking tap?) and then take off that to a jfet buffer? This is built to be in line with a 'real antenna' like my 40m, but it sure would be nice if it could take a whip input too

vodkat
Jun 30, 2012



cannot legally be sold as vodka

Jonny 290 posted:

i read somewhere that LEDs can make for pretty fair ones too, i saw those 10 for a buck though and said what the hell

One of the projects is to make an electronically tuned preselector front end for the new rtl-sdr stick, so if we want to start ruminating on that, i will not object

cool :cool:. my rtf-sdr stick just arrived so i'd appreciate some advice on cool stuff to do with it. currently I was thinking of trying to write some sort of script to monitor pirate radio stations, maybe record a number stations or something and to do that thing where you can track planes? but idk it seems like there are 1000s of possibilities and i have no idea of what will be fun let alone achievable.

Trabisnikof
Dec 24, 2005

vodkat posted:

cool :cool:. my rtf-sdr stick just arrived so i'd appreciate some advice on cool stuff to do with it. currently I was thinking of trying to write some sort of script to monitor pirate radio stations, maybe record a number stations or something and to do that thing where you can track planes? but idk it seems like there are 1000s of possibilities and i have no idea of what will be fun let alone achievable.

You in urban/rural? Near and transportation infrastructure like the sea? You can pretty easily start decoding a variety of digital things if that sounds interesting (OOK like car remotes). Also, do you have a PC and/or Linux box? There are a ton of multi platform tools but some platform specific ones that are fun (SDR# on PC for example).

vodkat
Jun 30, 2012



cannot legally be sold as vodka

Trabisnikof posted:

You in urban/rural? Near and transportation infrastructure like the sea? You can pretty easily start decoding a variety of digital things if that sounds interesting (OOK like car remotes). Also, do you have a PC and/or Linux box? There are a ton of multi platform tools but some platform specific ones that are fun (SDR# on PC for example).

I'm living in London so i'm guessing there is tonnes of radio poo poo going on, although I'm sure lots of it is not meant to be listed to. On a mac at the moment but I was assuming I'd have to boot into a separate linux vm or partition to actually do anything with my new dongle.

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp

vodkat posted:

I'm living in London so i'm guessing there is tonnes of radio poo poo going on, although I'm sure lots of it is not meant to be listed to. On a mac at the moment but I was assuming I'd have to boot into a separate linux vm or partition to actually do anything with my new dongle.

Negatory. http://cubicsdr.com/ , plug in and go hog wild. :)

it's actually very competent and has a lot of good features

Trabisnikof
Dec 24, 2005

vodkat posted:

I'm living in London so i'm guessing there is tonnes of radio poo poo going on, although I'm sure lots of it is not meant to be listed to. On a mac at the moment but I was assuming I'd have to boot into a separate linux vm or partition to actually do anything with my new dongle.

Seconding the cubicsdr recommendation, and jonny is right this stuff is plug and play.

I'd also recommend trying gqrx which comes in an OS X binary too, see which you prefer.

Should be bunches of stuff to listen to!

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
Tried to breadboard up the oscillator stage of a Tuna Tin 2, i can't get it to oscillate, though it draws DC

so i went to the Mini Tuna

http://members.shaw.ca/ve7sl/tuna.html

this one works. even put a trimcap in series with the xtal and i can wiggle the freq.
just a very rough ballpark of power, meter wise, i'd say maybe 50-100 mW RF? Drawing 16 mA at 12 volts gives right around 200 mW draw so i don't think i'm too far off

first time winding inductors. Glad i found a purpose for this fistful of 37 and 50 size toroids i bought

There's space on the board for an amp and a 2907 for proper keying, so that might be next

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
welp haha




first successful breadboard -> proto transfer

:D

I added the LED VXO back in after i verified it worked, and got a nice t/r switch out of a gutted icom 2m rig




here's the schematic if you want to follow along, it's roughly mirrored. You should be able to clearly ID the base bias resistors, the inductor between the collector and 12v, the emitter resistor+cap combo, the crystal of course and then the output filter

Jonny 290 fucked around with this message at 10:02 on Sep 24, 2016

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
and here's what it looks like on the scope (which is apparently in excellent working condition)



top is the output after the filter, bottom is the oscillator before the filter. you can see the second harmonic loving stuff up there, but after the filter, it's relatively clean (you can see a little bit of non-sinusoidal nature, may need more filtering)

btw both are 1v/div so you can see how there's obviously lots more output voltage after the filter (higher impedance there)

e: is my math right? right around 3v P-P into 50 ohms, 2.12ish volts RMS, into 50 ohms should be 42 mA and that translates to about 89 milliwatts output?

Jonny 290 fucked around with this message at 08:41 on Sep 25, 2016

longview
Dec 25, 2006

heh.

Jonny 290 posted:

do i want to do a 50 ohm in, do an impedance transformation to a tank circuit (i've heard about 'loading' tanks so I'm thinking tap?) and then take off that to a jfet buffer? This is built to be in line with a 'real antenna' like my 40m, but it sure would be nice if it could take a whip input too

this is slightly outside of my comfort zone, for a single band input at HF you could probably ignore the termination and just build something to parallel resonate the voltage up. put in a resistor to de-Q the circuit as needed.
for an all-band tunable design i suspect a tank circuit will need to be made of a number of elements that can be switched in unless the antenna is known (knowing the source impedance of the antenna helps, and is probably why sony can do it in so few components, that loop is probably inductive on most bands).

matching FET inputs to 50 ohm using only reactive components is not very easy since impedance matching a preamp is all about delivering all the power into the load, in order to put any power into a gate it has to step up the voltage hugely, the "load" is a ~5 pF capacitor. that either means a transformer for wide-band, a tank circuit for narrow band, or a resistor if you don't really care that much.

using a bipolar in common base mode is the classic easy-mode 50 ohm input amplifier and might be a better choice for a 50 ohm input buffer actually

mini-whip input is highly sensitive to input capacitance; if the input capacitance exceeds 4-5 pF effective then your low band performance will be terrible. designing a matching filter between the antenna and FET with that kind of requirement could be difficult. basically anything except a FET input, series inductor and some series resistance destroys the performance. forget any inductance to ground or pin limiters/switches in that case. probably best to do these as two separate projects or include two different antenna inputs.



i don't get how that transistor circuit oscillates; there's no feedback except through the power supply leads and the transistor capacitance

assuming you terminated to 50 ohm when measuring the 3V peak-peak the RMS value is around 1.1V or 20-30 mW. working with peak-peak measurements is convenient but it's not the same as a peak-voltage value which is defined as peak deviation from the reference voltage.
i keep this table bookmarked to keep those things straight http://ifmaxp1.ifm.uni-hamburg.de/DBM.shtml

also i wish i knew CW so i could do anything useful with these super simple CW transmitter rigs. congrats on getting a working CRO, i wish i had room to keep my old HP in the apartment.

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
Maybe a little feedback off of C1? It's not a complete bypass. 470 pf -> ~48 ohms at 7.1mhz

I didn't know poo poo about common base amplifiers, read up on them, quite weird and interesting. do you need negative voltages for that to work? i'm confuse. the emitter's going to have to be biased to -0.7 volts or so relative to the grounded base, right?

Jonny 290 fucked around with this message at 11:20 on Sep 25, 2016

longview
Dec 25, 2006

heh.
could be emitter feedback yeah

you can get away with single supply, here's one i built for a low impedance microphone element:



since it's for AF it's a little different than RF versions

ignore Q7; that's a second footprint to support SMD resistor in addition to TO-92

the base only has to be AC grounded, so a capacitor to ground there accomplishes that while allowing DC biasing, C3 in this case.

RV3/R11 is the negative feedback input from the output (ac coupled off the emitter from Q6). for maximum gain with no feedback R11 is 0 ohm

U1 is a biasing amplifier, this is often replaced with a resistor divider from VCC. you'll want to use that to set the collector current to where the transistor has a lot of hfe (1-10 mA for 2n3904s/bc547s, 10-100 mA for 2222As, doesn't matter much for RF transistor since they're much less sensitive to gain-current variations).
it can also be replaced with a single transistor sensing the voltage across R1 through a low pass filter to get better thermal stability without adding a whole opamp.

Q4/Q3 is a current source, this provides the collector current while having a very high output impedance. since the voltage gain is set by the collector resistance/emitter resistance, putting a current source there increases the effective resistane of the collector while allowing a DC current to flow. that configuration is good to a few hundred kHz, above that you'd use a resistor + inductor instead.
C7 decreases the current source impedance above a few tens of kHz to reduce RF gain in this case.

for large voltage gains the collector impedance of Q1 is very high, so you'd need a FET buffer as a follower. i recommend using some kind of feedback for anything that has to be linear, but since R11 should be a low value you'll want to tap off the AC feedback after the buffer amplifier.

longview
Dec 25, 2006

heh.
http://www.qsl.net/yu1aw/LNA/bfg135aeng.htm

here's another HF preamp solution i've been thinking about building, designed for SMD primarily though

i really hate when builders spec number of turns (especially on a given core) without also including the actual inductance though

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
that's really cool, thanks for the lesson. normally i'm so used to the "here slap a fet on the input it has 1000000000000 ohms impedance, just tune some poo poo on the front end who cares, get some voltage gain off that, then buffer with an emitter follower" thing.

longview
Dec 25, 2006

heh.
that still more or less applies up the point where the 2-5 pF of the gate starts affecting the tuning.

for example, the BF996 dual gate FET gate 1 is actually pretty close to 50 ohm resistive at 100 MHz, and turns into a ~30 ohm fully capacitive input at 1 GHz. at DC or low frequencies it's basically an infinite impedance.



here's an example of a 200 MHz gain controlled buffer from the BF998 datasheet, here C1, L1 and D1 control the input matching.

--

before the summer i actually drew up plans for a HF transverter to use with my bladeRF/other SDRs. if i build it that design will include a BFG135 based HF preamp and a LTC5510 high linearity active mixer, upconversion will be to around 900 MHz to allow the use of approx. 25 MHz wide cellular SAW filters on the output to reject the LO and mixer images. i also have two boxes of VCOs suitable for that frequency range so might as well use some of them.

including an automatic gain control amplifier could be a useful addition; a BF998 might be suitable for this job, and building a tunable preselector should be possible without too much trouble at the same time.
direct coupled resonator filters (http://www.changpuak.ch/electronics/Direct-Coupled-Resonator-Bandpass.php) seem like a decent way to design narrow filters without too many components and there's a lot of tuning room in just changing capacitor values. i might look into what it would take in terms of tuning range to make a 1 MHz wide full band tunable filter. it might even be possible to change the width electronically...

AGC would probably be done by measuring the RMS input level over a fairly wide bandwidth and using a servo to adjust the gain to whatever average level doesn't cause massive IMD. the AD8307 RMS detector is pretty decent and cheap so i'd go with that as the detector.

would still include a pin limiter using the HSMP-4820 though, just to clamp peak levels without distorting too much. the only thing required is a pin diode with a very sharp turn-on curve and a suitable inductor.

another thing i looked at was setting up a bank of standard SDR dongles hooked to this output to a few raspberry pis and accessing them over ethernet. boxing it up with 3-4 channels would let me monitor several bands at a time and i could provide a more stable clock (28.8 MHz; just double a 14.4 MHz TCXO, i have 399 left so shouldn't be a problem) for narrow band monitoring. would also give me a place to put my ADS-B receiver.

Captain Foo
May 11, 2004

we vibin'
we slidin'
we breathin'
we dyin'

I'm still :eyepop: at 12 million line csv

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
going through my new pile of counterfeit chinese semiconductors (lol)

picked up a strip of these, Ft looked pretty good and theyre big enough to be easy to work with

http://www.nxp.com/documents/data_sheet/BFG135.pdf

also i have 1SV149 tuning diodes now, ~30-400 pf or so. dang

what else. restock on 2n5109's....couple handfuls of j310s and 2n5486's.....just random small signal RF poo poo. Hundred 2sc3355's, looked like a decent general purpose bipolar for RF business on the cheap

oh and a 10tube of SA612's and i'm itching to make a lovely little direct conversion receiver

Jonny 290 fucked around with this message at 08:49 on Sep 29, 2016

longview
Dec 25, 2006

heh.
i have a pile of BFG135s from a danish ebay seller, i think it was 25 or 50 for $1-2. it's nominally part of a design i haven't built yet for a coax stub resonator oscillator around 500 MHz.

BFG97s are used in the stabilock IF section, they're slightly strange since they have dual emitters, but afaik they can be used as a single emitter device without having to worry about it. they can be had fairly cheaply from hfo_poland on ebay.

since we were discussing common base amps, i found one in the stabilock schematics:



input is from the rx mixer at 450 kHz, presumably it's CB since the mixer expects a low impedance load. there's some matching near the mixer as well. schlumberger designed it with closed loop feedback to the emitter and included a push pull stage for the output.

notice how it's drawn upside down, i don't know why. without worrying too much about the details R233 and R295 is the main feedback network, with T203 switching in R236 to reduce the gain as needed.

control inputs in this part of the unit swing from ~-12 to +12V, they use NE5534 opamps as comparators to interface the 5V control lines. a fairly expensive solution (especially when an LM358 or even 741 could do the job much cheaper) but it probably simplified design of the rest of the circuitry.

last night i finished recapping the boards, no difference on the self test fault. plan is to make extender cables for the RF and LO boards so i can do an alignment and check there.
additional testing suggested the self check generator may have some AM it shouldn't have which could be the cause of the extra ripple seen on the scope trace, but might as well do a full alignment when i have it set up.

Jonny 290
May 5, 2005



[ASK] me about OS/2 Warp
code:
 #import derail_from_another_thread.c

creationist believer posted:

but 2.4 GHz is unlicensed

:eng101: wifi channels 1-6 exist in a part of the spectrum shared with the 13 centimeter amateur band. if you stick to "ham radio rules" (ID with your callsign, nothing offensive, no crypto and no for-profit traffic), you are allowed to use up to your license limit, which for all classes in the US is 1500 watts peak power and no limit on antenna gain, which is a bit more than the watt or two you get with 'regular wifi' :eng101:

the callsign thing is nicely dealt with by making your SSID your callsign, it goes out in the clear every packet

spankmeister
Jun 15, 2008






Jonny 290 posted:

code:
 #import derail_from_another_thread.c


:eng101: wifi channels 1-6 exist in a part of the spectrum shared with the 13 centimeter amateur band. if you stick to "ham radio rules" (ID with your callsign, nothing offensive, no crypto and no for-profit traffic), you are allowed to use up to your license limit, which for all classes in the US is 1500 watts peak power and no limit on antenna gain, which is a bit more than the watt or two you get with 'regular wifi' :eng101:

the callsign thing is nicely dealt with by making your SSID your callsign, it goes out in the clear every packet

nothing offensive? well then your posting is out of the window

Adbot
ADBOT LOVES YOU

Cocoa Crispies
Jul 20, 2001

Vehicular Manslaughter!

Pillbug

Jonny 290 posted:

code:
 #import derail_from_another_thread.c


:eng101: wifi channels 1-6 exist in a part of the spectrum shared with the 13 centimeter amateur band. if you stick to "ham radio rules" (ID with your callsign, nothing offensive, no crypto and no for-profit traffic), you are allowed to use up to your license limit, which for all classes in the US is 1500 watts peak power and no limit on antenna gain, which is a bit more than the watt or two you get with 'regular wifi' :eng101:

the callsign thing is nicely dealt with by making your SSID your callsign, it goes out in the clear every packet

shame about the encrypted traffic rule though, basically means you're stuck with websites that don't use https then?

  • Locked thread