locked :-:)

There are libraries for parsing command-line arguments you know!!

On a sunny day (Wed, 24 Oct 2018 01:48:52 -0400) it happened bitrex wrote in :

Really, what are you trying to say? Read the source.

is a much more civilized way of organizing complex set of command line options, but if you must be strict-POSIX then I guess one must live with switch statements.

at least it's better documented than most C source:

bitrex wrote

Since it was a 'hack' to get the LMX2332 to be programmed from Linux, as I could not access the National 'code_loader4' software, there is a command line option for 'everything' (unless I forgot something).

No library will make the code shorter or faster. If you problem is to remember some options, that works like this:

cat test2: ./LMX2332_test -x 6 -s 1 -t 0 -u 0 -o 0 -p 3

So typing ./test2 (after making it executable) runs it with those option.

In real life just use command line history, 'cursor up' right!

(cat because you like cats obviously).

From the level of blah blah politics, opinions, planes, and hamburgers, this is an electronics group, was just an attempt to publish

1) some new circuit. 2) some related code.

Better than 'how do I cure my epoxy', there is plenty of that in the boating group I follow.

Thanks for reading. The next part of the circuit is already on the same A4 piece of paper, PIC, ethernet (for control),

25 MHz oscillator, oscillator select with maybe a HC4053 (if I can get that chip to pass that frequency, have only tried those to 10 MHz), switching supply to see if it can be powered from the LNB (motor) power cable, RF cable driver, check LNB mod,.... should all go on the same board.

Sorry for mentioning electronics.

grrrr

A little exposition helps a lot, I dunno what a LMX2332 is off the top of my head or why you would want to use one, or for what purpose. The post jumps right into the heavy duty technobabble which is fine in general as this is a technical newsgroup but people probably tend to tune out without a "hook" know what I mean? I like technobabble as much as the next guy but a little "motivation" helps a lot

This is the US politics rant/what I ate for dinner last night newsgroup now, didn't u get the memo? :(

It's good to see an on-topic post! High praise for your source code, which is well documented and very easy to understand. You can prettify your source code further if you just drop the commented-out "print_usage();" lines altogether. IMHO, it's better for users to see the specific cause of their error instead of reiterated usage instructions. Your schematic looks great too (no flux stains on it). Well done.

Thank you, 73,

--
Don Kuenz KB7RPU 
There was a young lady named Bright Whose speed was far faster than light; 
She set out one day In a relative way And returned on the previous night.

Don Kuenz wrote

You are welcome. Today soldered and tested the second, the 25 MHz, crystal oscillator. The 24 MHz will be used for 10.4 GHz reception, the 25 MHz is for normal TV, and does not really need to be locked to my Rubidium, but added the circuit anyways, the LMX2332 has 2 completely separate PLLs, so it is possible. Just not sure if the divide by 25 works, will test that later.

Top:

SMD side:

About almost 30 SMDs now, can you find those? These board are great for SMD work, those fit exactly between 2 isles.

No RF instability if you wire it right, the LMX2332 will oscillate at GHz if you do not decouple it, or drive the inputs from a high impedance, that is why the JFET source followers after the oscillators.

II bought a batch of 10 of these chips for 10 dollars on ebay, can envision many fun projects with those chips, for example a very high frequency prescaler for a frequency counter.

The Tecsun SSB radio as frequency detector works great, its PLL is of only a few Hz from the Rubidium reference at 10 MHz, so I adjust for zero beat with the BFO in mid position, real fun. As that way you need no probe to connect to the circuit, the probe capacitance effect is gone. Much better than probing with a few pF frequency counter.

wrote in news:pqqakq$10j5$ snipped-for-privacy@gioia.aioe.org:

You are still messing around with analog TV signals?

I have some boards like these, but with a 0.05" hole spacing. You can bridge two with an 0805 of course, but it means it works for smaller components as well.

No ground plane needed in your design? I'm surprised.

Clifford Heath.

Clifford Heath wrote

hehe, yes, there is so much bull as to PCB layout, look at this at about 1 GHz:

scroll down for the wiring side of the PCB.

I have made so many board like that. You need to know the wavelength versus wire issue.

I was worried a bit for the LMX2332 adaptor, but as you can see here the same sort of adaptors work fine to 1 GHz. There are several decoupling SMDs under the adaptors.

But then RF is a bit of an art.

The whole board will fit in a metal case (it is 'euro card' format 100x160 mm), with the board far away from the front / back plates.

For this one if you look at the phase comparator error voltage it is very very stable, moving your hand towards the oscillator circuit you see it compensate a bit for the increased capacity. I am planning to order some more trim caps... Maybe these:

Now that is value for money!

Anyways, for a 1 off, build the whole thing in a few hours maybe, I try to avoid pee see bees. From an art POV this manual made 21th century RF circuit may be a million dollar collectors item one day, it does not even shred itself... or will it? ;-)

And it is 'learning by designing' as opposed to 'learning by destroying' as somebody else here uses as sig. Once I know how top play the LMX2332, more projects with it become possible. There is not so much difference with learning to play a musical instrument. I have stated many times: You can give a good musician a kids flute, and he will play wonderful music on it. you can give a kid a Stradivarius and he will destroy it. Getting into filosphy now, better stop here ;-)

Any tones I stepped on again today, live with it.

With the image fully zoomed-in, my eyes see about 10 SMDs. That's a good board for rapid prototyping.

This is good to know. It ought to come in handy with other designs.

Yes. It's always satisfying to re-engineer dated instruments to make them more state-of-the-art.

Interesting.

Thank you, 73,

--
Don Kuenz KB7RPU 
There was a young lady named Bright Whose speed was far faster than light; 
She set out one day In a relative way And returned on the previous night.

LMX2332_with_24MHz_25MHz_RF_switch_IMG_6647.JPG So that 74HC4053 as 24 MHz / 25 MHz switch has been tested now, works perfectly.

Now the essential part to be added is the cable driver to the LNB, few hundred mVpp.

The big potmeter on the left will go, was for manual tuning, the PIC socket is already there, probably will add the ethernet module (right) for control. as I already wrote all that code, and that gives POE possibility.

The little blue switch is to test the 74HC4053 will go too, chip will be controlled form a PIC I/O pin.

I ordered some trimpots too:

Unbelievable prices, 50 for < 5$ 25 turns 10k cermet. Here maybe they charge you 5 Euro for one plus shipping..

Needed one for output level....

So, after let's see writing 10172 lines of PIC 18 asm without any testing. and fixing 1 error (oops, forgot to disable some interrupt), the PIC 18F14K22 controls the LMX2332:

Frequency is stable and locked to Rubidium (RED connector bottom right) The Lipo 'simulates' the POE that goes into that internet connector.

reference frequency and PLL control voltage, clean halfway 3.35 V supply.

The program words for the LMX2332 are now stored in EEPROM, and the idea is that you can set those via ethernet, Thing has its own IP address, MAC and port. You can select frequency and also read status of course. and save to EEPROM if you want it kept. All remotely. The output frequency of this circuit goes via a cable to the LNB. The black label is to protect me from staring into a way too bright red LED. The ethernet module is from ebay, contains just an ENC28J60 and RJ45 connector, is controlled via SPI from the PIC. Replaced the connector by a POE capable one.

Its beginning to come together...

Using a LM317 to make a stable ripple free 3.35 V supply, it gets hot in POE,

150 mA, 200 mA or so, now on a small heatsink, maybe I will change that to a switcher, or add a switcher for the power hungry ENC28J60 chip, and keep the VCO supply on the LM317.

Added some stuff, switcher to reduce current consumption, and some emitter follower buffers to get better stability with the LMX2332, variable output level gain stage, cable buffer.

The diagonal wires go from the extra buffers to the 74HC4053 frequency select switch Maybe add a lowpass against harmonics in the 24 / 25 MHz output later.

The circuit diagram should be redrawn, but won't. And it is still changing...

Time to build it all in a box, drill holes etc.. POE has been tested, UDP commands it accepts. You can even do 2 way UDP communication with it by using 2 instances of netcat each one in it own terminal:

The one in the small terminal lets it know who is calling, after that everything can be done in the main terminal. Normally it only gets UDP commands to change frequency. The asm has not changed much except for some typos in the help menu:

Oh, and I removed the IC socket (strip) for the 74HC4053 as its pins seem to be thinner and it made bad contact. Temperature test on the XVCOs was good too. Will test again in a closed box.

And there it is:

Now if I only had a 3 mm LED for that LED holder... Just drilled a hole for it, no idea what to use the LED for, power?

Anyways this is why I like those eurocards so much, perfect fit for these boxes.