At $59, this will be going out to a lot of new hams worldwide. ?Like it or not. ?Some in rural areas without much of a support network or money for additional materials. ?The majority will just open the box and wire it up to be on the air. ? But others won't. ?We at least need better debug instructions. Much of the trouble seems to be in the final. ?Perhaps it's good enough to just post DC voltages around the three transistors there, as Raj did a couple months ago, ?Then what, they speak into the mike and measure the RF power out on their Bird wattmeter? ?That assumes an awful lot of stuff is suddenly working, and is not very concise. ?There should be a recommended way to inject a known signal into the final and measure the results. ?One way to generate a known signal level into the final might be to unbalance the modulator with a DC voltage and move the BFO down to where it's centered in the filter passband, as I have been recommending for CW transmisssion. ?(Ideally, the Si5351 is also used for the BFO.) ? That still assumes half the rig is working, though with an RF probe you could work your way from BFO to the IRF510. ?For example, varying the BFO frequency and inspecting the signal going into the diode ring mixer with the RF probe on transmit would verify that the crystal filter is working, and show the center frequency.

Alternately, a digital pin shorted to ground at power up on the Nano tells it to turn on the Si5351 B channel with a 7.2mhz signal, goes out through two series resistors to ground to attenuate the 7.2mhz to about 0.1v pk-pk for injection into the base of Q13. ?That's my RF signal generator, and it would cost about 4 cents for those two resistors. ? If I were to kit those resistors up with a nano, si5351, display, pcboard, a few other parts, autoinsertion, mailing, dealing with folks saying it doesn't work or didn't arrive, my price would be something like $50. ?Much better to simply put those two resistors down on the main board for an additional 4 cents. ?A schottky diode, cap, and a resistor or two would bring that up to about 10 cents, and provides the RF probe to dig through the final with. ?If it's on the main board, then everyone has it, and we can have a common set of instructions for how to go about debug, and can get measurements that should agree. ?With some work, this could be a real educational tool for the new ham rather than a frustrating puzzle.

The problem with the diode RF probe is it doesn't have much dynamic range. ?However, given how gain is distributed throughout the Bitx40 RF stages, we may not need much dynamic range. ?Especially if instead of 2 resistors, we have a string of 3 or 4 at the Si5351's B output, providing RF signals at various levels. ?Might consider having the AD8310 available as a $10 option, which would also give us a very very nice RF probe, a calibrated S meter, and sweeperino like capabilities.

The STM32 processor boards mentioned elsewhere are interesting. ?For about an extra dollar over the Nano, you should have enough power to send and receive some of the digital modes. ?The ADC resolution goes up to 12 bits. ?Counter-timers can directly count at 30mhz+ for a very nice freq counter. ?More communication channels for perhipherals. ?And enough flash that this debug stuff could get quite sophisticated without getting in the way.

Creeping featurism run amock. ?I'm ok with the Nano, and instructions for dead-bugging a diode RF probe. ?It's all Ashhar's call. ?

Jerry, KE7ER

?On Sat, Feb 4, 2017 at 10:56 pm, Todd K7TFC wrote:

开云体育

You probably won't like this but that code just isn't worth it, it's all about display.? You would be better deciding what you want and write your own or hack the display code about.??? Heck, use the Raduino code but rewrite for tft display,

Trouble is display coding for a tft display is not trivial compared to an lcd, even at 20x4 line lcd, that's easy !

Use Jacks rotary speedup for just one band use along with a simpler uncluttered tft display.? I know the filter stuff etc looks nice but it's not relevant to a BitX.

I can send you some code already done for tft's for you to look at and hack.? I use a 'basic' style of coding which is easy to read and follow.

One tip I find helps whenever you get code to look at, is to first segment the functions with the single comment and hashes (pounds). i.e.?? //#######################

That way you get a clear idea of program flow.

Better stop now as I suppose this isn't relevant to BitX group.
--
Rob G4NQX

Jack,

No, whenever I use an 8 bit display I automatically use a Mega for the pins.? I've done all sorts of tricks before to use Nano or Uno even,? but it's not worth the hassle as they are cheap enough.

Apart from my portable rigs I rarely try to shoehorn or miniaturise projects.? If I use smd for something they automatically become smaller anyway and in fact I like smd work.

The signal generator project uses a Mega, a 3.5inch 8 bit mcufriend display, a 4x3 keypad for direct frequency entry to suit either of the modules, si5351 or ad9850(1)

An ad8703 log amp as power out measurement. Future work is to add switched lpf's to cater for ham bands. (I've already done this on another project) Amp and output leveller for the ad9850. Then switched attenuators.

Features as now are,? single frequency - frequency sweep between ranges - power out measurement - select either rf board - frequency limits to suit.

I can send you the sketch if you want a look.

--
Rob G4NQX

开云体育

Thanks Hans,

It's an odd one then.? I certainly uploaded the Tips doc ok, then next time it was no upload button.

Needs a mail to groups.io methinks.
--
Rob G4NQX

John, It sounds like the screeching and dimming is caused by the power supply falling over when the IRF starts drawing heavy current from it. It may be due to the gate bias being set a little high and the IRF going into avalanche. (that's why it stopped when you turned the bias down) Is your power supply beefy enough? Although you probably won't need more than a couple of Amps it's nice to have some headroom. Bear in mind many 'Overseas' brands rather overrate their specs! PC power supplies are a great, cheap, alternative to expensive lab supplies. but be carefull, they can supply a LOT of current if you short something! (Fuses are your freind!)

Also, I'm not sure your ammeter really has the resolution to accurately set the bias current. It might be worth using a current shunt resistor to measure it with. Dig around in the junk box (or somebody elses!) for a 1R 15W power resistor (or something close) stick it in the PA power line and measure the voltage between the resistor's terminals. Using the magic of ohm's law (I=V/R) you have your current measurement.

Hope that might be helpull, I'm no expert by a long way, but i'll help you out if i can.

开云体育

Sent from my iPhone

My suggestion would be get a novice licence to start. Kb6nu has excellent courses for all exams.

That would be a good introduction to what you're getting into.?

Paul kk4ptv

Art,

Are you using 3.5 inch or 2.8 inch display. Code is written for 2.8 and will take some scaling to fit properly.

I'm using a 3.5 inch mcufriend here on a twin signal generator project and it's 480 x 320. (Notoriously difficult to photo)

--
Rob G4NQX