Ok. I feel like I'm being pulled in different ways....told one thing, told something different. Told to check one thing...then told it won't help.

I guess that's the nature of forums, but I will ship it off to Gordon and he has much more experience than me.

The supplied jack assumes that you will have a? 3.5MM stereo plug for the mike/PTT connections.? Use??shielded wire for the mike connections that has two wires in the shield.??If you are using the supplied microphone cartridge, you attach the outer case of the mike (small metal connections running from the solder pad to the case) to the tip connection of the plug via one of the center conductors of the shielded wire.? The other solder pad attaches to the sleeve connection through the shield of the wire.? One side of the PTT switch goes between the ring? connection of the plug while the other side attaches to the shield connection at the mike.

¿ªÔÆÌåÓý

But Q70 gets powered up on RX, and Q6 gets powered up on TX.
That power switching is really disruptive.
Might leave them both powered up, then the DC bias resistor should work fine, if you somehow kill the microphone

But given what we got, I'm leaning toward the FET clamp on rx audio.
Reducing the size of C52 on Q70 might help get the pop over with quickly.

In addition to the audio pop in the phones on rx, the uBitx may have the Bitx40's? carrier burst on tx.?
? ??/g/BITX20/message/33707

Allard has been fixing some of this with delays in firmware.
If we can get by without a special interrupt library in the code, the sketch will be much easier to
understand, add features to, and port to other processors.?
So I'd prefer to fix the rx pop and tx burst with hardware hacks.

Jerry, KE7ER

¿ªÔÆÌåÓý

I know you said you wanted to ship it to Gordon, but just for future reference: DC (or AC) measurements are fine as long as you don¡¯t deal with RF - that¡¯s where a really simple RF Probe is a good tool to have. You can build one with parts you have in your junk box:?
--
Karl Heinz - K5KHK

¿ªÔÆÌåÓý

Hi All,

I have done over 6 projects with Arduino/I2C adapter/1602 display.
I remove the I2C pullups from the ebay PCF8574 I2C adapter assembly, and let the
pullups to 3.3v (located somewhere else, like near the SI570)
do the work. The PCF8574 easily copies the 3.3V I2C
signals, as they rise above the approx 2.5v threshold for a 0-1 transition.

Works like a charm.

Rod KM6SN

I bought a 4x20 display that's I2C and was really surprised that what they
used to convert the parallel to I2C was a PCF8574, an 8 bit i/o expander. 
I use the PCF8575, the 16 bit version, when I need more DIO pins.

Vince.

¿ªÔÆÌåÓý

I think I misunderstood the video and should try it tonight? I noticed last night the speaker hiss now appears to have a crackle in it and has a whine. That wasn't present before the pop, flash and smoke

PCF8575 happily works from 2.5-5.5 volts so it can co-exist with the
Si5351 at 3.3 without issue.

Vince.

--
Michigan VHF Corp.

¿ªÔÆÌåÓý

?John -- there are two capacitors in series at that point in the receiver --- are you talking about using a resistor to keep the center point at ground?? ?

I'm interested to see if this will work, because when the driving transistor is suddenly eneregized, I think you're STILL going to see a spike AC signal propagate through the two capacitors unless you roll off the high frequency response significantly.? ?

Gordon

¿ªÔÆÌåÓý

?I might add that the "prodding video" is actually FUNCTION TESTING, stage by stage, and the fact that they let you see and hear what to expect is TREMENDOUSLY HELPFUL.

Further, I haven't looked to see if every bidirectional amplifier is identical, but if they are, once you have seen the voltages in one of them, the voltages in the next one of the same type (RX for RX,? TX for TX) should all be the same.? ?

When you're in the low signal receive stages the RF portion is so tiny that it won't (hopefully make much difference).? ?

In the later stages of the transmitter there could be 2 volts of RF or more.

gordon

Ok thanks. Sending it off tomorrow ha

Thanks to everyone on this. Hopefully I learn something from this and can eventually do it on my own.

Jack's idea is good, we need more pins (if not for a keyer, then for something else)
and driving the display from the same i2c pins as the Si5351 is using should work great.

Except that the Si5351 is a 3.3v device with 3.3v pullups on those i2c lines.
Most of the Arduino world is stuck in a 20 year time warp at 5v.
So this could get messy.

Best solution might be to buy a new display that operates only from 3.3v,?
and includes the i2c adapter.
And a 3.3v regulator, since I doubt the USB chip on the Nano can give us much more
current on the 3.3v rail.

Alternately, create a new Raduino running the ATMega328p (or Teensy, etc) at 3.3v.
That's perhaps the best long term solution.

Might be some hash in the Si5351 clocks when updating the LCD that needs to be cleaned up.

The current sketch only drives data and clock from Nano to Si5351, this works great.
Si5351 register reads also work, since the Nano's input cell has a minimum high
of something below 3v, though this likely wouldn't work well at an i2c speed of 400khz.
?
Jerry, KE7ER

¿ªÔÆÌåÓý

?OK.....I would caution that the parts are IN A CIRCUIT.

So things are more complicated than when they are out, unconnected to anything else.

Most of the rig is common emitter amplifiers, which are arranged in forward and backward pairs, making a bidirectional amplifier.

If I remember properly, one set gets energized with the RX line gets +12 volts;? the other amplifier instead gets energized when the TX line gets the +12volts.

Trying to measure resistances in connected circuits is more work than I'm willing to do, because you have to take into account everything in parallel all the way around the various loops around each component --- which could include transitors (effectively, diodes) and real diodes, and transformers and other resistors.

What is more useful generally to me is function testing:

1)? ?we suggested a simple way to do that for large sections by simply looking for specific oscillator signals.

2)? ?the alternative is to look for proper voltages in each common emitter amplifier.? ? Take Q1 for example ---?

it is fed through a 220 ohm resistor.? ?On one side should be nearly +12 volts, and on the other side would be a lower voltage, allowing you to calculate and approximation of the??collector current via Ohms law.? Because of the Diode D1 there, the actual collector voltage should be 0.6 volts lower than what you measured on the transistor side of the 220 ohm resistor.?

The base voltage is provided by a voltage divider from the collector, with 1000 and 2200 ohms and generally the base current isn't as large as the voltage divider current, so the voltage on the base should be somewhere at or below about 1/3 of the collector voltage.? ?

The emitter voltage should be right on 0.6 volts LESS than the base voltage, and the dc part of that flows through a 100 ohm resistor, so again, you can calculate the current through the emitter--- which should be almost identical to the colleector current.??

If the emitter were at 0 volts and the collector at 12 volts that would mean the transistor is completely non conducting....probably not a good sign.

If the collector were at 0.2 volts higher than the emitter, that would mean the transistor was in full saturation.....not a good sign either, something much worse is going on....

If the base isn't exactly .6-0.7 volts higher than the emitter, and the circuit is powered, something ver bad is wrong with the circuit or the transistor.

IF the ciruit is an oscillator, it becomes much harder to measure becaure there is RF running arund and it may confuse your voltmeter, or your voltmeter leads may foul up the oscillator; so easier to check initially by hunting for their signal with a sniffer (insulated) wire.? ?

Others may have a brighter way to test than that, but that is sorta what I do.? ?I first understand how the circuit is supposed to work, and then i look for voltages to make sure it all fits.

Gordon

¿ªÔÆÌåÓý

On looking at the schematic... I think the solution might be simpler: use a high-value resistor to keep the DC level of the capacitor at the output of Q70 at ground. I saw this referred to a while back in an audio mixer circuit as "zero point switching" to avoid pops. Let me see what I can kludge in without messing up the beauty of the board :-)

John K0JD

¿ªÔÆÌåÓý