Be sure you don't have split enabled.??????????? N3SB

Gordon - did you do something unusual or different to put together the build environment?

I'm asking because the RPI4 pin functions on the sbitx schematic don't match the functions that the RPI foundation specifies for the RPI4.

If you install and use the raspi-gpio tools from the command line to manipulate the I/O lines coming out of the RPI4, it becomes obvious pretty quickly that there are some difference. It may be due to the specific I/O library that Farhan is using. I haven't found the time to research this further - especially since the build is working fine using Farhan's stock image.

73; Steve, N3SB

Also - with your rebuild of the sbitx application - did you remember to set a lower value of cwdelay to suit your desired breakin behavior?

73; Steve, N3SB

Gordon - if you installed the solid-state TR relay mod kit such that the white coax is connected to the fat PCB trace that runs from the PA section, next to the lowpass filter relays, then you are already bypassing the Reed TR Relay K9 and the TR switching is fully solid-state. If you connected that white coax to a thin trace that heads towards U3, then the reed TR Relay is in-line. It is a pretty fast relay, however. I have my Solid State TR Relay connected in series through the original reed TR Relay to add more isolation.

The lowpass filter relays get configured when you change bands, and retain that state for as long as you are on that band. Note that in the process of fixing another bug, Farhan found that the lowpass filter relays were not getting updated. V2 owners never noticed this, but the Solid State TR Relay mod implemented in the DE actually listened through the lowpass filters, so it was causing a problem for DE owners. Farhan fixed this bug pretty quickly and the update was put on the sbitx github.

73; Steve, N3SB

I've installed a new Nano so I can load the N8ME software without impacting the original build and I am having a hard time getting the frequency to calibrate.? I've tried with the original uBITX software as well as the N8ME software with no luck.? I have even resorted to using a signal generator, which I can hear just fine with the original Nano, but I can't seem to be able to dial in anything recognizable with the frequency calibration setting.? The settings seem to have an extraordinary range, both positive and negative, and I can't figure out what they correspond to.? They also jump by large increments.??

I've dug through the posts but I'm not able to find anything that helps me.? Any suggestions?? What is a reasonable range (+/-) for the offset in the frequency calibration configuration?
-----------------------------------
73, Greg, KM5GT

To be certain I wasn't missing a end-transient, I reshot the 80 meter photo with the horizontal sweep at the same speed as Farhan had (5 usec per major division) -- and you still see nothing on the drain of the PA Mosfet.

I do not understand why Farhan's scope shows an incredibly fast transition of the TX line, whereas mine shows a discharge with a time constant of 5-6 microseconds.? ?
The TX line is being pulled down by the RPi output, pulling against mosfet gate capacitances mainly.? ? Not quite sure why mine looks slower than his.? But that could be partly why he saw oscillations??

Gordon?

Definitely the right file. Starting to agree with the folks that are wondering whether there is a hardware problem.?

73
Mark
--
AJ6CU
KD8CEC 2.0, Nextion Screens,? and open source uBITX Raduino boards for Arduino IOT, BLE, RP2040, Teensy 4.0, and Raspberry Pi Pico
https://github.com/aj6cu

Sorry Evan it does not. When I try that the cross does appear but it does not register when I touch it. The touch screen function is not working.?
--
'72

Looking for advice / wisdom.

I loaded this version 2 update.? I didn't understand there were more instructions so took me a while to find them and get sqlite etc -- but compiled fine then and runs well.

UNFORTUNATELY for me at least, with a relay-bandpass-filter Development Edition, the solution for the end-transition oscillation causes the loss of the relay-free "semi breakin" CW operation that was so precious about this radio.? ?(It should be fine with the diode switched bandpass filter).??

So I set about to better understand both the issue (oscillations at the end) and the solution.
Curiously, even before I added the 12V zeners to the gates of the PA transistors, I never lost a PA transistor.? ??
The likely mechanism explained by Ashhar is that the T/R circuit connects output of PA, to input of chain (at Q11 base).? ?Makes good sense.?
Until bias & supply filtering capacitors C12, C23, and? C105 are discharged by Qmumble, the PA transistors and the chain are fully able to amplify -- and oscillate!

However, another way to attack this feedback is to DELAY the passage of signal through the T/R circuit, which is accomplished by R5 and C7 of the upgrade TR circuit. (Onset of block is sped up by diode D10, but decay of block is delayed by R5/C7).? ?Since the observed oscillations are a few microseconds, the delay posed by R5 and C7 needs to be GREATER than the decay time of C12, C23, C105 etc.? ?The time constant of R5 / C7 is almost 500 microseconds, so one might think this is adequate to prevent oscillations lasting microseconds.

Hoping to justify taking out the Low Pass Filter Relay switching FOR CW, (I don't mind it in any other instance), I attempted to MEASURE the same transients that Ashhar presented.? ?I am certainly not as good at managing scope probes and grounds as he, but in the accompanying photos, the purple trace is the TX line (grabbed from the green wire going to the TX board) and the yellow is RF from one of the PA drains by touching a 10X probe to the tab.? ?cwdelay is set to 100msec? ( I dont know what to say about the RF visible on the TX line (purple trace).? ?I don't know if that is artifact or real.)?

I looked at three different bands, 80/20/10? ?I'm nervous still about high power so the power loads were more like 15 watts on the lower bands and lower still on 10 meters? Nevertheless, I can see NO indication of a trailing oscillation.? ?

COMMENT?? ?Would it seem reasonable on this particular DE radio to eliminate the relay switching on CW?

THANKS,
Gordon KX4Z

Attached is a better .jpg of the XLoader screen.

I will try and install the original screen and firmware.

I hope you guys are not wasting too many brain cells on this one.? I figured it out.

I had two things, no, three things wrong.

First, I was confused where to set PA Bias.? It is set after the initial PTT power indication.? Then you increase it 200mA more. Next, you perform metercal and txcal to set the power settings.? Finally, and most importantly, you have to make sure your power connection is solid.? I was fooled thinking I had killed it again when in reality, at high power demand, the power connector fails.? When it cools off, it makes contact again and you think you broke stuff.? Tomorrow, I will replace the connector with a XT-60 (or similar current device).? Note to self: Next time Allison makes a suggestion... take it!?

Thanks guys.? Sorry to waste your time.

Cheers, Scott

Hi Scott;

You should definitely see a voltage drop across R77 when in transmit. When transmitting, one side of R77 is connected to +12 Volts, and the other side is connected to D6, a 5.6 Volt Zener diode. Please check again. It's possible that the Zener diode has opened up, or that some other part in the PA bias circuit got fried when the FET failed.

Adjusting the PA Bias pot will have no effect on current until you get close to the point where the FETs begin to turn on - then current will begin increasing rapidly. Your description of the PA Bias pot behavior sounds correct.

Transistors Q12 - Q16 are still in the transmit chain on a fully modified DE. They are known to get toasty hot when in transmit. I can see them accounting for the 300 mA current increase. (I haven't bypassed Q16 and Q17 driver FETs in my DE, so I can't compare results with your DE.) You can measure the voltage drop across R19, R20, R45, and R48, and determine the current being drawn by each transistor using ohm's law. Don't forget that these four resistors got replaced with new resistor with a lower value (The V2 has 4.7 Ohm resistors in these four locations.) Gordon KX4Z and others have provided voltage drop measurements at these four resistors here on the BITX20 forum, so you should be able to compare your readings to those previously stated here.

73; Steve, N3SB

Hi Guys,

Thanks for the feedback.? Here are some more notes.

Hi Gyula,

Yes sir.? The reception current draw is approximately 680mA @ 13.8VDC.? When I key the mic (PA Bias pot fully counter clockwise) current jumps to approximately 1.05A.? It's as if there is a partial short in the bias circuit.? I'm just not finding it.

Yes, I've got a 50ohm dummy load connected.

Hi Evan,

Right.? I expected an increase and if memory serves. when I did this the last time, the increase was 10's of mAs, not 100's.??

Power supply is a 20A linear source.? It's pretty darn solid.

Hi Steve,

Actually, this is a DE machine where Q16 and Q17 are isolated and the circuit is bypassed (reference?).
If I'm reading this right, removing C30, C32, C100 & C101 isolate the original driver circuit.? The only component that has not been removed is R50.? That's the 12VDC source for the now removed Drive Bias circuit. I've not removed it because I see no voltage drop at Q9, R10, R50 or R77 when transmit is asserted.? That leads me to think there is something wrong after the power amplifier.? Perhaps a short in the output filter somewhere.? A sticking relay?? My SWR bridge?? I kind of think it's not in the bridge because it's a 100W device and if it were shorted, I'd not get reception.

What's weird is the adjustability of the PA Bias doesn't work as it has.? At PTT activation it jumps 300mA.? Is that right?? It doesn't seem so.? Starting at CCW, I'd expect the current to only increase 10's of mAs and after a few degrees of rotation, it would start to climb.? It doesn't react until about 50-60 degree of rotation and then is climbs very quickly above the 300mA initial jump.

Maybe I'm doing this wrong.? I know the bias is to be approximately 200mA above the CCW level but is this to be 20mA above the initial PTT increase?? Maybe it is and I'm just confused.

Anyway, thanks for your help, Scott??

?

Frank,
you can read about the options here:?

--
Gyula HA3HZ

I don't own a sBitx (V2), but the project looks very promising!

What could pull me over the line to purchase a ready build or pcb:
-Frequency range from 80~6m?(even better 160~4m) (including my favourite band 60m !!!)
-A better equal output power over all bands!
-More information about the size of the PCB (including height of the used components)
-Sales of the compleet sBitx / sBitx PCB from a?distributor within the Euroean Union (making a purchase more?affordable including taxes & shipment) for European customers!?

73 de Frank PH2M

Scott - it's normal to see some increase in current when switching to transmit - even with the PA Bias pot fully counterclockwise. The driver stages draw some current when you switch to transmit. The 300 mA increase may be normal.

With the PA Bias pot fully counterclockwise, measure the gate voltage at Q16 and Q17 when you switch to transmit. It should be zero. If it's not, then the PA FET failure damaged the bias circuit, which is quite possible.

If you do see zero volts at those gates, then slowly turn the PA Bias pot clockwise and confirm that the voltage slowly increases. Don't exceed a Volt or so unless you are monitoring the current as well. Drain current increases rapidly as the FETs start to conduct.

73; Steve, N3SB


73; Steve, N3SB

开云体育

Chuck did you ?ever live in Colorado Springs?

ps.: In the meantime, don't forget to have an artificial load on the antenna connector.
The reception current consumption of the radio (in any state) =/= during transmission is not the same as the current consumption at rest (without a signal).
--
Gyula HA3HZ