开云体育

Hi Jonathan,?

Impressive mods to your rig.??

I would add to Raj's suggestion that you check around the L5 and L7 mod.? Since CW does not use the BiDi amps, and you did modify the connection from the BiDi amps to the mixer, you may have gotten something misconnected, shorted, or open in that area.

Have fun finding the issue.
73
Evan
AC9TU

Hi everyone,

I applied some mods to my ?BitX v4 tonight.? In general they seem to have worked, but I also seem to have lost the ability to transmit using SSB (CW seems to work beautifully).? I was wondering if anyone could help me debug.

More details follow on the mods I applied and the specifics of the problem I'm now encountering.

=== What I did ===

I applied to following mods, in the order listed, testing CW into a power meter?+ dummy load after each step:

1. Replaced the four relays near the antenna with the exact model of Axicom relays suggested ().

This resulted in slightly higher power output (10-20%) on 160m through 20m, and less or no increase in power on 17m-10m.

2. Replaced Q90 and R81 ()

This increased power a bit on some bands (especially lower ones) and decreased it a bit on other (especially higher ones)

3. Replaced C81 (per above) and L5 & L7 ().

Resulted in increased measured power on all bands, except no effect on 30m

4. Replaced the 6 2N3904s with TO18 2N2222As (checking specs for both to ensure proper orientation) and doubled up the 4 adjacent resistors (per Allison's fix above)

Resulted in much higher measured power, nearly twice previous measurements on some bands, especially 10m, 17m, and 40m.? Max I measured is ~26W at 1900kHz and minimum I measured is ~4.3W at 21MHz (although 21.2MHz was at ~9.3W).

I measured power at a list of 29 frequencies each time, and can send numbers and/or a graph if anyone's interested.? I can also send pictures of the board.? (Despite it basically being my first time soldering SMD, I think I did a decent job with all of it.? And my solder removal skills definitely got some practice!)

Unfortunately I can't currently measure spurs or harmonics, but the hope is that these fixes reproduce others' results.

=== Afterwards ===

After I tested power output with CW, I hopped around 80m, 40m, and 20m, listening to FT8 and WSPR, and tried transmitting in both modes on 40m.? After my WSPR transmit didn't seem to be picked up by anyone, I put my power meter inline again and found that SSB doesn't result in any power out (the relay clicks and the LED wired to the relay lights up, but there's no RF), with either my computer hooked up (using e.g. WSJT-X's "tune" button) or a microphone (+whistle test).? The CW key still results in the power levels I measured after all the fixes.

I also noted that there is a new smoky smell coming from the board....? Upon further inspection I noticed that some of the 2N2222As were a bit warm, but nothing else seemed out of the ordinary.

=== What I've checked ===

Besides a visual inspection that didn't yield much, I checked the line from the microphone itself to C60, and it does indeed conduct.? Which leads me to believe that the problem is something on the board.

Note that I'm running the ?BitX on a 3A 13.8V power supply.? I have a 3A fuse between power in and the on switch, which has never blown—but 2.5A fuses used to blow any time I had a little too much reflected power on some of the hotter (per original power measurements) bands (not as much of a problem with my current antenna on most bands).

I don't have much test equipment—mostly just a simple multimeter.? I might be able to get access to an oscilloscope if really needed.

Any suggestions for things to check are welcome!

--

Jonathan, KD5CFX

Hi everyone,

I applied some mods to my ?BitX v4 tonight.? In general they seem to have worked, but I also seem to have lost the ability to transmit using SSB (CW seems to work beautifully).? I was wondering if anyone could help me debug.

More details follow on the mods I applied and the specifics of the problem I'm now encountering.

=== What I did ===

I applied to following mods, in the order listed, testing CW into a power meter?+ dummy load after each step:

1. Replaced the four relays near the antenna with the exact model of Axicom relays suggested ().

This resulted in slightly higher power output (10-20%) on 160m through 20m, and less or no increase in power on 17m-10m.

2. Replaced Q90 and R81 ()

This increased power a bit on some bands (especially lower ones) and decreased it a bit on other (especially higher ones)

3. Replaced C81 (per above) and L5 & L7 ().

Resulted in increased measured power on all bands, except no effect on 30m

4. Replaced the 6 2N3904s with TO18 2N2222As (checking specs for both to ensure proper orientation) and doubled up the 4 adjacent resistors (per Allison's fix above)

Resulted in much higher measured power, nearly twice previous measurements on some bands, especially 10m, 17m, and 40m.? Max I measured is ~26W at 1900kHz and minimum I measured is ~4.3W at 21MHz (although 21.2MHz was at ~9.3W).

I measured power at a list of 29 frequencies each time, and can send numbers and/or a graph if anyone's interested.? I can also send pictures of the board.? (Despite it basically being my first time soldering SMD, I think I did a decent job with all of it.? And my solder removal skills definitely got some practice!)

Unfortunately I can't currently measure spurs or harmonics, but the hope is that these fixes reproduce others' results.

=== Afterwards ===

After I tested power output with CW, I hopped around 80m, 40m, and 20m, listening to FT8 and WSPR, and tried transmitting in both modes on 40m.? After my WSPR transmit didn't seem to be picked up by anyone, I put my power meter inline again and found that SSB doesn't result in any power out (the relay clicks and the LED wired to the relay lights up, but there's no RF), with either my computer hooked up (using e.g. WSJT-X's "tune" button) or a microphone (+whistle test).? The CW key still results in the power levels I measured after all the fixes.

I also noted that there is a new smoky smell coming from the board....? Upon further inspection I noticed that some of the 2N2222As were a bit warm, but nothing else seemed out of the ordinary.

=== What I've checked ===

Besides a visual inspection that didn't yield much, I checked the line from the microphone itself to C60, and it does indeed conduct.? Which leads me to believe that the problem is something on the board.

Note that I'm running the ?BitX on a 3A 13.8V power supply.? I have a 3A fuse between power in and the on switch, which has never blown—but 2.5A fuses used to blow any time I had a little too much reflected power on some of the hotter (per original power measurements) bands (not as much of a problem with my current antenna on most bands).

I don't have much test equipment—mostly just a simple multimeter.? I might be able to get access to an oscilloscope if really needed.

Any suggestions for things to check are welcome!

--

Jonathan, KD5CFX

Several have posted voltage charts for various versions of the uBITX and BITX.

Your next step might be to compare what the voltages are versus what they

should be.? Please do the voltages comparison and post that back here so others

can look at that.

I'm sure you are aware that the SSB and audio part goes through a different path

than the CW signal path.

Arv

_._

On Thu, Jul 15, 2021 at 2:48 AM Jonathan Washington <kd5cfx@...> wrote:

Hi everyone,

I applied some mods to my ?BitX v4 tonight.? In general they seem to have worked, but I also seem to have lost the ability to transmit using SSB (CW seems to work beautifully).? I was wondering if anyone could help me debug.

More details follow on the mods I applied and the specifics of the problem I'm now encountering.

=== What I did ===

I applied to following mods, in the order listed, testing CW into a power meter?+ dummy load after each step:

1. Replaced the four relays near the antenna with the exact model of Axicom relays suggested ().

This resulted in slightly higher power output (10-20%) on 160m through 20m, and less or no increase in power on 17m-10m.

2. Replaced Q90 and R81 ()

This increased power a bit on some bands (especially lower ones) and decreased it a bit on other (especially higher ones)

3. Replaced C81 (per above) and L5 & L7 ().

Resulted in increased measured power on all bands, except no effect on 30m

4. Replaced the 6 2N3904s with TO18 2N2222As (checking specs for both to ensure proper orientation) and doubled up the 4 adjacent resistors (per Allison's fix above)

Resulted in much higher measured power, nearly twice previous measurements on some bands, especially 10m, 17m, and 40m.? Max I measured is ~26W at 1900kHz and minimum I measured is ~4.3W at 21MHz (although 21.2MHz was at ~9.3W).

I measured power at a list of 29 frequencies each time, and can send numbers and/or a graph if anyone's interested.? I can also send pictures of the board.? (Despite it basically being my first time soldering SMD, I think I did a decent job with all of it.? And my solder removal skills definitely got some practice!)

Unfortunately I can't currently measure spurs or harmonics, but the hope is that these fixes reproduce others' results.

=== Afterwards ===

After I tested power output with CW, I hopped around 80m, 40m, and 20m, listening to FT8 and WSPR, and tried transmitting in both modes on 40m.? After my WSPR transmit didn't seem to be picked up by anyone, I put my power meter inline again and found that SSB doesn't result in any power out (the relay clicks and the LED wired to the relay lights up, but there's no RF), with either my computer hooked up (using e.g. WSJT-X's "tune" button) or a microphone (+whistle test).? The CW key still results in the power levels I measured after all the fixes.

I also noted that there is a new smoky smell coming from the board....? Upon further inspection I noticed that some of the 2N2222As were a bit warm, but nothing else seemed out of the ordinary.

=== What I've checked ===

Besides a visual inspection that didn't yield much, I checked the line from the microphone itself to C60, and it does indeed conduct.? Which leads me to believe that the problem is something on the board.

Note that I'm running the ?BitX on a 3A 13.8V power supply.? I have a 3A fuse between power in and the on switch, which has never blown—but 2.5A fuses used to blow any time I had a little too much reflected power on some of the hotter (per original power measurements) bands (not as much of a problem with my current antenna on most bands).

I don't have much test equipment—mostly just a simple multimeter.? I might be able to get access to an oscilloscope if really needed.

Any suggestions for things to check are welcome!

--

Jonathan, KD5CFX

Looks fine, Q94/95 are the finals.

Remove L5/7 and put back the original toroids and see. Some SMD inductors dont work well at that frequency.

Send us a photo of the L5 and 7 you used.

Raj

At 17/07/2021, you wrote:

Thank you all for your quick and helpful thoughts and suggestions!

Raj and Ashhar asked about RX.?? The answer is that RX does "work", but it is MUCH quieter than before.?? I was able to decode some FT8 signals with the volume cranked way up, and can hear static changing when I tune with a headset and the volume cranked way up.?? But something's definitely wrong here too.

Evan asked about the area around L5 and L7.?? As far as I can tell, everything everywhere on the board looks okay upon visual inspection—no cracked components, no sloppy solder pooints, no stray bits of solder in bad places, etc..?? There are two exceptions to this: the resistors I stacked for R941 and R942 look considerably darker than they did in the pictures I took immediately after completing the modifications, and than their siblings on R911 and R96.?? If I were to guess the origin of the burnt component smell based solely on visual inspection, I would guess those might have something to do with it.

Rajesh, yes, I did replace C81 along with Q90.?? I did some testing (CW into dummy load) after I replaced Q90 and R81, but before I replaced C81—thinking about it now, it seems that perhaps thiss wasn't so sharp of me.

Raj, I do get voltage on TX on the bidi transistors.?? Per Arv's suggestion, I'm attaching a voltage table.?? The results are comparable to the tables Raj uploaded in 2019, taking into consideration that I'm feeding it 13.8V instead of 12V (though I had trouble identifying Q94/95 and so didn't measure that).

If anything looks off, though, or if there are any follow-up suggestions, I'm all ears!

--
Jonathan, KD5CFX

On Thu, Jul 15, 2021 at 1:30 PM Arv Evans <arvid.evans@...> wrote:

Several have posted voltage charts for various versions of the uBITX and BITX.

Your next step might be to compare what the voltages are versus what they

should be.?? Please do the voltages comparison and post that back here so others

can look at that.

I'm sure you are aware that the SSB and audio part goes through a different path

than the CW signal path.

Arv

_._

On Thu, Jul 15, 2021 at 2:48 AM Jonathan Washington <kd5cfx@...> wrote:

Hi everyone,

I applied some mods to my ??BitX v4 tonight.?? In general they seem to have worked, but I also seem to have lost the ability to transmit using SSB (CW seems to work beautifully).?? I was wondering if anyone could help me debug.

More details follow on the mods I applied and the specifics of the problem I'm now encountering.

=== What I did ===

I applied to following mods, in the order listed, testing CW into a power meter? + dummy load after each step:

1. Replaced the four relays near the antenna with the exact model of Axicom relays suggested ().

This resulted in slightly higher power output (10-20%) on 160m through 20m, and less or no increase in power on 17m-10m.

2. Replaced Q90 and R81 ( )

This increased power a bit on some bands (especially lower ones) and decreased it a bit on other (especially higher ones)

3. Replaced C81 (per above) and L5 & L7 ().

Resulted in increased measured power on all bands, except no effect on 30m

4. Replaced the 6 2N3904s with TO18 2N2222As (checking specs for both to ensure proper orientation) and doubled up the 4 adjacent resistors (per Allison's fix above)

Resulted in much higher measured power, nearly twice previous measurements on some bands, especially 10m, 17m, and 40m.?? Max I measured is ~26W at 1900kHz and minimum I measured is ~4.3W at 21MHz (although 21.2MHz was at ~9.3W).

I measured power at a list of 29 frequencies each time, and can send numbers and/or a graph if anyone's interested.?? I can also send pictures of the board.?? (Despite it basically being my first time soldering SMD, I think I did a decent job with all of it.?? And my solder removal skills definitely got some practice!)

Unfortunately I can't currently measure spurs or harmonics, but the hope is that these fixes reproduce others' results.

=== Afterwards ===

After I tested power output with CW, I hopped around 80m, 40m, and 20m, listening to FT8 and WSPR, and tried transmitting in both modes on 40m.?? After my WSPR transmit didn't seem to be picked up by anyone, I put my power meter inline again and found that SSB doesn't result in any power out (the relay clicks and the LED wired to the relay lights up, but there's no RF), with either my computer hooked up (using e.g. WSJT-X's "tune" button) or a microphone (+whistle test).?? The CW key still results in the power levels I measured after all the fixes.

I also noted that there is a new smoky smell coming from the board....?? Upon further inspection I noticed that some of the 2N2222As were a bit warm, but nothing else seemed out of the ordinary.

=== What I've checked ===

Besides a visual inspection that didn't yield much, I checked the line from the microphone itself to C60, and it does indeed conduct.?? Which leads me to believe that the problem is something on the board.

Note that I'm running the ??BitX on a 3A 13.8V power supply.?? I have a 3A fuse between power in and the on switch, which has never blown—but 2.5A fuses used tto blow any time I had a little too much reflected power on some of the hotter (per original power measurements) bands (not as much of a problem with my current antenna on most bands).

I don't have much test equipment—mostly just a simple multimeter.?? I mmight be able to get access to an oscilloscope if really needed.

Any suggestions for things to check are welcome!

--

Jonathan, KD5CFX

Hello! Given Jonathan’s mods making 26w (under some conditions), are there any toroid ?heat management issues at >20 watts through the various LPFs and output transformer?

And if he were to drive the finals with (say) 24v, then the finals might produce >30watts? Wow.

I’ve run into heat dissipation “issues” at >10watts — so I’ve called in the mega heat sinks, fans, liquid nitrogen cooling, etc.? At these levels generally, the worry about spontaneous Chernobyl events ?adds an element of excitement. And I discovered that at certain power levels, you can melt the little three pin power connector on the board.? Fortunately, any servo connector will fit the board jack, you do lose the polarity feature of the proper connector, which gives you another chance for excitement.?

Jonathan — what is your approach to heat management at your prodigious power levels?

Looks like what I used. Should be no problem.

A picture is worth a 1000 word.. You soldered L7 in the wrong place...



Now fix it and test and open a bud and drink it on my behalf!

Cheers
Raj

At 17/07/2021, you wrote:

Thanks, Raj.?? I'll put the original toroids back and report back, but today is YL's birthday, so it might not be until tomorrow...

The inductors I used are these:

Attached is a picture I took after installing them.?? I suppose it's also possible I damaged them while soldering them in—as I said, this was my first time doing SMDD soldering.

--
Jonathan

Hello! Given Jonathan’s mods making 26w (under some conditions), are there any toroid ?heat management issues at >20 watts through the various LPFs and output transformer?

And if he were to drive the finals with (say) 24v, then the finals might produce >30watts? Wow.

I’ve run into heat dissipation “issues” at >10watts — so I’ve called in the mega heat sinks, fans, liquid nitrogen cooling, etc.? At these levels generally, the worry about spontaneous Chernobyl events ?adds an element of excitement. And I discovered that at certain power levels, you can melt the little three pin power connector on the board.? Fortunately, any servo connector will fit the board jack, you do lose the polarity feature of the proper connector, which gives you another chance for excitement.?

Jonathan — what is your approach to heat management at your prodigious power levels?

Don KM4UDX

My suggestion would be to not get too excited about over power beyond the

design specification.? Take a look at the datasheet for IRF510 devices to see

where the best position on the load-line is for each feed voltage, and stay

within that range.? Operating at over-voltage stresses some components and

causes splatter from spurious signals.? Over-driving any transmitter is not

a good idea.? Higher power within the absolute limits of the uBITX brings the

finals into flat-topping mode where one or the other, or both, sides of the

modulation are clipped when the signal reaches positive voltage limit or reaches

ground.? After all you will get less than one S-unit increase in signal at the

other end, no matter how much you overdrive the uBITX.?

Overdriving your transmitter does nothing to help you receive better.? It is

much more productive to work on antenna effectiveness and help your

signal on both receive and transmit side of things.

Arv

_._

On Sat, Jul 17, 2021 at 8:44 AM Don - KM4UDX <dontay155@...> wrote:

Hello! Given Jonathan’s mods making 26w (under some conditions), are there any toroid ?heat management issues at >20 watts through the various LPFs and output transformer?

And if he were to drive the finals with (say) 24v, then the finals might produce >30watts? Wow.

I’ve run into heat dissipation “issues” at >10watts — so I’ve called in the mega heat sinks, fans, liquid nitrogen cooling, etc.? At these levels generally, the worry about spontaneous Chernobyl events ?adds an element of excitement. And I discovered that at certain power levels, you can melt the little three pin power connector on the board.? Fortunately, any servo connector will fit the board jack, you do lose the polarity feature of the proper connector, which gives you another chance for excitement.?

Jonathan — what is your approach to heat management at your prodigious power levels?

Arv - so on my own, locked in the embrace of my own inexperience and lack of training, I had a slowly dawning revelation — improving my antenna system (transformers, chokes, feed lines, radiating elements, etc.) produced more better net benefits than torching my IFR510s (yet again).?

So I can humbly confirm from direct hands-on repeated experience…get ready…you are right.?

Some hard earned knowledge is more precious than ?academic insight. Especially for the slow learners. Hahah

And when I started on my uBITX radio journey, I watch my signal in horror as I experimented. Turns out the key to a clean waveform is a reduced drive level. Or as my Dad used to tell me repeatedly, “turn that darn thing down!”

But - my question stands, at higher power levels, is there any benefit from upgrading any of the various post final toroids??