Topics

Spurious RF at beginning of transmission

Bob
 

Ashhar Farhan VU2ESE's BITX40 is an amazing QRP SSB Radio and has set an unbelievable price-performance point.  And many folks are adding new bells/whistles every day (like Allard and Jack).   I've been working with a friend of mine, Willy W1LY, and his BITX40.  Willy had installed the version 1.13 software from Allard and the HW modifications needed for CW.  We have both noticed some interesting RF behavior when the radio first enters TX, even when on LSB (just PTT, with no audio).

After some investigation we are reasonably certain that the burst of RF is coming directly through the balanced mixer which is unbalanced for about 60 mS due to the microphone amplifier stage being switched on for transmit.  If one starts with the radio in RX, then C124, the 47uF cap in the mic amplifier is sitting at zero volts.  When the radio goes into TX, that cap has to charge up via R127 -- which takes a LONG time.  During that time, Q12 is also powering up and C1

Bob
 

Arrghh -- sorry, phone call and I hit the wrong button.Continuing ...

When the radio goes into TX, that cap has to charge up via R127 -- which takes a LONG time.  During that time, Q12 is also powering up and C122 has been sitting at zero volts, so it starts charging as well.  This is the response of the Microphone Amplifier to the rising edge of the TX signal -- and the net result is that the Balanced Mixer (T4/D15/D16) is knocked out of balance until the amplifier stabilizes.  The result of that is a spurious RF transmission that looks like the one I have attached as a file.  The RF rises almost immediately to about 5-8 watts, within 5 ms drops down to about 3 watts, builds back up to 5-8 watts in another 5 ms, then slowly decays over a period of 60 ms decays down to close to zero.  (All of this observed with the PTT line held).

That's pretty much the equivalent of a DOT at 20 WPM.  You can hear that TX burst clearly on a monitor receiver.

When sending CW that initial burst may be masked by the first element as long as the speed is slow enough (< 20 WPM).

Since the unintended RF burst is caused by the transient response of the stage being powered up, a more optimal solution would be to keep that stage powered at all times, but shunt the audio to the base of Q12.  Reducing the size of C124 would help shorten the burst.

73, Bob, WB4SON

Jerry Gaffke
 

Folks have been reasonably certain of that on Bitx rigs for about 15 years now.  A simple fix for the kerchunk into the headphones when entering receive was recently posted here, and there have been many others.  The transmit kerchunk you describe has not been addressed so much.  Easier to fix now that we have the Raduino, could shut down the Si5351 VFO buffer at CLK2 until that kerchunk is over with.  Though 60ms is a rather long time, if using VOX that timeout I'm suggesting could easily clip off something significant.  Let us know when you find a better fix.


On Wed, May 31, 2017 at 02:07 pm, Bob wrote:
the burst of RF is coming directly through the balanced mixer which is unbalanced for about 60 mS due to the microphone amplifier stage being switched on for transmit.

 

Bob
 


Incidentally I posted a photo of the output spectrum as well.  Right at the US FCC limit of no spur more than 43 dB below the carrier.:  https://groups.io/g/BITX20/photo/6276/1?p=Name,,,20,1,0,0

73, Bob, WB4SON

Joe
 

Hi,

I think PTT noise is an issue which most everyone if not all have noticed.  There are several solutions. I finally decided on a circuit which uses a FET and transistor to cure the problem, I have done it on both  of my BITX40 and it cleared the problem. 

Here is the article:Email -- Subject -  PTT Thump mute

On Sun, Apr 30, 2017 at 6:28 PM, John Pieper <j.pieper@...> wrote:

Hi all, this is John, AD0RW, new to the group. I have built my BITX40 board into a full-featured transceiver,

adding CW (a must for me), AGC, XIT/RIT, built-in keyer and SWR monitor, among other things. One of the

last issues I needed to nail down was the PTT thump, which was a real annoyance. I came up with a circuit

that does the job nicely and is much simpler than the muting board that was offered for the 3B/3C. Here it is:

The concept is simple: put a short across the LM386 input very quickly when PTT goes low, before the relays

have time to close causing the audio noise. C1 and R2 hold the muting on for a short time after PTT goes high

to cover the relays' opening. R2's value was arrived at by experiment, and a 470k should work just as well

but may be easier to find. R3 and R4 are only needed if you have a CW sidetone that you need to inject during

transmit. The adjustable sidetone level is valuable because with this arrangement the volume control has no

effect on it.

With this addition, the rig is now a real pleasure to operate.

73, John

-- 
73, Joe W3TTT




Hope this what you were looking for
Joe
VE1BWV

On Wed, May 31, 2017 at 6:07 PM, Bob <wb4son@...> wrote:

Ashhar Farhan VU2ESE's BITX40 is an amazing QRP SSB Radio and has set an unbelievable price-performance point.  And many folks are adding new bells/whistles every day (like Allard and Jack).   I've been working with a friend of mine, Willy W1LY, and his BITX40.  Willy had installed the version 1.13 software from Allard and the HW modifications needed for CW.  We have both noticed some interesting RF behavior when the radio first enters TX, even when on LSB (just PTT, with no audio).

After some investigation we are reasonably certain that the burst of RF is coming directly through the balanced mixer which is unbalanced for about 60 mS due to the microphone amplifier stage being switched on for transmit.  If one starts with the radio in RX, then C124, the 47uF cap in the mic amplifier is sitting at zero volts.  When the radio goes into TX, that cap has to charge up via R127 -- which takes a LONG time.  During that time, Q12 is also powering up and C1


Bob
 

Hi Joe,

That's a good cure for the RX thump, but it doesn't deal with the TX spur that is about 60 mS wide.

73, Bob, WB4SON


On Wed, May 31, 2017 at 6:12 PM, Joe <joeman2116@...> wrote:
Hi,

I think PTT noise is an issue which most everyone if not all have noticed.  There are several solutions. I finally decided on a circuit which uses a FET and transistor to cure the problem, I have done it on both  of my BITX40 and it cleared the problem. 

Here is the article:Email -- Subject -  PTT Thump mute

Kelly Jack
 

Message 22961 proposed a solution to this. I implemented it thinking it was a solution to the audio pop and so unimplemented it when it didnt solve for problem. Unfortunately i didnt test to see if it solved the carrier burst issue.

Regards 

Simon VK3ELH

 

At 01/06/2017, you wrote:

Ashhar Farhan VU2ESE's BITX40 is an amazing QRP SSB Radio and has set an unbelievable price-performance point. And many folks are adding new bells/whistles every day (like Allard and Jack). I've been working with a friend of mine, Willy W1LY, and his BITX40. Willy had installed the version 1.13 software from Allard and the HW modifications needed for CW. We have both noticed some interesting RF behavior when the radio first enters TX, even when on LSB (just PTT, with no audio).

After some investigation we are reasonably certain that the burst of RF is coming directly through the balanced mixer which is unbalanced for about 60 mS due to the microphone amplifier stage being switched on for transmit. If one starts with the radio in RX, then C124, the 47uF cap in the mic amplifier is sitting at zero volts. When the radio goes into TX, that cap has to charge up via R127 -- which takes a LONG time.

Jerry Gaffke
 

A very nice $0.01 solution.
Except it kills how Allard is doing CW.
He unbalances the carrier by putting 5v into a 10k resistor that goes to the top of C107.

Maybe fix this PTT pop somehow in microphone preamp so we can continue to use Allard's CW method? 

Could transmit CW like Farhan is doing it on the uBitx:  Move the VFO to the transmit freq and unbalance the first ring mixer.  Those three mixer ports got stirred up considerably on the uBitx, though it is otherwise the same mixer as on the Bitx40.  Is there a some simple way to do this without a significant hardware hack to the existing Bitx40 boards?

Don Cantrell was injecting RF at the transmit freq from a spare Si5351 output,  Could key it by writing to an Si5351 register, so could be done with just a resistor.  But this method does not offer an easy method for key shaping, whereas the mixer unbalance schemes used by Allard and Farhan do.  With Allard's it's C107 that gives key shaping, on Farhan's uBitx it's C1.  Though I think the uBitx's C1 may need to be considerably larger (maybe 1uF, though I'm not sure how the BiDi amp affects this) to give appropriate rise and fall times to the keying envelope.  Allard's original 47k resistor was probably about right for the 0.1uF C107, the move to 10k makes the edges a bit fast, but we could go back to 47k if the BFO is driven from CLK0 and thus can be centered in the crystal filter passband.  Keying envelopes may not matter much for a one off rig at 5W, but for wide distribution and the inevitable use with QRO amps, I'd want clean CW key shaping.

Jerry, KE7ER



On Wed, May 31, 2017 at 08:13 pm, Raj vu2zap wrote:
See my solution:

https://groups.io/g/BITX20/topic/ptt_pop_hack/4460108?p=,,,20,0,0,0::Relevance,,PTT+pop+raj,20,2,0,4460108

 

 

Jerry,

We could put an audio isolation transformer between the mic/preamp stages and the mixer.
The mixer side will need to be ground isolated then Allards CW will work and probably the carrier leak on SSB will be minimum.

Some thing like these transformers:
http://www.ebay.com/itm/2pcs-600-Ohm-Audio-Telephone-Radio-Coupling-Isolation-1-1-Transformer-US-Seller-/321917511674

I have some isolation transformers, will experiment when I get back to town next week.

Cheers
Raj

At 01/06/2017, you wrote:
A very nice $0.01 solution.
Except it kills how Allard is doing CW.
He unbalances the carrier by putting 5v into a 10k resistor that goes to the top of C107.

Maybe fix this PTT pop somehow in microphone preamp so we can continue to use Allard's CW method?

Could transmit CW like Farhan is doing it on the uBitx:  Move the VFO to the transmit freq and unbalance the first ring mixer.  Those three mixer ports got stirred up considerably on the uBitx, though it is otherwise the same mixer as on the Bitx40.  Is there a some simple way to do this without a significant hardware hack to the existing Bitx40 boards?

......snip........

Jerry, KE7ER



On Wed, May 31, 2017 at 08:13 pm, Raj vu2zap wrote:
See my solution:

https://groups.io/g/BITX20/topic/ptt_pop_hack/4460108?p=,,,20,0,0,0::Relevance,,PTT+pop+raj,20,2,0,4460108

 

Jerry Gaffke
 

Possible.  Wants $1.39 + $3.15 for shipping, so the transformer is kind of painful.
Especially with a thousand or more rigs spread around the world now.
I'd probably advise that most just live with the pop and use Allard's scheme for CW
Those that aren't interested in CW can use your easily scrounged 0.1uF fix.
 


On Wed, May 31, 2017 at 09:39 pm, Raj vu2zap wrote:
We could put an audio isolation transformer between the mic/preamp stages and the mixer.

 

 

Same transformers are available from China, 10 for 3$ with free shipping or so.
Some sellers with 2/5 pcs even cheaper.

Raj


At 01/06/2017, you wrote:
Possible.  Wants $1.39 + $3.15 for shipping, so the transformer is kind of painful.
Especially with a thousand or more rigs spread around the world now.
I'd probably advise that most just live with the pop and use Allard's scheme for CW
Those that aren't interested in CW can use your easily scrounged 0.1uF fix.
 

On Wed, May 31, 2017 at 09:39 pm, Raj vu2zap wrote:
We could put an audio isolation transformer between the mic/preamp stages and the mixer.

Jerry Gaffke
 

I like Raj's 0.1uF fix, very simple and elegant.  Oh, and very cheap.
Could add a jumper across it when operating CW.
Also, remove a new jumper that supplies power to the mic amp, not needed for CW.
That fixes the transmit pop on both CW and SSB without too much trouble.

If willing to do a significant hack to avoid messing with jumpers,
rearrange the first mixer for CW injection as on the uBitx and add Raj's 0.1uF modulator cap.




On Wed, May 31, 2017 at 09:58 pm, Jerry Gaffke wrote:
I'd probably advise that most just live with the pop and use Allard's scheme for CW
Those that aren't interested in CW can use your easily scrounged 0.1uF fix.

 

Timo
 

Hi

I just become a proud owner of BITX40. After 2 evenings tinkering, I got "first smoke" out (as literally, not virtually , whew). I found this RF spur issue, with among some others, but I think this must be solved. It is unacceptable.

After two seconds of thinking, I'm afraid that the audio trafo between the mike amp and mod won't do any good. When the power is applied to the mic amp, this will cause sudden current flow through the primary, and what does the secondary ?
The spike will be there present there too.

Instead of sticking in this perculiar way of generating CW, one could think alternative methods. Raduino's extra oscillator maybe and for those w/o , injecting sine wawe to mic will work also.

Timo,OH7LMQ

Bob
 

Just to report back, changing C124, originally a 47 uF cap, to 4.7 uF, reduces the duration of the TX burst on PTT from 50 mS, to less than 25 mS.  In fact, it is now so short, that it has essentially zero impact on any CW speed -- in other words, the burst is covered up by the first dot or dash that is sent.

Real easy fix to swap that out before the board is installed in the case, as it is a thru-hole component.  My thanks to Willy, W1LY, for making this mod on his Bitx40.

73, Bob, WB4SON

terry hughes
 

hi

my way of generating cw is a completely separate cw transceiver.

terry gm4dso