Topics

Always unpleasant noise pop T/R


Gerard
 

Hello,
Despite having modified my mounting and added some elements (Diode + resistance +capacitor) on the mosfet from the foot 25 of the arduino (T/R), I still have unpleasant noise during reception/ emission swap tests.


I looked at the oscilloscope to see how the RX and TX power supplies were switching.
I realize that, in fact, the rise in tension is almost instantaneous when we flip, but the descent in tension is not immediate, especially for part TX.
Which means that at each flip, we can have the RX powered, and again the TX for for several hundred milliseconds (Probably discharge capacitors)
So it translates as a shot of a gun or flush in the top (Each chooses. LOL)
I do not know if it is due to fat that the PA part is not cabled ?? I don't made emmision at this time.
So my question is: How to make the voltage on the TX alim drop very quickly.
Cabling resistance in parrallele on each chemical or other? A system with diodes?

Then you can say, that it is useless when we do not make a broadcast, (So no flip) but it is for the principle of finding a solution. I don’t know if the problem is 100% solved for you.

Thank you for your answers.
nb: The control of the tx/rx relay is perfect
cdt


Evan Hand
 

Gerard,
If you go back to message #82660:
https://groups.io/g/BITX20/message/82660

In the diagram, you will note that the gate of Q74 is turned on quickly with the 5 volts from the T/R signal from the Raduino/Nano through the diode D14, and then discharged through R78 (100k ohms) when switching back so that the turn-on is quicker than the mechanical relay, and releases slower than the transfer from transmitting to receiving.  This is the pop fix that works on the 3 uBITX that I have.

In your circuit, the Q74 equivalent is also part of the CAG (AGC) control.  That can be delaying the turn on, as well as discharging the capacitor too quickly.  I would suggest adding another transistor like Q74 (2n7000) to fix the pop, and use what you have for the gain control.

Just suggestions
73
Evan
AC9TU


Gerard
 

Evan,
Yes, I understand the principle. I have look that with oscilloscope.
I can test with a 2nd Mosfet
But in fact the T/R signal coming from the arduino is more "short" as the "time of descent" of the TX power supply during the rollover.
So in a little time your have the 2 supplies.
The simplest test to prove that this comes from is simply disconnecting the TX from the relay.
we’ll see if that unpleasant noise goes away then.
Other solution increase the T/R time by soft. This may be possible.
Otherwise, you did not tell me if you also had this noise when turning the encoder. It is also a 2nd problem a bit annoying.

cdt


Bob Lunsford
 

Gerard, try putting a 1K resistor the line to ground that feeds the transmitter and see if this drops the Transmit line enough to cause it to drop faster than the receive line. It will be only a sign that this is what is happening although 1K will probably not fix the problem. Conversely, you can put a 200 Ohm resistor in series with the line that feeds the receiver with a 50 MFD cap from the receive line to ground. Theoretically, this will cause a slight delay in the receive circuit. These should not be permanent fixes, though, just indicators that you are on the right path to making the problem go away. Just an idea.

Bob — KK5R


On Wednesday, October 21, 2020, 12:26:47 PM EDT, Gerard <kabupos@...> wrote:


Hello,
Despite having modified my mounting and added some elements (Diode + resistance +capacitor) on the mosfet from the foot 25 of the arduino (T/R), I still have unpleasant noise during reception/ emission swap tests.


I looked at the oscilloscope to see how the RX and TX power supplies were switching.
I realize that, in fact, the rise in tension is almost instantaneous when we flip, but the descent in tension is not immediate, especially for part TX.
Which means that at each flip, we can have the RX powered, and again the TX for for several hundred milliseconds (Probably discharge capacitors)
So it translates as a shot of a gun or flush in the top (Each chooses. LOL)
I do not know if it is due to fat that the PA part is not cabled ?? I don't made emmision at this time.
So my question is: How to make the voltage on the TX alim drop very quickly.
Cabling resistance in parrallele on each chemical or other? A system with diodes?

Then you can say, that it is useless when we do not make a broadcast, (So no flip) but it is for the principle of finding a solution. I don’t know if the problem is 100% solved for you.

Thank you for your answers.
nb: The control of the tx/rx relay is perfect
cdt


Bob Lunsford
 

Good suggestions, Evan...

On Wednesday, October 21, 2020, 12:43:14 PM EDT, Evan Hand <elhandjr@...> wrote:


Gerard,
If you go back to message #82660:
https://groups.io/g/BITX20/message/82660

In the diagram, you will note that the gate of Q74 is turned on quickly with the 5 volts from the T/R signal from the Raduino/Nano through the diode D14, and then discharged through R78 (100k ohms) when switching back so that the turn-on is quicker than the mechanical relay, and releases slower than the transfer from transmitting to receiving.  This is the pop fix that works on the 3 uBITX that I have.

In your circuit, the Q74 equivalent is also part of the CAG (AGC) control.  That can be delaying the turn on, as well as discharging the capacitor too quickly.  I would suggest adding another transistor like Q74 (2n7000) to fix the pop, and use what you have for the gain control.

Just suggestions
73
Evan
AC9TU


Gerard
 

Evan,

Thanks for the advice, I’ll look at it tomorrow
i add:
On the M0THY diagram, the capacitors of the final power stage are composed of 2 capacitors of 470 µ on the TX alim (Like  V.6) On its assembly, the power supply used for the power is the TX power supply. Since the power transistors are not mounted, the discharge time may be longer (TBC)
A test would be to delete them to see.
Moreover on the diagram  of the ubitx V6, the stage is powered by PA-PWR. Surely the possibility of adding an external power supply stronger than the TX. There may also be the ability to use the same TX alim with a jumper?

But, to talk about something else, you still haven’t told me if your rotary encoder  is making noise..

to morrow
Bye


Evan Hand
 

Gerard,
I do not have the noise when tuning with the encoder.  As stated before, there are some who believe that it is from the Si5351 frequency changes.  I do not know those details.  My guess would be that the encoder circuit may be stepping forward and back then forward again.  That is potentially caused by not enough time delay in the contact debounce of the encoder code.  Again, I do not have this issue, nor am I sure where it comes from.  RF from the encoder does not make sense to me. Bouncing contacts causing rapid frequency changes back and forth do.  You could try capacitors on the encoder contacts to see if that reduces the contact bounce.  You might also see it on a dual-trace oscilloscope on the encoder contacts. 

Let me explain in more detail the pop fix of Q74.
1 - On the transfer from RX to TX the Nano sets the TX/RX signal high (+5volts) from the Arduino/Nano
2 - Q15 is turned on that starts the K1 relay to pull in.  At the same time, the gate of Q74 is turned on shorting the audio input to the audio amplifier.  This is through the diode, bypassing the resistor.
3 - K1 completes the contact closure and energizes the transmitter sections and de-energizes the receiver sections.
4 - The transceiver is now in TX mode with audio muted.
5 - When the TX/RX signal is turned off by the Nano, Q15 is turned off and K1 starts to open.  The gate of Q74 is held high by the capacitor as the diode will not allow the charge to drain through it to the now low TX/RX signal from the Nano
6 - K1 completes turning off and removes power from the TX stages and applies it to the RX stages.  Meanwhile, the gate of Q74 is still held High until C79 discharges through the 100kohm resistor.  While still high, the audio is being muted.
7 - The gate voltage drops to the point where Q74 is no longer on, the audio unmutes, and the transceiver goes back to RX.

Again, just suggestions as I do not have the issue with my three uBITX.
73
Evan
AC9TU


Gerard
 

Evan,
Yes, I’ll look at all of that.
Then I had another idea.
Use the TX voltage to block the amplifier.
Example, make from the TX power a bridge to have 5vTx and control a 2nd Mosfet. (This would avoid, resistances on capacitors of the PA, etc.)
To be seen
cdt


Gerard
 

Evan,
I just did the test
If I disconnect one of the two capacitors (Ex ,c98 or c99 on the v6 diagram ) the noise is divided by 2.
If I disconnect the 2, no more noise by doing T/R or vice versa. Complete silence...
There is a conflict on the 2 alims (TX, RX) at this time when doing RX/TX (due to discharge of these capacitors)
Now the question is, if you don’t have this noise, is it due to the fact that the power transistors (which are mounted on your PCB) quickly absorb the discharge of these 2 capacitors?
Yes/no?
What to do then? Drive with a dividing bridge (to have 5v) the +TX with mosfet on the amplifier. (Same principle as the T/R)

Another question, when I look at the diagram, I have a hard time understanding how the final stage works.
It looks like the final transistor is driven by another as in parrallele, (Like Darlinton?) not see well on the diagram.
If you can explain.
cdt


Gerard
 

Evan,
I forgot to tell you that the resistance-to-discharge test, reduces, but doesn’t eliminate noise. If you go down too low in value.. It might smoke. Lol.
cdt


Evan Hand
 

Gerard,
I do not understand.  The pop fix that is used in the HFSignals version only impacts the receive function.  The transmitter is cut off when the bias voltage source is removed by the K1 relay.  That turns off the two IRF510s.  The +12volts to the drains is always on.  The intent again is to mute the audio until the tx is turned off by removing the bias voltage on the gates.  Is that not how your board is wired?  If you are turning on and off the power to the Drains of the IRF510s then you will get the popping noise as those capacitors charge and discharge.

I went back and looked again at the schematic for the board that you built.  It does have the drain power being turned on and off to the finals (In the TX supply).  That is NOT the way that the HFSignals version works.  As stated above, the power to the finals is always there.  It is through the bias voltage that they are turned on or off.  That does not have the large power surges that applying and removing power to the IRF510 drains will produce when charging and discharging the supply capacitors.

I see as your options:
1 - Modify the board to be similar to the HFSignals board and keep power applied to the IRF510s.  This would have an added advantage in that you could provide more than 12volts to the finals to get more power without damaging the rest of the board.  The PA power is separate from the board power.
2 - Try increasing the RC time constant of the R78 / C79 pair to extend the mute time beyond the discharge time.  I do not think that this will work as there will still be a large spike on the power rails when disconnected.  Also, there will be a race to see if the mute happens before the power application to the finals.

Remember that you WILL need to remove power from the bias regulator to have the IRF510s turn off if power is always on the drains.

73
Evan
AC9TU


Gerard
 

Evan,
Yes, actually on my diagram,the power supply is the TX for the transistors of the last power stage. It may be a bad design of this model. I looked at the v3 diagram, V4, the alim is still the PA-PWR.
So I would modify the circuit accordingly or disconnect the 2 capacitors because i made only reception at this time.

otherwise, there is some software to pilot ubitx
From a historical perspective, TSW had a release that supported remote control and mouse
control. 
 
User Manual:
 
Custom control program:
 
cdt


Robert Freeman
 

Hi Gerard,

I was seeing a similar issue and found this discussion.  Installing the diode made a big difference on my V4 rig going from RX to TX.  Maybe this is  related to your problems? 
73's, Robert NQ0t


On Thu, Oct 22, 2020 at 1:12 PM Gerard <kabupos@...> wrote:
Evan,
Yes, actually on my diagram,the power supply is the TX for the transistors of the last power stage. It may be a bad design of this model. I looked at the v3 diagram, V4, the alim is still the PA-PWR.
So I would modify the circuit accordingly or disconnect the 2 capacitors because i made only reception at this time.

otherwise, there is some software to pilot ubitx
From a historical perspective, TSW had a release that supported remote control and mouse
control. 
 
User Manual:
 
Custom control program:
 
cdt