Topics

D4D mods

Steven Weber
 

The SA602 only requires a 50 mV p-p signal to fully unbalance it. Larger signals create all kinds of unwanted mixer products. Since the VOX requires a 1200 mV p-p signal to trigger it, there needs to be a lot of attenuation between the audio input and the mixer, of about a 25:1 ratio. A 22K/1K divider would be close enough.  

The apparent reduction of power is the result of a cleaner, linear output and the fact that a watt meter reads the average power of the DSB signal.  If nothing else PLEASE MAKE THIS MOD!

The best way to go would be to add a trimmer resistor to the mixer input. Adjust the sound card output level until the board trips into transmit, then turn up the audio level to the mixer with the trimmer until the power output no longer increases. Now back down the level until the power drops off by a little more then 1/2. You are now operating the mixer in it's linear mode and not "flat topping". If you can monitor the output signal with a Scope, this is obvious. 

You can make the VOX more sensitive by adding a resistor from the base of Q5 to ground. A 150K resistor would make the VOX much more sensitive. The idea is to pre-bias the base just below the point at which it starts to turn on, with a little margin for supply voltage and temperature variations. If you do this, then the attenuation to the mixer can be reduced. 

Unequal loads on the outputs of the SA602 will unbalance the mixer and allow carrier to leak through. This happens when Q1 is turned off during transmit. All of a sudden all the current needed for Q1's emitter load is being supplied by the mixer output. To fix this, add a diode between the +5 supply and the collector of Q1 (cathode to Q1). It would probably also help to make R2 and R3 the same value to further balance the mixer output loads, but just adding the diode should be enough. This will significantly reduce or eliminate the carrier.  

You can test if your transmitting a lot of carrier by manually switching the board into transmit by grounding the relay at Q7with no audio input. 

73, Steve KD1JV 

Trent Trent
 

Have u done the mods yet Steve?
Btw have u got a AM BCB filter on yours and is it my mixer or is urs also seem a bit deaf on RX?

On Tue, 30 Jul 2019 at 08:22, Steven Weber <steve.kd1jv@...> wrote:
The SA602 only requires a 50 mV p-p signal to fully unbalance it. Larger signals create all kinds of unwanted mixer products. Since the VOX requires a 1200 mV p-p signal to trigger it, there needs to be a lot of attenuation between the audio input and the mixer, of about a 25:1 ratio. A 22K/1K divider would be close enough.  

The apparent reduction of power is the result of a cleaner, linear output and the fact that a watt meter reads the average power of the DSB signal.  If nothing else PLEASE MAKE THIS MOD!

The best way to go would be to add a trimmer resistor to the mixer input. Adjust the sound card output level until the board trips into transmit, then turn up the audio level to the mixer with the trimmer until the power output no longer increases. Now back down the level until the power drops off by a little more then 1/2. You are now operating the mixer in it's linear mode and not "flat topping". If you can monitor the output signal with a Scope, this is obvious. 

You can make the VOX more sensitive by adding a resistor from the base of Q5 to ground. A 150K resistor would make the VOX much more sensitive. The idea is to pre-bias the base just below the point at which it starts to turn on, with a little margin for supply voltage and temperature variations. If you do this, then the attenuation to the mixer can be reduced. 

Unequal loads on the outputs of the SA602 will unbalance the mixer and allow carrier to leak through. This happens when Q1 is turned off during transmit. All of a sudden all the current needed for Q1's emitter load is being supplied by the mixer output. To fix this, add a diode between the +5 supply and the collector of Q1 (cathode to Q1). It would probably also help to make R2 and R3 the same value to further balance the mixer output loads, but just adding the diode should be enough. This will significantly reduce or eliminate the carrier.  

You can test if your transmitting a lot of carrier by manually switching the board into transmit by grounding the relay at Q7with no audio input. 

73, Steve KD1JV 

Adam Rong
 

Hi Steve,

Thanks for your sharing. I will test them. One thing to confirm: If we add a diode between +5V and the collector of Q1, when 12V_R is present, the diode will be reverse biased? If not, I suppose it will mute the audio output.

Thanks,
Adam

Steven Weber
 

Yes, Q1’s collector should stay above 5V when 12_V is applied.

Steve

 

 

From: Adam Rong
Sent: Tuesday, July 30, 2019 10:33 AM
To: crkits@groups.io
Subject: Re: [crkits] D4D mods

 

Hi Steve,

Thanks for your sharing. I will test them. One thing to confirm: If we add a diode between +5V and the collector of Q1, when 12V_R is present, the diode will be reverse biased? If not, I suppose it will mute the audio output.

Thanks,
Adam

 

Adam Rong
 

Sounds good. I will modify the emitter resistor from 1.5k to 680 ohm and give it a try later this week.

Just FYI: After applying the diode modification to reduce NE602 power supply voltage from 5V to 4.3V suggested by BH4DCL, the carrier leakage is greatly reduced. Maybe your modification will help further.

Thanks,
Adam


On Wed, Jul 31, 2019 at 4:36 AM Steven Weber <steve.kd1jv@...> wrote:

Yes, Q1’s collector should stay above 5V when 12_V is applied.

Steve

 

 

From: Adam Rong
Sent: Tuesday, July 30, 2019 10:33 AM
To: crkits@groups.io
Subject: Re: [crkits] D4D mods

 

Hi Steve,

Thanks for your sharing. I will test them. One thing to confirm: If we add a diode between +5V and the collector of Q1, when 12V_R is present, the diode will be reverse biased? If not, I suppose it will mute the audio output.

Thanks,
Adam

 



--
Thanks,
Adam Rong

Adam Rong
 

Thank you, Steve. Your diode mod is effective. In my test, the carrier leakage improved by about 5 dB, as a side effect, both USB and LSB signals increased about 1 dB which is a surprise to me. I also applied the BH4DCL diode mod, and it improved the carrier leakage further for another 3 dB, and there was no noticeable difference on the USB and LSB signal level.

I did try to change the emitter resistor from 1.5k to half and it won't help too much, so I will not change it.

So, my recommendation is that we apply both diode mods, hihi.

Thanks,
Adam

On Wed, Jul 31, 2019 at 4:36 AM Steven Weber <steve.kd1jv@...> wrote:

Yes, Q1’s collector should stay above 5V when 12_V is applied.

Steve

 

 

From: Adam Rong
Sent: Tuesday, July 30, 2019 10:33 AM
To: crkits@groups.io
Subject: Re: [crkits] D4D mods

 

Hi Steve,

Thanks for your sharing. I will test them. One thing to confirm: If we add a diode between +5V and the collector of Q1, when 12V_R is present, the diode will be reverse biased? If not, I suppose it will mute the audio output.

Thanks,
Adam

 



--
Thanks,
Adam Rong

Adam Rong
 

And the photo follows. The left side is the KD1JV diode mod, and the right side is the BH4DCL diode mod.
微信图片_20190804230531.jpg


On Sun, Aug 4, 2019 at 10:59 PM Adam Rong via Groups.Io <rongxh=gmail.com@groups.io> wrote:
Thank you, Steve. Your diode mod is effective. In my test, the carrier leakage improved by about 5 dB, as a side effect, both USB and LSB signals increased about 1 dB which is a surprise to me. I also applied the BH4DCL diode mod, and it improved the carrier leakage further for another 3 dB, and there was no noticeable difference on the USB and LSB signal level.

I did try to change the emitter resistor from 1.5k to half and it won't help too much, so I will not change it.

So, my recommendation is that we apply both diode mods, hihi.

Thanks,
Adam

On Wed, Jul 31, 2019 at 4:36 AM Steven Weber <steve.kd1jv@...> wrote:

Yes, Q1’s collector should stay above 5V when 12_V is applied.

Steve

 

 

From: Adam Rong
Sent: Tuesday, July 30, 2019 10:33 AM
To: crkits@groups.io
Subject: Re: [crkits] D4D mods

 

Hi Steve,

Thanks for your sharing. I will test them. One thing to confirm: If we add a diode between +5V and the collector of Q1, when 12V_R is present, the diode will be reverse biased? If not, I suppose it will mute the audio output.

Thanks,
Adam

 



--
Thanks,
Adam Rong



--
Thanks,
Adam Rong

Steven Weber
 

I would remove the diode in series with the 5V to the SA602. The minimum operating voltage for the SA602 is 4.5V and performance degrades quickly below 4.5V.

 

 

Thank you, Steve. Your diode mod is effective. In my test, the carrier leakage improved by about 5 dB, as a side effect, both USB and LSB signals increased about 1 dB which is a surprise to me. I also applied the BH4DCL diode mod, and it improved the carrier leakage further for another 3 dB, and there was no noticeable difference on the USB and LSB signal level.

I did try to change the emitter resistor from 1.5k to half and it won't help too much, so I will not change it.

So, my recommendation is that we apply both diode mods, hihi.

 

Thanks,

Adam

 

On Wed, Jul 31, 2019 at 4:36 AM Steven Weber <steve.kd1jv@...> wrote:

Yes, Q1’s collector should stay above 5V when 12_V is applied.

Steve

 

 

From: Adam Rong
Sent: Tuesday, July 30, 2019 10:33 AM
To: crkits@groups.io
Subject: Re: [crkits] D4D mod

 

Hi Steve,

Thanks for your sharing. I will test them. One thing to confirm: If we add a diode between +5V and the collector of Q1, when 12V_R is present, the diode will be reverse biased? If not, I suppose it will mute the audio output.

Thanks,
Adam

 



--

Thanks,

Adam Rong

 

Adam Rong
 

Steve,

Thanks for reminding. We studied the spec while we apply the 4.3V power supply. The main impact parameters are:
Conversion gain, Noise figure, and IP3i

I believe the major impact is IP3i which will cause worse BCI.

With your modificaiton, I think it is okay not to add BH4DCL diode mod to avoid the impact to the receiver performance. The carrier leakage is now about 10 dB under USB and LSB signals even I applied 10 dB attenuation to the audio input.

Thanks,
Adam

Bryan Brauer
 

I just received my d4d kit this week, and wish to incorporate all recommended mods.  Will there be a file that lists all mods, or an update to the instructions that includes all?

Bryan B. Brauer, PE
Consulting Engineer


On Sun, Aug 4, 2019 at 8:21 AM Adam Rong <rongxh@...> wrote:
Steve,

Thanks for reminding. We studied the spec while we apply the 4.3V power supply. The main impact parameters are:
Conversion gain, Noise figure, and IP3i

I believe the major impact is IP3i which will cause worse BCI.

With your modificaiton, I think it is okay not to add BH4DCL diode mod to avoid the impact to the receiver performance. The carrier leakage is now about 10 dB under USB and LSB signals even I applied 10 dB attenuation to the audio input.

Thanks,
Adam

Adam Rong
 

Modification is not necessary if you want to get it work first. After you have time, you can consider these two modifications:
1. Replace C5 by a 105 monolithic cap (tolerance < 20%) to improve IMD at the cost of half RF power at the same audio input.
2. Add a diode per KD1JV to improve carrier leakage.

If your sound card output is too low to drive VOX, consider KD1JV improvement to add a 150k resistor. This mod will not become standard as it is related with transistor parameter and power supply voltage. You make your call to modify it or buy a USB sound card along.

Thanks,
Adam

On Mon, Aug 5, 2019 at 12:29 AM Bryan Brauer <bryan@...> wrote:
I just received my d4d kit this week, and wish to incorporate all recommended mods.  Will there be a file that lists all mods, or an update to the instructions that includes all?

Bryan B. Brauer, PE
Consulting Engineer


On Sun, Aug 4, 2019 at 8:21 AM Adam Rong <rongxh@...> wrote:
Steve,

Thanks for reminding. We studied the spec while we apply the 4.3V power supply. The main impact parameters are:
Conversion gain, Noise figure, and IP3i

I believe the major impact is IP3i which will cause worse BCI.

With your modificaiton, I think it is okay not to add BH4DCL diode mod to avoid the impact to the receiver performance. The carrier leakage is now about 10 dB under USB and LSB signals even I applied 10 dB attenuation to the audio input.

Thanks,
Adam



--
Thanks,
Adam Rong

Simon C <symen79@...>
 

What diode is recommended for Steve's Mod?

Adam Rong
 

In my test, I used 1N4148 but any diode will do, like 1N914, 1N4001-4007, 1N5817 and etc.

On Mon, Aug 5, 2019 at 8:12 AM Simon C via Groups.Io <symen79=yahoo.com@groups.io> wrote:
What diode is recommended for Steve's Mod?



--
Thanks,
Adam Rong

Jim WB2LHP in MI
 

I applied the following changes...Changed C5 to 105 and C6 to 103 (capacitive vs resistive divider) and applied audio drive just above the threshold for reliable VOX PTT...My DSB output went down to 400mW so I added another cap in parallel to C11 and was able to increase the RF drive to the PA...Adding 56pf in parallel rose the power output to 900mW...That made the difference between almost no contacts and routine contacts...Just FYI...Jim WB2LHP

Jim WB2LHP in MI
 

I also applied the KD1JV diode mod and VOX resistor mod...

Bryan Brauer
 

I made the changes suggested by WB2LDH.  The most power out, with both WSJT-W and soundcard at max, is 600mW as measured by my Elecraft W1 power meter.

Can you suggest a value at for capacitor parallel to C11 other than 56pf that might boost this a bit.

Bryan B. Brauer, PE
Consulting Engineer

Jim WB2LHP in MI
 

Sorry...Been away for a bit...Try anything you have between 50 and 100pf...It's not that critical but 82pf would be optimal...Your target output should be 1 watt or so...I'm sure it will do more but you don't want to become non-linear or create unwanted signals...Jim WB2LHP

Adam Rong
 

Well, you'd better observe the spectrum. In my case, if the capacitor is too big, there is a peak at 1/2 signal frequency.

73 de BD6CR
Adam

Lawrence Macionski
 

Does anyone know if or when the mods you speak of will be incorporated in the kit. We are planning on a proposed group kit building for our club, and some members have concern regarding tested and true versus experimental designs including actual cost per unit. Some members do not control their purse strings other members are not degree-d engineers. Obviously, there will be a number of entries into the "group build" debate and the selection of project will ultimately be the project  with the best salesman. 

Advise,

Larry W8LM

Larry Lovell
 

From my experience, there are many mods that people are making.  Some better than others.  If you purchase the kit and assemble it with the Rev D assembly manual, things should work very well.  All parts are included in that kit.


On Wed, Jan 29, 2020 at 3:47 PM Lawrence Macionski via Groups.Io <am_fm_radio=yahoo.com@groups.io> wrote:
Does anyone know if or when the mods you speak of will be incorporated in the kit. We are planning on a proposed group kit building for our club, and some members have concern regarding tested and true versus experimental designs including actual cost per unit. Some members do not control their purse strings other members are not degree-d engineers. Obviously, there will be a number of entries into the "group build" debate and the selection of project will ultimately be the project  with the best salesman. 

Advise,

Larry W8LM



--
Larry Lovell
73's N7RGW
http://QRVTronics.com
Cell: 214-697-1729