Topics

Are these output watt rating crazy?

Mark Hatch
 

Just got my V4 uBitx operational. Currently, as designed with no mods.

Got a new Diamond Watt/SWR meter (SX200) and fired it up using CW into a dummy load. Got the following...

160M   20w
80        15
40          7
30          8
20        15
17          8
15          5
13          7
10          5

No difference (as expected on pep vs average)

The higher frequencies seem to be similar to what others have reported, perhaps a little high.  But 160, 80 and 20 just seem far to high and makes me wonder whether the "new" meter has a calibration issue.

73
Mark
AJ6CU

Adrian Chadd
 

.... how'd you get it to tx on 160? :)


-a


On Thu, 25 Apr 2019 at 07:42, Mark Hatch <mark2382@...> wrote:
Just got my V4 uBitx operational. Currently, as designed with no mods.

Got a new Diamond Watt/SWR meter (SX200) and fired it up using CW into a dummy load. Got the following...

160M   20w
80        15
40          7
30          8
20        15
17          8
15          5
13          7
10          5

No difference (as expected on pep vs average)

The higher frequencies seem to be similar to what others have reported, perhaps a little high.  But 160, 80 and 20 just seem far to high and makes me wonder whether the "new" meter has a calibration issue.

73
Mark
AJ6CU

Mark Hatch
 

LOL...  Did not know it wasn't supposed to...  It did into a dummy load. Have the latest KD8CEC installed if that matter. 

Mark

MadRadioModder
 

It will transmit there… just no output filter…

 

 

From: BITX20@groups.io [mailto:BITX20@groups.io] On Behalf Of Adrian Chadd
Sent: Thursday, April 25, 2019 11:12 AM
To: BITX20@groups.io
Subject: Re: [BITX20] Are these output watt rating crazy?

 

.... how'd you get it to tx on 160? :)

 

 

-a

 

 

On Thu, 25 Apr 2019 at 07:42, Mark Hatch <mark2382@...> wrote:

Just got my V4 uBitx operational. Currently, as designed with no mods.

Got a new Diamond Watt/SWR meter (SX200) and fired it up using CW into a dummy load. Got the following...

160M   20w
80        15
40          7
30          8
20        15
17          8
15          5
13          7
10          5

No difference (as expected on pep vs average)

The higher frequencies seem to be similar to what others have reported, perhaps a little high.  But 160, 80 and 20 just seem far to high and makes me wonder whether the "new" meter has a calibration issue.

73
Mark
AJ6CU


Virus-free. www.avg.com

--

…_. _._

 

Hi,

The output will vary depending on the supply voltage, and on the setting of the driver potentiometer.  I doubt whether the latter is set precisely at the factory, so the boards will probably vary initially on their output.  On CW, some bands have more harmonics than others, so this will also show as variation in output wattage (since wattmeters are not frequency selective).  So considering these factors, I don't think your numbers are crazy.

73,


Mark.

 

Hmm,

On second thought, adding up all the harmonics probably won't make much of a difference.  More likely the variations in the individual bands is due to uneven reponse in the stages and mixer prior wrt frequency, and maybe too if some of the LPF's are too close to the band edge.

73,


Mark.

Mark Hatch
 

Thanks!

Ignoring 160,  I was just surprised to see 15w on 80 and 20. I know I didn't buy a Bird meter, but the spec on the 20w scale is 7.5%....

Mark

Jerry Gaffke
 

If transmitting on 160 meters, the uBitx would have the 80 meter transmit LPF filter selected.
Must have an external 160 meter low pass filter between rig or antenna,
or you will be sending some nasty 2'd and 3'rd harmonics (almost a square wave!).

Firmware should probably display a warning when 160m is selected.

Jerry, KE7ER


On Thu, Apr 25, 2019 at 09:23 AM, Mark Hatch wrote:
LOL...  Did not know it wasn't supposed to...  It did into a dummy load. Have the latest KD8CEC installed if that matter. 

Mark

Curt
 

Nice power readings - but i would not take them to the bank.  If you have another QRP transmitter, that might let you know if the wattmeter is a tad high.  Honestly a difference of 20% in transmit power isn't noticeable most of the time.  yes don't overdrive the PA to aid reducing the spurious - and nearly every v4 needs 2 spur fixes - one for PA filter network and one of the final mixer (SSB only). 

On 160m I suspect the 80m LPF should handle the third harmonic.  If you might be using a short vertical like me - you may find your 3.6 MHz VSWR into it to be around infinity -- therefore I don't sweat putting 10 watts into 160m.  (Other situations could dreadfully vary). 

Enjoy your new rig, and prepare to do some spurious mods.  Back down the transmit drive a little - and you should be 'acceptable' on 80 and 20 meters. 

73 Curt

Jerry Gaffke
 

The 80m LPF is also used for 60m, so the cut-toff freq is high enough to pass the third harmonic of 160m.
This is not a sine wave:
    https://groups.io/g/BITX20/message/48122


On Thu, Apr 25, 2019 at 05:35 PM, Curt wrote:
On 160m I suspect the 80m LPF should handle the third harmonic.

Vic WA4THR
 

Very similar results on my V4, a little less than yours on 20 (12w), a little more on 40 (9w), but the pattern is the same and a result of the filters selected on each band relative to the frequency. The rig will transmit anywhere, but should only be used between 80-10m as that is what the filter design is for.

=Vic=

Mark Hatch
 

Thanks to all the replies! After performing the mod to even up the bands, I suspect that I will need to turn the drive down. And I will keep the watt meter, it might be a tad high, but it is not way off either. I am wondering whether my dully load has a flat impedance response curve too. That could vary the output.

73
AJ6CU
Mark

Adrian Chadd
 

Which mod did you use to even up the bands?


-a


On Fri, 26 Apr 2019 at 06:30, Mark Hatch <mark2382@...> wrote:
Thanks to all the replies! After performing the mod to even up the bands, I suspect that I will need to turn the drive down. And I will keep the watt meter, it might be a tad high, but it is not way off either. I am wondering whether my dully load has a flat impedance response curve too. That could vary the output.

73
AJ6CU
Mark

Mark Hatch
 

None yet. Was planning on using the ones recommended over ubitx.net unless I saw another idea appear I. The next month or so. 

73
AJ6CU

 

Jerry Gaffke
Apr 25  
The 80m LPF is also used for 60m, so the cut-toff freq is high enough to pass the third harmonic of 160m.This is not a sine wave:    https://groups.io/g/BITX20/message/48122
Hi,

I just entered the lowest ubitx LFP into a filter program and see that it is actually a 40m LPF.  It has a pretty good cutoff, so it cuts off what it need to for 60m and 80m, but would do almost nothing for 160m as it won't cutoff anything until the 4th harmonic and will only do about a 13dB cut there so not really effective until the 5th harmonic.  So basically useless on 160m.

Besides that, the input impedance to the filter on 160m comes out to about 10 ohms so not what the PA is designed to see.  On 60m it shows as 20 ohms a little better, but not great.

73,


Mark.

 

Jerry Gaffke
 

Mark,

I assume you posted to this old thread because
you wanted to tell us about your simulation results.

Yes, the uBitx should not be used on 160m without a suitable LPF
inline with the antenna.  Without that LPF the third harmonic will be
nearly as strong as the fundamental.

Not sure exactly how you went about simulating the filter
or exactly what corner frequency you saw.
Regardless of what your filter program thinks, the 80m LPF
is being used in a more or less 50 ohm environment.
Changing the impedances will change the frequency response.

Attenuating some harmonic by 13dB may or may not be sufficient.
If that particular harmonic coming out of the final amp is already 30dB down
from the fundamental, then the extra 13dB of attenuation from the filter
will be sufficient to meet FCC regs.  Due to the push-pull configuration
of the IRF510 final, it is mostly the 3rd harmonic that gives us grief.

Would be interesting to measure (not simulate) the response of the 80m LPF
using something like the Antuino, driving the filter from a 50 ohm source
and terminating it into 50 ohms.

I believe the 3dB corner frequency of the 80m LPF is around 5mhz,
so if you tried to use it on 40m you would not be putting out much of a signal.
 
Here's the post you quoted    https://groups.io/g/BITX20/message/68292
It happens to be in the same thread as our current conversation,
though this thread hasn't been active for over 6 months.,

Jerry, KE7ER


On Sun, Nov 10, 2019 at 05:30 PM, Mark - N7EKU wrote:
Jerry Gaffke
Apr 25  
The 80m LPF is also used for 60m, so the cut-toff freq is high enough to pass the third harmonic of 160m.This is not a sine wave:    https://groups.io/g/BITX20/message/48122
Hi,

I just entered the lowest ubitx LFP into a filter program and see that it is actually a 40m LPF.  It has a pretty good cutoff, so it cuts off what it need to for 60m and 80m, but would do almost nothing for 160m as it won't cutoff anything until the 4th harmonic and will only do about a 13dB cut there so not really effective until the 5th harmonic.  So basically useless on 160m.

Besides that, the input impedance to the filter on 160m comes out to about 10 ohms so not what the PA is designed to see.  On 60m it shows as 20 ohms a little better, but not great.

73,


Mark.

 

Hi,

I used the AADE filter simulation/design program.  I posted because I had seem some comments on the power outputs and was curious to simulate the low filter to see if that might impact some of the power levels.  I simulated the lowest filter I saw on the ubitx v3 schematic.  Did I miss a lower one?  This one seems to be the lowest and used for both 40m and 80m with a cutoff very close to 7MHz (actually showed -2dB at 7.3MHz).

Yes the design environment is 50 ohms.  That does not mean the filter will show that for all frequencies.  At the design frequency, they should show that (as this one does for 40m).  At other frequencies they can present higher or lower impedances which will affect power output.  This one does also show around 50 ohms at 80m.  However at 60m it shows to be about 20 ohms and at 160m about 10 ohms.  Even using another filter on the output, with this filter inline it may cause trouble.  So it would be best to have a designed 160m filter linked directly to the output of the PA if one wants to use that band.  My thoughts anyway.

Of course it would be good to see an actual measurement of the filter with a VNA etc.  My equipment is packed away from a move at the moment so I have to rely on software at the moment.

I just thought people may be interested in what I had found.   I can work on my own no problem.




Cheero,


Mark.

Jerry Gaffke
 

Mark,

Interesting.   

You might be right that the filter with the 900nH inductors has an incorrectly implemented corner frequency.
Worth exploring further.
Your simulation shows a corner at around 7mhz, not the expected 5mhz.
That would let a 3.5mhz second harmonic leak by.
The second harmonic will be quite weak due to the push-pull topology of the IRF510 finals,
this 7mhz corner may not matter.

However, the stock uBitx definitely does not have a low pass filter appropriate for 160m !!!!!
We've had a number of folks here who assumed it did, since the firmware allowed them to spin down and transmit on 1.8mhz.
The third harmonic of 1.8mhz at 5.4mhz will go out on the air almost as strong as the fundamental.

The filter with the 900nH inductors that you simulated is definitely the lowest LPF on the uBitx.
When I said it had a 5mhz corner, I was going by Farhan's circuit description notes.
I had assumed this had been gone over thoroughly, since lots of folks in the forum
were trying to figure out how to reduce the problematic v3 harmonics from leaking past the relays.

I don't think the apparent 7mhz corner you found has anything to do with lack of power on 40m as seen by some.
I think it is more likely that some of the uBitx boards have incorrectly wound 750nH inductors.
When operating on 40m, both the v3 and v5 versions of the uBitx are using the LPF with the 750nH inductors.
At least, when operating with Farhan's code as archived on github.
There are other firmware releases out there, but I suspect they mostly copied
the code to control the LPF relays from Farhan.



Here's the v3 schematic:
  https://github.com/afarhan/ubitx/blob/master/ubitxv3.pdf

And here's the v3 code from
  https://github.com/afarhan/ubitx/blob/master/ubitx_20/ubitx_20.ino
that switches the relays around to select an appropriate LPF:

void setTXFilters(unsigned long freq){

  if (freq > 21000000L){  // the default filter is with 35 MHz cut-off
    digitalWrite(TX_LPF_A, 0);
    digitalWrite(TX_LPF_B, 0);
    digitalWrite(TX_LPF_C, 0);
  }
  else if (freq >= 14000000L){ //thrown the KT1 relay on, the 30 MHz LPF is bypassed and the 14-18 MHz LPF is allowd to go through
    digitalWrite(TX_LPF_A, 1);
    digitalWrite(TX_LPF_B, 0);
    digitalWrite(TX_LPF_C, 0);
  }
  else if (freq > 7000000L){
    digitalWrite(TX_LPF_A, 1);
    digitalWrite(TX_LPF_B, 1);
    digitalWrite(TX_LPF_C, 0);
  }
  else {
    digitalWrite(TX_LPF_A, 1);
    digitalWrite(TX_LPF_B, 1);
    digitalWrite(TX_LPF_C, 1);
  }
}
 
So when operating on 40m (unless you are tuned to be exactly on the lower band edge, don't do that!),
the signals A,B,C  (driving the KT1,KT2,KT3 relays) are driven with 1,1,0 respectively.
KT1 and KT2 are in bypass mode (relay on), and KT3 (off) selects the 7mhz LPF (has 750nH inductors).
So no, on a v3 uBitx, the LPF with the 900nH inductors is only used when operating below (or exactly on) 7mhz.

The v5 uBitx wires up the relays much differently to avoid having harmonics leak around the relays.
    file:///home/chronos/u-ba688dba5e13d507cb8c8d472169a2b5e212d72b/MyFiles/Downloads/ubitx_v5.pdf
The code for v5
    https://github.com/afarhan/ubitx_v5/blob/master/ubitx_v5.1_code.ino
drives the LPF relays with A,B,C == 0,1,0  when operating in the 40m band:

  else if (freq > 7000000L){
    digitalWrite(TX_LPF_A, 0);
    digitalWrite(TX_LPF_B, 1);
    digitalWrite(TX_LPF_C, 0);
  }
This powers the middle relay on the v5 to select the filter with the 750nH inductors,
the other two relays are left unpowered in bypass mode.
Note that the highest freq LPF with the 330nH inductors (for 15m,12m,10m) is always inline on the v5.

Jerry, KE7ER




On Sun, Nov 10, 2019 at 09:08 PM, Mark - N7EKU wrote:
I used the AADE filter simulation/design program.  I posted because I had seem some comments on the power outputs and was curious to simulate the low filter to see if that might impact some of the power levels.  I simulated the lowest filter I saw on the ubitx v3 schematic.  Did I miss a lower one?  This one seems to be the lowest and used for both 40m and 80m with a cutoff very close to 7MHz (actually showed -2dB at 7.3MHz).

Bill Cromwell
 

Hi,

I haven't had any luck on 160 meters so far due to a compromised antenna. But rest assured that I am using and external low pass filter that starts to rolloff a little above 2 MHz. The filters inside the radio are definitely *not* for 160 meters.

It hears pretty well on 160 for all I can tell:)

73,

Bill KU8H

On 11/11/19 12:07 PM, Jerry Gaffke via Groups.Io wrote:
---snip-----
However, the stock uBitx definitely does not have a low pass filter appropriate for 160m !!!!!
We've had a number of folks here who assumed it did, since the firmware allowed them to spin down and transmit on 1.8mhz.
The third harmonic of 1.8mhz at 5.4mhz will go out on the air almost as strong as the fundamental.

--
bark less - wag more

 

Hi Jerry,

Thank you for the information on the code.  Here is a simulation of the 750nH filter:



It looks like a good filter for 30m and 40m.  Input impedance showed 50 ohms on 30m and 40 ohms on 40m.

Sorry the image isn't bigger -- in linux the filter program only prints to either a printer or to "Microsoft XPS Document Printer" which I don't have going in linux so I have to do a screenshot.

Cheers,


Mark.