On Sat, Sep 8, 2018 at 06:11 PM, RCBoatGuy wrote:

Great find!? That information is not in the most recent datasheet.

Welcome,?

It can be annoying when data sheets get revised and important information left out.?

Just found a 1978 Fairchild diode data book online. While using such an old data book may have hazards of its own, it was amazing how much more characterization information was there. I guess solid state diodes were still new and exciting back then.

Tom, wb6b

Hi Allison,

I agree the drivers are a good idea, would be a way to increase the drive to the mixers, especially if the driver is placed closer to the mixer, and reduce crosstalk.

My comments on the current limiting was referencing the concerns about drawing excessive current form the Si5351 outputs.

Tom, wb6b

I assume that's T9 to T10, not T9 to T11.
Would be surprising if you could still make it work well 80-10m after removing the driver stage.
It takes significant power to drive the capacitive load presented by the IRF510 gates,
especially given the drain-to-gate Miller capacitance.

Your SA screen shot is showing 20mhz, if it's giving 5W at 20mhz I'm plenty surprised right there.
?

"I got better results by replacing C22 with filter and leavening R27 intact!"

Raj

---------------------------------------------------------------------------------------
???? ? I wrote that response after reading the above comment from Raj. I'm talking about the second 45MHz filter Raj added.

73 Kees K5BCQ

Allison,

Sorta..... Removing that R3 from the board frees up several cubic inches in that area and moving the Tx/Rx function to another "R3" on the LPF relay board greatly reduces the RF wiring lengths. That leaves the M1 to M2 connection to handle
and your suggestion to use a 2N7000 (faster than a speeding relay) to shunt the point to ground during Tx, especially if driven by the actual T/R line, helps reduce anti-thump-pop too. I see that was done on the V4.

73 Kees K5BCQ

So ..Thinking of this and that ...Came to the blinding revelation that the IRF150 is NOT a current controlled device, but a voltage controlled device ...With that in mind, I bypassed the main current drivers in the pa Q92,93,96,97 and connected directly from T9 to T11...Circuit alterations in pix as well as sweep of PA pass-band ...Output seem on the order of 5 watts ...Spur checking tomorrow

Jim

开云体育

Ah, Siri!

Ricbard, you might consider a Magnetic Loop. It will be hard to get high efficiency on 40 metres, but at 20 metres I have had success with a 1 Metre diameter MFJ loop and 10 watts. If you have a large window though, I would try for a dipole, either straight or bent as Curt suggested.?

开云体育

Timothy, that book from the ARRL is extremely nice

I would recommend putting the voltage regulator in the supply line to the TDA2822.? I did that before applying power to my kit and have not had a problem since then.? I have very high volume levels if I need them.? The TDA2822 is a good amp when supplied this way. I have also put a mod in to minimize the click and pop produced between transmitting and receiving, and an AGC circuit with RF gain attenuation control that has been a great help.? To my ears the receiver is very acceptable now.??
73, Dan, W2DLC

Here is the link to that regulator mod:
/g/BITX20/message/44099

Tom,

Go back and read what I've posted.? Any load that creates a current flow increases the crosstalk.
Using a CMOS buffer with minimal capacitive loading to the 5351 as we want NO current flow.

Actual testing with 50, 100, 240 ohms all proved it got better but 1k and higher was much better.

Allison

Great find!? That information is not in the most recent datasheet.? Current limited output buffers would explain a lot.? However, having a low Rload with current limited outputs would mean slower edge rates on the output clocks.

I knew that the output drivers could be programmed to 2, 4, 6, or 8mA (which oddly is in an app note, not the datasheet), but the latter 2 settings are inconsistent with the current data sheet's claim of an Ioddx max of 5.6mA per output buffer.? I'm not sure what to make of that...

The good news is that the current limited buffers makes excessive ground bounce from power starvation unlikely.? So Allison's test results may, indeed, be valid.?

73,

Carl, K0MWC

I would try any of the mobile whip antennas mounted to the balcony railing and use a quarter wave wire spiral wound on a 4'X4' wooden or PVC frame as the counterpoise.? Just notch the frame to wind the wire in a spiral from center out to the end of the arms.? You could wind one for the lowest band with jumpers at the quarter wave length for higher bands, just open the jumper to go to a higher band.? The mobile whip can be a single band or a multi-band, just change the counterpoise to match the frequency.
I have used this kind of counterpoise in mobile installations to get a good match with the shortened antennas involved.

Allison,

I agree with everything you have said here.

I really don't have anything new to offer at this time. I've moved on
to other, more pressing projects. Perhaps I'll get back to the ubitx as
the weather gets colder and the yard needs less work.

tim ab0wr

On Fri, 07 Sep 2018 10:58:54 -0700
"ajparent1/KB1GMX" <kb1gmx@...> wrote:

Tim,

Known problems:
Amplifier power vs frequency.
IMD extremely poor for power over 2W
Output Low pass filters.
Spurs, unwanted and unfiltered mix products that show above 20mhz.
Carrier leakage that is not related to modulator or 12mhz filter.

I was the first to try adding an mmic after Q90 to

help with making output more constant if you remember. Because of
physical limitations all I could get it to do was cause an oscillation
that destroyed the IRF510.<<

Both Jerry and I tried that.... instability was rife. I used my HP
current limited power supply didn't fry things.

FYI the IMD issue is related to the first item on the list.? The amp
is struggling to work because of poor choice of transistors for most
of the stages before the finals. Put decently capable parts with
reasonable bias elements and you get a significant improvement.

Do offer what you have as its part of a total set of things that need
attention. However fixing it tends to be for the brave as things like
cutting traces, rewinding coils, are needed.

Hope that helps.

Allison

Yep! That's the whole point of a single point ground! Run the negative
lead from the radio to the battery!

tim ab0wr

On Fri, 7 Sep 2018 12:44:35 -0700
"iz oos" <and2oosiz2@...> wrote:

Ok Tim, in that case the chassis of your rig is not connected to the
chassis of the car. If bonded to the chassis of the rig and the
negative of the battery one would create a ground loop. I think you
connected directly the chassis of your radio to the negative of your
battery with a wire of at least the same size of the 'red' cable.

Il 07/set/2018 20:49, "Tim Gorman" <tgorman2@...> ha scritto:

Respectfully, having fought noise problems in several different cars
and trucks, dependence on the dash to provide a ground or even
depending on the floor pan of the vehicle is a good way to introduce
noise into the rig. Not all body parts are properly connected
together and then connected to the battery and physical stress
during car movement can cause noise generation because of high
resistance grounds. Similar to single point grounds in places like
telephone switches or power substations, a negative lead from the
radio directly to the battery has always worked best for me.

tim ab0wr

On Fri, 7 Sep 2018 09:40:13 -0700
"iz oos" <and2oosiz2@...> wrote:

Ok. Think to an audio amplifier in a car. Often you need just the
red wire to turn it on. When the metal case of the amplifier is
connected to the chassis of the car which in turn is connected to
the negative of the battery (please correct me if I am wrong, I
think it is correct). If you add another ground wire (the black
in our example) is a nuisance that may make a ground loop. The
same applies in the uBitx, if the case is all metal some wires
are nuisances that may lead to ground loop issues.

when use metal case then jack sleeves have ground at case an
yellow wire isn't need any more and use thick black ground wire
and and and..then big written 'be aware of ground loop

Il 07/set/2018 17:50, "iz oos" <and2oosiz2@...> ha scritto:

Not in Germany but not that far away, I have qsos on VHF with
German stations only during contests with big big stations like
DR9A... I mostly use qrp and minimal equipment on higher bands as
well. Not every time an increase of noise level is due to ground
loops. But it happened to me I solved noise issue using for
instance isolating audio transformers between the rig and
external accessories and especially PC soundcards. My case is all
metal except for the front and the back panels so it is surely
different from yours. I didn't notice annoying noise that could
be due to ground loops. All installations are different. Ground
loops may become a problem when too many things are connected
together and are not bond together to a common ground point. I am
not an expert on this, read the ARRL or equivalent handbooks that
treat this topic.

Il 07/set/2018 16:36, "sdr freak" <sdrfreak55@...> ha
scritto:

ok thank you!

would you life in germany, too? if yes, we should try to
transmit with the vhf radio when you want it too?

ok by the way, in your post you mean the increased noise is a
sign of ground loops.. ok something new i have learned by
this.. very thankfully about these kind of knowledge!

when the ground loops you say, isn't the problem in ubitx by
wire up the v1.9 shematic in a metal/aluminum case without
extra ground cable to the case and less cable for the sleeves,
how i should i understand the words in shematic v1.9 which says
the things that makes me crazy.. there is written when use
metal case then jack sleeves have ground at case an yellow wire
isn't need any more and use thick black ground wire and and
and..then big written 'be aware of ground loops'...

this is the problem i doesn't understand and how have you make
your project by this, by the using of or not using these whole
points sayes in shematic v1.9? you dont have a metal box only
front is metal ?

hm here again the shematic v1.9 which i have to understand..
/g/BITX20/photo/49276/0/uBITX_wireup_v1.9.
png?p=Name,,,20,1,0,0



G
M
T


English German

English German









Text-to-speech function is limited to 200 characters


Options : History : Feedback : Donate
<

hosted_button_id=GD9D8CPW8HFA2>

Close

Here is a slightly older revision of the Si5351 data sheet before some marketing person possibly demanded that the data sheet had too much technical gibberish and scary numbers in it, and that needed to be fixed.?



The output drivers are current limited with programmable levels of 2, 4, 6, and 8 MA.

Also, the spec sheet claims the DC output impedance is 85 ohms when the 8MA drive is selected. Maybe they think that is close enough to the 50 ohm transmission line they use in the examples that it provides a source side impedance match. In moving data across transmission lines, ribbon cable, twisted pairs and such (usually more like 100+ ohms) you only need to match the source with a series "termination" resistor or the load with a parallel terminating resistor.?

Tom, wb6b

Richard

I would see what HF antennas might fit on or next to your largest window. A dipole with ends bent, or a moxon come to mind. Try to google high rise indoor antennas for ideas. You will be limited to the direction of your window, but your antenna would be much higher than most of us. If you don't have blockage from another large building you might have a decent circumstance.

Curt

Daniel

Great questions, a ham license is an opportunity to keep learning. Good answers here.

Best to get 50 ohm coax, but as others suggest 75 ohms will work. I have never used a balun on a dipole. You can make a center support from a scrap of PVC, as strain relief between coax and the wires is vital.

A dipole interacts with the ground and everything else around, so cutting it for 468 divided by frequency in MHz provides total length in feet, but almost always needs some adjusting. Make wire a few feet longer, and wrap the extra back around to shorten it. You will need a swr meter to adjust this, see what you can find or borrow from a nearby ham.

Be patient with the low power bitx but you can make contacts.

73 Curt

Daniel

Great questions, a ham license is an opportunity to keep learning. Good answers here.

Best to get 50 ohm coax, but as others suggest 75 ohms will work. I have never used a balun on a dipole. You can make a center support from a scrap of PVC, as strain relief between coax and the wires is vital.

A dipole interacts with the ground and everything else around, so cutting it for 468 divided by frequency in MHz provides total length in feet, but almost always needs some adjusting. Make wire a few feet longer, and wrap the extra back around to shorten it. You will need a swr meter to adjust this, see what you can find or borrow from a nearby ham.

Be patient with the low power bitx but you can make contacts.

73 Curt