Date   
Re: #bitx40help BITX40 LSB Winlink #bitx40help

M Garza <mgarza896@...>
 

Daniel,
Your understanding is correct.
We have a 12 MHz IF.  5 MHZ BFO gives us a 7 MHZ LSB signal and a 19 MHZ BFO gives us a 7 MHZ USB signal.  The IF & filter are constant.  In our case, the BFO is moved to either side of the ladder filter to change side bands.

Marco - KG5PRT 

On Tue, Mar 27, 2018, 8:47 AM Daniel Crane <danjamitch@...> wrote:
Thanks Joe.  Yup, Allard's sketch.  That's what I got from Marco and Gordon's posts.

Just to recite my new understanding, for those like me whose facility with superhet design strays just into and out of visibility -

The magic is in the ladder/BFO combination, which shears off the carrier and unwanted sideband, while passing the wanted sideband.  This is not touched in Allard's sketch.  So...  if you use, say, a 10MHz IF, then  3 MHz BFO will get you 7MHz LSB and a 4 MHz BFO will get you 14MHz USB.  Of course you have to filter to make sure you have signal content where you want it to look.

To get 7MHz USB you need a BFO at (drumroll) 17 MHz.  You have to point the ladder at -7Mhz in order to get the mirror image sideband.  Am I right?

Dan

On Mon, Mar 26, 2018 at 9:49 PM, Joe Puma <kd2nfc@...> wrote:
I thought the b40 has USB with the latest version Allard’s? sketch. I’ve used my bitx40 with wsjtx which needs USB too. Winlink should work for you. 



On Mar 26, 2018, at 8:39 PM, Daniel Crane <danjamitch@...> wrote:

I am trying to find a way to connect to Winlink with my BitX40.

I have one of the last Bitx40's shipped, in late November, just before the ubitx started to arrive.  Another fellow whose name I cannot recall got one at the same time and had a lot of difficulty getting it going.  The group here helped him along and over a period of weeks got him up and running.  Just at the conclusion of that process he reported that he had that day made successful Winlink connections.  From the context and timing he would not have converted his BitX 40 to USB.

Does anybody here know if it is possible to connect to Winlink on lower sideband?  I've tried WINMOR, ARDOP, and VARA, but no joy.  No RMS answers my calls.  I have made a number of PSK-31 QSO's between here (LA) and stations as far away as Washington state, so I know my radio works.  I decode RTTY and Olivia as well.

I posted this question on the Winlink group and got the school answer, "Winlink requires USB".

I uptune so as to match 1500Hz center frequencies.  And I uptune 700Hz more because my BitX isn't calibrated.  Occasionally I get lucky and see an RMS engaged  in a session and center that signal. I patiently tune in 100Hz increments.

But still no joy.  Any ideas, other than move on to USB?

Thanks/Dan
KM6GRV




Re: #bitx40 #ubitx #bitx40 #ubitx

Tim Gorman
 

On Mon, 26 Mar 2018 22:10:07 -0700
"Jerry Gaffke via Groups.Io" <jgaffke=yahoo.com@groups.io> wrote:


If the audio is not loud enough then increase the control voltage, up
to a max of 10 or 12v. Should be easy enough with an op-amp.
Though I have not done this, or ever designed an AGC system.
So take this advice with a large grain of salt. 
How do you decide if the audio is not loud enough? Especially in a
simple receiver like this? AGC usually works just opposite of this. It
is based on audio being too loud.



If for some reason you can't use the 10v swing, then could add a 5dB
amp prior to the attenuator. Or maybe goose up the bidi amps for a
little bit more gain.
This would seem to be eminently doable.


I'm thinking the attenuator should be in series with R37, 
the BAP64Q is designed for use in a 50 ohm environment.
R17 might be better. The 45Mhz filter is pretty broad. A mismatch here
might affect the receiver operation less.

Re: FS:Ubitx and Bitx40 ++

_Dave_ K0MBT
 

Email sent to n8dah@... 
Re ubitx.

Re: #bitx40 #ubitx #bitx40 #ubitx

Tim Gorman
 

Don,

I echo your concern about using a non-linear device in series with the
RF. It becomes a simple unbalanced mixer.

I was referring to the datasheet of the pin attenuator.
www.nxp.com/docs/en/data-sheet/BAP64Q.pdf

Here is an article that has an interesting take on using AGC:

http://ke3ij.com/AGC8040.htm

tim ab0wr

On Tue, 27 Mar 2018 06:05:07 -0700
"Don, ND6T via Groups.Io" <nd6t_6=yahoo.com@groups.io> wrote:

After sleeping on it I've got a fresh perspective: Why PIN diodes?
We're not at a significant power point. Why a bridge arrangement? At
this position we are not concerned with maintaining impedance at idle
conditions, they become less important as signals are being
attenuated. So, with 1N4148 switching diodes I found that just the
two series diodes will impose an idle loss of 3 dB and a control
range of 62 dB. The idle current is around 8.3 milliamps. Of course,
going to a single diode (things are getting better) I measured an
more agreeable 1.4 dB insertion loss with a 6 ma. bias and a control
range of 49 dB at 7 MHz, 36 dB at 30 MHz. My principle concerns with
using a device in series with the receive path is the generation of
IM components, especially when biased for partial conduction. BCI
could be a real concern. That was one of the beauties of the shunt
transistor. Tim, the limiting factor of the dynamic range is the
finite Rds. Yes, I was surprised also, that I got 20 dB. This is one
of the reasons that I question most everything. Sometimes I am
rewarded by a boon. At the least I receive an education. Oh, and what
article are you referring? -Don

Re: FS:Ubitx and Bitx40 ++

 

Both radios and digi interface are sold. Tuner is now sold.

I2C radiuno is still for sale
--
David

 N8DAH
Kit-Projects.com

Re: #bitx40 #ubitx #bitx40 #ubitx

Tim Gorman
 

Jerry,

I forgot to reply to this.

Would this really help? I'm going to put together Jerry's agc circuit.
It wouldn't take much to drop one 7002 on top of the other.

Is this a voltage thing instead of a current thing? Two parallel
resistors pull more current but don't change the voltage drop across
each. If the attenuation works by dropping the RF voltage seen by the
receiver than adding a second 7002 probably won't change anything.

tim ab0wr

On Mon, 26 Mar 2018 17:25:39 -0700
"Jerry Gaffke via Groups.Io" <jgaffke=yahoo.com@groups.io> wrote:

Maybe two 2n7002's in parallel?

Re: #bitx40 #ubitx #bitx40 #ubitx

Jerry Gaffke
 

Yes, I'd be concerned about non-linear behavior in a 1n4148 up front there
where we have a whole world of RF coming in.
Maybe a single PIN diode?
Though PIN diodes also start becoming non-linear as the frequency falls,
could still have trouble from some blowtorch at 550khz.
The PIN diode bridge's suggested configuration has a 1k resistor on the control voltage,
so that current would be around 1ma at 10v, 0.5ma per diode.   
Absolute max forward current on the BAP64Q is 100ma.

I think it's a minor consideration, but the 30mhz LPF may behave differently
at other than 50 ohms.

If you do wind up with a good attenuator controlled by DC,
having DC through the RF gain pot would make wiring it up less critical.

I've got some BAP64Q's on order, will try them after the 12mhz amp.
Had been planning to do this 6 months ago.

Jerry 


On Tue, Mar 27, 2018 at 06:05 am, Don, ND6T wrote:
After sleeping on it I've got a fresh perspective: Why PIN diodes? We're not at a significant power point. Why a bridge arrangement? At this position we are not concerned with maintaining impedance at idle conditions, they become less important as signals are being attenuated. So, with 1N4148 switching diodes I found that just the two series diodes will impose an idle loss of 3 dB and a control range of 62 dB. The idle current is around 8.3 milliamps. Of course, going to a single diode (things are getting better) I measured an more agreeable 1.4 dB insertion loss with a 6 ma. bias and a control range of 49 dB at 7 MHz, 36 dB at 30 MHz. My principle concerns with using a device in series with the receive path is the generation of IM components, especially when biased for partial conduction. BCI could be a real concern. That was one of the beauties of the shunt transistor.
Tim, the limiting factor of the dynamic range is the finite Rds. Yes, I was surprised also, that I got 20 dB. This is one of the reasons that I question most everything. Sometimes I am rewarded by a boon. At the least I receive an education. Oh, and what article are you referring? -Don

Re: #bitx40 #ubitx #bitx40 #ubitx

Jerry Gaffke
 

I said "not loud enough" because the PIN diode bridge takes a higher control voltage to give less attenuation,
backwards from the FET attenuator we usually see in this forum.
You decide if it is "not loud enough" by comparing the DC from the detected audio against some threshold voltage
with an op amp or similar.  Seems like this would work well enough if you want the audio to always have the
pretty much the same volume at the speaker, though I suppose it would be preferable to have really loud
incoming signals come through a bit louder at the speaker in a well defined manner.
Don's 1n4148 attenuator won't be doing any better in this respect.

Figure 5 of the BAP64Q datasheet shows an input return loss of better than -10dB
across all control voltages of interest when in a 50 ohm environment   
So might be good enough for 12mhz.
But you're right, putting near the 45mhz filter would be better.

Jerry



On Tue, Mar 27, 2018 at 07:45 am, Tim Gorman wrote:
How do you decide if the audio is not loud enough? Especially in a
simple receiver like this? AGC usually works just opposite of this. It
is based on audio being too loud.

Re: The issues of the TDA2822

wb5klj@...
 

I have just gotten my ubxi40 up and running about a 5 days ago. Has the FCI TDA2822 chip. Running the radio on 13.5V. I did place a 10 ohm resistor in series between the output and the headset/speaker jack, so far no issues.

Ron P.

On Mon, Mar 26, 2018 at 2:08 PM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:
My impression is that the WX failure mode is one of too high a supply voltage
suddenly puncturing a barrier somewhere.  
There may be no noticeable heating prior to the failure.
Perhaps those cut from the periphery of the silicon wafer are more likely to fail.

Raj reports that with very loud audio coming through while running at 12v,
he can short the output of his FCI TDA2822 and have it get hot but not fail. 
There are likely other poor quality clones beyond the WX,
but TDA2822's from reputable vendors don't seem to fail even when abused a bit.

Jerry

On Mon, Mar 26, 2018 at 12:39 pm, Arvo KD9HLC wrote:
On Mon, Mar 26, 2018 at 12:00 pm, Tim Gorman wrote:
I have run the tda2822wx for 24/7 over the past week. No problem. The
chip never even gets warm on the strongest signals. This is with 12v
applied to the 2822.
Hi, Tim.

This is consistent with what''s been reported, that some WX chips seem fine while a few suffer early failure.


Re: #bitx40 #ubitx #bitx40 #ubitx

Jerry Gaffke
 

It's an Rds(on) thing, the 2n7002 is spec'd for a maximum on resistance of 7.5 ohms with Vgs at 5v.
Don found he could cherry pick 2n7002's for better attenuation, they likely had a lower Rds(on).
Assuming a 50 ohm series resistance at the source and a 50 ohm load, a 2v source voltage
gives 1v at the load.  With 7.5 ohms in parallel with the load, I get
    7.5 ohms in parallel with 50 ohms = 6.52 ohms, and a load voltage of (2v/(50+6.52))*6.52=0.23v
Would need down around 0.1v at the load to get Don's 20dB attenuation, but that 7.5 Rds(on)
is a worst case over process and temperature.  So we're in the ballpark.

Putting two 2n7002 in parallel will cut Rds(on) in half, giving you an extra 6dB of attenuation.
It also doubles the capacitance seen at the drain, this will hurt our minimum possible attenuation,
especially at 30mhz.

My bet is on the BAP64Q.

Jerry


On Tue, Mar 27, 2018 at 08:21 am, Tim Gorman wrote:
 Jerry,

I forgot to reply to this.

Would this really help? I'm going to put together Jerry's agc circuit.
It wouldn't take much to drop one 7002 on top of the other.

Is this a voltage thing instead of a current thing? Two parallel
resistors pull more current but don't change the voltage drop across
each. If the attenuation works by dropping the RF voltage seen by the
receiver than adding a second 7002 probably won't change anything.

tim ab0wr

Re: Placed order #ubitx

 

Receive this Feb 2nd after placing my order January 25th:
Dear Customer,

We have received your payment of USD 119 on January 25, 2018 through Paypal. Your payment has been included in the 'Pre order' list. The orders are being shipped in the order of the payment received. It is difficult for us to estimate the shipping date for your order. However, we are putting our best efforts to ship them asap. The tracking details of the shipment will be updated on the Paypal transaction once the order is shipped.

We are overwhelmed with your response for micro BITX and we are encouraged to do better and roll out new products.

Please cooperate with us for the delay in shipping your orders. We look forward to your continued patronage. In case, you have any issues, get in touch with us at hfsignals@...

HFSignals Team

Ubitx.net last week stated that orders of 24 January were now being shipped. Of course my order was the 25th.

Have hear that there have been problems with parts supply such as the audio IC that is no longer in production and that there is now worldwide shortage of torrid cores, So I would guess that production of the ubitx may have been delayed.

Re: Any gotchas replacing the TDA2822 with an LM386? #ubitx #ubitx-help

Sajid Rahum
 

Hi John
Same here; i ordered the 8 pin socket plus NJM2073D.  I will desolder the WX chip first and put in the socket.

I had ordered TDA2822 from two seperate sources + NJM.  Will plan to use try out either to see if they work.

Sajid

Re: #bitx40 #ubitx #bitx40 #ubitx

Tim Gorman
 

Jerry,

Since my planned use for my first ubitx is as a battery operated radio
for emergency nets, e.g. traffic/weather nets, on 80m and 40m I'm not
too concerned about a small amount of attenuation that can't be
removed. Comms on 80m/40m are typically noise limited and not minimum
discernible signal limited.

I don't have a good way to test Rds on sot-23 parts but I might check
out the 2n7000's I have to see if I can find some with low Rds. I don't
have to use smd stuff, plenty of room for thru-hole components.

tim ab0wr

On Tue, 27 Mar 2018 08:41:09 -0700
"Jerry Gaffke via Groups.Io" <jgaffke=yahoo.com@groups.io> wrote:

It's an Rds(on) thing, the 2n7002 is spec'd for a maximum on
resistance of 7.5 ohms with Vgs at 5v. Don found he could cherry pick
2n7002's for better attenuation, they likely had a lower Rds(on).
Assuming a 50 ohm series resistance at the source and a 50 ohm load,
a 2v source voltage gives 1v at the load.  With 7.5 ohms in parallel
with the load, I get 7.5 ohms in parallel with 50 ohms = 6.52 ohms,
and a load voltage of (2v/(50+6.52))*6.52=0.23v Would need down
around 0.1v at the load to get Don's 20dB attenuation, but that 7.5
Rds(on) is a worst case over process and temperature.  So we're in
the ballpark.

Putting two 2n7002 in parallel will cut Rds(on) in half, giving you
an extra 6dB of attenuation. It also doubles the capacitance seen at
the drain, this will hurt our minimum possible attenuation,
especially at 30mhz.

My bet is on the BAP64Q.

Jerry

On Tue, Mar 27, 2018 at 08:21 am, Tim Gorman wrote:


 Jerry,

I forgot to reply to this.

Would this really help? I'm going to put together Jerry's agc
circuit. It wouldn't take much to drop one 7002 on top of the other.

Is this a voltage thing instead of a current thing? Two parallel
resistors pull more current but don't change the voltage drop across
each. If the attenuation works by dropping the RF voltage seen by
the receiver than adding a second 7002 probably won't change
anything.

tim ab0wr

Re: #bitx40 #ubitx #bitx40 #ubitx

K9HZ <bill@...>
 


You may be able to get 60 db or more if you drive Vctrl negative and have the ability to adjust the VCC point and Vctrl differentially.


Dr. William J. Schmidt - K9HZ J68HZ 8P6HK ZF2HZ PJ4/K9HZ VP5/K9HZ PJ2/K9HZ

 

Owner - Operator

Big Signal Ranch – K9ZC

Staunton, Illinois

 

Owner – Operator

Villa Grand Piton - J68HZ

Soufriere, St. Lucia W.I.

Rent it: www.VillaGrandPiton.com


email:  bill@...

 


On Mar 27, 2018, at 10:27 AM, Jerry Gaffke via Groups.Io <jgaffke@...> wrote:

I said "not loud enough" because the PIN diode bridge takes a higher control voltage to give less attenuation,
backwards from the FET attenuator we usually see in this forum.
You decide if it is "not loud enough" by comparing the DC from the detected audio against some threshold voltage
with an op amp or similar.  Seems like this would work well enough if you want the audio to always have the
pretty much the same volume at the speaker, though I suppose it would be preferable to have really loud
incoming signals come through a bit louder at the speaker in a well defined manner.
Don's 1n4148 attenuator won't be doing any better in this respect.

Figure 5 of the BAP64Q datasheet shows an input return loss of better than -10dB
across all control voltages of interest when in a 50 ohm environment   
So might be good enough for 12mhz.
But you're right, putting near the 45mhz filter would be better.

Jerry



On Tue, Mar 27, 2018 at 07:45 am, Tim Gorman wrote:
How do you decide if the audio is not loud enough? Especially in a
simple receiver like this? AGC usually works just opposite of this. It
is based on audio being too loud.

Re: #bitx40 #ubitx #bitx40 #ubitx

Tim Gorman
 

Jerry,

You don't really need an op amp or anything complicated. Take the
positive dc signal from rectifying the audio and apply it to the base
of a PNP transistor. The collector voltage will go down as the base
voltage goes positive (i.e. a stronger audio signal). Set the bias on
the PNP transistor so that it sits at 2v-3v with any dc voltage less
than .7v, i.e. less than the rectifier "on" voltage. Then stronger
audio signals will overcome the diode threshold and push the collector
voltage down and cause the pin diode attenuation to go up.

Assuming the control voltage will remain around 2v-3v most of the time
the return loss should be at 15db or better most of the time at 45mhz.
As the attenuation goes up (i.e. control voltage goes down) the return
loss gets worse but do we really care at that point, especially after
the 45mhz stage?

I realize the diode threshold is not a cliff but a knee but the concept
should still work. If you need a lower turn on voltage you could always
use Schottky diodes.

tim ab0wr

On Tue, 27 Mar 2018 08:27:13 -0700
"Jerry Gaffke via Groups.Io" <jgaffke=yahoo.com@groups.io> wrote:

I said "not loud enough" because the PIN diode bridge takes a higher
control voltage to give less attenuation, backwards from the FET
attenuator we usually see in this forum. You decide if it is "not
loud enough" by comparing the DC from the detected audio against some
threshold voltage with an op amp or similar.  Seems like this would
work well enough if you want the audio to always have the pretty much
the same volume at the speaker, though I suppose it would be
preferable to have really loud incoming signals come through a bit
louder at the speaker in a well defined manner. Don's 1n4148
attenuator won't be doing any better in this respect.

Figure 5 of the BAP64Q datasheet shows an input return loss of better
than -10dB across all control voltages of interest when in a 50 ohm
environment So might be good enough for 12mhz.
But you're right, putting near the 45mhz filter would be better.

Jerry

On Tue, Mar 27, 2018 at 07:45 am, Tim Gorman wrote:


How do you decide if the audio is not loud enough? Especially in a
simple receiver like this? AGC usually works just opposite of this.
It is based on audio being too loud.

Re: Any gotchas replacing the TDA2822 with an LM386? #ubitx #ubitx-help

James
 

I had the misfortune of destroying my WX chip (which spectacularly caught fire) after 13v was applied.

Have dropped in an lm386 using an adaptor and had no problems until a week later when I developed problems on transmit. I now have a loud tone/?feedback both through headphones and RF output. 

I'm not sure that this is related to the failure of the tda2822 but have yet to find the fault. 

Anyway, the lm386 via adaptor was working very well until this new fault.

73
James

Ubitx variable RF out?

Mike
 

Group:

  Is there an easy way to make the RF output variable on a Ubitx?
I now have a LDMOS amp that can only take 3 watts max for 1 KW out.

Thank you!

Mike WA3O

Re: #bitx40 #ubitx #bitx40 #ubitx

Dexter N Muir
 

>How do you decide if the audio is not loud enough? Especially in a
simple receiver like this? AGC usually works just opposite of this. It
is based on audio being too loud.<

This is NOT Audio - it's IF - an RF AGC. AM superhets rectify the IF and feed the 'average' DC forward to control the RF gain. Here we're talking about boosting the RF / IF to provide that DC (output) to feed forward and control the input. This supposedly means less Audio distortion and greater control range, a more constant 'average' signal level into the 'detector'. AM has its Carrier to determine this 'control voltage'. SSB doesn't, so needs more RF to average.Well, that's my understanding anyway.

73
Dex, ZL2DEX

Re: Ubitx variable RF out?

Jerry Gaffke
 

Easiest is to adjust RV1, turning clockwise decreases drive.

If SSB (not CW), could reduce mike gain somehow at Q6.

Could add an attenuator between the uBitx and your big amp.
This might be the best solution as it can make the amp more stable
to be driven from a well behaved 50 ohm source.
 
I'd suggest you make sure your uBitx is clean first.
Spurious outputs that at 10W are barely noticeable will be 
pink ticket bait at 1000W.

Jerry



On Tue, Mar 27, 2018 at 01:28 pm, Mike wrote:
  Is there an easy way to make the RF output variable on a Ubitx?
I now have a LDMOS amp that can only take 3 watts max for 1 KW out.

Re: Ubitx variable RF out?

Mike
 

Jerry ,

 Thanks for the response. Can I remove RV1 and just move it to the front panel?

Thanks!

Mike WA3O