EMRFD Message Archive 9307

Message Date From Subject
9307 2013-11-04 17:32:48 rcbuckiii DBL diode mixer LO level
I am using a SBL-1 as the front end mixer for a 40 meter receiver. I built a crystal oscillator to feed the LO port of the SBL-1. I used the crystal oscillator circuit that Wes used in the Progressive Receiver from back in the 80s.
 
I have a question concerning the levels into the mixer LO port. Unloaded the oscillator puts out 1.2 volts rms when peaking the output. When I connect it to the SBL-1 LO pin, I am only able to peak the output to 183 mV rms. That is about -2dBm into the LO pin.
 
I am feeding the RF port with a 5.2 MHz signal from a VFO I built. The VFO signal into the RF pin is .156 volts rms (-3dBm). The output from the mixer is .025 volts rms (-19 dBm). That is a 16 dB loss.
 
The above readings were all taken with my Tektronix 2445 and the math done to convert from peak voltages to rms voltages.
 
If I feed the three signals into my KA7EXM power meter the VFO and output signals almost agree with the scope reading. However, the LO signal is drastically different. On the power meter the VFO signal reads -5.3 dBM. The output signal reads -18.9 dBm. The LO signal reads +3.6 dBM versus -2dBM on the scope. Does the SBL-1 mixer have that much of an effect on the LO signal?
 
My crystal oscillator is running at 2 MHz. For the output transformer I made the primary 87 turns tapped at 12 turns. For the secondary I used 8 turns.
 
I think the large loss through the mixer is due to the LO signal being so low. According to the Progressive Receiver notes, the oscillator output should be +10dBm. Could my -2 dBm output be caused by the output transformer?
 
Ray AB7HE
9309 2013-11-04 19:27:42 Ashhar Farhan Re: DBL diode mixer LO level
can you place a 6 db attenuator in the output of the crystal oscillator and then make these measurements? It is vital that power measurements on the scope be made not with a regular probe but a 50 ohms cable that is attached to a 50 ohms load the scope end ('at the scope end' is an important detail). 

- f


9312 2013-11-04 20:36:28 rcbuckiii Re: DBL diode mixer LO level

If I do that my LO power will be down to -3 dBm. Is that acceptable? My scope has a 50 ohm position

9313 2013-11-04 21:46:41 Ashhar Farhan Re: DBL diode mixer LO level
Yes, the 50 ohms termination will allow you to measure the power
accurately. So, if-3 db is what you are getting, then that's what the
osc is putting out at 50 ohms. Try increasing the turns in the output
link to see if you get more power. But the measurements must always
happen at 50 ohms.
- farhan

On 11/5/13, rcbuckiii@yahoo.com <rcbuckiii@yahoo.com> wrote:
> If I do that my LO power will be down to -3 dBm. Is that acceptable? My
> scope has a 50 ohm position
9314 2013-11-05 02:27:12 Roelof Bakker Re: DBL diode mixer LO level
Hello Ray,

Which type of core did you use for the output transformer?

73,
Roelof, pa0rdt
9315 2013-11-05 03:20:19 Ashhar Farhan Re: DBL diode mixer LO level
Almost any type will do. I have a large supply of binocular cores that
were meant for 300 ohms to 75 ohms baluns for tv, back when tv was
wireless. But I also rebuild most circuits with FT 37-43.
I haven't built the progressive receiver crystal oscillator. It is a
bit of a visual trick with the feedback.
- farhan

On 11/5/13, Roelof Bakker <roelof@ndb.demon.nl> wrote:
> Hello Ray,
>
> Which type of core did you use for the output transformer?
>
> 73,
> Roelof, pa0rdt
>
>
> ------------------------------------
>
> Yahoo Groups Links
>
>
>
>

--
Sent from my mobile device
9317 2013-11-05 05:29:41 rcbuckiii Re: DBL diode mixer LO level

I am using a T50-2 core since the output needs to be resonated. The original Progressive Receiver cores were T50-6. But since this oscillator is at 2 MHz, I used a -2.

 

Also, with a 50 ohm load the oscillator would stop. So I added a 300 pF cap from emitter to base so it would functi

9322 2013-11-05 18:30:53 rcbuckiii Re: DBL diode mixer LO level

The problem with the 2 MHz oscillator either has to do with the output transformer or with the actual oscillator itself. I think I'll breadboard up the circuit shown in Fig 4.16 in EMRFD. That may function better at 2 MHz than the Progressive Receiver oscillator.

I built a Progressive Receiver oscillator running at 9 MHz as a test. It puts out +8.5 dBm into a 50 ohm resistor. When I connect it to the SBL-1 the output drops to +4 dBm. But I'm not concerned as that is normal according to the two paragraphs under "LO DRIVE LEVEL" on page 5.6 of EMRFD. That is exactly what I am seeing, a highly distorted, lower level signal into the LO port of the SBL-1.

Ray, AB7HE 



9324 2013-11-05 23:39:02 Ashhar Farhan Re: DBL diode mixer LO level
Distortion is normal when it connects to a diode mixer. The power may
very well still be +7dbm. The power meter you are using is expecting a
sine wave, I presume.
The actual test of 'enough' BFO injection is just your own ears. If
the signal is clean, without splatter from the adjacent signals then
you are doing good.
- farhan

On 11/6/13, rcbuckiii@yahoo.com <rcbuckiii@yahoo.com> wrote:
> The problem with the 2 MHz oscillator either has to do with the output
> transformer or with the actual oscillator itself. I think I'll breadboard up
> the circuit shown in Fig 4.16 in EMRFD. That may function better at 2 MHz
> than the Progressive Receiver oscillator.
>
> I built a Progressive Receiver oscillator running at 9 MHz as a test. It
> puts out +8.5 dBm into a 50 ohm resistor. When I connect it to the SBL-1 the
> output drops to +4 dBm. But I'm not concerned as that is normal according to
> the two paragraphs under "LO DRIVE LEVEL" on page 5.6 of EMRFD. That is
> exactly what I am seeing, a highly distorted, lower level signal into the LO
> port of the SBL-1.
>
> Ray, AB7HE
>
>
>
>
9325 2013-11-06 08:44:31 William Carver Re: DBL diode mixer LO level
Per Farhan....you test your LO to see that it delivers +7 dBm into a 50
ohm resistive load. Amplitude/purity, etc, are evaluated with that 50
ohm load load. The diodes in the mixer will clip (round off) the top and
bottom of the sine wave, making LO amplitude evaluation "iffy" after its
connected to the mixer.

It is also desirable to have the LO source look like a 50 ohm source to
absorb spurious mixer signals that appear there. Ulrich Rhode wrote
about this issue in Ham Radio in the 70s. This is why many designs have
an amplifier stage between the LO oscillator and the mixer: not
necessarily only to get more power, but also to more closely approximate
a 50 ohm source. Even driving it with a square wave from a flip flop the
mixer benefits from putting a 50 ohm resistance (minus the source
impedance of the logic stage) between the mixer LO port and the logic.

I know this is dumb to mention since you're fighting to get enough LO
amplitude already, but after you get it mixing with +7 from your
existing LO you might get better third-order intercept by putting an
amplifier after the oscillator, lowpass filtering it to remove even
order harmonics (2nd, 4th, etc) then having a 6 dB attenuator between
the filter and the mixer.

One step at a time!

W7AAZ






On Wed, 2013-11-06 at 13:09 +0530, Ashhar Farhan wrote:
>
> Distortion is normal when it connects to a diode mixer. The power may
> very well still be +7dbm. The power meter you are using is expecting a
> sine wave, I presume.
> The actual test of 'enough' BFO injection is just your own ears. If
> the signal is clean, without splatter from the adjacent signals then
> you are doing good.
> - farhan
>
> On 11/6/13, rcbuckiii@yahoo.com <rcbuckiii@yahoo.com> wrote:
> > The problem with the 2 MHz oscillator either has to do with the
> output
> > transformer or with the actual oscillator itself. I think I'll
> breadboard up
> > the circuit shown in Fig 4.16 in EMRFD. That may function better at
> 2 MHz
> > than the Progressive Receiver oscillator.
> >
> > I built a Progressive Receiver oscillator running at 9 MHz as a
> test. It
> > puts out +8.5 dBm into a 50 ohm resistor. When I connect it to the
> SBL-1 the
> > output drops to +4 dBm. But I'm not concerned as that is normal
> according to
> > the two paragraphs under "LO DRIVE LEVEL" on page 5.6 of EMRFD. That
> is
> > exactly what I am seeing, a highly distorted, lower level signal
> into the LO
> > port of the SBL-1.
> >
> > Ray, AB7HE
> >
> >
> >
> >
9326 2013-11-06 14:20:06 rcbuckiii Re: DBL diode mixer LO level
Bill, the only issue I'm having is getting the 2 MHz oscillator working. The same circuit works fine at 9 MHz. I can load the 9 MHz one with 50 ohms and I get +8.5 dBm. The 2 MHz one actually dies if I connect a 50 ohm load to the output. It was putting out +3.6 dBm when it was connected to all the other hay wired stuff on the bench. Once I broke it out separate by itself, it works until I put a 50 ohm load
9327 2013-11-06 14:48:48 William Carver Re: DBL diode mixer LO level
In concept you "should" be able to build the oscillator for any
frequency (within the limits of the active device) by duplicating the
impedance of the L's and C's that surround the oscillator.

So to move from 9 to 2 MHz you would want to multiply capacitor and
inductor values by 4.5. Tank capacitors just multiply by 4.5.

Your -2 core v.s. -6 core permeability gets you 25% of the way on the
transformer, leaving 4.5 divided by 1.25, or 3.6 to go. Since inductance
is proportional to turns squared, and the square root of 3.6 is 1.897, I
would expect if you multiplied the -6 transformer turns by 1.9 then the
transformer will be about right.

Those are a hipshot guess. Oscillators are pretty forgiving, they WANT
to oscillate, so I would expect that hipshot to work if the original
design worked. Especially since the transistor gain will be higher at
2.5 MHz.

Let's say it doesn't work....in fact you're no dummy so I kind of expect
it still doesn't work! So now what? I'd measure the voltage on the
transistor base and emitter. The difference should be 0.65-0.7 volts if
its not oscillating. Divide the emitter voltage by the emitter resistor
to get the dc current.

Is it a reasonable current? Should be a handful of mA, I would think.
For example in Solid State Design for the Radio Amateur, figure 31 on
page 129, the oscillator there should be drawing 6.1 mA if the xtal is
removed.

But looking at figure 31 brings up another possibility. In that circuit
L6 and the 180 pF tuning capacitor are tuned to 5.2 MHz. But not
exactly. I believe it has to be tuned either a bit higher, or a bit
lower for the oscillator to run. So it brings up the question of whether
you are able to tune any resonances in YOUR oscillator above and below
the 2 MHz frequency?

If after everything has been tried, the next thing I'd do is drop the
number of turns on the link to the 50 ohm load. This reduces the load
seen by the oscillator transistor. If that does the trick, but is not
enough to drive the diode mixer directly, you have a couple of choices:
first is to boost the voltage and/or current so the transistor can
generate more power. IE, drop the emitter resistor to get double the dc
current for starters. The other choice is to add your buffer stage to
get more mixer power that way.

Although there's nothing that says you cannot directly drive a diode
mixer with the oscillator, I've always had a buffer stage in between.

On Wed, 2013-11-06 at 14:20 -0800, rcbuckiii@yahoo.com wrote:
>
> Bill, the only issue I'm having is getting the 2 MHz oscillator
> working. The same circuit works fine at 9 MHz. I can load the 9 MHz
> one with 50 ohms and I get +8.5 dBm. The 2 MHz one actually dies if I
> connect a 50 ohm load to the output. It was putting out +3.6 dBm when
> it was connected to all the other hay wired stuff
9329 2013-11-07 09:20:05 Richard Hayter Re: DBL diode mixer LO level
Ray

is it that the L/C ratio needs adjusting in the tank?  Scaling up by a multiplier of 4.5 from the 9Mhz version gives 14uH and 450pf. On T 50-2 this is 54Turns + 12 Turns for a total of 66 turns (as opposed to 75. Then you would need 450pf to resonate or thereabouts as opposed to the max you have which is 350pf?

73 Dick


9331 2013-11-07 19:20:43 rcbuckiii Re: DBL diode mixer LO level

Dick,

I think I've come to the conclusion you can't build a 2 MHz single stage oscillator and get any output with a 50 ohm load. In the original Progressive Receiver oscillator they had a 3.3 MHz oscillator. That oscillator used a T68-2 core and had 65 turns. Using the reactance of that transformer (499 ohms) I scaled a transformer for 2 MHz. I wound that transformer this afternoon (84 turns on a T68-2 core) and put it on the oscillator board. It resonates with maximum output at approximately 160 pF.

With no load, the oscillator puts out 1.8 V rms. When I connect a 50 ohm load to the output, the oscillator dies. I put a 500 ohm pot on the output and started adjusting downward in value. At 140 ohms the oscillator dies. Based on that result, I don't think I can use a 50 ohm input feedback amp as the oscillator will not be able to drive it.

So I either have to go to an active mixer (NE602) or try using a Colpitts oscillator to see if that will work into 50 ohms at 2 MHz.

Ray,
AB7HE



9332 2013-11-07 19:25:44 William Carver Re: DBL diode mixer LO level
Ray, I've lost track of the exact oscillator circuit. But certainly
there HAS to be a way to "tap down" on the tank to deliver power to a 50
ohm load without killing oscillation.

What is the circuit again?

W7AAZ


On Thu, 2013-11-07 at 19:20 -0800, rcbuckiii@yahoo.com wrote:
>
> Dick,
>
> I think I've come to the conclusion you can't build a 2 MHz single
> stage oscillator and get any output with a 50 ohm load.
9333 2013-11-07 19:53:43 rcbuckiii Re: DBL diode mixer LO level

Bill,

It is the crystal oscillator used in Wes' Progressive Receiver published in Nov 1981 QST. The oscillator circuit is

9334 2013-11-07 20:58:39 William Carver Re: DBL diode mixer LO level
Got it from QST archives.

The turns ratio of the transformer (primary turns minus "tap" turns,
divided by secondary turns) varies between 5+ to 4.5:1. 5:1, when
terminated by 50 ohms that would make a 1.25K load on the collector. For
the 3.3 MHz values the collector tank Q is about four.

Lets say you were able to get +10 dBm into 50 ohms, as he suggests.
That's 14.1 mA (if it was a sine wave) in the load, and 3.1 mA RMS at
the collector. I'm guessing the collector dc current is about 5.4 mA,
and those numbers seem consistent. Are you getting about 2.1 volts at
the emitter of the oscillator transistor?

In concept, you could reduce the secondary turns to increase the
impedance seen by the collector when you put a 50 ohm load on the
secondary. But upon casual inspection that doesn't seem like it should
be necessary.

But lets cut through the crap: I looked in my junkbox and have some
1.832 and 2.000 MHz xtals there. Unknown quality/Q, etc. The series
resistance of low freq crystals rises as the frequency goes lower. I
have several irons in the fire, but I'll try to throw and oscillator
together tomorrow with those two crystal frequencies and see what
happens.

Bill




On Thu, 2013-11-07 at 19:53 -0800, rcbuckiii@yahoo.com wrote:
>
> Bill,
>
> It is the crystal oscillator used in Wes' Progressive Receiver
> published in Nov 1981 QST. The oscillator circuit is
9335 2013-11-08 05:18:36 rcbuckiii Re: DBL diode mixer LO level

Just for reference this is how I wound my transformer:
T68-2 Primary 84 T, tap at 17 -- Sec 13 T -- I used #32 wire

 

You may find you have to use a feedback cap between the emitter and base to get the circuit to oscillate. I found anything from 330-680 pF works.

 

Ray



9336 2013-11-08 06:36:29 William Carver Re: DBL diode mixer LO level
OK Ray....
As I was dozing off to sleep last night I was thinking about xtal series
resistance. It gets higher as you go lower in frequency, and that could
be why its not working at 2 MHz.

Suggestion: bypass the xtal with, say, 10 ohms. It should then oscillate
at a frequency controlled by resonance of the transformer and its tuning
capacitor. You should be able to tune it above and below the xtal
frequency a bit.

Hafta run car to dealer for a recall check...low oil level (as shipped!)
in the differential causes the back wheels to lock up while you're
driving.....

Bill


On Fri, 2013-11-08 at 05:18 -0800, rcbuckiii@yahoo.com wrote:
>
> Just for reference this is how I wound my transformer:
> T68-2 Primary 84 T, tap at 17 -- Sec 13 T -- I used #32 wire
>
>
>
> You may find you have to use a feedback cap between the emitter and
> base to get the circuit to oscillate. I found anything from 330-680 pF
> works.
>
>
>
> Ray
>
>
>
>
9337 2013-11-08 07:32:06 John Marshall Re: DBL diode mixer LO level
Bill's suggestion is described on page 4-15 of EMRFD (1st ed) as a way of aligning a very similar oscillator for overtone operation. That page also shows a 10 MHz version of the Progressive Receiver oscillator. Another thing you could try is to remove the 2.2K resistor shunting the transformer. That will reduce the load on the oscillator some and may compensate for a crystal with higher series resistance.

John, KU4AF
Pittsboro, NC

9338 2013-11-08 19:50:03 rcbuckiii Re: DBL diode mixer LO level

Bill, I built another 2 MHz oscillator today. It is the circuit shown in Fig 1, pg 8 of QRP Classics. I don't see the same circuit in EMRFD.

However, the circuit shown in EMRFD, pg 4.14, Fig 4.26 is close. Lift the grounded end of the crystal and return it to the collector through a .1 uF cap. Make C3 100 pF and ground it. Place a tuning capacitor across the primary of the transformer. You then have the circuit from QRP Classics that I built.


I wound the transformer with 73 turns on the primary and 13 turns on the secondary. My tuning cap consists of a 180 pF cap in parallel with a 5-65 pF variable cap. The primary of the transformer measured 29 uH on my ADE LCR meter.

It puts out 1.95 volts Peak (1.38 V rms) into a 50 ohm resistor. That leads me to believe the Hartley circuit used for the Progressive Receiver oscillator is not really a good choice for a 2 MHz oscillator.

Ray, AB7HE



9339 2013-11-09 06:13:01 William Carver Re: DBL diode mixer LO level
VOILA! That is almost +16 dBm.

But...but...the verbal description of what you built sounds like the
SAME circuit as what you had before except the ground is in a different
point. So you have created a mental puzzle for me, hi.

I suspect Wes put the ground where he did so that the capacitor in
series with the xtal has one side grounded, making mounting easier if
you wanted it to be slightliy variable (like USB/LSB for the BFO).

As the old saying goes, "All's well that ends well". Congrats.

Bill





On Fri, 2013-11-08 at 19:50 -0800, rcbuckiii@yahoo.com wrote:
>
> Bill, I built another 2 MHz oscillator today. It is the circuit shown
> in Fig 1, pg 8 of QRP Classics. I don't see the same circuit in EMRFD.
>
> However, the circuit shown in EMRFD, pg 4.14, Fig 4.26 is close. Lift
> the grounded end of the crystal and return it to the collector through
> a .1 uF cap. Make C3 100 pF and ground it. Place a tuning capacitor
> across the primary of the transformer. You then have the circuit from
> QRP Classics that I built.
>
>
> I wound the transformer with 73 turns on the primary and 13 turns on
> the secondary. My tuning cap consists of a 180 pF cap in parallel with
> a 5-65 pF variable cap. The primary of the transformer measured 29 uH
> on my ADE LCR meter.
>
> It puts out 1.95 volts Peak (1.38 V rms) into a 50 ohm resistor. That
> leads me to believe the Hartley circuit used for the Progressive
> Receiver oscillator is not really a good choice for a 2 MHz
> oscillator.
>
> Ray, AB7HE
>
>
>
>
>
9341 2013-11-11 19:24:15 rcbuckiii Re: DBL diode mixer LO level

As a follow up to this discussion, I built the circuit shown in Fig 4.26 in EMRFD as a learning exercise. I used a 330 pF capacitor for C3 as suggested in the text. That circuit puts on 2.20 volts peak (1.55 V rms) into a 50 ohm load. This is right at 17 dBm.

I used the same T68-2 transformer with 73 T primary, 13 T secondary that I had used in my previous oscillator that had almost 16 dBm output. I initially had problems with the oscillator starting with a 50 ohm load. I could start it with no load, connect the 50 ohm resistor, and it would continue putting out about 1.5 volss peak. I replaced the 50 ohm resistor with a 500 ohm pot. I found that as long as I stayed above 140 ohms the oscillator would start on power up. Calculating the turns ratio vs load indicated the 2N3904 wanted to see a 4400 ohm collector impedance. That seems a little high but may be due to the 2 MHz frequency.

I calculated that for a 4400 ohm primary and 50 ohm secondary, the secondary winding needed to be 8 turns. I took 5 turns off of the secondary and the oscillator starts up at full power immediately with a 50 ohm resistor

9342 2013-11-11 20:03:16 William Carver Re: DBL diode mixer LO level
It's nice to close the loop and get it to work. The series resistance of
xtals goes up as the frequency drops and that resistance is in series
with the emitter: sounds like the transistor and circuit have marginal
gain with a 2 MHz xtal with the 50 ohm load.

Glad to hear things all kind of make sense after a fashion, and you're
able to get the drive you need for the mixer.

Bill


On Mon, 2013-11-11 at 19:24 -0800, rcbuckiii@yahoo.com wrote:
>
> As a follow up to this discussion, I built the circuit shown in Fig
> 4.26 in EMRFD as a learning exercise. I used a 330 pF capacitor for C3
> as suggested in the text. That circuit puts on 2.20 volts peak (1.55 V
> rms) into a 50 ohm load. This is right at 17 dBm.
>
> I used the same T68-2 transformer with 73 T primary, 13 T secondary
> that I had used in my previous oscillator that had almost 16 dBm
> output. I initially had problems with the oscillator starting with a
> 50 ohm load. I could start it with no load, connect the 50 ohm
> resistor, and it would continue putting out about 1.5 volss peak. I
> replaced the 50 ohm resistor with a 500 ohm pot. I found that as long
> as I stayed above 140 ohms the oscillator would start on power up.
> Calculating the turns ratio vs load indicated the 2N3904 wanted to see
> a 4400 ohm collector impedance. That seems a little high but may be
> due to the 2 MHz frequency.
>
> I calculated that for a 4400 ohm primary and 50 ohm secondary, the
> secondary winding needed to be 8 turns. I took 5 turns off of the
> secondary and the oscillator starts up at full power immediately with
> a 50 ohm resistor