SSB Carrier Supression

[Space Varmint]

Anyway, here is the schematic of the modulator. I cut out the up converter since it is not quite done yet. The diodes are 1N914s. I measured the forward resistance on them and they are all the same. That is under minute light because the light will effect the resistance. The silicon in a glass case will produce electrons if struck by light.

 

[Mikebits]

I would recommend not running your op amp Vcc supply from your Q1 collector.

 

[xanadunow]

dear audioguru

With that extra squack, you do sound like a duck ( just a joke - I am a poet and don't even know it).

The SSB is about the efficiency - it was and is the original purpose. Can you endavour passing "nothing" at the qualiy that exceeds your own requirements - comparing with, "passing the message" at the lesser quality but getting it through?

I am sure you do undersand but only "playing" your case as a "joke"..

Since I have visited here last time, everyone else had and does display very good understanding of all, this is about.

Except you, that is.

Does it mean that we are all wrong?, that you are joking?, or plainly - do not understand?

Regards,
xnadunow

 

[xanadunow]

Circuit, Vcc, resistor.

There is only 10 oHms from the collector to Vcc and the noninverting input seems to be at quite stable point. It appears, the the designer of this part of the circuit did see it as some form of a negative feedback. Pretty unussual, but not harming.

Regards,
xanadunow

 

[Space Varmint]

Hey, I do appreciate you guys

Anyway I found out that my driver amp I had built years ago is a class C.....

So I found an old linear I built and it is the worst piece of junk. I mean it was built years back. Anyway it's got some nice transistors so guess what I got to do now?

But I am very surprized that when I turn on the modulator, it comes up right on freq. I sealed it up. All I need is this linear and I will give you some signal reports.

Back to work......

 

[kchriste]

I would recommend not running your op amp Vcc supply from your Q1 collector.

The amplifier, made up of Q1 (Assuming beta = 200) and it's associated components, improves the input impedance of the OpAmp circuit from 47KΩ to slightly less than 43KΩ. It also adds much needed noise to the high performance 741. I wonder how much residual FM this circuit produces, considering that there is no buffer between the oscillator and the mixer.

 

[Leftyretro]

Hey, I do appreciate you guys

Anyway I found out that my driver amp I had built years ago is a class C.....

So I found an old linear I built and it is the worst piece of junk. I mean it was built years back. Anyway it's got some nice transistors so guess what I got to do now?

But I am very surprized that when I turn on the modulator, it comes up right on freq. I sealed it up. All I need is this linear and I will give you some signal reports.

Back to work......

Well one big deficiency in your drawing is the output of your balanced mixer is wired directly to your crystal filter. A passive diode mixer must have it's output matched to a wide band 50 ohm load. This can be done with either passive or active circuits, but is a requirement for decent performance. A mini-circuits application notes for balanced mixer goes into this subject extensively. The input of your crystal filter is going to be a very complex impedeance and be very frequency dependent, anything but being a good wideband 50ohm match.

Lefty

 

[xanadunow]

sucked in..

That is how I fell Space Varmint.. you do not need any help. But it is a nice joke anyway.

Regards,
xanadunow

 

[Space Varmint]

Well one big deficiency in your drawing is the output of your balanced mixer is wired directly to your crystal filter. A passive diode mixer must have it's output matched to a wide band 50 ohm load. This can be done with either passive or active circuits, but is a requirement for decent performance. A mini-circuits application notes for balanced mixer goes into this subject extensively. The input of your crystal filter is going to be a very complex value and be very frequency dependent, anything but being a good wideband 50ohm match.

Lefty

Tell ya what Lefty. As I got more & more power I did hear a slight carrier. But I got two stages done on a 3 stage linear amp and put in some filtering along the way to where any crud from the up converter mixer is about all gone and as I come up even more the signal is about to destroy my receiver. If I zero beat I get feedback with max receiver attenuation. That being said, I cannot even pick up any carrier even with the receiver gain
turned up. It appears that noise from the mixer in the up conversion can slip through the filter initially and can be picked up as carrier. As the signal increases and is cleaned up it has disappeared. I have just finished two of the 3 stages and with no matching network ans the coax ungrounded, just feeding the center conductor through a capacitor to antenna, I am measuring 1 watt. If I give it a good whistle in the mic. That will be the final's driver stage. Unfortunately the final I have on hand is only 40 watts. An MRF477. Now I got an MRF320 rated at 80 watts but the range is up to 175 MHz. Would probably feedback if I try to use it. Any opinions before I put in this 40 watt? Would rather have 80 watts but might not be feasible.

 

[Mikebits]

I hope you have a LPF before your antenna.

 

[xanadunow]

I do like most (ok.. ALL) of your comments "lefty" but.. it also makes me feel very "retro".. and.. "mike" does always have his watchfull eye about it Plese - let us enjoy this Xmas..

 

[Mikebits]

I do like most (ok.. ALL) of your comments "lefty" but.. it also makes me feel very "retro".. and.. "mike" does always have his watchfull eye about it Plese - let us enjoy this Xmas..

Sorry Xanu, not quite sure I follow what your trying to say. Hope you do not think I was trying to nitpick. Lefty and I just added our 2 cents as it seems SV would not post if he did not want a comment.

 

[xanadunow]

not at all

Sorry Xanu, not quite sure I follow what your trying to say. Hope you do not think I was trying to nitpick. Lefty and I just added our 2 cents as it seems SV would not post if he did not want a comment.

I am finding the two of you as the most "down to earth" fellows that I do respect.

Altough, your big eye - does scare me - a bit..

Regards,
xanadunow

 

[Space Varmint]

I hope you have a LPF before your antenna.

Always Mike. I usually like a "T" network right there.

 

[Space Varmint]

Well one big deficiency in your drawing is the output of your balanced mixer is wired directly to your crystal filter. A passive diode mixer must have it's output matched to a wide band 50 ohm load. This can be done with either passive or active circuits, but is a requirement for decent performance. A mini-circuits application notes for balanced mixer goes into this subject extensively. The input of your crystal filter is going to be a very complex impedeance and be very frequency dependent, anything but being a good wideband 50ohm match.

Lefty

But.....I will bear this in mind. I always look for results. It just turned out that it came out real good! Got lucky that time . The last diode ring I made worked pretty dern good too. But the voice was a bit choppy like the diodes were'nt turning on exactly at the right time. So I got to admit, all of Nigel's bashing got through to me...lol. I did not realize the diodes had to be closely matched like that. So here it goes....

Thank You Nigel

I think it contributed to a tad extra carrier too.

Where I had my biggest problem this time was that VXO. I spend about 4 days on that sucker. I had no problem warping the frequency up, but it did not want to warp down under the crystal's frequency. Even now it is not a smooth adjustment but what it does is sort of repel the crystal's frequency. It will adjust from well under the crystal's frequency, right up to...within, about 1KHz and as I continue to rotate the cap, it will stay right there until I go far enough and then it locks right on and if you keep turning the cap it stays on frequency. But, if I adjust the other cap, it will start right on frequency and adjust smoothly upward and has a strong lock.

All I know is this. Everytime I turn it on, it is right on frequency. So, like I said, I look for results. Hey, I bet that's the first Hartley crystal oscillator you'all have seen....huh? It has one beautiful output too! I mean perfectly sinusoidal.

 

[Space Varmint]

Holy Komoly! That just reminded me of something. If anybody knows anything about PLL then you know that a broader resolution is easier to filter off the phase sidebands. In turn you get a harder lock and even broader bandwidth. I like working with about 10K resolution. I found that by changing the resolution to like 11KHz and 12 and even 9KHz and using a processors internal timer as the reference frequency and jam loading the timer I can achieve multiple frequencies inside the 10KHz steps still using 10KHz reference of 9 thru 12 or 13 which is even cleaner and better. But there are some freqs that are not possible to derive even using this combination of schemes. Well, I like using a Hartley as the VCO in my PLLs so I could place a varactor in place of one of the variable caps to warp the frequency in a VXO and use this as the processors clock, since the internal timer is reference from the uProc clock and use it as an external clock source. What that would do would allow me to achieve every square inch (figuratively speaking) of bandwidth in a sigle loop PLL!!!

Awesome!

 

[Leftyretro]

Holy Komoly! That just reminded me of something. If anybody knows anything about PLL then you know that a broader resolution is easier to filter off the phase sidebands. In turn you get a harder lock and even broader bandwidth. I like working with about 10K resolution. I found that by changing the resolution to like 11KHz and 12 and even 9KHz and using a processors internal timer as the reference frequency and jam loading the timer I can achieve multiple frequencies inside the 10KHz steps still using 10KHz reference of 9 thru 12 or 13 which is even cleaner and better. But there are some freqs that are not possible to derive even using this combination of schemes. Well, I like using a Hartley as the VCO in my PLLs so I could place a varactor in place of one of the variable caps to warp the frequency in a VXO and use this as the processors clock, since the internal timer is reference from the uProc clock and use it as an external clock source. What that would do would allow me to achieve every square inch (figuratively speaking) of bandwidth in a sigle loop PLL!!!

Awesome!

Yes, that is not a method often seen (VXOing the reference) in commerical useage but for ham homebrew it certainly simplifies a PLL system. I've not played with PLL much, the post phase detection filtering requirement calculations always scared me

If ever get around to building my 'holy grail' HF receiver I would most likely go with a DDS generator for the LO, I need something that goes from 40.455mhz to 70.455mhz. Just wish those Analog Devices's DDS chips came in packages that I could actually solder and work with

Lefty

 

[Space Varmint]

Yes, that is not a method often seen (VXOing the reference) in commerical useage but for ham homebrew it certainly simplifies a PLL system. I've not played with PLL much, the post phase detection filtering requirement calculations always scared me

If ever get around to building my 'holy grail' HF receiver I would most likely go with a DDS generator for the LO, I need something that goes from 40.455mhz to 70.455mhz. Just wish those Analog Devices's DDS chips came in packages that I could actually solder and work with

Lefty

What is that? 6 meters? Yeah I like DDS. You might want to look at this chip. I believe it comes in DIP:

But hey, start with VFO, and modularize your design so that you can replace it with something digital. That's what I did. I made my receiver to be 3 to 30 MHz but I have a snap in module that has just a 40 meter PLL.

I know what your saying about the calculations. Like I said, I always look for results. The end result is the FCC requirement of I believe it is 60db (full quite) down on the phase sidebands. I like more than that. More like 120db down but then you start loosing bandwidth. But PLL is really pretty easy with some of the chips out there. I think it is Landsdale that picked up the Motorola line. My particular favorite is the MC145170. Yeah all thoughs calculations for a capacitor and a resistor are really over kill.

 

[Space Varmint]

I have a question:

If the Motorola data sheet "test circuit" is using a diode to ground and a 10 ohm resistor to ground on the base, and they have a switch input for CW and SSB where one says, "SSB bias input" then they are saying they expect an LM317 adjustable regulator to provide the bias, right? Because I ran out of room on this board to do that and I just want to use resistor bias. So I won't be using the 10 ohm and the diode if I try this, right? Because if I use 10 ohms to ground with it locked to .6V, I would need 20 to 40 ohms to supply. That would be a rather large IDC for a 40 watt transistor.

 

[xanadunow]

You have been spotted SV