Continue to Site

Welcome to our site!

Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

  • Welcome to our site! Electro Tech is an online community (with over 170,000 members) who enjoy talking about and building electronic circuits, projects and gadgets. To participate you need to register. Registration is free. Click here to register now.

SSB Carrier Supression

Status
Not open for further replies.
Yeah but your power output will be the same. It may take more power to produce AM high level modulation, but the net result in the final is exactly the same power dissipation.

High level AM modulation produces considerably higher output levels than low level AM as per SSB type techniques, or cathode modulation etc.

However, you're not really gaining anything because you're using the extra power to modulate it.
 
That was my point, although there are extremly sharp (and expensive)bandpass filters using many stages that can have almost straight "brick wall" passband shape. However I've never seen a SSB modulator/demodulator circuit without either a balanced modulator or using phasing circuits to cancel the carrier frequency.

Lefty

No, I'm gonna use a ballanced mod. I wish I had gotten a CA3039.
 
Say ya'll, If anyone has the time or interest....can you post a differential amplifier based balanced modulator? I really can't visualize how this thing may work. When I look for some circuits, I always pull up the MC1496. I don't want an IC for theoretical purposes. If I understood the chip better, I might attempt it, but I would like to look a a basic discreet Xsistor circuit first.

At this point I have achieved best results from the diode ring mixer using 1N914's. So I got some here where I measured the forward resistance (out of light) and I cannot even read the reverse resistance. Out of range. But I have some that look right on. I may just go back to that.
 
The MC1496 is simply two differential amplifiers with their outputs connected together and their inputs connected back-to-back. The transistors are well matched because the IC is monolithic.
The inputs are biased at 0V when the supply is dual polarity or the inputs are biased at a few volts when the supply is single.
 
mixers

Yes Varmint, the diod ring produces the best results with the lowest amount of pain and let me quote from the good site you have mentioned earlier (**broken link removed**): "..don't even consider using 1N914 types if you want results.."

Regards,
xanadunow
 
Last edited:
Yes Varmint, the diod ring produces the best results with the lowest amount of pain and let me quote from the good site you have mentioned earlier (**broken link removed**): "..don't even consider using 1N914 types if you want results.."

Regards,
xanadunow
You think so? Nigel says the 1496 is better. I have seen other sights saying the '941s are fine. So far I have gotten best results from the '914s in a diode ring but I got a 1496 chip wired up to try. I got the design from the ARRL handbook but I will be using 6 volts. So I may do a comparison but I really wish I had a schematic of a discreet transistor circuit of the differential amps. Thanks for trying to explain it guru.
 
You think so? Nigel says the 1496 is better. I have seen other sights saying the '941s are fine. So far I have gotten best results from the '914s in a diode ring but I got a 1496 chip wired up to try. I got the design from the ARRL handbook but I will be using 6 volts. So I may do a comparison but I really wish I had a schematic of a discreet transistor circuit of the differential amps. Thanks for trying to explain it guru.

The diode rings have the advantage that they will work at substantial signal levels, so are often used for frontend receiver mixers, as they are pretty well immune to cross modulation. 1496 are better balanced, and probably better for generating SSB, but not so good for a receiver frontend mixer.

You posted a circuit a while back using just two diodes, and a preset to adjust the balance - have you tried making one like that?.

If it's of any interest?, my long ago receiver used a dual-gate mosfet front end, with delayed AGC applied to the second gate. It was fed from a tuned pre-scaler, with a manual attenuator on the front to prevent overload and crossmod. The mixer was another dual-gate mosfet, and the oscillator used bi-polar transistors. The IF used CA-something or other IC's (they've been mentioned somewhere in these threads, can't remember the number - I've still got a couple left somewhere?), feeding a 1496 product detector. I used a 9MHz crystal filter, with matching upper and lower sideband crystals, I used two separate oscillators, and switched power between them, making the switch non-critical, and outside the screened box.

Like I've mentioned previously, my original intention was to make a transceiver from it, but I never passed my morse test, so never bothered. I've still got the Plessy SL6-something IC I bought for the mike preamp, which did all the processing in a single IC.
 
The diode rings have the advantage that they will work at substantial signal levels, so are often used for frontend receiver mixers, as they are pretty well immune to cross modulation. 1496 are better balanced, and probably better for generating SSB, but not so good for a receiver frontend mixer.

You posted a circuit a while back using just two diodes, and a preset to adjust the balance - have you tried making one like that?.

If it's of any interest?, my long ago receiver used a dual-gate mosfet front end, with delayed AGC applied to the second gate. It was fed from a tuned pre-scaler, with a manual attenuator on the front to prevent overload and crossmod. The mixer was another dual-gate mosfet, and the oscillator used bi-polar transistors. The IF used CA-something or other IC's (they've been mentioned somewhere in these threads, can't remember the number - I've still got a couple left somewhere?), feeding a 1496 product detector. I used a 9MHz crystal filter, with matching upper and lower sideband crystals, I used two separate oscillators, and switched power between them, making the switch non-critical, and outside the screened box.

Like I've mentioned previously, my original intention was to make a transceiver from it, but I never passed my morse test, so never bothered. I've still got the Plessy SL6-something IC I bought for the mike preamp, which did all the processing in a single IC.

Plessy made some great RF and audio application ICs back then but I never found an easy source for them so didn't get to play with them.

Lefty
 
Wow! Did you see that 14MHz band pass filter? That thing is pretty hairy.

Well a mixer produces more then just the sum and difference frequencies, but also the harmonics of the sum and differences as well as a small amount of the 2 fundmental frequecies being mixed as no mixer has perfect balance.

As I'm sure you are aware the FCC has rules about the amount of harmonic power content a transmitter can output and as this filter is for the transmit chain it is to keep the output signal FCC happy.

Lefty
 
I think Nigel has nailed it on the head. The immunity to the cross modulation of the Rx stages is paramount and the mixer in particular must be able to handle it all across all voltage levels finding their way to the antenna's Rx Input.

AGC is helpfull, but when you are trying to listen to a weak station and suddenly a very strong signal appears nearby, it has to be reduced if you want to continue to receive the weak station, if you do not - you will loose the weak station because AGC will kill it by reducing the gain. This is where the immunity to cross modulation plays it's role. If it is poor, the strong signal will sweep-out the weak one (in the mixer) and everything is lost, but with good immunity, even knowing the annoyance of the strong signal is present - the weak one is still there (in the receiver), and only - just harder to receive. At this point we employ all the tricks of trade adjusting the selectivity/trimming the strong signal out od the IF band, doing the IF shift (if we have one), employing notch filter in the IF (if we have one), and finally - doing just about all the same in the audio stages of the receiver. But this would only work if the mixer is immune to the cross modulation and the balanced/double balanced diode mixers do the job here the best.

When transmitting, the mixer is just as important but we are no longer concerned with the cross modulation. We know exactly the optimal level values of frequencies arriving at the mixer (and their ratio) and we can make these signals just right.

Balancing the mixer in the most "balanced" manner achieve the best possible rejection of unwanted signals and the IC here is sure a winner. This is all true as Nigel stated, but lefty was very quick to point out everything there is to be had about harmonics and these are the two main concerns when talking about balanced mixers when used in the transmitter; the balance and the harmonics.

All to easy I came to the conclusion that introducing an op-amp with a number of non-linear components is a recepie in a mixer to produce more signals we do not want, more than balanced diode mixer will. Maybe the balance will not be so perfect with a diode ring, but I am pretty sure - we will be getting more "nasties" out of the way when producing the signal. Unlike the OpAmp, the diode ring has only 4 non-linear components.

The implication here is that one has to be very carefull when designing an op-amp mixer, nothing more.

On the other hand, the balanced diode ring mixer is easy to make. If you already made one for the receiver - it is conceivable that you will also make one for your transmitter.

One not mentioned here thing is the bandwidth of a diode ring mixer. No matter what the IF is, the balanced diode ring will do the job, you don't really have to guess.

You do not have to guess if you are using quality diodes and toroids "fit" to work at the employed frequency range. The rest is simple and the results are there.

And what for those who want the best of the double balanced diode ring mixers?

The industry has responded long time ago. We do not have to play with diodes and toroids. We do look at "minicurcits". They are hybrids premade pretested and with repeatable set of parameters. And yes, they do have a matched set of diodes. A good way to go if you want good results.

I have not looked it up on the internet because I have these pdf files stored on the computer. I am sure you can still "google" them. They relate to two minicircuits:

SBL-2-1 (bandwidth 5MHz to 1 GHz), double balanced, +7 dBm LO

and

SBL-1H (bandwidth 1-500MHz), double balanced, +17dBm

These would be pretty old by now but I have a few of them in my hands and if I ever build a balanced mixer again - this is what I will be using.

Regards,
xanadunow
 
Last edited:
Cross modulation, by that I assume you mean Intermodulation distortion as I think they are similar.

I have been preaching Mini-Circuits for some time now to no avail (See this thread) https://www.electro-tech-online.com/threads/shure-mike-model-444-freq-scheme-ssb.85841/
but this is a great off the shelf solution for a Rx front end mixer.
Mini-Ckts has real good specs for intermod as well as Noise figure. I think you could almost build an entire RX unit from Mini-Ckt parts. They have amps, mixers, filters, and VCO's.

For the TX modulator, Mini-Circuits does not spec their parts below 500KHz so I am not sure how well the would work at audio modulation frequencies, but I suspect they would be fine. Maybe someone could test it and report the results.
 
Last edited:
definitions

Hi Mikebits, I had to read up on this in wiki (so it was quick) and I am now not too sure which is the right term. Therefore, I need to state not so much on the definitions (however the cross modulation sounds as a right one), but more - on the symptom.

The symptom is overloading of the mixer that practically "wipes out" what I was trying to listen to. This overloading is caused by the neighbouring strong signal at the front end of the receiver.

I also need to state that most of my statements do come from what I do remember and are not supported by the many hours of research prior to posting. For that I know, I will always be "easy picking" on definitions but always hopefull of preserving the merit of the matter.

I am most glad to see this support for those minicircuits. I know it is said that you can possibly tune the most out of the well designed discrete assembly but I feel that for practical reasons I am better of tunning to the best what is the simple rather than to get lost in complications of what is impracticably perfect.

The KIS (keep it simple) principle is the one that works for me :)

Regards,
xanadunow
 
Last edited:
Man, this diode ring with '914s is just does a great job! I put another filter on it. But I still got problems with that oscillator.
 
Status
Not open for further replies.

Latest threads

New Articles From Microcontroller Tips

Back
Top