Use one single fader to atenuate 5 10mhz 0..5V analog differents signals

[Axxel]

Dear friends.

I want to control the atenuation factor of 5 separate signals with a single potentiometer.

Those signals will vary form 0 to 5Volts max, at a frequency of 10Mhz max.



I thought I will use an operational amplifier in substraction mode, but it will not give me the correct result, because what I really need is a DIVISION.

And a division is made with a voltage divisor, we don't need a op-amp.

Exactly like an AUDIO VOLUME CONTROL, like double potentiometers, here is the same but not double, five-time!

The idea is generate a voltage reference with this pot and use after 5 opeational amplifiers to divide each one of the input signals?

[rstofer]

AD539 analog multiplier?
https://www.analog.com/media/en/training-seminars/tutorials/MT-079.pdf

Something similar?

The ones I use are 4 quadrant and they accept +-10V inputs and give +-10V output.  If the control input is 1V then the output for a +-10v input is just +-1V output.  This is caused by the scaling factor K.  Were it not for K, the output could be saturated fairly easily so they divide the real product by 10.

[Axxel]

AD539 analog multiplier?
https://www.analog.com/media/en/training-seminars/tutorials/MT-079.pdf

Something similar?

The ones I use are 4 quadrant and they accept +-10V inputs and give +-10V output.  If the control input is 1V then the output for a +-10v input is just +-1V output.  This is caused by the scaling factor K.  Were it not for K, the output could be saturated fairly easily so they divide the real product by 10.

Analog devices ICs are GOODS but expensive and too exclusive.

I found quad op-amps circuits with the lm324 configuration and 10mhz bandwith like MC33079  https://www.onsemi.com/pub/Collateral/MC33078-D.PDF and the idea is using a classic op-amp...

[vk6zgo]

5 motorised pots, run from a single controller.

[not1xor1]

Dear friends.

I want to control the atenuation factor of 5 separate signals with a single potentiometer.

Those signals will vary form 0 to 5Volts max, at a frequency of 10Mhz max.

I do not know if there are any parts able to get up to 10 MHz, but you might use digital potentiometers or OTAs
You might even control the signals by PWM or use photoresistors...
just some ideas

[mikerj]

Not a simple problem.  Since this is apparently a DC coupled application, you would need to see if the DC performance of analog multipliers is suitable, they tend to suffer from DC offsets that shift with temperature etc.  High performance analog multipliers don't tend to come cheap.

It's unlikely you'll find a digipot capable of your revised 10MHz bandwidth, there are a few that will manage 5MHz if you go for the lowest end to end resistance parts.  You'd need a micro to control them as well.

CDS cells controlled by LEDs will give headaches getting 10 channels matched (and keeping them matched), and you may need two cells per channel driven differentially to get a useful dynamic range and control law.

Another option would be multiplying current output DACs which have the bandwidth you need, and can provide good DC accuracy.  The downside will be cost, they aren't cheap parts and you'd have to add suitable op-amps to convert the outputs back to voltage, plus a micro to control the DACs.

[macboy]

Multiplying DACs come in voltage output type as well. The reference voltage input would be the signal to be attenuated, and the output becomes that signal, attenuated by the DAC setting.

Another option is five digital potentiometers. A similar device is a PGA or Programmable Gain Amplifier. These are commonly used for volume control of audio. Notice that all of the above (DAC, pot, PGA) require a simple microcontroller circuit to read the master analog potentiometer, then set the five active devices accordingly.

You might run into issues with the 10 MHz bandwidth requirement. It is still possible, but not nearly as easy as 1 MHz or less.  I would definitely use a buffer (op-amp voltage follower, gain = 1) at the input and at the output, for each signal. This will provide both the low impedance source and a high impedance load for the attenuating device, whichever you choose. In addition to reducing distortion, this will also reduce high frequency attenuation (or maybe peaking) due to stray capacitance and other factors.

[mikerj]

Multiplying DACs come in voltage output type as well.

But typically not at the bandwidth required by the OP.

[jhpadjustable]

I would like to know the "why" of this project, and about the signal that needs to be attenuated (sine wave, NRZ, RZ, etc.), and the impedance characteristics of the source and sink, and how important dc preservation is, and what levels of accuracy and precision are required in the attenuation, and all price/availability constraints. In other words, complete requirements. Otherwise, it sounds too much like a homework assignment without showing existing work.

There are many RF attenuators that have flat passbands well in excess of 10MHz, are rated down to dc or close to it, and only cost about $3 each thanks to mass production of mobile handsets. Peregrine Semiconductor and M/A-COM are manufacturers to search for.

[Axxel]

OK. Basically I want to control the level of a RGB Signal. The global luminosity.

I made an error, I don't need 5 wires, only 3 (Vsync and HSync).

Unfortunaly, most of analog video circuits are obsolete, so I want to make a design with sure-value components.

I started this design:



It works! But unfortunately, its TOO much components, repetitive, a big PCB.

Now, i'm going there:

https://www.renesas.com/in/en/www/doc/datasheet/x9c102-103-104-503.pdf

[jhpadjustable]

Thank you.

On that digipot, watch out for the charge pump noise spec. Depending on your video levels you might see some herring-bone effects.

A simple JFET shunt attenuator would be fully analog and tiny, if you don't mind generating a negative supply, but might require hand selection of components to match channels.

[Axxel]

Thank you.

On that digipot, watch out for the charge pump noise spec. Depending on your video levels you might see some herring-bone effects.

A simple JFET shunt attenuator would be fully analog and tiny, if you don't mind generating a negative supply, but might require hand selection of components to match channels.

The rgb signals are form 0 to 5V, just before attacking the NTSC encoder I will pass by a divisor bridge to get the 0 to 0.7V.

I will buy those chips on moday, these are disponible in the shop. I will try this and let's see what about the noise...

Negative, no. Only from 0V to 5V. JFET shunt attenuator I never built one.



https://www.nutsvolts.com/magazine/article/fet_principles_and_circuits_part_2

This should be a great solution, and as you said, FULL ANALOGIC! I will try this configuration too...