Rebuilding my old analog console

[Kevint]

Quote: "Now I'm pretty sure that isn't what you are planning" - LOL! don't be too sure

I just measured the direct output voltage from a 4233 and it's 20V P-P. The power supply rails on the op amp are +/-15V. I was planning of summing 4ea of these signals together. So does that mean I would have a max of 80V? If that is the case what options are there?

[lasermaster1977]

Quote: "Now I'm pretty sure that isn't what you are planning" - LOL! don't be too sure

I just measured the direct output voltage from a 4233 and it's 20V P-P. The power supply rails on the op amp are +/-15V. I was planning of summing 4ea of these signals together. So does that mean I would have a max of 80V? If that is the case what options are there?

Ok, 20V P-P centered around Ground, right? Since the instantaneous phase differences might be unpredictable, you always want to look at worst case, so 4 x 10 = 40V peak...ain't good.
If the train is too big to go through the tunnel, make the train smaller until it fits.

Can you reduce the signal gain of the 4233's so that 4 x nV peak <= 10V peak? Or each 4233 output voltage signal swing is + or - 2V peak.

Its kind of hard to say what the best solution is not knowing the purpose and configuration is of the 4233's.

Are these the MAX4233 op amps. Their working voltage range is 2.7v to 5.5v or +and- 1.35v up to +and- 2.75v, so how are you getting 20V P-P as an output??

Feel free to email me.

[laserist]

Quote: "Now I'm pretty sure that isn't what you are planning" - LOL! don't be too sure

I just measured the direct output voltage from a 4233 and it's 20V P-P. The power supply rails on the op amp are +/-15V. I was planning of summing 4ea of these signals together. So does that mean I would have a max of 80V? If that is the case what options are there?

Build the summing amp with 25% gain. 10k feedback resistor and 40k input resistors.

[Kevint]

The 4423's are my burr brown quadrature oscillators. There running them on +/-15V rails with an output voltage of +/-10V. I'm summing 4 of these together and maybe a couple more signals in the future.

[Kevint]

I could put a resistor voltage divider or a trim pot in the signal path to reduce the signal.

[lasermaster1977]

I could put a resistor voltage divider or a trim pot in the signal path to reduce the signal.

Thanks for that Burr Brown clarification. My dyslexia seems to get worse in my old age.

The key to the possible solutions is understanding the simple relationship of op amp's Rf / Rin ratios. The result can be >1 or <1.

So laserist's suggestion is the best and simplest way to go. So, if you go with the 10K input resistor solution for the four Rin values then Rf would be 2.5K or 2.4K (for 25% attenuation on worst case values for the 4 inputs of 10V, plus or minus) thereby allowing you to leave the 4423's the way they are.

[laserist]

Reducing the feedback resistor value rather than increasing the input resistor values could result in too much load on the output stage of the opamp. (Depending on the opamp...)

[lasermaster1977]

Reducing the feedback resistor value rather than increasing the input resistor values could result in too much load on the output stage of the opamp. (Depending on the opamp...)

Good point. With the TL084 keeping Rf above 2K should be fine with light loads ~10K.

[Kevint]

Using the op amp wins. I will give it a try. Thank you all.

[laserist]

One thing about analog consoles is you tend to modify them over time. For example the output of the CYGN-B summed with the CYGN-A originally. Then it was added to Fixed Cycloid 7 & 9, Spiral & Inverse Spirals, and maybe the clipper. That was just up to when I left in 1984. Why not design giving yourself some headroom...