Fred Forsell passive balanced summing mixer questions

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Bill Kahler

Member
Joined
Nov 24, 2008
Messages
22
In the process of building one of these according to Dave's schematic. 8 ins, 2 outs, balanced.

I noticed a post from a guy who used 10k resistors instead of the 5k value on the schematic. Also he used 220 ohm shunts. Are both of these changes recommended? I will be coming out of a Digi 003 and going into a pair of Seventh Circle pres.

Also, attach ground to chassis or no?

Thanks in advance for the help.

Bill
 
Which mixer are you referring to?  A schematic or thread link would be helpful. 

Dave designed an 8 X 2 low impedance stereo mixer - 5K dual pots but no low resistance shunts. Maybe something else you're looking at.


 
It's the 8 channel passive summing mixer listed here as the 8 channel summing buss:
http://www.forsselltech.com/schematics.shtml

Thanks again.
 
Bump. So I've decided to go ahead with the 10K resistors, figuring this will allow me to use more of my Seventh Circle pres for make up gain. Looking forward to seeing the difference in my mixes through this gizmo.
 
That is Fred Forssell's summing mixer. It likes to see ~10K load on the output.
NYD designed summing mixer  uses mic pre amps for the make up gain.
 
Bill Kahler said:
In the process of building one of these according to Dave's schematic. 8 ins, 2 outs, balanced.

I noticed a post from a guy who used 10k resistors instead of the 5k value on the schematic. Also he used 220 ohm shunts. Are both of these changes recommended? I will be coming out of a Digi 003 and going into a pair of Seventh Circle pres.

Also, attach ground to chassis or no?

Thanks in advance for the help.

Bill



The Hi-Z version of NYDave's stereo mixer - at least the one I have in front of me at the moment - uses:

transformer > 10Klog pot > 10K series iso resistor > pan network 50Kdual/24K slugs > mix resistor network > make up amp > Hi-Z pot(50K i think) > buss amp.

Maybe that's the one you were looking at.  As for 220r shunts - that sounds more like the shunt resistor in a U-pad.

Some discussion here:

http://www.groupdiy.com/index.php?topic=21450.0


 
Bill Kahler said:
It's the 8 channel passive summing mixer listed here as the 8 channel summing buss:
http://www.forsselltech.com/schematics.shtml

Thanks again.
This one is not passive; the first one, listed as " 8 Channel Passive Summing Busses " is passive.
The "8 channel summing buss" has its own virtual-ground summing amps, so a shunt resistor is not needed (in fact not desirable).
 
abbey road d enfer said:
Bill Kahler said:
It's the 8 channel passive summing mixer listed here as the 8 channel summing buss:
http://www.forsselltech.com/schematics.shtml

Thanks again.
This one is not passive; the first one, listed as " 8 Channel Passive Summing Busses " is passive.
The "8 channel summing buss" has its own virtual-ground summing amps, so a shunt resistor is not needed (in fact not desirable).

We may be looking at different schematics but both Forssell buses (bus correctly spelled with one 's') use a passive summing topology with the latter example followed by a non-inverting active buffer/make up gain stage. This is arguably not passive, and six of one half dozen of another compared to the virtual earth topology which as you surely know uses an opamp in the inverting mode (thus the name).

It is a common myth that the make up gain of passive + gain stage is somehow less offensive than a well executed virtual earth sum bus with it's very similar noise gain. On paper the noise gain of a virtual earth is N+1 vs. N for passive plus make up. For more than a small handful of inputs that +1 term is insignificant.

A completely passive sum bus should have no sound character of it's own, so will take on the sound character of the following gain stage (if that is not neutral). If you want a sum bus that sounds like xyz mic pre or gain stage, this is how to get that.
======

Regarding the OP's question, if feeding into a "Seventh Circle pres", which I ASSume is a mic preamp, it will be expecting a "microphone like"  source impedance (150-200 ohms).  There are two issues with using a mic preamp for make up gain. They will be optimized for low source impedance and operating at high gain. Using a mic preamp with higher than microphone source impedance and lower gain will result in higher than expected noise floor from the mic preamp.

One solution is to terminate the passive sum which drops the source impedance and level to make a better match to the high gain stage. It should be apparent that you are throwing away signal needlessly only to boost it back up again (much more than N+1). A simple virtual earth will require less make up gain so in principle deliver better performance.   

Of course if you desire the specific sound character of the Seventh Circle pre, this is how. I would suggest perhaps experimenting with best termination for S/N.

I can't answer your question about ground wiring, without looking at NYDs schematic, it should offer some specific advice. In general chassis would connect to pin 1, and balanced outputs to 2 and 3, if unbalanced outputs signal ground goes to pin 3 directly or through a build out impedance similar to hot leg source impedance. 
 
JR

 
Thanks for all the replies everybody. Here is one of the original threads that started me down this project: http://www.groupdiy.com/index.php?topic=15854.20 and another: http://www.groupdiy.com/index.php?topic=16731.0

In fact it is the Forsell schematic and in another thread NY Dave suggested 10k resistors and the 220 ohm shunts.

Having finished it and done some testing it is true that the noise floor has come up too much and I think I would be better off using 5K resistors as stated in the original Forsell schematic. I did use chassis ground and I read somewhere that it's probably not needed for this design.

I guess the other thing to try would be to clip the shunt resistors and see what kind of difference it makes.

Yes the Seventh Circle pres are for microphones. I ran a mix through the C84s (lost a lot of bass), and the J99s (sounded kind of wooly), and the N72s which were the closest to adding just a bit of character but clean and I hate to say "warmer." I also have a pair of Brent Averill 1272s but they are similarly designed as the N72s.

I am interested in experimenting further because when I first hooked it up, the difference was not subtle at all. In fact the right/left separation and three dimensionality of the mix made me think at first there was some kind of phase issue but in mono it still sounded fine. By the way, I'm mixing a couple singer songwriter projects with sparse tracks and very acoustic.

Hey JR, what do you mean by "A simple virtual earth will require less make up gain so in principle deliver better performance"? 
 
Bump.

So I replaced the 10K resistors with 5K, seems like it still takes the level down too far as the result is noisy. I am planning on trying 2 K and maybe 1K as well.  I am wondering if I am thinking correctly on this since JR said this:

"Regarding the OP's question, if feeding into a "Seventh Circle pres", which I ASSume is a mic preamp, it will be expecting a "microphone like"  source impedance (150-200 ohms).  There are two issues with using a mic preamp for make up gain. They will be optimized for low source impedance and operating at high gain. Using a mic preamp with higher than microphone source impedance and lower gain will result in higher than expected noise floor from the mic preamp."

How does Rolls do this with the Folcrum? They must have the impedance correct. . . .

 
Bill Kahler said:
Hey JR, what do you mean by "A simple virtual earth will require less make up gain so in principle deliver better performance"? 

The virtual earth sum amp, will exhibit a noise gain or n+1 (which I have stated before). Using a termination resistor to pad down the output of a passive sum, to better match a mic preamp, will by definition involve operating at higher gain to make up for the additional loss. For a modest number of inputs a virtual earth sum bus with decent quality opamps, should be lower noise than a good mic preamp operating at high gain.

It is rarely specified what the S/N of mic preamps are at low gain, since they are typically optimized for high gain, so results may vary. 

Sorry if I wasn't clear.

JR
 
Hi JR,

Thanks for the reply. While a fun experiment, I guess for a while I will continue to mix in the box!

Bill
 
Hey J.R. , an ignorant question here but
do you mean one resister to ground right before the amp ?
could one try a trimmer in this position ?

I was planning on trying out a passive sum box but
haven't got to it yet
 
I am not giving practical design advice for a passive mixer... I am speaking of noise relationships wrt to termination etc.

Let someone more enthusiastic promote this stuff...

JR
 

Latest posts

Back
Top