D
D
designer.zang
New member
J
jimmy77611
New member
Is there an advantage to this setup as opposed to summing in the box and sending out a stereo mixdown? Just curious.
mshilarious
Banned
-4dB output? Summing loss in a ten channel network should be 20 * log ( 1 / 10 ) = -20dB.
D
dintymoore
Guest
Excellent job!
I am designing one and am a bit confused about the summing resistor values. The popular values I've read people recommending are 1K, 2K, 4.7K, 5K and 10K.
There's little info I've found on which I should use, and the pros and cons on the various values.
The main reason I want to do this is so I can use a Summit tube eq on my lead vox track, as well as to get rid of my antiquated DMP11 mixer that runs at 16/44.1.
Q1: As far as the resistors, am I correct in saying that the higher values will give more protection against interaction between the inputs but bring the level down more at the output, thus requiring more makeup gain?
My guess is that no one can tell me what values to use and I will probably mock the thing up and find out that way.
Q2: would adding an effects send pot to some inputs be opening a can of worms?
Q3: I'm planning on using my TC Finalizer for makeup gain. Is there any reason you can think of that you couldn't use a pair of Summit TLA-100A's?
thanks in advance
I am designing one and am a bit confused about the summing resistor values. The popular values I've read people recommending are 1K, 2K, 4.7K, 5K and 10K.
There's little info I've found on which I should use, and the pros and cons on the various values.
The main reason I want to do this is so I can use a Summit tube eq on my lead vox track, as well as to get rid of my antiquated DMP11 mixer that runs at 16/44.1.
Q1: As far as the resistors, am I correct in saying that the higher values will give more protection against interaction between the inputs but bring the level down more at the output, thus requiring more makeup gain?
My guess is that no one can tell me what values to use and I will probably mock the thing up and find out that way.
Q2: would adding an effects send pot to some inputs be opening a can of worms?
Q3: I'm planning on using my TC Finalizer for makeup gain. Is there any reason you can think of that you couldn't use a pair of Summit TLA-100A's?
thanks in advance
mshilarious
Banned
Somewhat. You have to look at the circuit from the point of view of each node on the summing network. It sees a load of its summing resistor + the parallel load of all the other nodes in series with the output impedance of their devices. Practically speaking, you can treat that as the summing resistor value. That sets a low end constraint on the summing resistor value; too low and you'll excessively load your output device. 1K is probably the minimum you want to deal with. If you go low, you need to calculate power dissipation in the resistors: +10dBV is about 3VRMS, that's only 9mW, but at +20dBV, it's 10VRMS for 100mW. So you'd want to make sure you were using 1/8W resistors (not a big deal, but I did see SMT resistors above so worth thinking about).
But you don't want to go too high, because high resistance means more thermal noise. So 1M ohm resistors are probably out.
The next concern is what is the network feeding? If it's a high input impedance makeup gain stage, then no worries, you are done. It actually doesn't change the network loss to change summing resistor values.
But mic amps aren't high input impedance. Now you have to think some more. Tossing a 1K load on the entire network could cause a significant drop, depending on the network size and summing resistor values. If you are switching nodes in and out of the network and you care about keeping levels the same, you want to go small in that case. Or if you simply want to minimize network loss and added noise, you'd also want to go small.
How about a 10K line input impedance? That's just like another node on the network, really.
Some people will add a shunt resistor to a network of a low value, say 100 ohms even. That can greatly increase network loss (which is bad), but it has the effect of stabilizing the network loss irrespective of number of channels connected to the network. It also lowers the output impedance of the network, if that is a concern.
What I tended to do was 3k3-ish for a 4 node network into a mic amp, and twice that for 8 node. That keeps the output impedance of the network below 1K, the output devices lightly loaded, and not too much added network loss from connecting to a mic amp. For larger networks, I don't think there's any point going over 10K.
Bottom line, in the range of values you listed it doesn't really make a lot of difference.
Put the pot and/or insert before the summing resistors, no trouble there. You are just replacing one source device with another, or adding network nodes if you like. You can even simply have a pot in series with the inputs for a blend or level control, it screws with the network loss a little but not enough to worry about. The pot should be the same value or less than the summing resistors if it is directly connecting to the summing network.
Don't know those devices but the primary concern is a) the amount of makeup gain required and b) the input noise of the makeup device. The input noise needs to be less than the output noise of the network (output noise of device plus thermal noise of summing network, less network loss).
D
designer.zang
New member
thanks for the comment.
to jimmy77611,
i have do a rough test with acid pro6 demo song.mixdown via software itself,rewire,analog summing.i think soft mixdown hide some thing.
to mshilarious,
you are very professional!
for this box,the ten channel does not send to one bus.there are four bus for diffent input.switched TRS,unbalance,mono,stereo.
to dintymoore,
Q1:i use 10k 1% smt resistor in the signal and 150k 1/2w film resistor between the tip,sleeve&ring.i am going to add a balance gain stage for a decent output level.
Q2:i think simple is the best.if you have to then do it anyway
for me,i prefer a pot free box.
Q3:i do not know the device you said.is it a gain or a buffer?
i design a small balance +10dB gain+buffer module,i have not test it under the summing box.photo attached.
sch file
to jimmy77611,
i have do a rough test with acid pro6 demo song.mixdown via software itself,rewire,analog summing.i think soft mixdown hide some thing.
to mshilarious,
you are very professional!
for this box,the ten channel does not send to one bus.there are four bus for diffent input.switched TRS,unbalance,mono,stereo.
to dintymoore,
Q1:i use 10k 1% smt resistor in the signal and 150k 1/2w film resistor between the tip,sleeve&ring.i am going to add a balance gain stage for a decent output level.
Q2:i think simple is the best.if you have to then do it anyway
for me,i prefer a pot free box.
Q3:i do not know the device you said.is it a gain or a buffer?
i design a small balance +10dB gain+buffer module,i have not test it under the summing box.photo attached.
sch file
Attachments
D
dintymoore
Guest
mshilarious and designer.zang thank you very much for your responses, all this info is invaluable.
When I made my other summing box, I use 10K resistors, and my new box will likely have 12 inputs, so it sounds like 10K's would be good.
I didn't know that there was an advantage to the resistors having lower wattages, like 1/8W or the 1% SMT's used here.
I might try and find some multi-position swithches as I'd rather have them for the effect sends than pots actually - hi, med , and low and fine tune it on the fx box.
The TC Finalizer is a multi-function mastering tool and among other things provides a gain stage that hopefully will be sufficient.
Thanks
When I made my other summing box, I use 10K resistors, and my new box will likely have 12 inputs, so it sounds like 10K's would be good.
I didn't know that there was an advantage to the resistors having lower wattages, like 1/8W or the 1% SMT's used here.
I might try and find some multi-position swithches as I'd rather have them for the effect sends than pots actually - hi, med , and low and fine tune it on the fx box.
The TC Finalizer is a multi-function mastering tool and among other things provides a gain stage that hopefully will be sufficient.
Thanks
mshilarious
Banned
There is no advantage to smaller watt resistors other than size. In theory, bigger is better, as there will be lower temperature (same heat dissipated over a larger area) and therefore less thermal noise. But unless you grossly undersize a resistor, that is not a practical problem. It's not an issue at all with 10K summing resistors.
Resistor tolerance matters in a balanced summing network but makes no difference in a unbalanced network. A 5% tolerance in an unbalanced network could cause two nodes nominally at unity to be off half a decibel, but practically speaking you are doing a mix so you would compensate by changing the output to the summing network. On the other hand, a 5% tolerance could seriously degrade common mode rejection in a balanced network.
D
dintymoore
Guest
Thanks mshilarious.
Let's say I make a 12-in 2-out summing box, if I use 10K resistors, will I have to use more makeup gain than if I used 5K resistors?
Let's say I make a 12-in 2-out summing box, if I use 10K resistors, will I have to use more makeup gain than if I used 5K resistors?
mshilarious
Banned
Again, depends on the load on the network. If it's a 10K load, less than 1dB difference (I assumed a six node network = 12/2). A 1K load, about 3dB. As the network gets larger, the difference between the two resistor values is smaller (although the total network loss is greater).
D
dintymoore
Guest
Thanks.
It sounds like calculating it on paper would be more of a side track than just mocking it up, so that's what I'll do.
Heck, the guys at NASA calculate stuff on paper, push the button and loose a billion dollar rocket so my risks are a bit less than that.
It sounds like calculating it on paper would be more of a side track than just mocking it up, so that's what I'll do.
Heck, the guys at NASA calculate stuff on paper, push the button and loose a billion dollar rocket so my risks are a bit less than that.
D
designer.zang
New member
i do a test of digital vs analog.i find something.during a vocal mix with a drum or a wooden guitar,the analog show a nature stage,very clean of each voice.
who can explain why the expensive software was defeat by few resistors?
who can explain why the expensive software was defeat by few resistors?
mshilarious
Banned
Clearly it was not defeated by resistors, as the audio probably passed through many resistors on the way in. So whatever flavor the resistors imparted would have already been characteristic of the signal.
More likely, there is the much more significant effects of DAC, makeup gain amplifiers, and ADC.
I used to do a summing cable, resistors built right into a cable. So naturally I looped the cable from DAC to ADC (the loss of a four node network didn't require makeup gain, I could compensate by calibrating the converter). The difference? Noise (thermal, shot, and some induced hum), distortion, and a bit of HF attenuation due to converter filter behavior.
The resulting analog mix was not distinguishable from a converter loop of a digital mix, and I'm willing to bet a straight digital mix with the same level and type of noise and distortion observed added with DSP.
Fishmed
Well-known member
Here are some good theads on the subject:
my-diy-summer-done
Yet another passive summing box!
These are where I got my designs from.
my-diy-summer-done
Yet another passive summing box!
These are where I got my designs from.
D
dintymoore
Guest
Yes, those are good sites.
On this one, what are the resistors I've circled:
for an in a 8 to 2 summing box using 10K summing resistors would you use 10K resistors for them too?
Just for the record, I know a lot of people read threads on "passive preamps" with the same attitude I have about shows about Roswald. I too, think that most of this is nuts and people really have to look at why they're doing it, realize that there never was/will be a "passive preamp", and that as soon as you use makeup gain it's an active mixer.
For me, here's my reasons:
* to use a $3K tube eq on my vocal track
* there's so much gain in my system that I can afford to get rid of my DMP11's and substitute a passive sum box and upgrade from 16/44.1 to 24/48 and probably end up with a few hundred bucks in my pocket profit
* I will be using less gear and have a shorter signal path
* I tried just my synth module and right off the bat it's obviously better
* I wouldn't doubt that what I've read is true, that often times you'll end up with something that is wrong but musically more preferable because of what scientists would call errors in design.
On this one, what are the resistors I've circled:
for an in a 8 to 2 summing box using 10K summing resistors would you use 10K resistors for them too?
Just for the record, I know a lot of people read threads on "passive preamps" with the same attitude I have about shows about Roswald. I too, think that most of this is nuts and people really have to look at why they're doing it, realize that there never was/will be a "passive preamp", and that as soon as you use makeup gain it's an active mixer.
For me, here's my reasons:
* to use a $3K tube eq on my vocal track
* there's so much gain in my system that I can afford to get rid of my DMP11's and substitute a passive sum box and upgrade from 16/44.1 to 24/48 and probably end up with a few hundred bucks in my pocket profit
* I will be using less gear and have a shorter signal path
* I tried just my synth module and right off the bat it's obviously better
* I wouldn't doubt that what I've read is true, that often times you'll end up with something that is wrong but musically more preferable because of what scientists would call errors in design.
mshilarious
Banned
Yes, those are good sites.
On this one, what are the resistors I've circled:
for an in a 8 to 2 summing box using 10K summing resistors would you use 10K resistors for them too?
Just for the record, I know a lot of people read threads on "passive preamps" with the same attitude I have about shows about Roswald. I too, think that most of this is nuts and people really have to look at why they're doing it, realize that there never was/will be a "passive preamp", and that as soon as you use makeup gain it's an active mixer.
For me, here's my reasons:
* to use a $3K tube eq on my vocal track
* there's so much gain in my system that I can afford to get rid of my DMP11's and substitute a passive sum box and upgrade from 16/44.1 to 24/48 and probably end up with a few hundred bucks in my pocket profit
* I will be using less gear and have a shorter signal path
* I tried just my synth module and right off the bat it's obviously better
* I wouldn't doubt that what I've read is true, that often times you'll end up with something that is wrong but musically more preferable because of what scientists would call errors in design.
I wouldn't say it's wrong, but it might be understood for an incorrect reason. If we decide that resistors are good, then we could simply conclude that we need to run our signal through as much resistance as possible. Obviously, that's not a great idea. So we might decide to run our signal through a series of gain stages followed by passive attenuators. On some level, that is what a large console must do. But is that the sole reason people choose to use consoles? I don't think so.
An important topic is whether "passive" summing followed by makeup gain is better than a true active topology, with the summing network feeding an inverting (current) input. The active network doesn't fear any of the source of interactions I have described, and it's what the vast majority of mixers use. If flexibility is called for in configuration of the network, it will probably be selected rather quickly.
Anyway, in analogland if you need to mix, you need resistors. Ain't no way around that, it's what happens afterwards that gets interesting.
Fishmed
Well-known member
The two main reasons why I like the passive design are:
1) I can try various amps for my makeup gain
2) It was cheap and easy to build
D
dintymoore
Guest
Yes, plus you can custom make it for your situation and leave out that 90+% of the mixer that you never use.
To me it's largely about using the least amount of stuff.