Leslie West
Well-known member
- Joined
- Dec 16, 2004
- Messages
- 101
Oh I forgot to tell you that this summing amp sounds great!...... :grin:
V475 translation + summing network
- Thread starter rafafredd
- Start date
Help Support GroupDIY Audio Forum:
> RE is always listed as 2x5k1 or 2x 10k or 2x1k2, but in the last page it´s really clear that I can feed as many channels I want, so, should the value of RE be the same disreguard the number of inputs used 
?
Yes.
This is an Active summer. The gain from any one input does not depend on how many other inputs are connected.
I too have trouble reading the German (and poor scan), but I'd just try 10K input resistors and omit the optional RG gain resistors. Adjust the 10K up or down to get unity gain. While they give you some gain-range, it clearly is not the main gain-amp or output stage, it should just mix and deliver a signal for further fading and boosting.
I'm sure it performs spectacularly; the religous issue of Passive vs Active is bogus.
BTW: you can probably connect it to a mixbus computed for "passive" and it will work fine.
?Yes.
This is an Active summer. The gain from any one input does not depend on how many other inputs are connected.
I too have trouble reading the German (and poor scan), but I'd just try 10K input resistors and omit the optional RG gain resistors. Adjust the 10K up or down to get unity gain. While they give you some gain-range, it clearly is not the main gain-amp or output stage, it should just mix and deliver a signal for further fading and boosting.
I'm sure it performs spectacularly; the religous issue of Passive vs Active is bogus.
BTW: you can probably connect it to a mixbus computed for "passive" and it will work fine.
OK, thanks to Leslie I get the idea.
It is zero-impedance input. Without Rg, it has gain of "20KΩ", so 2*5K input resistors gives gain of 6dB. This is probably suitable for 1 to 8 inputs. When you have lots of inputs all running at once, the sum is greater than the channels, so either the summer output overloads or the channels have to work at low level. Increasing the summing resistors increases noise. Instead you add RG to reduce summer gain. With RG=short, the gain is "2.5KΩ". One input with 2*5K will come out -12dB, but (assuming 3dB increase for each doubling of live inputs) 64 inputs come out a nice level. Up past 100 inputs it won't sum accurately (you go to sub-summers and a master summer). For around 16 inputs, I'd try RG=~13K. Unless you are a German Engineer, the exact value of RG isn't too critical, but the matching of the two RGs should be good (1%).
If you are not a German broadcaster, you can yank the filter board. Run an ohm-meter across the pins: some will be "shorted" (there is a low-DCR choke across them). Dummy-up a plug with jumpers on those pins. Or leave the card in and carefully tack-solder jumpers on the pins or chokes.
It is zero-impedance input. Without Rg, it has gain of "20KΩ", so 2*5K input resistors gives gain of 6dB. This is probably suitable for 1 to 8 inputs. When you have lots of inputs all running at once, the sum is greater than the channels, so either the summer output overloads or the channels have to work at low level. Increasing the summing resistors increases noise. Instead you add RG to reduce summer gain. With RG=short, the gain is "2.5KΩ". One input with 2*5K will come out -12dB, but (assuming 3dB increase for each doubling of live inputs) 64 inputs come out a nice level. Up past 100 inputs it won't sum accurately (you go to sub-summers and a master summer). For around 16 inputs, I'd try RG=~13K. Unless you are a German Engineer, the exact value of RG isn't too critical, but the matching of the two RGs should be good (1%).
If you are not a German broadcaster, you can yank the filter board. Run an ohm-meter across the pins: some will be "shorted" (there is a low-DCR choke across them). Dummy-up a plug with jumpers on those pins. Or leave the card in and carefully tack-solder jumpers on the pins or chokes.
Rafa, stuff you quoted was related to a bit different subject (putting
trafo before active summing).
I am not a freak of this thing. As well as input trafos used in 672's
can be wired in different way and amp can be arranged without this
freaky feedback across input trafo.
trafo before active summing).
I am not a freak of this thing. As well as input trafos used in 672's
can be wired in different way and amp can be arranged without this
freaky feedback across input trafo.
> this freaky feedback across input trafo.
It is freaky. And I would not try to build one.
But current-driving a transformer has some big advantages. The voltage across the windings is almost zero, so distortion is almost zero. Inductance can be low without hurting bass, leakage can be high without hurting treble, bandwidth can be wide. Because it will work with poor coupling, you can get very high CMRR. The "cost" is that you need a good amplifier, but they are now cheap.
It is freaky. And I would not try to build one.
But current-driving a transformer has some big advantages. The voltage across the windings is almost zero, so distortion is almost zero. Inductance can be low without hurting bass, leakage can be high without hurting treble, bandwidth can be wide. Because it will work with poor coupling, you can get very high CMRR. The "cost" is that you need a good amplifier, but they are now cheap.
rafafredd
Well-known member
NICE, guys. Thanks a lot! I´m now working my mind on a schem for a pan circuit:
I did it based on the NYDAVE notes about passive networks, so I´m not sure it will work, but I guess it will.
I did it based on the NYDAVE notes about passive networks, so I´m not sure it will work, but I guess it will.
Leslie West
Well-known member
- Joined
- Dec 16, 2004
- Messages
- 101
Oh, btw. When you want to build a amp like this, you can order these transformers at http://www.haufe-uebertrager.de/.
Tell them its from an old Neumann amp and give them the Transformer number. You find the number in the schematic....
Tell them its from an old Neumann amp and give them the Transformer number. You find the number in the schematic....
rafafredd
Well-known member
Note that the V475-2C use different trabsformers. I think it´s an update.
Hey,
This looks kinda tasty... I'm thinking mebbe a ext stage of this with the 24chan passive mixer I'm planning (in that old yam mc2404 frame i picked up) could be good...
Raf, can you post the -2C bits still?
cheers! :thumb:
This looks kinda tasty... I'm thinking mebbe a ext stage of this with the 24chan passive mixer I'm planning (in that old yam mc2404 frame i picked up) could be good...
Raf, can you post the -2C bits still?
cheers! :thumb:
Leslie West
Well-known member
- Joined
- Dec 16, 2004
- Messages
- 101
[quote author="rafafredd"]Note that the V475-2C use different trabsformers. I think it´s an update.[/quote]
yes, perhaps some little changes.....but if you have the transformer number it shouldn´t be any problem....
yes, perhaps some little changes.....but if you have the transformer number it shouldn´t be any problem....
rafafredd
Well-known member
I´ll post it. Have to scan first.
What do you think about my panpot network? Will it work ok?
Also, can I use a 2pole swicth for Rg?
What do you think about my panpot network? Will it work ok?
Also, can I use a 2pole swicth for Rg?
NewYorkDave
Well-known member
Rafa, is this what you had in mind?
rafafredd
Well-known member
Yeah... I hope it will work in an active mixer...
NewYorkDave
Well-known member
Yes, it will work fine. Since the active mix buss is "zero ohms", so to speak, just make RL = VR1.
rafafredd
Well-known member
Thanks Dave. I got a little confuse about R5, R6, R7 and R8.
As your schem is for unbalanced input, I let R5 and R6 as 0.5RL but made R7 and R8 as 0.1RL instead of the 0.2RL. Does it looks OK?
Here it is again for the sake of confort:
As your schem is for unbalanced input, I let R5 and R6 as 0.5RL but made R7 and R8 as 0.1RL instead of the 0.2RL. Does it looks OK?
Here it is again for the sake of confort:
NewYorkDave
Well-known member
To do this style of panpot in a balanced mix buss, you would need a four-gang pot. The panpot you drew will not work as you intend.
rafafredd
Well-known member
I hate asking you this, but would you make a drawing, Dave? PLEASE???
rafafredd
Well-known member
shamed bump
Samuel Groner
Well-known member
You need two NYD-panpots for your app. One for "hot" and one for "cold". The four pots should be ganged.
Take the left-hot and left-cold out to your left bus and the right-hot and right-cold to your right bus.
Does this help?
Samuel
Take the left-hot and left-cold out to your left bus and the right-hot and right-cold to your right bus.
Does this help?
Samuel
NewYorkDave
Well-known member
Sorry, this thread dropped off my radar until today.
Unless you can come up with a four-gang pot, it might be a better idea to use a "modified Orban" arrangement for panning, if you insist on the mixer being balanced throughout. The Orban-type panpot uses a single pot section, so it could be modified for balanced use with a dual pot.
But if you can come up with a four-gang pot with decent tracking between sections, go for it, since the "NYD" panpot has half (or less) the loss of the Orban.
Honestly, I've only read a portion of the messages in this thread, so I'll have to go back and review the messages to get an idea of exactly what you want to accomplish.
Unless you can come up with a four-gang pot, it might be a better idea to use a "modified Orban" arrangement for panning, if you insist on the mixer being balanced throughout. The Orban-type panpot uses a single pot section, so it could be modified for balanced use with a dual pot.
But if you can come up with a four-gang pot with decent tracking between sections, go for it, since the "NYD" panpot has half (or less) the loss of the Orban.
Honestly, I've only read a portion of the messages in this thread, so I'll have to go back and review the messages to get an idea of exactly what you want to accomplish.
