Re: Transformer outputs and volume control Q
You need to maintain the balanced signal all the way to the amp so noise is minimized. Your transformer should go in your steel box at the amp end.
Also you're better off using the JT-11p-1.
check this out:
http://www.jensen-transformers.com/as/as089.pdf
The yellow wire (after the damping network) would then go to one leg of your pot. the wiper to the input of the amp and the other leg of the pot to ground. Any pot from say 10k upto 100k will work. The difference being how much control resolution you want and what you have available.
weinstro said:
You need to maintain the balanced signal all the way to the amp so noise is minimized. Your transformer should go in your steel box at the amp end.
Also you're better off using the JT-11p-1.
check this out:
http://www.jensen-transformers.com/as/as089.pdf
The yellow wire (after the damping network) would then go to one leg of your pot. the wiper to the input of the amp and the other leg of the pot to ground. Any pot from say 10k upto 100k will work. The difference being how much control resolution you want and what you have available.
Re: Transformer outputs and volume control Q
I'm not sure what you mean by 6-gang MOP. So I'm going to guess you are talking about a 6-gang motor driven pot.
check here for one that can be built into a remote kit:
http://cgi.ebay.com/ALPS-6-gang-mot...58164QQcmdZViewItemQQ_trksidZp1742.m153.l1262
weinstro said:
I'm not sure what you mean by 6-gang MOP. So I'm going to guess you are talking about a 6-gang motor driven pot.
check here for one that can be built into a remote kit:
http://cgi.ebay.com/ALPS-6-gang-mot...58164QQcmdZViewItemQQ_trksidZp1742.m153.l1262
Re: Re: Transformer outputs and volume control Q
Correct. MOP = Motor Operated Potentiometer.
The link you provided is the one I'm eyeballing. The big worry with this is that Alps specifies a gang-to-gang variance of 5 dB. And that's for the high quality RK27 series. The linked item is the smaller RK16 series, and the spec sheet doesn't provide a gang-to-gang variance figure. 5 dB is a bit more than what I'd like to plan for in a crossover network.
But, it's simple, it doesn't care what the voltage is, and I can walk up to the box and twist a volume knob - I don't like push button volume controls.
ScottGardner said:
Correct. MOP = Motor Operated Potentiometer.
The link you provided is the one I'm eyeballing. The big worry with this is that Alps specifies a gang-to-gang variance of 5 dB. And that's for the high quality RK27 series. The linked item is the smaller RK16 series, and the spec sheet doesn't provide a gang-to-gang variance figure. 5 dB is a bit more than what I'd like to plan for in a crossover network.
But, it's simple, it doesn't care what the voltage is, and I can walk up to the box and twist a volume knob - I don't like push button volume controls.
Re: Re: Transformer outputs and volume control Q
Thanks for the response!
Hmmm. I think I need something that will drop the output from a max of +22dBU to +4dBU, hence the 4:1 stepdown. You differ?
Where I'm really unclear is how the impedance will be seen by the Behringer if the volume control is on the primary or secondary side of this transformer. I'm also unclear if the net input impedance of the amplifier is such that frequency response will be degraded.
There's a wide passage way separating the main speakers, and the amplifiers are mono units. I think I need to mount the volume control in a central location, like where the DCX is. The transformers can be installed at each amp, in which case the volume control is on the transformer primary. Or, the transformers can be installed with the volume control in which case there's freedom where to place the volume control (i.e., primary or secondary side). Shielded twisted pair cable is already installed in wall. Under these conditions, does your suggestion change?
Thanks, again!
Thanks for the response!
ScottGardner said:
Hmmm. I think I need something that will drop the output from a max of +22dBU to +4dBU, hence the 4:1 stepdown. You differ?
Where I'm really unclear is how the impedance will be seen by the Behringer if the volume control is on the primary or secondary side of this transformer. I'm also unclear if the net input impedance of the amplifier is such that frequency response will be degraded.
ScottGardner said:
There's a wide passage way separating the main speakers, and the amplifiers are mono units. I think I need to mount the volume control in a central location, like where the DCX is. The transformers can be installed at each amp, in which case the volume control is on the transformer primary. Or, the transformers can be installed with the volume control in which case there's freedom where to place the volume control (i.e., primary or secondary side). Shielded twisted pair cable is already installed in wall. Under these conditions, does your suggestion change?
Thanks, again!
Re: Re: Re: Transformer outputs and volume control Q
The impedance on the secondary (pot & amp input) transforms to the primary by the square of the turns ratio. You go from 4:1 which is 4 times looking back to the primary. A pot of 10k on the secondary makes the primary load, what the DCX sees, 16 times 10k. No sweat for the DCX.
The pot effective resistance with the amp input capacitance form a low pass filter. The max effective resistance of a 10k pot is 2.5k (when the wiper is at midpoint, because then both halves of 5k are effectively in parallel, seen from the amp). With an input cap of up to 1000pF (which already is quite high), the lp filter sets in at around 66kHz. Again, no sweat.
You may even want to go to a lower pot like 5k or even 2k because that also minimizes noise contribution from the pot, but I wouldn't bother.
Jan Didden
weinstro said:
The impedance on the secondary (pot & amp input) transforms to the primary by the square of the turns ratio. You go from 4:1 which is 4 times looking back to the primary. A pot of 10k on the secondary makes the primary load, what the DCX sees, 16 times 10k. No sweat for the DCX.
The pot effective resistance with the amp input capacitance form a low pass filter. The max effective resistance of a 10k pot is 2.5k (when the wiper is at midpoint, because then both halves of 5k are effectively in parallel, seen from the amp). With an input cap of up to 1000pF (which already is quite high), the lp filter sets in at around 66kHz. Again, no sweat.
You may even want to go to a lower pot like 5k or even 2k because that also minimizes noise contribution from the pot, but I wouldn't bother.
Jan Didden
You are starting to push my level of understanding but Jan says I’m bragging when I say I’m a newbie so I’ll take a shot at answering and hopefully someone will jump in and correct me where/if I’m wrong.
First of all there are problems with putting the volume on the primary. The primary side is balanced so you can’t just attenuate one of the legs. You have to attenuate both legs and this attenuation has to be very precise between the legs otherwise noise will be introduced.
Second the transformer in the circuit is there to convert the balanced signal to a single ended signal. After this conversion has taken place the volume does the level matching (+22db or +4db to -10db and lower).
I understand that in general you want low impendence feeding high impedance.
With the JT-11p-1 you get:
DCX(low -150) -> TransPri(Hi-10K)::TransSec(Hi-10k) -> Amp(Very Hi- 40k)
If you use the JT-10KB-D you have:
DCX(low -150) -> TransPri(VeryHi-40K)::TransSec(Med Low-2.4k) -> Amp(Very Hi- 40k)
So the problem I see is that in second diagram you have a high impedance primary that doesn’t give you any benefit and may actually affect high frequency performance because of your cable length. This is where I’m a bit unclear myself but I’m pretty sure the first diagram is the preferred way to go.
This creates other problems for you though. If you look at the circuit again you will see that the length of cable between the secondary and the amp should not exceed 2 feet. This is why I was thinking you may want to use a remote.
Maybe if you gave a more complete picture of what you are trying to do I/we/someone would be better able to help.
Scott
EDIT: Jan beat me to the post thanks Jan BTW how far off am I?
First of all there are problems with putting the volume on the primary. The primary side is balanced so you can’t just attenuate one of the legs. You have to attenuate both legs and this attenuation has to be very precise between the legs otherwise noise will be introduced.
Second the transformer in the circuit is there to convert the balanced signal to a single ended signal. After this conversion has taken place the volume does the level matching (+22db or +4db to -10db and lower).
I understand that in general you want low impendence feeding high impedance.
With the JT-11p-1 you get:
DCX(low -150) -> TransPri(Hi-10K)::TransSec(Hi-10k) -> Amp(Very Hi- 40k)
If you use the JT-10KB-D you have:
DCX(low -150) -> TransPri(VeryHi-40K)::TransSec(Med Low-2.4k) -> Amp(Very Hi- 40k)
So the problem I see is that in second diagram you have a high impedance primary that doesn’t give you any benefit and may actually affect high frequency performance because of your cable length. This is where I’m a bit unclear myself but I’m pretty sure the first diagram is the preferred way to go.
This creates other problems for you though. If you look at the circuit again you will see that the length of cable between the secondary and the amp should not exceed 2 feet. This is why I was thinking you may want to use a remote.
Maybe if you gave a more complete picture of what you are trying to do I/we/someone would be better able to help.
Scott
EDIT: Jan beat me to the post thanks Jan BTW how far off am I?
Hi Weinstro,
A small remark: A 4:1 only gives you 12dB lower level. You need a 8:1 to loose 18db.
Keep the pot value as low as possible if you are to place it on the send side. This to reduce the possibility of hum and noise pickup. Worst case Rout from a pot is R/2 when the pot is its middle position.
A small remark: A 4:1 only gives you 12dB lower level. You need a 8:1 to loose 18db.
Keep the pot value as low as possible if you are to place it on the send side. This to reduce the possibility of hum and noise pickup. Worst case Rout from a pot is R/2 when the pot is its middle position.
ScottGardner said:
Scott, good point about the cable length, I missed that. If you place the xformer far away, you see the high impedance at the primary in parallel with the relatively high cable capacitance. There may be some hf attenuation you dont want, indeed.
Then again, why not use a 1:1 xformer with a passive 12dB resistor network at the primary? That lowers the impedance at the prim side and negates the cable capacitance largely.
You *can* place a (low value, say 1k) pot at the prim side. Use it as a variable resistor (wiper connected to one end point) fed by two resistors from each DAC output, but it would probably not have a nice control 'feel'.
Jan Didden
First, thanks for your patience, and taking the time to respond.
I promise to avoid asking the obvious questions.
I would place the pot across the +/- outputs of the DCX, rather than from leg to ground. A series resistor needs to be placed in each leg to keep the resistance minimum within spec for the DCX. This approach would also keep signal current out of the ground. I only found one digital pot that could work with the signal levels here, though.
I'm not quite sold on not incorporating level matching with the balanced to unbalanced conversion. As an example, Jensen markets transformers with a 4:1 turns ratio for this purpose. But, I'm listening.
Good things to think about. Regarding the cable, the manufacturer spec for the capacitance is such that I should expect about 2000 pF for the length I have.
The transformers I've sourced are actually rated 10k:600 ohms. Keep in mind that this is really just a turns ratio.
Recall that physically I have speaker/amplifer, an 8 ft. entry, then another speaker/amplifer. I have pulled speaker wire and AES/EBU cable to drive two channels for each corner (4 in total).
Here's how I see my options:
1. DCX(Low -150) -> volume control -> TransPri(Hi-10K)::TransSec(Low -600) -> 75 ft. of cable (low level) -> Amp(Very Hi- 40k)
2. DCX(Low -150) -> TransPri(Hi-10K)::TransSec(Low -600) -> volume control -> 75 ft. of cable (low level) -> Amp(Very Hi- 40k)
3. DCX(Low -150) -> volume control -> 75 ft. of cable (low level) -> TransPri(Hi-10K)::TransSec(Low -600) -> Amp(Very Hi- 40k)
4. DCX(Low -150) -> TransPri(Hi-10K)::TransSec(Low -600) -> volume control -> Amp(Very Hi- 40k)-> 75 ft. of cable (speaker)
I would probably use Option #4 to drive the woofers, to avoid potential saturation issues with the transformers. I also have a K5600 6-channel volume control which uses LM1973's. I could use this in Option #2, and #4 above. The pot values are 40k.
Maybe this can be simplified a bit.
Let's say for the moment that the transformer is gone, the DCX magically outputs a max of 2.5 Vrms, and now the output Z of the DCX is 160/4*4 = 10 ohms (or is it 2560?). I still have a volume control to install, and 2000 pF for the cabling in parallel with the input Z of the amplifier. Where does the volume control go, and what is the optimal value?
This has got to be a common problem that pros deal with.
I suppose that high frequency roll-off could be compensated for in the DCX, too.
ScottGardner said:
I promise to avoid asking the obvious questions.
I would place the pot across the +/- outputs of the DCX, rather than from leg to ground. A series resistor needs to be placed in each leg to keep the resistance minimum within spec for the DCX. This approach would also keep signal current out of the ground. I only found one digital pot that could work with the signal levels here, though.
I'm not quite sold on not incorporating level matching with the balanced to unbalanced conversion. As an example, Jensen markets transformers with a 4:1 turns ratio for this purpose. But, I'm listening.
Good things to think about. Regarding the cable, the manufacturer spec for the capacitance is such that I should expect about 2000 pF for the length I have.
The transformers I've sourced are actually rated 10k:600 ohms. Keep in mind that this is really just a turns ratio.
Recall that physically I have speaker/amplifer, an 8 ft. entry, then another speaker/amplifer. I have pulled speaker wire and AES/EBU cable to drive two channels for each corner (4 in total).
Here's how I see my options:
1. DCX(Low -150) -> volume control -> TransPri(Hi-10K)::TransSec(Low -600) -> 75 ft. of cable (low level) -> Amp(Very Hi- 40k)
2. DCX(Low -150) -> TransPri(Hi-10K)::TransSec(Low -600) -> volume control -> 75 ft. of cable (low level) -> Amp(Very Hi- 40k)
3. DCX(Low -150) -> volume control -> 75 ft. of cable (low level) -> TransPri(Hi-10K)::TransSec(Low -600) -> Amp(Very Hi- 40k)
4. DCX(Low -150) -> TransPri(Hi-10K)::TransSec(Low -600) -> volume control -> Amp(Very Hi- 40k)-> 75 ft. of cable (speaker)
I would probably use Option #4 to drive the woofers, to avoid potential saturation issues with the transformers. I also have a K5600 6-channel volume control which uses LM1973's. I could use this in Option #2, and #4 above. The pot values are 40k.
Maybe this can be simplified a bit.
Let's say for the moment that the transformer is gone, the DCX magically outputs a max of 2.5 Vrms, and now the output Z of the DCX is 160/4*4 = 10 ohms (or is it 2560?). I still have a volume control to install, and 2000 pF for the cabling in parallel with the input Z of the amplifier. Where does the volume control go, and what is the optimal value?
This has got to be a common problem that pros deal with.
I suppose that high frequency roll-off could be compensated for in the DCX, too.
Re: Transformer outputs and volume control Q
Check out Fig. 3.2 on this link:
http://www.jensen-transformers.com/an/an003.pdf
This is what I'm trying to do. I just don't know where the best place to stick the volume control is.
Check out Fig. 3.2 on this link:
http://www.jensen-transformers.com/an/an003.pdf
This is what I'm trying to do. I just don't know where the best place to stick the volume control is.
Re: Re: Transformer outputs and volume control Q
Looks good to me... (Who am I to argue with Mr. Jensen?)
Here is a good thread on balanced attenuators:
http://www.diyaudio.com/forums/showthread.php?threadid=97991
The best place to put the volume is where it solves your problem the way you want it solved. All of the choices you've got should work. A few guidelines are:
Keep the impedance low for long runs.
The simplest choice is usually the best one.
Try to save some money. (It's DIY remember)
weinstro said:
Looks good to me... (Who am I to argue with Mr. Jensen?)
Here is a good thread on balanced attenuators:
http://www.diyaudio.com/forums/showthread.php?threadid=97991
The best place to put the volume is where it solves your problem the way you want it solved. All of the choices you've got should work. A few guidelines are:
Keep the impedance low for long runs.
The simplest choice is usually the best one.
Try to save some money. (It's DIY remember)
this thread is just ...ouch
can't read all those pages ..crazy
neway !!!
Has anyone figured out what can be used
for a multichanel all digital PRE + Xover
in conjunction with the DCX ?
I plan on having a 5-6 channel HT setup
and i have still found nothing that would accept
multichannel digital signal with a few different
format and split it to digital single channels
to feed multiple DCX for XOVER
A consumer HT receiver would be the best,
but i know of none with multiple digital output...
What i have thought of for an all digital signal processing:
COMPUTER/BR/DVD SOURCE ->
DIGITAL SIGNAL SPLITTER ( that can accept
multiple format, refit the multichannel to the X channels required for one specific setup ..and may include volume control if requried ) --- >
multiple DCX ( one for each channel or pair of channel )---->
External standalone super high qualit DAC
( probably 1794 or SABRE powered ) -- >
AMPS
what do you think ?
can't read all those pages ..crazy
neway !!!
Has anyone figured out what can be used
for a multichanel all digital PRE + Xover
in conjunction with the DCX ?
I plan on having a 5-6 channel HT setup
and i have still found nothing that would accept
multichannel digital signal with a few different
format and split it to digital single channels
to feed multiple DCX for XOVER
A consumer HT receiver would be the best,
but i know of none with multiple digital output...
What i have thought of for an all digital signal processing:
COMPUTER/BR/DVD SOURCE ->
DIGITAL SIGNAL SPLITTER ( that can accept
multiple format, refit the multichannel to the X channels required for one specific setup ..and may include volume control if requried ) --- >
multiple DCX ( one for each channel or pair of channel )---->
External standalone super high qualit DAC
( probably 1794 or SABRE powered ) -- >
AMPS
what do you think ?
I have an Echo Audio Gina3g. When I put the Digital output into ADAT mode I get 8 digital channels. The eight channels come out the single optical port. If I use RME's ADI-8 QS I can take those 8 ADAT channels and turn them into 8 (4 pairs) AES/EBU outputs to feed 4 DCX's.
Some of their other stuff puts this all in one package.
Some of their other stuff puts this all in one package.
You could also use this Aphex unit with the Gina3g to get 8 aes digital channels and its cheaper than the RME.
http://www.proaudiosolutions.com/product-p/aphex-144.htm
One other thing to keep in mind: Whatever sound card you choose will need to support windows multichannel surround. All the Echo Audio products do - even in Vista x64.
This unit from Behringer will take ADAT and convert it to 8 analog outs:
http://www.behringer.com/ada8000/
http://www.proaudiosolutions.com/product-p/aphex-144.htm
One other thing to keep in mind: Whatever sound card you choose will need to support windows multichannel surround. All the Echo Audio products do - even in Vista x64.
This unit from Behringer will take ADAT and convert it to 8 analog outs:
http://www.behringer.com/ada8000/
Pulse-R said:
Since you brought this up. What's wrong with digital volume control? Granted if the unit is just stripping bits that would be bad.
But what if the unit was doing 32 bit calculations internally wouldn't that preserve the dynamic range?
I understand that Vista now does floating point calculation for volume in the PC. Which is supposedly lossless.