It's a bit long ago but it does ring a bell. I made a prototype, then fixed a few things, then published the design. I wouldn't be surprised if I managed to publish the schematic of one version and the board of another. I'll snout around in my pile o' files and report back.
edit: I believe I wanted to eliminate one trace jump. I'm an enormous stickler for keeping the ground plane solid and I habitually put stuff on a single layer that would take most people two or three but it does mean I'm constantly swapping pins etc.
edit: I believe I wanted to eliminate one trace jump. I'm an enormous stickler for keeping the ground plane solid and I habitually put stuff on a single layer that would take most people two or three but it does mean I'm constantly swapping pins etc.
Last edited:
Ok, I've seen it: it's at the input relays. They're swapped as well for each channel
Well, if it eliminates vias trace jumps onto the ground plane, might as well
Regarding the 3rd relay, I presume that's also driven from the negative side to balance the loading on the regulators? First two relays are driven from +VOP. But third relay is driven between the full rails with a negative bias from the D5 18V zener? If you've got +/- 12V rails, why drive between full swing and use 18V zener when you could just drive between one rail, ground and clamp with a 6V zener?
Regarding the 3rd relay, I presume that's also driven from the negative side to balance the loading on the regulators? First two relays are driven from +VOP. But third relay is driven between the full rails with a negative bias from the D5 18V zener? If you've got +/- 12V rails, why drive between full swing and use 18V zener when you could just drive between one rail, ground and clamp with a 6V zener?
I think the output chip is specified to drive down to 600ohms. And do that at quite high voltage.
Most Receivers have an input impedance >10k and some are >100k
A pair of 10k loads comes to an effective 5k and that is way above 600ohms.
What is the impedance (and RF capacitance) of the Hypex?
Most Receivers have an input impedance >10k and some are >100k
A pair of 10k loads comes to an effective 5k and that is way above 600ohms.
What is the impedance (and RF capacitance) of the Hypex?
Well, if it eliminates vias trace jumps onto the ground plane, might as well
Regarding the 3rd relay, I presume that's also driven from the negative side to balance the loading on the regulators? First two relays are driven from +VOP. But third relay is driven between the full rails with a negative bias from the D5 18V zener? If you've got +/- 12V rails, why drive between full swing and use 18V zener when you could just drive between one rail, ground and clamp with a 6V zener?
What does K3 switch? Answer that, and you will realize why the different switching arrangement.
Jan
Hi,
Sorry for the poor questions, I know it's a balanced Output AND Input , but is there a possibility to use it with an unbalanced connection at input with for instance a XLRtoRCA adaptator while staying balanced at the output (as JG writted in the beginning of the thread)?
I mean without sacrificing the excellent possibilities of the design ?
Is there the power supply in the BOM to achieve the best performance from Putzey d's design, i.e choosing both the powertraffo with the right VA and typology : R-core, torroid, etc ?
Thank you in advance
best regards
Sorry for the poor questions, I know it's a balanced Output AND Input , but is there a possibility to use it with an unbalanced connection at input with for instance a XLRtoRCA adaptator while staying balanced at the output (as JG writted in the beginning of the thread)?
I mean without sacrificing the excellent possibilities of the design ?
Is there the power supply in the BOM to achieve the best performance from Putzey d's design, i.e choosing both the powertraffo with the right VA and typology : R-core, torroid, etc ?
Thank you in advance
best regards
the B.Putzeys Balanced attenuator is actually a balanced input that is converted to unbalanced to feed through to an unbalanced vol pot, with some optional gain, and that output is converted back to balanced.
If you are feeding in an unbalanaced Source and connection you will end up with
unbal > bal > unbal > bal and this feeds a balanced connection into a balanced receiver.
The first two stages are completely wasted. unbal to bal back to unbal !
If you are feeding in an unbalanaced Source and connection you will end up with
unbal > bal > unbal > bal and this feeds a balanced connection into a balanced receiver.
The first two stages are completely wasted. unbal to bal back to unbal !
Sorry Andrew I do not agree. Don't fall in the trap to think that balanced means two signals with equal magnitude and opposite phase - it is not.
It is about cancelling noise and hum. It works perfect with a SE RCA input - just connect one input to RCA hot, the other to RCA gnd and off you go with all advantages of balanced!
Jan
It is about cancelling noise and hum. It works perfect with a SE RCA input - just connect one input to RCA hot, the other to RCA gnd and off you go with all advantages of balanced!
Jan
Thanks Jan, but I still struggle to understand. Yes, the relay shorts the hot / cold outputs. Is is so that should one of the power rails fail, the outputs are shorted effectively removing any "wacky" output? In other words, the relay open to "unshort" the output only if both voltage rails are OK?
Exactly! And short the outputs immediately at shut-off. The zener takes care of that - the output is shorted before the total power supply has collapsed completely.
Both cases avoid switch on/off plops.
Jan
I cannot fall into that trap. I advise repeatedly to those Members who do fall into that trap. I use the full phrase "balanced impedance connection" to show it is the impedance that is balanced. This has NOTHING to do with equallity of inequality of the signals level on the signal poles.
The advantage of the balanced impedance rejection of interference only applies INSIDE the enclosure of the Balanced impedance attenuator.
The unbalanced source and unbalanced interconnect do not get that benefit.
The input up to the balanced impedance enclosure are limited to the interference rejection of standard unbalanced connection.
But in the meantime there are all the electronics to convert the balanced impedance input to unbalanced impedance to feed into the vol pot.
This section is not needed. The unbalanced source can use an unbalanced interconnect to feed direct into the unbalanced vol pot, using all the conventional unbalanced impedance interference attenuation that works adequately well for our domestic short cables in lowish interference domestic surroundings.
It's long cables in complex set ups in high mains fields of the PA environment that NEED the balanced impedance connections.
There is still the option of using the unbal to bal at the output of the vol pot, if he has a balanced impedance at the input to his Receiver.
Last edited: