Prog should be a necessary part of every collection!
Don't understand your comment on bridged connection. The amp has a balanced input whether bridged or not. Can you explain what you meant.
Higher crest factors are generally present in better recordings, orchestral arrangements and such. Lower crest factors occur when the music is dense and or compressed so the demand for current rises. My guess is that even in the case of playing Metallica all day you won't need more than an additional 40VA over the stated recommendation. Any distortion introduced by current starvation on those recordings won't even get noticed.
Differential inputs provide many advantages. The biggest one is that any ground loop is removed from the signal path. This means with the differential input, you get a 80-90 dB attenuation of any hum in the ground loop. The differential input also provides 80-90 dB attenuation EMI coupled onto the input cable. In addition to the electrical advantages, the differential input provides the practical advantage of easy bridging of multiple boards. The differential input can be connected to a single-ended source.
I build all my gear with differential I/O. It measures better and it sounds better.
I'm rather impressed by their line-up. The chips are useful and offer stellar performance. Their precision trimmed on-die resistors are very, very nice. That's the way they can build a differential receiver with transformer-like CMRR.
~Tom
Also a huge one is that you can ground the shield properly to the metal enclosure at the connector without ground loops problems, and not have your hifi remind you every time that your cellphone is accessing the network.
I have never heard "zzzzt-zzt-zzt-zzt" noises coming out of loudspeakers at concerts... that would be quite annoying since probably 10 people in the audience post to facebook every minute these days...
The last statement is the crux of it for me. None of my other components have differential I/O, so the advantages of having it on this amp are (currently) nil.
On the other hand, as you say, the differential input can be connected to a single-ended source. As components in my system get replaced/upgraded/reverse-engineered perhaps some day I will utilize differential I/O. I wouldn't be saving a whole lot of money by omitting the THAT1200, after all.
Perhaps all this talk of the benefits of differential I/O will inspire me to research how to retrofit a differential output on my pre-amp (if that is even possible).
Not necessarily. If you can keep the output and ground potential of the source within the input common mode range of the THAT1200, you can actually get nearly the full 80-90 dB of hum rejection (common-mode rejection) at DC and around 50 dB rejection at 20 kHz (numbers from memory of a measurement I took yesterday). It hinges on that "if". I need to think through this and perhaps come up with a better solution than simply shorting Pin 1 to Pin 3 in the XLR connector the way XLR-RCA adapters work.
That is my point exactly.
Oh, it's possible...
~Tom
Thanks for the tip! I have found AN-003 - I assume you are referring to the method outlined in 2.4 "A Simple Alternative"?
Oh, it's possible...
~Tom
One project always leads to the next. But I think I am getting ahead of myself.
Thanks Andrew! Here's the link to Jensen's AN-003: http://www.jensen-transformers.com/wp-content/uploads/2014/08/an003.pdf
I would say that if your source has single-ended out, build the adapter cable shown in Figure 2.1. With a 0 ohm output impedance, the difference in impedance between the shield and the center conductor will degrade the CMRR performance slightly. This is still orders of magnitude better than omitting the THAT1200, which will give you 0 dB CMRR...
A better solution would be to add a differential output to the source. Figure 2.4 in the Jensen app note would be a simple way to accomplish this. A better way would be to use a THAT1606 or THAT1646.
~Tom
I would imagine that you could fit that in the existing source, assuming you have a suitable power supply available. A small piece of vero/vector board with the recommended external components (decoupling caps, etc.) would be a good way to start.
~Tom
I've been staring at figure 2.1 for a while now and it just isn't making much sense to me at all. Typical shielded coax only has two conductors. At the unbalanced output end, the diagram shows both the signal and ground of the jack connected to the center conductor (short?!) of the cable, but also the ground of the jack is connected to the shield of the cable. How is this even possible? I've shorted the ground and signal conductors of coax before and the result is pretty unpleasant.
Then at the balanced input end, it shows both a hi and lo signal popping out the center conductor. Magic?
P.S. Please forgive my ignorance of electronics. Some stuff is very intuitive for me, but this scenario certainly isn't.
Attachments
Last edited:
Two words: Microphone cable...
It contains two conductors and a shield.
An easy way to build the cable shown in Jensen's app note is to buy an XLR patch cord, lob off the female connector, and solder an RCA connector on the cable following Fig. 2.1 in AN-003.
~Tom
An externally hosted image should be here but it was not working when we last tested it.
It contains two conductors and a shield.
An easy way to build the cable shown in Jensen's app note is to buy an XLR patch cord, lob off the female connector, and solder an RCA connector on the cable following Fig. 2.1 in AN-003.
~Tom