[CSMR]

I don't know what the cause of the noise is. Some complex situation inside the amp caused by linking the negative terminals. You can easily confirm that this is the cause by hooking up your speakers without playing anything and linking the terminals with wire. If you get noise then you know what the problem is.

 

[blumenco]

I wish that all headphones were balanced. it would give one more band on a headphone plug.

 

[blumenco]

it was very hard to discern with speakers but I think that that is the problem. I will try some things out tomorrow. going to bed now. I will try and sleep on it. till then, I take no responsibility for injury, death, alienation from significant other from trying this project. It does sound quite good, but this problem does need to be solved in order that optimal quality can be acheived. I do not suggest trying this project until there has been a solution reached. i repeat - this is a work in progress - try at own risk - have fun.

till tomorrow.
Clark

 

[blumenco]

if this ends up requiring that one use balanced headphones - which i have a hunch will, does anyone know a source for those little sennheiser connectoid thingies for the headphones. I may as well make my own cable. this would end up being a little more complecated for the layperson, but nevertheless still incredibly cheap depending on the cable material, etc.

my true motivation with this project is to get a nice cheap headphone amp to use when i am walking out in the wheatfields here watching a sunset or something.

 

[utep10]

I've taken my stock HD600 cable and cut off the 1/4 plug and connected the four wires into the T-amp. It sounds really good, alot of resolution. I do hear the low level hiss that was mentioned before and I'm using batteries.

I have also experiemented with some power resistors hooked across the speaker terminals of the T-amp, in parallel with my HD600s to somewhat simulate a speaker load (not exact). I've used a 10ohm (10W) and a 4ohm (10W). The 10ohm didn't do much for the sound, almost the same as without, but the 4ohm really warmed to sound, rounding off the treble and making the bass somewhat muddy. Something in between should be about right for me.

 

[eweitzman]

Class D amps put out significant amounts of ultrasonic noise. The reference circuit for the TA2024 chip in the SI amp shows a filter between the amp and the speaker connectors. This filter will remove most noise above 20kHz when you connect a 4 or 8 ohm load as instructed by the manufacturer. I assume that the SI amp follows this design.

When you place a larger load (300 ohm HD650 + 150 ohm resistor) on it, the filter's behavior changes and there's a rise in the ultrasonics output until reaching a 10db boosted peak at 53kHz. If you value your phones and ears, I'd recommend against using the SI amp with headphones until someone comes up with definitive proof one way or the other.

There are two threads where I've argued this with several people.

http://www6.head-fi.org/forums/showthread.php?t=103035
http://www6.head-fi.org/forums/showthread.php?t=87285

Please refer to these threads for further info. I don't particularly want to rehash the arguments again, though new arguments are most welcome.

- Eric

 

[CSMR]

Quote:

Originally Posted by utep10 I've taken my stock HD600 cable and cut off the 1/4 plug and connected the four wires into the T-amp. It sounds really good, alot of resolution. I do hear the low level hiss that was mentioned before and I'm using batteries. I have also experiemented with some power resistors hooked across the speaker terminals of the T-amp, in parallel with my HD600s to somewhat simulate a speaker load (not exact). I've used a 10ohm (10W) and a 4ohm (10W). The 10ohm didn't do much for the sound, almost the same as without, but the 4ohm really warmed to sound, rounding off the treble and making the bass somewhat muddy. Something in between should be about right for me.

That's an interesting thought. Would this negate solve eweitzman's problem?
Are you really connecting the HD600 directly? That would give a lot of noise even if you managed to get volume under control.

 

[eweitzman]

Quote:

Originally Posted by CSMR That's an interesting thought. Would this negate solve eweitzman's problem?

It's not my problem, it's yours if you use the SI w/ phones

A parallel load resistor is one of four possible solutions I outlined in a post in one of the other threads. There are tradeoffs involved with this solution: noise from resistor, most of the amp's power getting dumped into the resistor requiring higher volume setting w/ higher distortion, etc.

 

[CSMR]

I see, of course.
Perhaps two impedence-matching circuits (not that I know what they are) would be best. I don't know how complicated that would be. (My ASL UHC adapter uses one of these, but unfortunately with a common ground.)

 

[Iron_Dreamer]

Quote:

Originally Posted by utep10 I've taken my stock HD600 cable and cut off the 1/4 plug and connected the four wires into the T-amp. It sounds really good, alot of resolution. I do hear the low level hiss that was mentioned before and I'm using batteries.

I have done the same and had the same results, albeit with a 12V 1A Ratshack PSU.

 

[eweitzman]

Quote:

Originally Posted by CSMR Perhaps two impedence-matching circuits (not that I know what they are) would be best. I don't know how complicated that would be.

Typically, transformers are used to match impedances when there is power transfer involved.

The alternatives are listed in this post. If you can come up with something else, I'll add it to the list.

 

[blumenco]

looks like the resistors might be the way to go (?) if the right ones could be found. that would definitely make the phones appear more like a speaker. the problem would be coloration, etc. because ALL resistor induct, have noise, etc.

I remember looking at the distortion graph of the t-amp. pretty much there is constant (low) distortion until a certain power output is reached, and then it starts distorting like hell. so loading the circuit would not be a problem unless it significantly interfered with power output (which may or may not happen depending on the ability of this amp to drive current rather than voltage which i think of it as more of a current amp than most).

the question is still raised though, will the resistors help the recombined ground issue (if it is an issue)?

clark

 

[sluggo]

The Tripath amp's outputs are configured in an H-bridge. There is no ground. If you measure the "black" output w/r/t the chassis, you will see that it is quite active. If you connect the two "black" outputs from the Teac together at the common ground connection of a female headphone jack, you are summing the negative outputs of the amp's two channels. From the standpoint of sound and amp reliability, this is a problem.

Since most of the energy in audio is in the low frequencies, and the low frequencies are most often in phase, you probably haven't damaged anything permanently. However, if you play a recording that has lots of high-energy, out-of-phase content, then you will end up with a lot of current through one channel's high-side mosfet and the other channel's low-side mosfet, via the "ground" connection on the headphone jack. Things could blow up.

The right way to drive headphones from a bridged output is to disconect the two "ground" leads from each driver at your headphone's plug so that you end up with 4 wires, the + and - leads from each driver. You then connect the headphones to the amp as you would a pair of speakers (except that you would add an appropriate resistance to each circuit).

Your amp willl thank you for it. Your noise problem should go away as well.

sluggo

ps Memory fades, but I believe there are 4-line 1/4" plug/jack sets which will work for this sort of thing. Don't think I've ever seen an 1/8" equivalent, though.

pps The tripath's output filter is tuned for a 4-6 Ohm load. Anything lower will cause the high frequency response to droop, anything higher will cause it to rise. The impedance of the headphones you're using is (I assume) much higher than 6 Ohms, and this will lead to a rising frequency response above 10k or so. To take care of this, you could parallel a 6 Ohm, 2 Watt resistor with each driver. The combined impedance will still be close to 6 Ohms, and you should have plenty of drive for your headphones.

 

[blumenco]

I will try all of those things. actually my amp is at a freind's house. So i am considering making a pretty straight forward adaptor for the phones. it will consist of a stock sennheiser cable, cut to get four wires, and six ohn resistors across the positive and negatives of the wires (essentially parrallel to the transducers). then i will make some nice little connectoid leads for them and plug them in. then I will bring them to a high frequency mic and see if that 53000 hz is coming through. (i will also look with an oscilliscope, electrically) if it is not, and i think that the sound is acceptable, then i am going to package the design on a post and tell people one of the best and cheapest headphone amp for sennheisers possible.

Thanks.

 

[eweitzman]

Quote:

Originally Posted by blumenco ... then I will bring them to a high frequency mic and see if that 53000 hz is coming through.

Most excellent! Someone with access to a good HF microphone! Please keep us posted about what you find.