UCD type headphone amplifier

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Ex-Moderator
Joined 2005
classd4sure said:
Remember the razer. That will better allow you to push the speed while keeping transient current draw as small as possible.
What you could use is a comparator with emitter followers, AC coupled to complimentary mosfet drivers.

Hi Chris. Forgive my ignorance but what's the "razer"?

Your idea about capacitvely coupling the comparator to the drivers just sunk in. It is definitely worth looking into. In fact I'm going to mess around in SwCAD right now.


classd4sure said:
There's just no need at all for such a thing, think about it, you don't even need a zobel with UCD.
It'll work the same for just about any given load, within reason.

I know, I was just wondering if he ran it without one. Some people don't know that the UCDs have load independent (within reason, as you stated) frequency response.
 
Poor spewlling, didn't mean to be cryptic, but you had just quoted the guy:

http://en.wikipedia.org/wiki/Occam's_Razor

Yup, you can use the caps as cheapo level shifters, and the followers can drive them good. Really all you need in a low power amp with complimentary outputs. The old SODA was done this way too. If I had some example around I'd post it but sadly, I've been concentrating on dual N for so long that I don't.

Just two caps, one for high side and one for low, drive them in the middle.
 
BWRX said:


Mourip, thanks for letting us know about that. One question though, did you use a loading resistor with the UCD like you had to with the TA2020 amp?

Nope. As someone mentioned I had read that the UCD amps worked well into high impedance loads. That was one reason I decided to try one. I thought that it would be more naturally suited for use with the AKGs. Turns out that was true :)
 
Ex-Moderator
Joined 2005
Time to resurrect this since I'm a little bored at work!

It's been a while since I worked on this but I found an old switcher cad schematic and got down to business. I came up with a slightly different scheme to link the comparator output to the complementary drivers for the N and P channel output fets that looks to work pretty well. The fets right after the comparator provide a little bit of timing delay because the comparator output has to cross the voltage threshold before they turn on.

The only problem I'm having is getting the thing to switch at a much higher frequency. Do I need to introduce more hysteresis somewhere? I thought less hysteresis would make the amp switch faster? Or am I missing something?

The amp also simulates better with a lower ohm load (why is that?). If you decrease the output filter cap you can get the switching frequency to go up, but so does the noise when you do an FFT plot of the output waveform. You can also tweak the value of the feedback cap and see what it does to the gain and the amount of residual riding on the audio waveform.
 

Attachments

  • ucdheadphone.zip
    1.4 KB · Views: 110
Ex-Moderator
Joined 2005
Good question! I meant better in terms of the FFT plot of the output waveform.

How do I idealize the level shift mosfets?

Using the level shifter mosfets to convert the comparator output voltage to control the gate drivers while providing good isolation from the output stages seems like a pretty good implementation to me. It shouldn't have any speed issues for one and current levels are fairly low in the drive circuit components.

It's not so much the performance of the level shifters and gate drive circuitry that I don't understand, I'm just not sure how to obtain a higher switching frequency while keeping realistic component values in the output filter. I know the switcing frequency depends on the vlaue of the components in the output filter, the feedback loop, and the delay between a change in the input to a change in the output but what else am I missing? Any tips or ideas?
 
Simulating for proper feedback values and a good FFT should be done with an idealized gain block or whatever, refer to spiceman, he's even provided a few for it. That gets you much closer to the right values and proper phase measurements and all that jazz.

Idealize the level shift with a set of spice ideal switches. I'd expect it to be instantly faster.

In this case I'm viewing it as self limiting, as you have a limited output current to drive the capacitive inputs of the mosfets with, all the way up to Vth. If you want higher frequency you will need more current capability to get to Vth faster. So why not just do away with the mosfet since you're not working with insane voltages, and drive those gate caps directly with an emitter follower buffer?

I honestly think that's as complicated as it needs to be. Higher speed will come with very good component selection. It does take some seriouse tweaking to get it all to look decent that way though, but so do all the others right?

Cheers
 
I like the aforemensioned idea of using a logic buffer/driver as the output stage. If you look at a chip such as the 74LCX244 or 74LCX16244, octal buffers capable of sourcing/sinking 24mA per pin. Parallel a few pins and you have decent current drive.

Furthermore, if you use a comparator with complementary outputs you could easily form a full bridge, giving benifits of both not needing a split supply, and doubling your output swing from a lower battery supply voltage.

The only cavet on my mind, is that a logic IC isn't designed for such operation and may not perform as expected.


In my mind, simply and cheap. Just my two cents however.
 
Power?

For driving headphones cleanly you don't need much power. In fact generally higher power means higher noise so is counterproductive. An amp capable of a couple of watts into 8 ohms is overkill. Doing a bridge design means you must modify the phones for separate right and left circuits to disconnect the common ground. Using a single ended design you could boost the voltage a bit to get the same effect and not have to deal with the ground issue or greatly increased noise.
Roger
 
BWRX said:
This is a continuation of the idea that originally formed in this thread: http://www.diyaudio.com/forums/showthread.php?postid=945463#post945463

After a little thought, the design could be quite simple. The high side would have a pnp connected to the rail and the inductor, its base connected to the collector of an npn through a resistor. The emitter of the npn would be grounded, and the comparator would drive the base through a resistor. Mirror that for low side (using npn's instead of pnp's and pnp's instead of npn's) and it should be very simple to build. The output stage parts would include 2 pnp's, 2 npn's, and 4 resistors. The comparator would be an IC, and the feedback loop/input parts would include 3 resistors and a capacitor. That could be a bare bones implementation. I'll try to draw up a schematic and maybe come up with some part values a little later tonight.

Everyone feel free to list what they'd like to see from a class d headphone amp!


I would love to see enough juice to drive 300ohm cans, balanced inputs with a balanced output to the cans, no shared ground.
 
mourip said:


I built a UCD180-AD amp specifically to power my AKG K1000 headphones after having such good luck with my Audiodigit 2020 amp. The Audiodigit was powered by SLA batteries. It was absolutely silent and had amazing detail.

The UCD is better. It has a much more relaxed upper end, way better bass, and there is no comparison in complex choral or orchestral passages. Of course the UCD costs 4 times what the Tripath 2020 did to build and has 180 watts instead of 8.

I tried my Sennheiser HD650's with both amps and was disappointed but mainly because the K1000s are so much better sounding to begin with.

AKG K1000s with the UCD are the best sound I have gotten so far as a primarily headphone user. My front end is "PC Audio" (Foobar and iTunes) to Scott Nixon USB DAC to Sowter TVA to Hypex UCD180-AD.

I have the T-amp for sale on EBAY right now and am keeping the UCD.

Hope this helps...




This is what I am interested in. My question now would be whether or not the HD650 sounded great and your K1000s are that much nicer sounding, or did that margin widen when driven by the UcD? I mean, if the HD650 are say, 80% as satifying on a normal head amp as the K1000, would you say that they are 80% as good on the UcD? Also, what did you use as a power supply and what was the rail voltage? Did you use two amp modules to run a balanced signal to your cans? How many watts do you think it was putting out at 8ohms?

Got any pictures to send me?
 
Judging by the length of this thread, my lack of knowledge about electronic design, and the debate of whether or not it is possible, I would love to take a poll of who wants a UcD headphone amp? Then, after generating enough interest, Jan from Hypex may help us out and design one for us. This would end the debate and allow me to get my hands on one instead of just reading about it and not understanding half of what is being said! Those of you here who are religious headphone guys should pipe up about the belief that headphones do not need lots of power. I have read many posts on headfi about people driving Stax and AKG cans off of home amps including Nuforce switching amps and had amazing succes! What I want is not to have to go out and buy new cans just to do it. Senn HD650(and its fore fathers) are arguably the most widely used high end can in the world so the design should keep that in mind. My previous amp had 2 watts/ch of SET power at 30ohms and that was just enough to drive the HD650 from some sources.

Maybe I am barking up the wrong tree, but one dude here is already running his AKG K1000's off of a UcD180 amp with great success. Why not take that ball and run with it. I know most of you guys love to tinker, but I would much rather Hypex tinker for me and I put it together. Opinions?
 
"...from some sources."

That implies that the amp was more than fully capable with other sources. That means your problem was not power, but gain. Unless your amp is clipping, you have enough power. If it is not loud enough, and it isn't clipping, that means you need more gain somewhere in the signal chain.
 
khundude2012 said:

This is what I am interested in. My question now would be whether or not the HD650 sounded great and your K1000s are that much nicer sounding, or did that margin widen when driven by the UcD? I mean, if the HD650 are say, 80% as satifying on a normal head amp as the K1000, would you say that they are 80% as good on the UcD? Also, what did you use as a power supply and what was the rail voltage? Did you use two amp modules to run a balanced signal to your cans? How many watts do you think it was putting out at 8ohms?

Got any pictures to send me?

I think that it might boil down to comparing apples and oranges. I use the HD650s with my gainclone amp for watching DVDs and love them. Nice full clear satifying sound. When I have used the K1000s with the Gainclone amp I did not like it. Ironically, with the UCD the HD650s sound muffled compared to the clear and wonderfully open K1000s.

Using the K1000s with the UCD is basically the best sound I have gotten so far in headphone audio. The detail is remarkable with good bass and a very balanced tonal presentation. If I were to search for any improvement at all it would be to tilt the sound a bit toward the "romantic" side. I would probably do this by adding a tube buffer after the DAC however I am concerned about loosing detail. Right now I am just enjoying the music. Might try the buffer this winter.

I would say that if I had never heard the K1000s I might be happy with the HD650s but for me they just did not seem to be a good fit with my UCD. I love them with the Gainclone.

Sorry I have not taken any photos yet. Might try this weekend...

On another note. I have never understood why folks are concerned about driving phones from a high(er) powered amp. The phones will only draw the power they need. You will never be putting full power into the phones. I think that many good choices for headphone listening are overlooked because of this odd idea off "overpowering" phones. My two cents :)

Best,

Paul
 
Linkwitz Lab has a comparison of various amps including a homebrew UcD 180 here: Distortion

Thanks. Linkwitz is someone who knows what he is doing...

Wow. I wrote that post 12 years ago. That Hypex amp is long gone. It was nice for clarity but was too dry and clinical for my tastes.

I use Linear Tube Audio equipment now. It is designed by David Berning and licensed to LTA. K1000s are long gone also. I still have my HD650s and now have some 800s's...
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.