Hypex revisited

[leok]

I may have made a mistake and buried this in an old thread that had irrelevent baggage.

For people who ar looking for an amplifier that will provide a natural sound with Klipsch speakers, lack of harshness and plenty of dynamic depth:

I have spent considerable effort in trying to achieve the same: and I don't have a product to sell.

My signature lists several previous efforts. My latest effort has used the Hypex module mentioned below. I find the Hypex approach to be unique in its effectiveness in reducing low power distortion: it is not conventional Class-D. Several off the shelf amps are available using the Hypex modules and for someone looking for very low distortion, natural sound, they should be considered. Following my entry in the older thread:

I finally completed the stereo amp based on the Hypex UcD180. I took all I'd learned from the Tripath experience and put it into the design. I'm satisfied. It blows away anything I've heard. It also convinces me that the key to good sound is low distortion at low power (<.03% THD & noise at 10 mW). Nice to have that clean low power in an amp that will deliver 30W (I kept my design as low power as possible).

I don't have any further plans for audio upgrades, but really wanted to finish that amp. I'm doing a lot of listening now.

Leo

[DrWho]

Of all the off-the-shelf digital amps I've heard, the UcD180 is probably my favorite. I especially like how it doesn't have the rising output impedance in the high frequencies - so your cymbals don't tizz like with the tripath types.

Just make sure you've got yourself a really stiff power supply, and keep your power isolated from between channels.

[leok]

I agree: two Tripath lessons.

In my Hypex amps I use 10,000 uF (4 total) low ESR supply caps with 5uF polypropylene film caps in parallel connected to the amps with 12 gauge wire under 2" in length wire. The design is completely dual mono, with the two amps sharing only the power cord and initial line rf filter.

An interesting thing about the Hypex approach is it IS a global feedback design and so it has supply noise rejection. What's different about the feedback is it doesn't drop off as the signal gets close to perfectly linear. Instead, it produces a relatively constant, high freq. dither around perfectly linear with the error pushed above the audio bandwidth.

Leo

[leok]

There are a few pictures at this site: http://picasaweb.google.com/martha.creedon/Amp#

The amp module on the top shelf contains power caps and the Hypex modules. DC enters one side and audio I/O is on the far side. The AC/DC and softstart module is the larger cabinet on the lower shelf. AC enters one side and DC exits the other. The cable is 7 14awg wires shielded and covered with a plastic mesh. I made no effort to have it look nice because it will eventually be located completely out of site.

[Audio Flynn]

Pictures are great. I do not find it as ugly as some other peoples home brew amps.

An interesting concept. In July and August I would like something in a 2 channel chip amp that runs cool, not that I am inside much that time of year to listen critcally anyway.

Thanks Leo!

[thebes]

I'm suprised Leo, given all the discussions on this subject over the last couple of years that there's been so little interest displayed in your post, particularly given the inroads digital amps are making into the world of Solid State.

I guess us tube guys really have conquered this place.

I do recall us discussing theTripath amp you had modified when you visited a couple of years ago. If I recall correctly, you seemed to think this topology wasn't for everybody, but was particularly suited for the type of music you played.

Please correct me if I'm wrong, and how do you think this stacks up against your tube gear, which I do believe is SET.

[leok]

Hi Marty,

Good to hear from you. The Hypex amp is such an improvement over any of my other amps, I'm not sure what my next move is. The Hypex is currently on the RF-7s replacing the SETs which had been the amp of choice. The Hypex sounds way more realistic and natural. I have yet to hear any distortion at all .. I mean this thing simply doesn't screw up. And there is much more dynamic capability into the lower amplitudes: hall ambience is much more meaningful. I can clearly hear echoes from side and rear walls. Images are 3 dimensional and rock solid. Differences in media are more obvious: records often handle sibilance better (cd isn't distorted, just doesn't seem to be all there .. maybe it's the player), but CDs have much more dynamic capability (again, here I'm referring mostly to dynamics into silence, not high acoustic power). Violins (single and massed) are soooo nice.

I'm going to try the Hypex with the Chorus-IIs next.

Really, I was wondering this morning if I was simply going to stop listening to the other amps entirely .. they just sound so flat and 2-dimensional. I'll keep the Tripath amps for various portable applications.

I don't know why there isn't more interest here except Hypex is not the lowest cost way to go (although it may be the lowest cost way to get such low distortion at low power), and I'm not sure that many people buy my proposal that low distortion at low power is the road to audio bliss. Building a good Hypex application takes a little skill. I think the cheapest commercial application is may 2 grand .. lemme check: CI Audio does make the D-100 .. this might use the same module. With my usual fear of trading low power clarity for higher power I would go for the -100 rather than the -200. Music Direct seems to sell a pair of D-100s for $1599 .. I might pick up a pair if the improvement with the Chorus-II is similar to what happened with the RF-7s.

I had just completed an upgrade to my PP amps and was so happy with them until I completed the Hypex.

Again, good to hear from you and I hope the holiday season goes well.

Leo

[maxg]

Hi Leo,

I did built a Hypex implementation a couple of years ago - but it was a much less expensive implementation than you are refering to and that might well have been the reason for the disappointing sound.

Right now - over here - these RingRex units appear to be all the rage - http://www.6moons.com/audioreviews/kingrex/t20.html - I have heard a few of them - including a side by side with a decent 300B implementation and reckon they might be one of the best value for money options out there right now (assuming reasonably sensitive speakers - which should be a safe assumption on here- right?).

Hae you had a chance to listen to these? Be interested in your opinion.

[leok]

Hello Max,

I wasn't aware of this Tripath implementation. I'll read over the reference and get back. There are a few details I want to check.

Thanks,

Leo

[leok]

Hello Max,

I had found that my preference for amplifiers increased as low power distortion went down. The distortion specs. for the TA 2020 used in the KingRex are almost identical to those of the TA 1101B I used in the P6Ds, so I would expect the performance to be very similar. What I found was the Tripath way out performed almost all ss amps available a few years ago, in the region of audio I care about most: low level detail and timbral accuracy. This made sense in that low power distortion of most ss amps at the time was through the roof. Current ss designs may have improved on this to some degree. The Tripath design still had a little grit, which shows up in the good, but not vanishing .03% distortion (and rising) at 0.5W. This predicts the amp has a noise floor and some audible distortion and listening bears that out. There is also some interaction with Tripath PWM and speaker at high frequencies. This never bothered me, but I did go for an output filter that was optimized for an 8 Ohm tweeter .. maybe that helped.

I moved on to no-feedback tube designs because they handled low level detail better and were even lower in grit (I would presume certain types of distortion) than the Tripath. The Tripath does have an op-amp input stage and the grit may be typical feedback and/or crossover noise.

I picked up a Crown D-45 because its low level distortion was an improvement over Tripath and it offered better damping which I wanted for the Chorus-II. This also worked as specs predicted: damping got rid of the mid bump in the Chorus with no-feedback tubes and the low level distortion was on a par with the tube amps.

I was attracted to the Hypex because the low level distortion was over an order of magnitude better than the Crown and it retained or improved upon the damping at all frequencies. Additional power was a bonus. Again, performance is as the specs predict: big improvement in low level detail, vanishing distortion (I have yet to hear any), and total control over the speaker.

I am interested in what it was about the performance of your Hypex amp that you didn't like.

[leok]

Hello Max,

Did you use a preamp with your Hypex amp? If so, it may have contributed to your disappointment. I have never seen preamp distortion as low as that of the Hypex modules, and if you were using a pre. that's probably what you were hearing, not the Hypex. The hypex input impedance is 100K Ohms so even if your phono pre is tube, a high resistance potentiometer will serve as a lower distortion amplitude controller than an active preamp.

Also, I've now tried the Hypex amps with the Chorus-IIs and the result is just as spectacular as it was with the RF-7s.

Leo

[maxg]

Hi Leo,

Wow - lots of questions. OK - first off I tested the Hypex a couple of years ago - so this is all from memory My main amp is a 200 wpc SS amp driving my 91 dB speakers (own design). I wanted to see how a Hypex implementation (30 wpc? Can't recall - maybe 45) would fare. It couldn't drive the bass properly - the soundstage was reduced and depth reduced.

It was run with a tube pre-amp (The Decware ZTPRE). I do recall it having its own volume control and 2 input selectors and I am fairly sure that I would have tried it in that way too. If memory serves it was better with the tube pre than without.

I think I now see why we have different experiences with the unit however, You are looking at the performance at very low output levels and I am listening to it in what are, for me, real world conditions.

Just to explain that my speakers are 91 db - I sit about 2.5 - 3 meters back and my prefered listening level is around 85 dB RMS at the listening position (rising to 95 to 100 on ocasion). In other words - I probably almost never pull less than a watt from the amps - and sometimes a LOT more.

Further - one of the most important things for me when listening is the ability of the amp to cope with dynamic range - as one might find in a decent orchestral performance and recording. I have many records with a dynamic range in the order of 40 dB (20 up, 20 down). These present dramatic and sudden shifts in power requirements from the amp. I have found that for me any amp with less than 50 wc just cant hack it - and really 70 wpc is a minimum I would now consider.

[leok]

Max,

Thanks for the response. I tend to use such low power I forget a lot of people use a lot more. Certainly if you need the power and it's not there the amp wouldn't work.

Leo