[Kevin Haskins]

For those who have read through the shoot-out Danny conducted our amp got simply thrashed in the subjective evaluation.   Since our experience has been decidedly positive with the Hypex UCD modules we where more than a little surprised.    

I withheld public comment on the entire affair.   I figured any comment by me was simply going to be interpreted as either a cover-up or defense of a poor product.   It is the unfortunate side effect of instant communications via the Internet that news travels fast.   A bad comment on a forum often has a negative effect on sales of that product.    A negative result in something as public as this just kills it.

Anyway.... I wanted to point out why I think the results of our amplifier in the event where so poor.    Danny has been kind enough to post this information on his web site but I'm afraid it may be too little too late.   I'll just copy and paste the same information.   For brevity though, the crux of the matter is that the amplifier was hooked up wrong during the evaluation.    Danny and I have determined this with 100% accuracy.   There is no doubt it was hooked up wrong.   There is still some question of how significant the mistake was to the final results, which I will follow-up on once I can make the measurements.

Anyway... without further delay.

Danny,

I’m aghast at how the listening trial came out.    It is so far from the results of our installed user base that I immediately started questioning the results.    

There is an absolutely HUGE installed user base of DIYers using the Hypex amplifiers.   The amp you used was nothing more than a stock UCD-400AD built per the Hypex specifications.    The almost universal feedback we get from these amplifiers, not to mention our personal experience using them at CES, building units for studios and using them ourselves, told us that something was not right.    

My first thought was damage in shipping but then I remembered our phone conversation about cables and it hit me like a ton of bricks.    I realized that you didn’t use our single-ended to balanced adaptor cables during the evaluation.   There is no magic in our cables.   They are a simple twisted pair design with a shield.   What is critical is how you wire the input stage of a balanced amplifier.    When you failed to use our adaptor cables (as per our instructions sent … shame on you! )  for the evaluation you inadvertently introduced noise at the input of the amplifier.    I’m unsure of exactly how much but my guess would be the neighborhood of 10db-20db.   I’ll have the amplifier in question sent back here and do some quick noise measurements with Praxis to illustrate my point.  

Let me explain a little about the issues of adapting a singled ended-source to a balanced load.

The Hypex amps have true balanced input stages.   To use them with single-ended sources you have to tie the inverting input and ground reference of the amplifier together to match the source ground reference.   Normally, you tie pins 1 & 3 of a typical XLR connector together to achieve this.   There are a couple of ways of doing that for single-ended sources.   I’ll leave out the option of using transformers (a great method) for brevity.

The preferred method of converting a single-ended source is to use an adaptor cable that ties the inverting input & the ground reference together on the SOURCE end of the cable.   In this manner you keep ground currents that may travel on the signal ground reference isolated to the source side of the connection.

The second and less desirable option is to tie the analog ground reference and the inverting input together just inside the RCA inputs on the amplifier.   The reason this is less desirable has to do with grounding differences created in ALL unbalanced audio systems.   Currents DO travel through ground references and the less we have flowing through the analog signal reference the better.    Most OEMs do this with a switch on the rear panel that allows the user to select between balanced or single-ended inputs.   The switch effectively ties the ground reference to the inverting input when switched for single-ended inputs.   Our amplifiers do not use such a switch.   We provide XLR shorting plugs which short those pins if customers require the less desirable conversion method.

With all of our demo amplifiers we build adaptor cables that utilize the first method of conversion.   We do this because we want to minimize the chance that someone will have a problem with background noise due to grounding issues and we want to show off our amplifiers in the best possible light.  

In the final analysis when you used your single-ended cables directly into the RCA inputs without the proper XLR shorting plugs you wired the input of the amplifier incorrectly.    The result is very predicable… noise and hence the poor performance in your listening trials.

I’ll get you the Praxis measurements once I get the amplifier back.

Best regards,

Kevin Haskins

[CornellAlum]

I think maybe you read a little too much into the results.  I for one didn't think anything negative nor positive when reading the results.  As I pointed out in the thread, is it really possible that amps using the same exact modules could sound so very different.  Nah!

[RAW]

Well put Kevin.
I know that the issue raised with the AMP is not a issue at all as we used Kevins 180 amp at CES.Why the 180 that was all Kevin had EXTRA as the Amps we had to use for CES, well lets say they were the amps that withdrew from this digital event!!
And we had no issue once we listened to them I walked over to KEVIN and said HELP PLEASE!!!!So we got to use his spare 180 amp he had brought.
Not a single issue with distortion at all.Very,very clean and no grounding issues at all.

We had a few people stunned to say the least when they got a price of the AMP and Pre Amp.Stans pre amp was worth over 2 times the cost of the 180 from DIY cable then last we put in the Audience power conditioner and that was almost 3 times the cost of the DIY amp.
Having comments over and over from the start of CES to the last day we had nothing but positive comments and shock as I said when people listened to Kevins amp.

A very good product for sure Kevin and keep your head up as we know that once people find out about this issue things will change!!

All else fails follow instructions    

[kfr01]

Kevin:

I believe your analysis is correct.  I had at least a handful of people listen to your amplifier when it was here on demo.  Nobody, and I mean nobody, heard any noise floor whatsoever.  It was black on black.  We were using the RCA-to-XLR cables.  

It also should be noted that my front-end isn't nearly as nice as the one danny used ($800 Parasound P3 Preamp, Stock Perp. Tech. P-3A DAC).

Nothing was wrong with the sound.  Indeed, it sounded like a world-class amplifier.  No noise, just black.  

Karl

[Kevin Haskins]

In Danny's defense the instructions sent are fairly vague.   I take full responsibility for the confusion.

We also should have bulletproofed the inputs so that there was no way anyone could hook them up incorrectly.    Both are issues I plan to fix.

[klh]

Has Danny publicly acknowledged an error in the wiring to the amp? I'm sure that would also help in a well deserved recovery... not to mention a rematch!

[Kevin Haskins]

Has Danny publicly acknowledged an error in the wiring to the amp? I'm sure that would also help in a well deserved recovery... not to mention a rematch!

Sure... Danny has posted the contents in the above post on several places on his site.  

Danny did nothing wrong.... it is just one of those things that happen.   I take responsibility for the error.

[CJ Paul]

This is one of those bummer situations.  Kevin is basically in a no-win situation.  Even the way he described the situation, some will still suspect ulterior motives, excuse making etc.  I have zero stake in either company but for all the Nuforce bashers out there that crucified them for pulling out, I hope you can clearly see why now.

[Hank]

I listened to Al's setup at CES and the sound of the electronics was very, very clean and the amp price seemed like quite a bargain to me.

BTW, Al's towers sounded great!

Kevin, you're coming across as a gentleman who is taking responsibility.  I predict you'll bounce back.  Don't "pull out" of future comparisons as some do.  There will always be people who hear differently.  That's why we have a great variety of products and circuit topologies in electronics and speakers.

Keep on truckin' guys.

[arthurs]

I for one hope Kevin has no negative impact as a result of the event.  He's been quick and forthright in pointing out an issue, open and honest about his role in the confusion, and willing to make course corrections to resolve.  Anyone making a buying or eliminating from consideration decision based on one comparison event, now known to have a flaw, shame on you anyway, especially if you weren't even there.....

I agree with Hank, keep participating Kevin, shows you have the nads to let your product be checked out, and I think your follow up to this has been commendable...honesty, integrity, and serious dedication to improvement are reasons to buy from someone

[miklorsmith]

If I were in the market for an amp in this category (I'm a small amp guy), these would be on my Very short list.

[skrivis]

Has Danny publicly acknowledged an error in the wiring to the amp? I'm sure that would also help in a well deserved recovery... not to mention a rematch!

Sure... Danny has posted the contents in the above post on several places on his site.  

Danny did nothing wrong.... it is just one of those things that happen.   I take responsibility for the error.

How much difference in noise level is there between the preferred method and the less satisfactory one? The quote seems to be saying that the 10-20 dB worse is if the ground and minus aren't tied together at all. So how about when you use an RCA-RCA cable with the XLR shorting plug inserted?

[DanWiggins]

There shouldn't be any real difference in noise floor in your case, Skrivis.  The input impedance of the Hypex modules is REALLY high - 100KOhms - and this is where I put most of the "blame".

The high input impedance is great for those who use preamps with high output impedances (like tube-based preamps) or for passive line-stages (where you can have 10-20K output impedance) - it keeps the signal high.

HOWEVER, the downside is that the inputs can float very easily without an input or a shorting bridge.  A 100K pull-down is pretty weak, and will not clamp most noise signals.  And since the Hypex modules are balanced-input only, you need to clamp both the positive and negative inputs or you get extra noise pickup.

Shorting the inputs will definitely kill the noise issues; however, my preferred solution would be for Hypex to make the input impedance 10K (best), or to add an external 12K pull-down in parallel with the inputs to create a basic 10K clamp.  That will lower the noise a good 20 dB over leaving them essentially open as-is.  Of course, that means you can't use some preamps, but there is a price to be paid for everything!

Bottom line - if you're using the shorting plugs, you're good to go.

Dan Wiggins
Adire Audio®

[brj]

Well said, Art, and I'd second his comments.

Kevin, I'm not really in a DIY kinda mode these days, but your continuously positive and informative comments are such that I try to make a point of reading your posts whether they are DIY related or not.  That kind of attitude goes a long way with me, as well as your willingness to solicit uncensored (constructive) feedback without tester restrictions.  When such feedback gets used for real product improvements, I'm even more impressed!  I would welcome the chance to hear your amp and other products again, and hope they make another pass through TX!

(Actually, I wanted to hear it again anyway, as yours was one of the amps that I really wanted to test with and without a power conditioner in the loop.)

[Dr. Krull]

my preferred solution would be for Hypex to make the input impedance 10K (best)

Isn't that how it used to be?

-Krull

[Kevin Haskins]

How much difference in noise level is there between the preferred method and the less satisfactory one? The quote seems to be saying that the 10-20 dB worse is if the ground and minus aren't tied together at all. So how about when you use an RCA-RCA cable with the XLR shorting plug inserted?

It is difficult to know.   It would depend upon the system and how much difference there was in the ground potentials between amp and preamp.

[Kevin Haskins]

Well said, Art, and I'd second his comments.

Kevin, I'm not really in a DIY kinda mode these days, but your continuously positive and informative comments are such that I try to make a point of reading your posts whether they are DIY related or not.  That kind of attitude goes a long way with me, as well as your willingness to solicit uncensored (constructive) feedback without tester restrictions.  When such feedback gets used for real product improvements, I'm even more impressed!  I would welcome the c ...

Thank you and others for the kind words.   I'm just doing my best to get it right.

[StevenACNJ]

"Dan's analysis of the input impedance is good also. For his studio monitors he has changed the input impedance on all of them to 10K. It helps shunt more of the noise to ground if your source can handle the lower input impedance. Most studios use preamps with a solid output stage that handle the 10K input impedance with ease. Only with tube preamps and passive line stages do you run into trouble with the lower input impedances and the 100K is nice to have. As Dan says.... there is no free lunch."

Question

I plan on running five UCD 400AD monoblock amps.

My pre/pro is a Meridian G68ADV with the following impedance specs:

1. analog output impedance - 47 ohm

Based on that, will I run into any noise problems?

Thanks

[Kevin Haskins]

You won't run into any noise issues EVER if you use the modules correctly.   All of this technical discussion shouldn't be taken out of context.    The Hypex modules when used correctly have NO NOISE!   They are dead silent under all operating conditions and if you have background noise either something is wired wrong or you have a system level problem.

I don't want people getting the idea that the bone stock modules have noise problems.   That just isn't the case.

[samplesj]

How much difference in noise level is there between the preferred method and the less satisfactory one? The quote seems to be saying that the 10-20 dB worse is if the ground and minus aren't tied together at all. So how about when you use an RCA-RCA cable with the XLR shorting plug inserted?

All of this depends on how the module is wired up.  If it is tied internally either directly or via a switch then there is no issue.  Its not a module issue, but a construction decision.  Kevin's method avoided an extra switch in the signal path which is a cool idea, but requires a certain style of connection.  Its all in the tradeoffs.

When I first hooked my modules up I just tied straight to the RCA jack and connected the -/gnd wires internally and aside from some hum from the avel transformer it was quiet.  I don't have any noise issues at all.  After I replaced the junky transformer it was dead quiet.

Of course my current 2channel speakers aren't especially efficent, but after I hooked up the transformers as a passive pre I actually cranked the volume all the way up (with paused signal of course) and stood with my ear as close to the tweeter on my A/V-3s (91db) as I could without actually crushing it and heard nothing.  Thats much queiter than I had even heard.  I've had one particularly flawed amp that had audible hissing/buzzing several feet away on ~87db sensitivity speakers.