• WANTED: Happy members who like to discuss audio and other topics related to our interest. Desire to learn and share knowledge of science required as is 20 years of participation in forums (not all true). Come here to have fun, be ready to be teased and not take online life too seriously. We now measure and review equipment for free! Click here for details.

Tom Christiansen Audio HPA-1 Headphone Amp Review

Azeia

Active Member
Joined
Jun 13, 2019
Messages
123
Likes
290
True of course. For me using XLR for both in and out does offer both advantages, as having the additional power on tap when needed is great. With three gain level settings, there really is no downside for me.
Yeah but the engineer could just have designed an amp with more power on SE headphone out in the first place. Like the Neurochrome HP-1 had enough power from SE to pretty much blow up any headphone driver out there, all from it's SE output, it could drive even the hardest to drive headphones in the world to really loud volumes.

So getting more power out of balanced headphone out is really just a design decision. Personally I prefer an amplifier that does not require me to mod any of my headphones that have fixed cables (or single-entry removable cables), just so I can get that extra power. This is why I prefer @tomchr's design on the HP-1, where you have the XLR headphone out for people who already have those cables and want to use them, but you're not getting shafted on power when using the SE port; I don't like to mod expensive things unless I absolutely have no choice, and I imagine I'm not alone here.
 

Celty

Active Member
Joined
Nov 4, 2019
Messages
251
Likes
161
Yeah but the engineer could just have designed an amp with more power on SE headphone out in the first place. Like the Neurochrome HP-1 had enough power from SE to pretty much blow up any headphone driver out there, all from it's SE output, it could drive even the hardest to drive headphones in the world to really loud volumes.

So getting more power out of balanced headphone out is really just a design decision. Personally I prefer an amplifier that does not require me to mod any of my headphones that have fixed cables (or single-entry removable cables), just so I can get that extra power. This is why I prefer @tomchr's design on the HP-1, where you have the XLR headphone out for people who already have those cables and want to use them, but you're not getting shafted on power when using the SE port; I don't like to mod expensive things unless I absolutely have no choice, and I imagine I'm not alone here.
Ah I was not intending to contrast the designs of the HP-1 (or the HPA-1 being reviewed in this thread) to my Monolith 887, just commenting on the advantages / disadvantages of XLR noted by Amir :). I do agree that the more connection options the better for my use, I love the flexibility.
The Hp-1 does look like an interesting product. Always a good thing to see well designed electronics coming out.
 

tomchr

Technical Expert
Technical Expert
Manufacturer
Joined
Nov 5, 2018
Messages
385
Likes
1,213
Location
Calgary, Canada
Yeah but the engineer could just have designed an amp with more power on SE headphone out in the first place.
Sure. Any engineer could "just" have designed the amp for higher power. This would "just" have meant higher voltage swing and output current, which would likely have required a discrete output stage. Now, of course, the engineer could "just" have designed such an output stage, but it involves greater complexity both electrically and mechanically, and often also requires a bias adjustment. This "just" drives up the assembly cost and manufacturing cost, which the customer will "just" have to pay. See. Nothing is ever "just". There's always a consequence. People wanted lower cost than the Neurochrome HP-1, so I designed the Tom Christiansen Audio HPA-1 to provide nearly the same performance at a lower cost. I did have to give up on output power and 4-pin output to meet the cost target.

Like the Neurochrome HP-1 had enough power from SE to pretty much blow up any headphone driver out there, all from it's SE output, it could drive even the hardest to drive headphones in the world to really loud volumes.
I agree. Yet people wanted 5 W into 50 Ω. The HP-1 could deliver 3 W. Now, any engineer can calculate that the ~2 dB difference between 3 W and 5 W is negligible, but most people aren't engineers, and they've been convinced that they need 5 W.

So getting more power out of balanced headphone out is really just a design decision. Personally I prefer an amplifier that does not require me to mod any of my headphones that have fixed cables (or single-entry removable cables), just so I can get that extra power. This is why I prefer @tomchr's design on the HP-1, where you have the XLR headphone out for people who already have those cables and want to use them, but you're not getting shafted on power when using the SE port; I don't like to mod expensive things unless I absolutely have no choice, and I imagine I'm not alone here.
I certainly agree with you on that. I expect things to work out of the box. I don't have the patience to modify expensive things to make them work well. I'd much rather just buy something that works well and be done. I apply the same philosophy in my designs.

The Hp-1 does look like an interesting product. Always a good thing to see well designed electronics coming out.
Sadly I had to discontinue the Neurochrome HP-1 as it was impossible to mass-produce at a reasonable cost. I've learned a lot between the time I designed the HP-1 and have applied those learnings to the HPA-1.

Tom
 

Azeia

Active Member
Joined
Jun 13, 2019
Messages
123
Likes
290
See. Nothing is ever "just". There's always a consequence. People wanted lower cost than the Neurochrome HP-1, so I designed the Tom Christiansen Audio HPA-1 to provide nearly the same performance at a lower cost. I did have to give up on output power and 4-pin output to meet the cost target.
I wasn't suggesting it was easy, I was mostly agreeing with you and others that "balanced" headphone out seems more of a gimmick than anything else, and that vendors seem all too keen to reinforce myths regarding balanced headphones.

Maybe I just read into things a lot, but when I hear certain people talk about the benefits of balanced headphones (and the amps that drive them), there is almost a sort of hidden implication that it's impossible to get that level of performance, or the higher power output, from an SE headphone out, even if you have a superior amplifier design. It seems to me that a lot of people think there is something inherent in the design of balanced headphone amps that makes them able to deliver those benefits, whereas other amp designs cannot. This is mainly what I was reacting to.
 

JJB70

Major Contributor
Forum Donor
Joined
Aug 17, 2018
Messages
2,022
Likes
4,145
Location
Milton Keynes, England
How much power does an amp need, really?

There are spectacularly inefficient and hard to drive headphones no doubt but they are outliers and even most non-mobile open back designs are pretty easy to drive now.

My DAP has an SE output of 132mW at 32 ohms I think (Shanling M3s), that drives all of my headphones to a perfectly good volume (admittedly none of them are particularly demanding to drive).

My JDS Atom will drive any of them to ear bleeding volume that I physically cannot endure with lots of room to spare.

I think many mobile outputs are a bit anaemic and some mobile amp/DACs are a bit gutless but some of the outputs I see on amps now are way way beyond what people need. Now too much is better than not enough but there is a sensible medium. And remember, you only get one pair of ears, wreck them by pumping 5W per channel of headphone amp power into your ears and you soon won't be doing much listening of any kind.
 

tomchr

Technical Expert
Technical Expert
Manufacturer
Joined
Nov 5, 2018
Messages
385
Likes
1,213
Location
Calgary, Canada
I wasn't suggesting it was easy
I didn't perceive it that way either. What I was trying to say was that a small change in the specs (2 dB higher power) can set off an avalanche of changes that causes the cost to explode. This, in turn, means the customers will have different expectations of your product (it's more expensive now), which sets off another avalanche of feature creep and increased cost.

Or going the other way: Reducing the output power from 3 W to 1 W allowed me to replace four expensive ICs with one expensive IC, which allowed me to offer the HPA-1 at its current price point. There were many other changes - the HPA-1 was designed from scratch for mass-production - but the output power was certainly one of the main cost drivers.

I was mostly agreeing with you and others that "balanced" headphone out seems more of a gimmick than anything else, and that vendors seem all too keen to reinforce myths regarding balanced headphones.
Agree wholeheartedly there. "Our amp is special. Look! It's so special that it has a large-and-in-charge 4-pin connector on the front panel! You should buy our amp." Sadly gimmicks sell, so if/when I offer a more expensive amp, it will need to have a 4-pin output or it won't be perceived as a serious amp worthy of a high price.

Maybe I just read into things a lot, but when I hear certain people talk about the benefits of balanced headphones (and the amps that drive them), there is almost a sort of hidden implication that it's impossible to get that level of performance, or the higher power output, from an SE headphone out, even if you have a superior amplifier design. It seems to me that a lot of people think there is something inherent in the design of balanced headphone amps that makes them able to deliver those benefits, whereas other amp designs cannot.
I don't think that's reading too much into it. It's a well-propagated myth. Myths and stories sell. There are also some designs (tube amps with high distortion, for example) where using a balanced circuit can reduce some of the distortion. Sadly, such designs reduce the even-order harmonics that most find pleasant-sounding while leaving the odd-order harmonics (that most find harsh sounding) intact.
For precision designs, I have yet to find a balanced cancellation scheme that improves performance. I find it to be a much better approach to "just" design for low distortion from the start. With modern opamps, that's less of a challenge than it used to be, but you still have to pay attention to the components used (passives can cause distortion), the PCB layout, the layout of the components within the chassis, etc.

How much power does an amp need, really?
Not much, really. As you point out, except for a few outliers, most headphones require very little power to drive. Thus, you don't need much power. However, people have been told that 5 W (50 Ω) is what they should have so they want that. Some are also looking to buy a "statement amp" or "endgame amp". They currently rock a pair of Sennheiser HD650, but "might at some point in the future" own a pair of headphones that are harder to drive and they don't want to buy another amp if/when that happens. So they aim for the 5 W (50 Ω) number.
It's the same with speaker amps. People have been told that they need 100 W (8 Ω) minimum. So 60 W amps don't sell much. 125 W amps do sell. Interestingly, 250 W amps don't sell either ("that's too much power, 100 W is the sweet spot" seems to be the thinking). Never mind that the amp is unlikely to see peaks much beyond 10 W in actual use, even when cranked. Also never mind that the actual output power of the amp depends on the setting of the volume knob and not on the max power spec of the amp. Humans are not rational critters.

Just to be clear: I'm not implying that consumers are of lower than average intelligence or that they have been brainwashed. I'm merely pointing out that we humans are not as rational in our decision-making as we like to think. To borrow the title from Dan Ariely's book on the subject, we're "Predictably Irrational". That's a good read, by the way. As a vendor of hifi equipment I have to work within this realm of irrationality. It's the perfect combination of psychology and engineering. I love it.

Tom (M.Sc.EE; BA Psych, Honours.)
 
Last edited:
Joined
Jan 11, 2020
Messages
67
Likes
40
@tomchr Thank you for answering my question earlier. I deal with currents issues in my day job, I work on GPU for supercomputers. You would be amazed at the amount of current a low voltage Modern GPU's produces, the more interesting stuff is the very fast onset peak currents, that put a very large stress on a power subsystem. It is also workload driven like Audio Amplifier. The last part I worked on was 13.2B transistors in a 7 nm process.

I was seeing some behavior on some amps with AMP like a current limiter was engaged. Volume gain was not linear more like on S-Curve. I guess I need to take Ocliscope out and fuction generator, and look more formally at there behavior at voltage vs current with headphone.

Well, your engineering work looks first-rate, it reminds me of work that Benson & Two-rock guitar amplifiers develop for my other hobby. I am happy to have my money go to a Boutique builder with passion and love of his craft like you have shown.
 
Last edited:

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
3,677
Likes
7,129
Location
Gold Coast, Queensland, Australia
Sadly gimmicks sell, so if/when I offer a more expensive amp, it will need to have a 4-pin output or it won't be perceived as a serious amp worthy of a high price.
Perhaps that ugly 4-pin will be dead and buried by then and the Pentaconn will be the latest de-riguer inclusion on "serious" HPAs? ;0

PS. Is that Pentaconn a licensing thing? Like MQA for plugs?
 

tomchr

Technical Expert
Technical Expert
Manufacturer
Joined
Nov 5, 2018
Messages
385
Likes
1,213
Location
Calgary, Canada
At least the Pentaconn doesn't blow up the chassis size like the 4-pin XLR does. But we'd need 5-pin inputs as well. Too bad Neutrik or Switchcraft don't make a 5-pin TRRRS connector.

Tom
 

JJB70

Major Contributor
Forum Donor
Joined
Aug 17, 2018
Messages
2,022
Likes
4,145
Location
Milton Keynes, England
I know proprietary formats have issues, but I think if people want a balanced headphone amp I think the pentaconn has a lot going for it. The 2.5mm is a bit small and weedy for my taste (3.5mm is bad enough, let alone 2.5) and the 4 pin XLR too far the other way. The pentaconn seems to be the Goldiloks porridge of balanced headphone jacks.
 

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
3,677
Likes
7,129
Location
Gold Coast, Queensland, Australia
The pentaconn seems to be the Goldiloks porridge of balanced headphone jacks.
I agree.

My first reaction was WTF, until I thought about it, and now I like the plug. May not be fun manually making up leads however...
 
Joined
Jul 4, 2019
Messages
53
Likes
40
At least the Pentaconn doesn't blow up the chassis size like the 4-pin XLR does. But we'd need 5-pin inputs as well. Too bad Neutrik or Switchcraft don't make a 5-pin TRRRS connector.

Tom
Yep, and the original socket is only available from Japan or China for crazy money. There are plenty of cheap options for 4,4mm TRRRS plugs, but no sockets. And for what the original costs over the dubious channels where you can get them, you can probably build a pretty decent discrete output stage.
 
Last edited:

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
3,677
Likes
7,129
Location
Gold Coast, Queensland, Australia
At least the Pentaconn doesn't blow up the chassis size like the 4-pin XLR does. But we'd need 5-pin inputs as well.
My suggestion would be to "future proof" the product with a cut out on the PCB and a header that supports whatever socketed mini PCB and jack the buyer wants. Upgrades could be a nice little earner.
 

restorer-john

Major Contributor
Joined
Mar 1, 2018
Messages
3,677
Likes
7,129
Location
Gold Coast, Queensland, Australia
Yep, and the original socket is only available from Japan or China for crazy money.
If it is indeed a licensing ripoff, that'd be sad, but I really don't know why the balanced crowd doesn't simply use twin 3.5mm TRS, one for left, one for right.
 
Joined
Jul 4, 2019
Messages
53
Likes
40
I obviously meant to say plugs, not jacks in that other post.

Surely if you buy upwards of a thousand, the prices might get better, but I doubt they ever get really attractive at this point. If I remember correctly it is a standard set by the JEITA, chinese companies could manufacture them just like they do with plugs. Even if there are licensing processes involved, they usually don't care I guess. So far the sockets seem to be exclusively manufactured by Nippon DICS. That being said, I couldn't resist buying two of them to deploy them in one of my projects. I'll attach a photo next to some Neutriks.

Edit: Checked TaoBao, sellers there want 250RMB (about 30€ a piece). Now that much I didn't pay (got mine from a Japanese eBay seller) and I'm sure a lot of that price is due to their low availability, but I really have my doubts about this connector becoming more common under these circumstances. Would be interesting to know what the purchasing price for a global distributor would be if they bought the manufacturer MOQ.

http://www.ndics.com/en/products/pentaconn/
IMG_2322.jpeg Pentaconn-Jack-Circuit.jpg
 
Last edited:

Azeia

Active Member
Joined
Jun 13, 2019
Messages
123
Likes
290
Perhaps that ugly 4-pin will be dead and buried by then and the Pentaconn will be the latest de-riguer inclusion on "serious" HPAs? ;0
I think part of the reason the desktop headphone amps use XLR4 is the psychological aspects (as Tom has spoken about many times). For one thing, XLR to a lot of laymen is associated with "better" when used for line out/in, so people see the XLR4 headphone connector and think it's better than other options as well. Another thing however, is that it's easier to recognize at a glance on a picture that an amp supports balanced, whereas when I first saw Sony's TA-ZH1ES in pictures, I thought I was looking at three 2.5mm jacks, one 3.5mm, and one 6.35mm (1/4") jack, and I was extremely confused as to why Sony was using 2.5mm jacks, because I didn't know there was such a thing as pentaconn at the time.

PS. Is that Pentaconn a licensing thing? Like MQA for plugs?
I'm not sure but I wouldn't think so. Licensing usually cannot be enforced unless you have patents or something like the DMCA to attack people who reverse-engineer the format/connector/etc, or you can use trademark law, preventing people from using the name/logo of the format or connector on their products. Copyright law doesn't apply here. With HDMI for instance, it is the logo, and the mandatory requirement of HDCP DRM (which reverse-engineering would violate the DMCA) that results in people having to license it. You could actually make an unlicensed HDMI device however that has no HDCP, and would only work with devices that do not enable content protection, but no one wants to do that because the user won't blame the DRM peddlers, they will blame the company in question for making a product that doesn't work with everything.

MQA as far as I know is guarded by patents; I would imagine the patents aren't super resilient or impressive, but who wants to spend money trying to invalidate those patents in court for such a niche format that is only there to bilk audiophiles out of their money? The current legal system in most countries isn't really equipped to deal with situations where you have frivolous patents on obvious things, but no one is willing to put money into challenging said patents in court. Most patent reviewers are also not well equipped to be reviewing software patents, so a lot of stuff just goes through regardless of how obvious it is. Honestly, software should not even be patentable in the first place as far as I'm concerned.

As for pentaconn, I think it would be extremely difficult to patent something that is basically just a fatter version of the standard 3.5mm jack, with some extra rings. Software patents can be complex and it's easier to confuse people in a court so they don't understand what they're looking at, but it'd be more difficult to convince people that adding rings on a connector or making it "bigger" is somehow novel.

If I had to guess, the name "pentaconn" is probably trademarked, so others would have to call it "4.4mm balanced" or something else. Other than that, I think the only reason we haven't seen more of it is that it's still somewhat more recent, and given that as mentioned earlier, balanced is more of a gimmick, vendors want to show off those big visible XLR4 connectors in product screenshots, so they're not in a hurry to adopt new standards, or go through the trouble of cloning this connector. Some companies may even see balanced as a fad that will run it's course and go away at some point.

At least the Pentaconn doesn't blow up the chassis size like the 4-pin XLR does. But we'd need 5-pin inputs as well. Too bad Neutrik or Switchcraft don't make a 5-pin TRRRS connector.
If chassis size is the issue, wouldn't using two 1/4" TRS balanced connectors make sense for the input, as seen on some pro audio gear?
 
Joined
Jan 11, 2020
Messages
67
Likes
40
The one thing I do like about XLR is a positive locking connector. Even the Mini XLR and LEMO connectors are much nicer to deal with on the Audio capture device when you are dealing with mic's and DI In and Out on the production side.
 

tomchr

Technical Expert
Technical Expert
Manufacturer
Joined
Nov 5, 2018
Messages
385
Likes
1,213
Location
Calgary, Canada
If chassis size is the issue, wouldn't using two 1/4" TRS balanced connectors make sense for the input, as seen on some pro audio gear?
The XLR connectors aren't the only driver of chassis size. The Alps RK271-series "Blue Velvet" also pushes the chassis size. As does mains voltage inside the chassis (clearances are needed).

In the past (with Neurochrome), I've used XLR-TRS combo jacks. Many didn't like them because they weren't "audiophile enough". 3-pin = audiophile. 6-pin that allows support for the same 3-pin connector != audiophile. Go figure. Humans... Not rational critters.

So 3-pin XLR it is. I would also think that many just do a quick scan of the rear panel pictures to see which inputs are supported. I know I do. In those cases, a 3-pin XLR will be easily recognizable as a balanced input, whereas a pair of 1/4" jacks won't be.

Tom
 

Azeia

Active Member
Joined
Jun 13, 2019
Messages
123
Likes
290
The XLR connectors aren't the only driver of chassis size. The Alps RK271-series "Blue Velvet" also pushes the chassis size. As does mains voltage inside the chassis (clearances are needed).
I may have misunderstood something, when you mentioned that it'd be nice if there were a 5-pin input connector as well, to complement pentaconn, I was assuming you meant for rear inputs because saving space on the front by swapping XLR4 wouldn't get you much if you still had large connectors on the rear. I'm a bit confused now.
 

tomchr

Technical Expert
Technical Expert
Manufacturer
Joined
Nov 5, 2018
Messages
385
Likes
1,213
Location
Calgary, Canada
Look. I have to be somewhat responsive to market demands. If the market wants a large connector for the balanced input AND there is a good technical reason for adopting such an input, then I will allow the chassis to grow. Consumers will then have to pay more as the product is now more expensive to produce.

There are many technical advantages (and some marketing advantages) of featuring a balanced input. In addition, many in my market survey indicated that they wanted a balanced input on an amp at this price point. They also wanted the Alps RK271 "Blue Velvet" pot. So I included those. To my surprise, very few indicated that they were interested in a balanced or 4-pin output on an amp at this price point, so away it went.

I am not aware of any technical advantage of a balanced headphone output. There may be some high-distortion topologies where you can gain some advantage on paper, but I don't operate there. I design ultra-low distortion headphone amps. In my designs, the 1/4" output and the 4-pin output (if there is one) will provide the exact same stellar performance.

As I have pointed out before, I am not willing to compromise on performance or cost to support a marketing fad with no technical advantage. In a relatively compact headphone amp such as the HPA-1, adding a 4-pin XLR connector would cause the chassis cost to explode as the chassis would need to be wider. The PCB would need to be wider as well, which impacts both the cost of the raw PCB and its assembly cost. With a more compact connector for the balanced output, it is conceivable that such an output could be added without impacting the cost (too much).

An even-higher-end amp would need to be larger or it won't be perceived as serious (marketing). It also needs to provide higher output power (mostly marketing, some technical advantage in limited use cases) and feature a 4-pin output (marketing). It would need a more expanded set of features (mix of technical advantages and marketing). In such an amp, adding the 4-pin output basically adds the cost of the connector plus a little bit for assembly. Also, consumers expect to find a 4-pin output on such an amp, so it is needed due to marketing demands. It may be a fad, but it won't add much to the cost of the even-higher-end amp, so why not support it. The even-higher-end amp would be more expensive to produce, thus, require a higher price tag than the HPA-1.

A still-good-but-lower-end amp would have the differential input removed. This would allow the chassis to shrink. Unfortunately, this would also mean that the RK271 pot would be gone as it wouldn't fit in the chassis. One could entertain the thought of using an external wall wart for the power supply. That would allow the PCB slots in the chassis to be used (much smaller clearance needed due to the lower voltage in the chassis), which would reduce the chassis production cost. The still-good-but-lower-end amp would be possible to produce and sell at a price point below that of the HPA-1.

Tradeoffs, tradeoffs...

Any real-world engineering problem is a multivariate optimization. It's sorta like pushing on a balloon. You push in one place and it expands elsewhere. There may be an infinite (or at least high) number of good solutions that are all different, and good in different ways. There may be one global optimum (I've never had that happen). Or there may be no solutions. Add all the variables related to market demands/needs, marketing (consumer psychology), and cost, and watch the decision space grow exponentially. You may find that optimizing for engineering (as I did with the Neurochrome HP-1) results in a product that's very far from optimum on the other parameters.
What I'm doing with Tom Christiansen Audio is to strike a good balance between "engineering optimization" and "market optimization". I will openly admit that I'm biased in the "engineering" direction. I'm at peace with that. :) I fully intend to keep my products rooted in science and engineering, but I will have to accept that to sell my products to people other than engineers, I will have to keep an eye on the marketing and consumer psychology aspects as well. I think the HPA-1 strikes a good balance between engineering and marketing. Others are free to disagree.

I hope I've shed some light on what goes into the specifications of an amp. If you're now even more confused, just remember that you're not alone. Multivariate optimization is not for everybody... :)

BTW: Tomorrow I'll increase the preorder price of the HPA-1 to $849. I expect to be able to ship before the end of the month. Once the HPA-1 is in stock, its price will settle at the final $899. There are 13 left at the preorder price. I'll start actively advertising these starting tomorrow as well.

Tom
 
Last edited:
Top Bottom