Project Ember

The output impedance of the amp can affect some headphones' FR. (Although many amp makers say that amps should be as close to zero as possible)

It's very subtle, but with some of the sharper headphones, it can take the edge off. The AKG's all sound better from 120 ohms out.

It's not night and day, but it's a subtle tweaking that the configurable amps allow you to do in order to match the headphone a little better.

Sometimes, a headphone with a 100 hz hump in the upper bass can have the hump reduced by changing the output impedance.

There is a trade off though - it means that your amp needs to be able to go louder with a higher impedance and many unfortunately run out of steam, so you can lose headroom.

Because the Ember has so much power, you can raise the output impedance in order to calm the treble (in most cases) and still retain enough power for lots of headroom.

Meier Audio demonstrates this on his site where he shows FR graphs for headphones at something like 150 output impedance and then 450 I think it is. (Very high) It's exaggerated, but you can clearly see that the FR changes on Beyers, I seem to remember.

It's a handy thing to have as long as the amp has enough guts. I set my Panda at 120 ohms output but it doesn't seem to have the same power as the Ember.

EDIT: found it here ...... www.meier-audio.homepage.t-online.de/

I think Headfi goes on about some ratio. I think it's better to just use your ears.

The AKG K701 sounded better from the Panda at 120 ohms out. For the T40, I use close to zero. AT's seem better at 120. Now I've found Grados seem to mellow slightly at 120.

Senn HD650 at zero, but funnily enough, I don't mind the HD600 at 120!! I think it reduces a mild mid bass hump - not sure. (Could be my head)

IMO, it's a 'suck it and see' thing. Some headphones seem more sensitive to it while others don't notice.

To work well at 120 ohms though, you MUST have an amp with lots of power for the headroom which can be a trade off if you set a high impedance for the output.

That's where the Ember does really well too.

Oh yes! The Ember goes from 0.1 to 35 to 100, with the input being adjustable from high to low. So for near zero, you might want low input gain and for 100, you might need to set it at higher gain to compensate so that the volume knob works in a 'useful' part of its travel.

Even at 100 ohms output impedance impediments is still plenty of headroom. That's partially why I like it so much.

Nope, there is no golden rule except for IEM's perhaps where the output R of the amplifier needs to be close to 0 Ohm (as is true for a lot of portable headphones).
The output resistance of an amp and the impedance of a headphone do NOT have to 'match' and have no relation in the sense that they are/have to be 'linked'.
They DO affect FR when the impedance of the headphone varies.

headphone A may sound 'best' on low output R and headphone B may sound best at medium output R while headphone C may sound best at a high output R.
For this reason all my designs (except portable ones) have a selectable output resistance so the amp becomes more universal.

The output resistance part is explained HERE.
Scroll down and click on the link: resistance, impedance and other issues

The relation between impedance and power is explained in this article (on the same page) headphone power and amplifiers

Ember has 3 output resistances that can be selected with jumpers.
0.1 Ohm (so very close to 0 Ohm)
35 Ohm (about the same as the Xcans series)
120 Ohm (a not very often used industry standard)

3 steps seems a bit coarse and one would expect the 35 to be 50.
BUT our hearing works in a log kind of way so between 1 and 100 a figure of around 33 is 'in the middle'.
The effect it has (aside from the obvious drop in SPL i.e. volume) depends on a few factors and not just one.


Use your ears to find what suits your headphone(s) and taste.

Some of the older speaker amps with a headphone out could have output resistances up to 600 Ohm.

There are no amplifiers around anymore that go above 120 Ohm and most are veering towards 0 Ohm.
Strangely enough there are only a handfull of amplifiers around that have a selectable output Resistance.

Well here it is my review of Project Ember.
I will admit it was hard writing it because I find it difficult to describe the sounds that I hear.
So in layman's terms here is what I think.

I decided to listen to the Ember purely from memory and I have not been comparing it to any other equipment.
Equipment used:
MF X-Ray CD player into
Project Ember into
Beyerdynamic DT770 or
MF XA200 mono amps into
Acoustic Energy AE1 speakers

FIRST IMPRESSION

The first impression you get when you have a really good look at the Ember is - this is an amplifier that screams quality. Quality components such as the gold plated headphone socket that has a real nice feel to it when you plug your headphones in, Nichicon capacitors used throughout with good quality phono sockets to the rear. The PCB is something else - 3mm thick with a black coating and what looks like gold plated component pads, it is just pure quality.  Also the build quality is second to none and like I said earlier the attention to detail is first class. I love the looks as well, a little miniature Taj Mahal sitting on my shelf glowing away a nice cool blue.
The power supply is a bit like a laptop psu with a little green light telling you it's on.


HEADPHONE LISTENING

One of the first things you notice when listening through headphones is the phenomenal level of detail with no hint of fatigue, you can listen for hours and never get tired. Everything is so musical with an extraordinary insight into the music. I have certainly heard things that I have never heard before. Transient attacks are lightning fast and everything is crystal clear with no colourations.
Also you notice how the music seems to slow right down - I always think this is the sign of a good amp, don't ask me why this happens but I think it is because the amp totally controls the music separating the musicians more than other amps giving you more air around the instruments. The silent passages are inky black. There is plenty of power to drive any headphone easily and I never got past 9 o'clock when I was listening through headphones.

I tweaked with the headphone gain jumpers and listened on all three settings. These are a doddle to use, you just jump them over to the next setting - even with the amp on and your headphones plugged in, keeping the volume pot at the same level to immediately hear the difference.
Low was very good; medium was far too loud and high was ok. Other members will have to try these settings out as well because Low is 0.1Ω - Medium is 35Ω and high is 120Ω gain, which didn't make much sense to me as I was expecting high to be the loudest. All I know is that I preferred listening on the low gain setting for my Beyer DT770's.

I also tweaked with the input gain and again preferred it on the low setting.

The sound through headphones is very similar to the Aune T1.

USED AS A PRE AMP.

I was very disappointed to begin with; the sound was ok but not what I am used to. I then moved the jumpers to by-pass the input caps and was immediately impressed with the transformation, this is more like it - just like I am used to listening through my MFV2. The sound stage opened right up and the sound became more dynamic and natural with a lot deeper and controlled bass. So for my liking just make the amp with the caps already by-passed. Picture below shows the top cover off and the input caps by passed with jumpers.

Listening volume as pre-amp was a little bit higher at about 11 or 12 o'clock.
I did some tube rolling and the auto bias facility is an absolute pleasure to use - just switch the amp off at the rear, swap tubes and switch back on to immediately hear the difference. The stock tube got the hottest but even so you could change the valves without getting burned.
I only have three tubes to play about with but swapping them is child's play with Fran's auto bias. Tube rolling was more revealing through speakers than through headphones but I did try it both ways.

The Stock tube is almost perfect for me; I suppose Frans or Jeremy chose this tube for a reason. The mids are superb with female voices sounding smoother and more natural.
The Russian Harma deep cryo  treated valve below, was a bit bright.

The Jan Philips - seemed flat and lack of dynamics, with a narrow soundstage.

The Russian 6H23N-EB just had the edge with more sparkle at the top end and a lower more controlled bass. So for me the 6H23N-EB would be my choice, having said that I would be extremely happy with the stock valve.

The unit runs super cool apart from the tube but that never gets really hot even after a long listen.
I did notice that if you turn the volume pot up it is very scratchy out of both channels but stops as soon as the required volume is reached.

I also swapped the standard mains lead to the power supply to an OFC cable and again was immediately impressed with the sound, a bit like by-passing the input caps, so to me a good quality mains cable is a must. The power supply light also takes about a minute to go out after you have switched it off.

Some people want more inputs to be added, well not for me - signal in and line out and headphone out is perfect for my use, this is how I use my MFV2.
A add on module or matching input selector unit would cure this - like Jeremy quoted earlier.
I agree with Ian that a gold knob is a must and adds to the overall quality feel and should be offered as an optional extra along with gold plated PCB stand offs & thumb screws.
Some people also want Alloy cases - DON'T DO IT and hide away all those beautiful components and workmanship, I don't know if it is feasible but a solid glass case with adequate ventilation would be perfect so you can still see everything. Just found this picture on RG that was posted not so long back with a CMOY encased in some sort of acrylic? just to give you the sort of idea that I would like to see the Ember encased in.

I also envy the last person to audition this amp as I am sure it will sound even more fabulous with more hours burn in?

CONCLUSION


This is a very revealing amp so make sure your system is up to scratch, if you have any weak links in your system it will soon show it up. Some people report RF interference from say mobile phones etc. I never once heard any interference.
At approx £220 + import taxes fully built and £160 + import taxes for a kit - it is an absolute bargain and insane value for money that will compete with amps costing 10x this value.

So I would now like to thank Jeremy & Frans for giving me the opportunity to listen to this stunning little work of art, you should be very proud of your achievements and I am very confident it will sell very easily. I do not have the funds to buy one at the moment but I am sure there will be some updates coming after all the members have had their say, hopefully then I may be in a better position to build a kit version in some sort of glass enclosure with a linear power supply.


FOR
Stunning Sound Quality
Exceptional build quality
Outstanding value for money
Simple auto bias tube rolling of 6 & 12v valves
Simple headphone gain settings
Output Power to spare
Good frequency response
Easy tweakable settings to suit you
Looks


AGAINST
Open frame
Scratchy volume pot

That's a thorough review and think Jeremy is likely to appreciate.



It's called 'gold wash' and is a very thin layer of 'chemical' gold which is 'deposited' on the pads via electrolysis.
Normally pads are either bare copper (which oxidises) or tinned which also oxidises, gold does not oxidise and only needs to be a very thin layer.
The thicker PCB is a nightmare to do repairs on or experiment with but is nice and sturdy for these open frame devices as it adds some rigidity and looks nice (as you can see the PCB)




It's a Switched Mode Power Supply that is heavily over-dimensioned. The one provided is of good quality with low 'leakage' from mains. A linear power supply is being developed. A good linear power supply isn't going to be very cheap nor cheap in shipping costs. The provided SMPS is a good trade off.


You mentioned 'gain' but the jumpers you refer to are output resistance settings. The higher the value the softer the headphone will sound when the volume control isn't moved. The effect is greater for low impedance headphones than for high impedance headphones.

The low impedance setting is most likely to sound best on the DT770. The bass might become bloating and or more boomy on higher output resistance settings.

The output Resistance setting is not of any influence in pre-amp mode as they are only in the headphone socket's pathway.




The Sunrise and Horizon only had the high-gain setting and some complained about the lack of travel on the volpot. For this reason (and to be more flexible) a gain switch was implemented. The gain is also dependent on the used tube b.t.w.
So depending on circumstances such as output level of the source, used tube, used headphone (efficiency and impedance) and needed 'travel' on the volpot so it adjusts more accurate the gain jumper can be set. A larger diameter volpot knob also helps objectively as well as subjectively. As it is a DIY amp and has a 'standard' volpot one can easily change the knob.


This only happens when the input caps are 'bypassed'. With the input caps in line the volpot won't (should not) make a scratchy sound while adjusting the volume. It is mentioned in the manual and caused by a DC grid current passing through the volpot when the input cap is bypassed. The volpot will not be damaged when bypassed.




Encased in solid acrylic it might be difficult to change jumpers, tubes, tweak components or (heaven forbid) do repairs
Also the heatsinks (responsible for the low running temps) wouldn't work any more.

Thanks for the nice write-up !