Project Ember

Over time we changed a lot of things along the way and I had to build and test some of the circuits before handing them over to Jeremy.
Most of the experiments were done on the OPA-88 amp I made a while back.


this kept us from 'finalising' the design.

Over time the following steps were made before we thought it could be called a product.

The first design was more or less a Horizon with the output stage changed to the buffer type output that is still in place. (see OPA-88 amp shown above)
Subsequently I added auto-bias.
Changed that circuit slightly in a few small iterations to make it perform better.
At this point the project Ember was more or less ready to launch we figured so hints in that direction were vented.

Next I thought it would be handy if the heater was selected automatically so a jumper was no longer needed and no errors can be made by users any more. It would also be easier to use tubes with no markings on them as the voltage didn't matter.

After that I thought it would be easier if the output R selector was simply done with 1 jumper in 3 positions instead of 2 jumpers that work intuitively 'strange' so meant a redesign.

Later I simplified and improved the auto heater select circuit

Changed the RCA output (pre-amp) voltage so it could be used better as a preamp.

Changed the output device configuration for better performance and used other devices but were hard to get so changed it back.

Added a gain select jumper as some found the gain to be on the side of too high (with a little travel on the volpot a too loud volume)

Changed the output configuration to something similar as we did on the Starlight and hoped to pump out 5W-6W on 32 Ohm this way.

a proto PCB was built.

Had some issues with the protection relay action which didn't switch as intended so made a small change.
The proto had some temperature issues and doubling the output power was not working as we wanted and were plagued by small oscillations.
This meant we fell back on the current output configuration and changed it a bit to deal with temperature issues when using normal 8-pin DIP output devices. In real life situations this worked perfectly and the temperature stayed low. On a test bench it became too hot.
I added thermal protection (which had to be tested thoroughly).
It was decided later on SMD devices were going to be used but meant these devices had to be pre-soldered.
Another change in the layout..

The layout was a constant re-arranging and optimising anyway (as it is with any design)

Indicator LED's were added and the output stage configuration was changed slightly again.

We still had some things we wanted to change but decided the amp is functioning as we wanted it too so it was finally launched.

The switch to the front was another idea that I wanted for a long time but because the DC input is at the back this meant long wires running up and down on the PCB so for optimizing power wire routing decided to keep those wires as short as possible.

Now it is up and running and performing even better than my 'proto' OPA-88 amplifier due to optimised PCB design.

We are planning to offer a linear power supply for the amplifier range (regulated 24V and 48V DC linear power supplies).

The only thing that gets warm (not hot) is the tube.
The output stage (actually the rest of the entire amp) will be room temp.
Perhaps when driving the likes of HE6 e.t.c. it might get warm when driving it loud.

The earlier prototype did not have any cooling and the output devices would get hot in this case.
In the production model the top and bottom layer of the PCB act as big heatsinks which would already be more than enough but Jeremy liked the looks and better cooling action of the small heatsinks on top.

These low temperatures will certainly help in longevity.
One of the main reasons (there were others as well) for it's existence.

The question is not silly.

Even though the OPA551 is rated for 200mA continuous it can easily deliver 300mA peak currents.
Music is all about dynamics so 300mA is max.
BUT this is a peak value and to calculate power you need the RMS value.
300mA / SQRT(2) = 210mA RMS.
It does need sufficient heatsinking though and is the reason we are using the DDPAK version instead of the DIP-8.

To calculate power P= U x I
Since the U = I x R we can also say P = I x I x R

In 32 Ohm this means: 0.21A x 0.21A x 32 = 1.44W
In 50 Ohm this means 0.21A x 0.21A x 50 = 2.25W
of course at a certain point the output voltage will become limiting factor.
so in a 120 Ohm based on the I x I x R we would be having 5.4W output power.
Alas the maximum output voltage is limited to 15V and because P = (U x U)/R this means (15 x 15)/120 = 1.9W

Because the output stage has an isolator resistor added that as a bonus lowers the dissipation when low impedance HP's are used power is slightly lower so just under 2W.

What I really like about all of these amps is how they are designed to get as good a sound as possible out of them for the money. They provide what headphone people want, in that there is some room for us to be able to tweak the configuration so that there's a better match with our headphones. Self-adjusting the bias. Longevity will be even better with less heat.

Also, with plenty of life there it will be able to provide lots of headroom which is really important. My Horizon also provides loads of headroom too and is a little beauty of an amp.

Jeremy has also done a great job, getting these things to the market and now providing us all with different options; either as a kit or pre-built.

Hi Lacey. Welcome to DIYAH.

You may well find Horizons selling second hand because some may want to go for the new Ember.

I currently have a Horizon and it drives the HD650 with absolute ease. It's a good amp with plenty of power as well. I'm currently listening via the Horizon into Beyer DT990. (250 ohms) The amp is at 8 'o clock and has plenty more volume to go - and it's not dead as you raise it. It keeps going to ridiculous levels.

I guess what I'm saying is that you may get a good deal on a Horizon if you want to save the pennies.

The Ember looks really nice as well though and it certainly has some power available for plenty of headroom.

the burning question for me is
will this amp drive the he-6 to a decent level?