Can you handle the Beast! Introducing the iDSD Diablo 2!!

[Tano]

If you are asking me, yes I can handle the “fake” specs.

This is just a hobby, I don’t see why I should see iFi’s advertising wording as “so sad”.

I am a grown adult with better things to concern myself with than an advertising spec being exaggerated. I listen to music from my Diablo to relax and remove myself from other stressors, it does a great job, that is all I need, if others want to get all concerned go for it, I just tried to inject some perspective.

Cool. That doesn't mean that it doesn't sound good or that you are cannot enjoy it. But you are paying for features that aren't there.
Marketing of the Diablo 2 is based on power, which is faked.

I think it's good to complain about this practices that are really bad for the hobby.

Anyways, enjoy your music

 

[BS5711]

Cool. That doesn't mean that it doesn't sound good or that you are cannot enjoy it. But you are paying for features that aren't there.
Marketing of the Diablo 2 is based on power, which is faked.

I think it's good to complain about this practices that are really bad for the hobby.

Anyways, enjoy your music

OK, we see it slightly differently but that is fine with me.

I listened to the Diablo before I bought it, I based my assessment of its value to me on its performance not the stated numbers. As such I was paying for what it did not what iFi said it did and I was good with that.

 

[GoldenSound]

So the Balanced Out is rather compromised and is best ignored as marketing feature, UNLESS the headphone's are high impedance and need a LOT of voltage (e.g. AKG K1000), in which case using the Balanced Out of course is the correct choice..

Due to lower noise, distortion and greater available power the SE Output will usually be the better choice both objectively and subjectively, as long as the more limited maximum output voltage (but not reduced Current or Power) is not a problem for the Headphone.

I would suggest that main "performance" measurements should be in SE and that "balanced" are better seen as " supplementary". So I would suggest you re-measure everything in SE, to illustrate to potential customers what is the best connection and operation of this device for maximum performance.

Hey Thorsten

I really appreciate your reply. At your recommendation I have gone back and conducted another set of measurements with the SE output.

It doesn't seem that the Diablo 2 provides more power on SE, the available power is about half that of the balanced output and will start to clip around 0.8W @ 32 Ohm:

For 4V output into a 32 Ohm load though the SE output does indeed perform better than the 4.4mm/Bal output. About 6dB if going solely on THD+N (4.4mm output was -94dB THD+N). So it seems that whilst less max power is available, if you are remaining under the max threshold you are better off using SE.

I'll update the post shortly and also include a full report with comprehensive tests on the SE output to compare to the 4.4mm output. (Though note I did have to turn the gain up to med to get to 4V, still performs better however).

Of course, it will still not come close to 5W RMS per channel, if measured SE.

And just to with, the difference in maximum sound pressure level with a sustained tone is around 5dB, it will be less with Music which has a greater peak to average rating.

Personally I would suggest that using AES-2/2012 is an appropriate means to measure both Speaker "Music Power" handling (which is what it specified for) and Amplifier "Music Power" output in a way that is more relevant in practice and accounts for the dynamic nature of music.

But it should be also referenced against "continuous power", aka RMS, but that's just me.

Thor

Yeah my issue isn't really with which particular power measurements are used. There are lots of valid reasons as to why alternative measurements might be more applicable to real world usage, and I personally appreciate when manufacturers provide extra info (I used the HM1 as an example as they provide industry standard RMS power, peak power, AND also a 1.5ms 'transient power' spec, so there's comprehensive info.)

The issue is simply that if not doing things in the same way that almost everyone else in the industry is, you need to make that clear to the people buying your product, otherwise people will understandably assume there's nothing different and that if your number is higher than others, they will reasonably assume your amp is more powerful even if that is actually not the case.

If a laptop manufacturer simply said their laptop was "4x faster than competitors!" with no further info, then people would assume that the product is 4x faster, and would likely be annoyed if they found out it was 4x faster.....for half a second and then it had to throttle down to a level that was half as powerful as competitors the rest of the time.

 

[Thorsten Loesch]

It doesn't seem that the Diablo 2 provides more power on SE, the available power is about half that of the balanced output and will start to clip around 0.8W @ 32 Ohm:

This is really odd. In SE the second (BAL -) output should be unloaded and only draw quiescent current. Thus all power should be available to the first (SE) output.

With the balanced output we are (for any given halve wave) drawing current on both positive and negative rail simultaneously.

This means our limited power supply current (~ 12W) is distributed across two channels and two rails, so each rail and channel has around 3W to work with. At the highest rail voltage (~ +/-16V) that is 0.1875A, so we expect 1.125W power from this.

Disconnecting the negative side and connecting the load to ground instead means only one rail will draw current for each halfwave, so current available is (theoretically) doubled to 0.375A which in turn should allow ~ 11V into 32 Ohm for ~ 3.8W output (output is voltage limited).

It's a while back, but I would have tested this on the original Red Label (Diablo) Prototypes for sure.

I wonder what is up there? It seems the Amplifier now wastes nearly 3/4 of the power that should be available somewhere? They are for sure not being send into the load.

Also note that the gain switch does not adjust gain, it also adjust the power rail voltages with (IIRC) +/- 6V for "eco" (now renamed to "normal" which incidentally has the HP Amp at unity gain), +/-9V for "normal" (now renamed to "Turbo" with 9dB gain) and +/-16V for "turbo" (now renamed "Nitro" with 18dB gain).

All these are from memory and on the prototypes, it is possible I misremember slightly or there were late changes going into production after I left.

That is just wrong. With the Power level at "Nitro", you should get > 11V. Are you measuring in "Turbo"? That would limit maximum output voltage to around 5.5V. Which seems in line with your measurements.

For 4V output into a 32 Ohm load though the SE output does indeed perform better than the 4.4mm/Bal output. About 6dB if going solely on THD+N (4.4mm output was -94dB THD+N). So it seems that whilst less max power is available, if you are remaining under the max threshold you are better off using SE.

Yes. As said, to make the balanced output bodged onto a single ended design that was already near finalising, the only option was double the Amplifier circuitry without further changes to the rest and use the second amplifier inverting from the output of the first amplifier. This invariably doubles distortion and increases noise.

The issue is simply that if not doing things in the same way that almost everyone else in the industry is, you need to make that clear to the people buying your product, otherwise people will understandably assume there's nothing different and that if your number is higher than others, they will reasonably assume your amp is more powerful even if that is actually not the case.

I could not agree more. Whenever I provided test results to S&M I made sure to be very clear what was measured and how.

What made it afterwards into the official spec and if this was edited is another story.

Thor

 

[GoldenSound]

That is just wrong. With the Power level at "Nitro", you should get > 11V. Are you measuring in "Turbo"? That would limit maximum output voltage to around 5.5V. Which seems in line with your measurements.

That's interesting, yep in high gain it does indeed go higher. I hadn't checked in high gain because in balanced/4.4mm the gain did not alter the power limit, it was the same in both high and medium. So it seems that the higher 'gain' setting does provide more power but only when on SE?

 

[Thorsten Loesch]

That's interesting, yep in high gain it does indeed go higher. I hadn't checked in high gain because in balanced/4.4mm the gain did not alter the power limit, it was the same in both high and medium. So it seems that the higher 'gain' setting does provide more power but only when on SE?

Yup. In balanced the higher settings are current starved, so the power supply rails collapse.

You can use 600 Ohm to test voltage limits and 32/16 Ohm to test current limits.

The highest power in Balanced will actually be delivered with an impedance greater than 32 Ohm, you should try 50 Ohm (many Planars) and also 80/120/300 Ohm.

This is still slightly low on expectations. Maybe moving the PSU from the dedicated PCB onto the main PCB they had cut down some parts? You used external power? Make sure the USB Power supply can deliver 5V/3A (so 15VA & native USB-C as minimum).

It is possible that that the Battery fitted is not the correct specification, it is 4,000mAh and should allow 2C discharge, but perhaps what is fitted had a specification creep and only allows 0.5C discharge. It would make it cheaper but would further gimp power output.

Thor

 

[GoldenSound]

This is still slightly low on expectations. Maybe moving the PSU from the dedicated PCB onto the main PCB they had cut down some parts? You used external power? Make sure the USB Power supply can deliver 5V/3A (so 15VA & native USB-C as minimum).

It is possible that that the Battery fitted is not the correct specification, it is 4,000mAh and should allow 2C discharge, but perhaps what is fitted had a specification creep and only allows 0.5C discharge. It would make it cheaper but would further gimp power output.

Thor

Yep this is with the included iFi iPower power supply which states 5V 2.5A.
Though the power output at 32 Ohm remains the same regardless of whether it's plugged in or unplugged as it turns out.

I also just checked with a 50W samsung USB C PSU and got the same.

 

[Thorsten Loesch]

Yep this is with the included iFi iPower power supply which states 5V 2.5A.
Though the power output at 32 Ohm remains the same regardless of whether it's plugged in or unplugged as it turns out.

I also just checked with a 50W samsung USB C PSU and got the same.

Hmmm, ok. If I actually particularly cared I'd go over the PCB pictures. But not that much.

Good I was able to help you get a more complete picture of the device.

Thor

 

[Thorsten Loesch]

Hmmm, ok. If I actually particularly cared I'd go over the PCB pictures. But not that much.

Ok, one last note, based on the measurements we saw, I'd expect maximum output power in SE into a 37 Ohm load in "Nitro" mode and BAL into a 150 Ohm load, by first approximation.

And I'd expect around 2 Watt into 50 Ohm (SE) which would be for old style Planars.

Using balanced would likely also limit at around 2 Watt with 50 Ohm. So there is no point using the Diablo 2 balanced with headphone impedances lower than around 150 Ohm.

So essentially use Balanced for the HD600 style headphones, HD800 etc etc al, AKG K-1000 and 600 Ohm K-240 Monitor and perhaps 150 Ohm Beyer Dynamics.

Stick to SE for all Planar's, low impedance headphones and in ears.

So the "beast" is more like:

Except less qute.

Time to go to the beach...

Thor

 

[Terr0rSandmannLuxx]

Full measurement post is up, video review to follow in a few weeks: https://goldensound.audio/2023/11/24/ifi-idsd-diablo-2-measurements/

TLDR:

- Overall performance/behaviour is effectively identical to the original Diablo.
- iEMatch works nicely to provide about a 7dB noise floor improvement (though of course will increase output impedance, possibly changing FR of IEMS, try with/without to see what works best on your IEMs)
- Power spec is indeed far below what is claimed. Actual power capability is about 1.6W, not 5W. If you want to go by peak power instead of RMS power then it's about 2.5W. iFi attained the '5W' spec using their own power test which is vastly easier than either a proper continuous power test (how basically all amplifier power specs are given) or peak power (which there is an industry standard set out for in CEA-2006.). If you were to test other amps in the same way a lot of '6W' amps might suddenly be 10W or 15W amps.

This is a good product, 1.6W in a portable (1.1W when unplugged) is still quite a bit, and the features of the Diablo 2 are very nice with many welcome changes over the original.

But in my view the way iFi is presenting the power capability of this device is simply misleading and unethical. They don't spec their other amps this way so I'm not sure why it's different here.

But this is not new, even if I do not like this marketing approach.
The micro iDSD Signature and micro iDSD Black Label were the same. They had approx. 1.5W and just 4W peak. But as far as I remember, once the original Diablo was introduced 2-3 years ago, ifi said that the Diablo can deliver the power permanently and not just in peak situations due to the fully balanced circuit.

Currently, I am using an Expanse which needs some juice to fully shine and thats what the Diablo delivers quite well. Thats why I always thought there are the full 4W-5W available. I do not even use the turbo mode, just normal mode.

 

[Thorsten Loesch]

Currently, I am using an Expanse which needs some juice to fully shine and thats what the Diablo delivers quite well. Thats why I always thought there are the full 4W-5W available. I do not even use the turbo mode, just normal mode.

The Expanse is around 23 Ohm and around 100dB/1V. Normal mode will actually give you more power than Turbo Mode.

You might want to try the SE Output in Turbo over the Balanced out in Normal. With Turbo you can get 120dB Peak and 117dB continuous SPL. Even Normal mode in SE might suffice in which case i would recommend that.

Thor

 

[Thorsten Loesch]

But this is not new, even if I do not like this marketing approach.
The micro iDSD Signature and micro iDSD Black Label were the same. They had approx. 1.5W and just 4W peak.

Yes. And it was stated in the specification:

But as far as I remember, once the original Diablo was introduced 2-3 years ago, ifi said that the Diablo can deliver the power permanently and not just in peak situations due to the fully balanced circuit.

Im wonder who said that and where they got that information from.

The power supply is the limitation and the power supply is basically the same in all the various iDSD micro and derivates.

Thor

 

[Terr0rSandmannLuxx]

The Expanse is around 23 Ohm and around 100dB/1V. Normal mode will actually give you more power than Turbo Mode.

You might want to try the SE Output in Turbo over the Balanced out in Normal. With Turbo you can get 120dB Peak and 117dB continuous SPL. Even Normal mode in SE might suffice in which case i would recommend that.

Thor

Currently, I have just the original Diablo here. I ordered the Diablo 2 to compare it. Why? Because on paper, the Diablo 2 is so close to be my perfect device (portable, powerful, Bluetooth, nice build quality, desktop mode). The only thing I am missing is Xbass, but with my current headphone and IEMs I do not need it. My plan was to sell my desktop and mobil setup and have just 1 device which can handle everything and every use case.

Since the last 2 pages, I am not sure anymore if the Diablo 2 is such device. It seems it has more technical flaws than I expected

The power supply is the limitation and the power supply is basically the same in all the various iDSD micro and derivates.

Thor

I would not put my hand into the fire for my statement. I just remember, but maybe I am wrong. But I am sure it was somewhere in the ifi announcement thread (if true) --> https://www.head-fi.org/threads/ifi-audio-idsd-diablo-a-portable-reference-done-our-way.951615/

 

[Thorsten Loesch]

Since the last 2 pages, I am not sure anymore if the Diablo 2 is such device. It seems it has more technical flaws than I expected

If all you need is a Diablo with Bluetooth in one box and you can stomach the rather substantial price increase, you will be fine.

The Diablo 2 simply shares the design limitations (I'd not call them flaws) of the original.

I would not put my hand into the fire for my statement. I just remember, but maybe I am wrong. But I am sure it was somewhere in the ifi announcement thread (if true) --> https://www.head-fi.org/threads/ifi-audio-idsd-diablo-a-portable-reference-done-our-way.951615/

I don't read anywhere a claim of 5W (5,000mW) continuous power. The comparison to a high performance sports car seems right, likely a 90's Jaguar on a German Autobahn (where the engine gives up after 10 minutes at top speed).

A lot of the copy is however not quite accurate. Like the Amplifier circuit is the polar opposite of how balanced is implemented in the Pro series. The stated SE power at 8.8V/32Ohm even matches Golden's results within 0.1dB, though Balanced is overstated if we consider continuous power.

Thor

 

[Loftprojection]

A few questions on Diablo 2 as a DAC.

In the first post @iFi audio mentions the DAC can go in NOS mode. I love NOS sound and my main DAC is a TDA1541A tube DAC so is this BurrBrown DAC chip able to work and sound like those old NOS chips?

Is the BB chip used the same as the one in other DAC/AMP from iFi and does it sound the similar when using the line out of the Diablo2? If so could reviews of older products bring similar expectations on the sound of the line out of Diablo2?

Is the line out dedicated with very good parts in its signal path or since most users use Diablo like device as DAC/AMP then it’s not worth using high quality parts for the line out?

I’ve been looking for a « battery transportable » DAC for a long time to feed my Woo WA8 transportable tube amp but I usually dislike the sould of ESS and AKM DAC chips so I’ve been waiting for someone to finally make an R2R transportable DAC (not dongle), doesn’t seem like this will ever happen!

Many thanks for your answers.