[eskamobob1]

Have you measured Diablo into 300 ohm

The ifi posted spec for 5W is at 32 ohm according to their own product page

 

[BS5711]

The ifi posted spec for 5W is at 32 ohm according to their own product page

Yes I can read.

Did you read the rest of my post for context ?

There was a reason I asked about 300 ohms regardless of what the stated numbers at 32 ohms are.

 

[GoldenSound]

Cameron,

Firstly I am not arguing with you here, I am just curious, and I in no way condone others claiming you are lying or have some other motive with your post, I think that is uncalled for.

Have you measured Diablo into 300 ohm ?

In comparing my Diablo to my Asgard 3 my impression was that the Diablo was well outclassed into 13 ohms DCA Aeon Noire but the tables were turned into 300 ohm HD600 where I felt like the Diablo actually bested the Asgard. The power/voltage numbers into 300 ohms for both products suggest that should be the case and I certainly don't hear any horrid distortion driving the HD600 to crazy volume even in Turbo mode.

The Diablo did well enough into the 13 ohm Aeon that within sane volumes the limitations didn't matter and I still preferred to listen with the Diablo over the Asgard 3 which I sold only weeks into owning it.

My impression from the Diablo and a few other ifI products over the last few years is they are designed with more of an emphasis on higher voltage than current output and so struggle into low impedance and low sensitivity loads but do disproportionately well into higher impedances.

I've not but I can do so tomorrow morning.

Generally speaking with 300/600ohm loads, you'll know if you're hitting the max or not. With a voltage limit like that it's typically a hard maximum where performance is basically the same upto that point and so if you don't hear any outright clipping you're in the clear.

With lower impedance loads is where it gets trickier. With current limiting it's a bit more complex cause

a) performance degredation can be more gradual rather than an absolute limit. An amp might have a max output of 6W and indeed be able to do 6W continuous, but may have 10x higher distortion at 1W than it does at 0.1W for example. The max output is often only part of the story at lower impedance loads.

b) continuous and peak specs will often differ more for lower impedance loads. With higher impedance loads where there's almost no current being pulled, a device will typically be able to just output a given voltage upto the maximum limit for as long as necessary. With current limiting/lower impedance scenarios, thermals or local energy storage can often be the issue. So the device might be able to do a high level very briefly but then will quickly reach higher levels of distortion.
The diablo is an example here actually. Where it'll do 4.98W for about 2 milliseconds but then quickly distorts more.

I imagine the 600Ohm spec for the diablo is correct but I'll check in the morning.

As to why you might prefer diablo over another amp, power is always just one part of the story. An immensely powerful amp can still sound terrible and a low power amp can sound great. There's lots of factors and preference is as always also a factor.

To be clear I'm not saying I don't like the diablo subjectively. I do. My subjective view hasn't changed at all. I'm just saying the power specs are considerably inflated and that's not OK.

The specs being wrong and the diablo sounding good can both be true

 

[eskamobob1]

Yes I can read.

Did you read the rest of my post for context ?

There was a reason I asked about 300 ohms regardless of what the stated numbers at 32 ohms are.

I just honestly don't see how the rest matters if ifi have flat out confirmed that the 32ohm spec is asbalutely wildly inflated and in no way actualy indicative of the power of the device. If the rest of the power outputs are real rms values, cool, but the 5w claim is litteraly the spear head of their marketing campaign for the produxt

 

[CH23]

it's interesting to see that as soon as someone puts numbers and graphs here, it's a certain kind of people who'll do their utmost best to discredit them with nothing but non-facts and non-logic.
@801evan i'm going to continue bringing this up because it entertained me so SO much, but you claimed putting a piece of rubber on top of your diablo would improve the way it sounded.

@BS5711 my message was not just directed at you, it's more that there's a LOT of pandering to companies going on, and I feel like I should encourage people not to do so, but instead help our community understand that as long as people neglect the negatives, then the companies won't learn from those mistakes, much like humans won't if you don't explain (or hold them accountable) for what they did wrong.

Companies are not people however, and if they want to remain relevant, then they should always strive to do better than the competition. Not by twisting the truth, but by making better products. Not by sleek marketing, but by word of mouth from actual people actually using their products.

Never be an apologist, but apply critical thinking, give credit where credit is due, and respectfully show them what could be improved.

 

[BS5711]

I just honestly don't see how the rest matters if ifi have flat out confirmed that the 32ohm spec is asbalutely wildly inflated and in no way actualy indicative of the power of the device. If the rest of the power outputs are real rms values, cool, but the 5w claim is litteraly the spear head of their marketing campaign for the produxt

Sure, I understand that 100% and have no issue with that being your opinion and perspective.

I am more interested in the technical performance of the Diablo in light of my own observations of its performance than I am bagging ifi about marketing claims, I trust that is Ok also.

 

[GoldenSound]

Sure, I understand that 100% and have no issue with that being your opinion and perspective.

I am more interested in the technical performance of the Diablo in light of my own observations of its performance than I am bagging ifi about marketing claims, I trust that is Ok also.

With the asgard 3 specifically I imagine it's mostly down to the fact that the asgard is a fairly warm sounding amp so it could be that your preference is toward the diablo signature.

As to why this might be, it's most likely the fact that the asgard's THD rises quite a lot into the higher frequencies if I remember correctly. This often seems to make stuff sound warmer/sweeter but potentially also softer

 

[BS5711]

I've not but I can do so tomorrow morning.

Generally speaking with 300/600ohm loads, you'll know if you're hitting the max or not. With a voltage limit like that it's typically a hard maximum where performance is basically the same upto that point and so if you don't hear any outright clipping you're in the clear.

With lower impedance loads is where it gets trickier. With current limiting it's a bit more complex cause

a) performance degredation can be more gradual rather than an absolute limit. An amp might have a max output of 6W and indeed be able to do 6W continuous, but may have 10x higher distortion at 1W than it does at 0.1W for example. The max output is often only part of the story at lower impedance loads.

b) continuous and peak specs will often differ more for lower impedance loads. With higher impedance loads where there's almost no current being pulled, a device will typically be able to just output a given voltage upto the maximum limit for as long as necessary. With current limiting/lower impedance scenarios, thermals or local energy storage can often be the issue. So the device might be able to do a high level very briefly but then will quickly reach higher levels of distortion.
The diablo is an example here actually. Where it'll do 4.98W for about 2 milliseconds but then quickly distorts more.

I imagine the 600Ohm spec for the diablo is correct but I'll check in the morning.

As to why you might prefer diablo over another amp, power is always just one part of the story. An immensely powerful amp can still sound terrible and a low power amp can sound great. There's lots of factors and preference is as always also a factor.

To be clear I'm not saying I don't like the diablo subjectively. I do. My subjective view hasn't changed at all. I'm just saying the power specs are considerably inflated and that's not OK.

The specs being wrong and the diablo sounding good can both be true

Thanks for the insight, that gels with my more amateur hour understanding.

I do think it would be appropriate to test into 300 and maybe 600 ohms just for perspective.

I would certainly be personally interested in the results but I also think if you are going to be so critical of one area of performance it would be diplomatic to test it at other areas to give some overall perspective.

You might just get more ammunition for your argument of exaggerating power output but, as you have already inferred, I suspect you will find that the Diablo is a lot closer to advertised numbers into higher impedance loads.

I for one would really appreciate the additional testing ..... 300 ohm at least as I don't have any headphones at 600 ohms

 

[GoldenSound]

Thanks for the insight, that gels with my more amateur hour understanding.

I do think it would be appropriate to test into 300 and maybe 600 ohms just for perspective.

I would certainly be personally interested in the results but I also think if you are going to be so critical of one area of performance it would be diplomatic to test it at other areas to give some overall perspective.

You might just get more ammunition for your argument of exaggerating power output but, as you have already inferred, I suspect you will find that the Diablo is a lot closer to advertised numbers into higher impedance loads.

I for one would really appreciate the additional testing ..... 300 ohm at least as I don't have any headphones at 600 ohms

Just got in and tested it.
600 Ohm power spec seems to be on the money, it'll get to almost exactly 0.6W @ 600 Ohm before hitting 1% THD:

300 Ohm:

Because of the fact that with higher impedance loads voltage is almost always the limit rather than current, if we plot the THD vs output level graph with output voltage instead of wattage on the X axis you can see it's basically identical:

(300 Ohm Red, 600 Ohm Purple)

If we add on the 32 Ohm test (Green) you can see that it maxes out about 10dB earlier, at about 15dBV (5.6V) instead of about 26dBV (19.9V)
This is normal/expected on basically all headphone amps as current is almost always the limiting factor as you go down in impedance. But showing the graph in this format is a good way to see how an amp might change in response to different loads, as wattage alone will change based on the load impedance and that can make comparisons a bit tricky.

So yeah the iDSD Diablo 600 Ohm spec is correct, but the 32 Ohm spec is about 6x too high

 

[stenog]

So what we basically has confirmed now is:

If you use 600 ohm(and maybe 300ohm) headphones you should get the Diablo. If not just get a Topping G5, 1 W at 16ohm and 1.2 W at 32ohm, confirmed. It's even a lot cheaper and has Bluetooth!

Ifi Audio, maybe you should edit the first post.

 

[GoldenSound]

So what we basically has confirmed now is:

If you use 600 ohm(and maybe 300ohm) headphones you should get the Diablo. If not just get a Topping G5, 1 W at 16ohm and 1.2 W at 32ohm, confirmed. It's even a lot cheaper and has Bluetooth!

Ifi Audio, maybe you should edit the first post.

With 300/600 Ohm headphones power is rarely an issue to be honest. And again due to the voltage limited nature of it, it's simply a case of 'Do you hear obvious clipping?' if not, you're fine.

It's not really like lower impedance planars etc where even if things aren't clipping some devices will be struggling/distorting more despite being under their max spec.

 

[BS5711]

Just got in and tested it.
600 Ohm power spec seems to be on the money, it'll get to almost exactly 0.6W @ 600 Ohm before hitting 1% THD:

300 Ohm:


Because of the fact that with higher impedance loads voltage is almost always the limit rather than current, if we plot the THD vs output level graph with output voltage instead of wattage on the X axis you can see it's basically identical:

(300 Ohm Red, 600 Ohm Purple)


If we add on the 32 Ohm test (Green) you can see that it maxes out about 10dB earlier, at about 15dBV (5.6V) instead of about 26dBV (19.9V)
This is normal/expected on basically all headphone amps as current is almost always the limiting factor as you go down in impedance. But showing the graph in this format is a good way to see how an amp might change in response to different loads, as wattage alone will change based on the load impedance and that can make comparisons a bit tricky.


So yeah the iDSD Diablo 600 Ohm spec is correct, but the 32 Ohm spec is about 6x too high

That is very interesting and aligns perfectly with my user experience as we discussed briefly yesterday, thanks for taking the time.

In the interest of science and to give ifI a fair go could I request one or two more tests ?

I’ll outline why below:

Some months ago I had a private message conversation with Thorsten Loesch who used to work for ifI and who designed the Diablo. He got in touch to give me some insight into a problem I was having with the Diablo shutting down when pushed hard with DCA Aeons.

Thorsten indicated that the problem was caused by the inability of the Diablo to provide the relatively high current that the Aeon drew as a result of their very low 13 ohm impedance and quite low sensitivity. He told me that the problem lay in the batteries inability to discharge fast enough to maintain the high current output, this was a problem he was aware of but he left ifI before the Diablo was released and someone else was left to assess the battery problem which he inferred was never actually implemented.

Thorsten told me that if running the Diablo plugged into a power supply it would bypass the battery and, if the power supply was of sufficient output capacity, would provide the current needed to in turn enable the Diablo to provide a higher output current via the headphone jack.

I experimented with several different USB chargers powering the Diablo when driving 13 ohm Aeon and pushing the volume hard on a very bass heavy track. On battery only the Diablo could be made to shut down if wound up near full volume even in Eco mode. It has to be connected to a power supply to reset the current limit cut out.

On a power supply the Diablo would go into hiccup protection if the power supply wasn’t of sufficient capacity rather than shutting down.

The supplied charger was actually worse than on battery power alone. The more powerful the charger the harder the Diablo could be run without going into hiccup protection. On a simple Samsung 25 Watt fast charger the Diablo could be run full volume on Turbo mode with going into hiccup protection. No I didn’t listen to it at that volume, obviously that is crazy loud, after a certain point I took the headphones off my head and cranked the volume some more, the DCA Aeon are robust !!

After discussing this with ifI earlier in this thread they acknowledged this was a problem and agreed that the Diablo would in fact output more power into low impedance loads if run off a robust power supply versus off battery power alone due to the exact reasons discussed above.

They also stated that into a very low impedance load off battery power the single ended output would effectively be higher than the balanced output, again due to battery discharge rate/current supply limitations.

It would be very interesting if you repeated the 32 ohm test with a minimum 25 Watt fast charge type power supply connected. I know my Diablo is an entirely different beast when connected to a good power supply. My guess is you will get a higher maximum power and less distortion since the amplifier circuits will be fed with sufficient power and since it is known that battery output is the bottleneck to the power output.

Also what would be interesting is a test at very low impedance, 13 ohms ideally but 16 ohms would probably be low enough. My guess is you would not actually be able to complete the test on battery power alone since the Diablo would shut down. On a robust power supply my guess is it would be quite a different picture, it certainly is based just on my user experience.

Appreciate the testing.

 

[eskamobob1]

Thorsten told me that if running the Diablo plugged into a power supply it would bypass the battery and, if the power supply was of sufficient output capacity, would provide the current needed to in turn enable the Diablo to provide a higher output current via the headphone jack.

So, just Fwiw, I asked GO to check the same thing when this first popped up. On the Diablo it didn't make a difference (he mentions he tested this case in the first post but it's easy to miss Fwiw). This is presumably because the Diablo specifically will only run off of battery Power even when plugged in (according to the ifi team last I asked) unlike the signature and micro BL which both have a battery bypass mode.

 

[GoldenSound]

That is very interesting and aligns perfectly with my user experience as we discussed briefly yesterday, thanks for taking the time.

In the interest of science and to give ifI a fair go could I request one or two more tests ?

I’ll outline why below:

Some months ago I had a private message conversation with Thorsten Loesch who used to work for ifI and who designed the Diablo. He got in touch to give me some insight into a problem I was having with the Diablo shutting down when pushed hard with DCA Aeons.

Thorsten indicated that the problem was caused by the inability of the Diablo to provide the relatively high current that the Aeon drew as a result of their very low 13 ohm impedance and quite low sensitivity. He told me that the problem lay in the batteries inability to discharge fast enough to maintain the high current output, this was a problem he was aware of but he left ifI before the Diablo was released and someone else was left to assess the battery problem which he inferred was never actually implemented.

Thorsten told me that if running the Diablo plugged into a power supply it would bypass the battery and, if the power supply was of sufficient output capacity, would provide the current needed to in turn enable the Diablo to provide a higher output current via the headphone jack.

I experimented with several different USB chargers powering the Diablo when driving 13 ohm Aeon and pushing the volume hard on a very bass heavy track. On battery only the Diablo could be made to shut down if wound up near full volume even in Eco mode. It has to be connected to a power supply to reset the current limit cut out.

On a power supply the Diablo would go into hiccup production rather than shutting down.

The supplied charger was actually worse than on battery power alone. The more powerful the charger the harder the Diablo could be run without going into hiccup protection. On a simple Samsung 25 Watt fast charger the Diablo could be run full volume on Turbo mode with going into hiccup protection. No I didn’t listen to it at that volume, obviously that is crazy loud, after a certain point I took the headphones off my head and cranked the volume some more, the DCA Aeon are robust !!

After discussing this with ifI earlier in this thread they acknowledged this was a problem and agreed that the Diablo would in fact output more power into low impedance loads if run off a robust power supply versus off battery power alone due to the exact reasons discussed above.

It would be very interesting if you repeated the 32 ohm test with a minimum 25 Watt fast charge type power supply connected. I know my Diablo is an entirely different beast when connected to a good power supply. My guess is you will get a higher maximum power and less distortion since the amplifier circuits will be fed with sufficient power and since it is known that battery output is the bottleneck to the power output.

Also what would be interesting is a test at very low impedance, 13 ohms ideally but 16 ohms would probably be low enough. My guess is you would not actually be able to complete the test on battery power alone since the Diablo would shut down. On a robust power supply my guess is it would be quite a different picture, it certainly is based just on my user experience.

Appreciate the testing.

I'll run this now and see what happens. The RMS power capability was the same when running it on an iFi iPower X PSU or my 50W Samsung charger, and afaik iFi has stated that the Diablo is always running on battery even when plugged in, unlike the iDSD Black Label and iDSD Signature which did have a sort of battery bypass.

I'll check it with a beefier PSU, I've got a Ferrum HYPSOS here which should be plenty (6A continuous so 30W). Will report back shortly.

As to the shutoff, it shuts off with 32 Ohm too if you push it past about 1W for any more than a few hundred milliseconds

 

[BS5711]

I'll run this now and see what happens. The RMS power capability was the same when running it on an iFi iPower X PSU or my 50W Samsung charger, and afaik iFi has stated that the Diablo is always running on battery even when plugged in, unlike the iDSD Black Label and iDSD Signature which did have a sort of battery bypass.

I'll check it with a beefier PSU, I've got a Ferrum HYPSOS here which should be plenty (6A continuous so 30W). Will report back shortly.

As to the shutoff, it shuts off with 32 Ohm too if you push it past about 1W for any more than a few hundred milliseconds

That would be interesting thank you.

I was advised that on a power supply the battery was bypassed and my experience certainly indicates that is the case. If it shuts down on Eco mode on battery but can literally be run full volume on Turbo off a power supply the power supply must be doing something other than charging the battery.

The supplied ifI power supply was actually worse than on battery alone.