ifi Zen Phono Review (phono stage)

[RandomEar]

Balanced out!

What is people's opinion on the headroom? Seems impressive until 5kHz and then starts to drop... what are the frequencies for which this is most important? Seems to me that a typical click/pop is in the middle low frequencies (by ear but no idea...)

I would dedact this from the Fourier Transform: The FT of a "pop-ish" function like 1/exp(|x|) peaks at zero and decays into higher frequencies. At the extreme end, the FT of an infinitely sharp unit impulse is just a flat line - that is, all frequencies are equally represented. I'd say a real pop is somewhere in-between, which would still mean that it looks qualitatively similare to the FT of 1/exp(|x|) and consists of more LF than HF stuff.

 

[AudioSceptic]

This is a review and detailed measurements of the ifi Zen Phono pre-amp. It is on kind loan from a member and costs US $199.
View attachment 214691
As you can see, it comes in a similar case to other (newer) ifi line. And as with those, the white labels are not legible in all but one or two angles. The left switch is power which has a delay in it (?). The right switch is a sub-sonic filter. The four dots are LED indicators for the four separate gain settings which you can select in the back:

View attachment 214693

Power is provided by an ordinary (non-ifi) 5 volt supply. Internally this is stepped up using a very high frequency switching regulator (running at 1.5 MHz). This should do away with mains spikes we see in many phono stages.

An unusual output is the 4.4 mm penticon balanced output. This is a very unusual connector for anything but headphones. It is nice to see it though if it provides a benefit in the form of cutting out ground loops. Internal architecture is balance as well (I think).

Company provides extensive measurements/specifications compared to what we typically see and speaks a lot to the low noise of the unit.

ifi Zen Phono Measurements
I stayed with the gain setting 1 which is suitable for Moving Magnet cartridges. Here is our dashboard using 5 mv input:
View attachment 214694

The spectrum of our 1 kHz signal is extremely clean. It is free of both power supply and distortion spikes. You would expect SINAD to excel but it does not:
View attachment 214695

The reason it doesn't is that noise floor is rather high. This did not add up with company specs/brochure and member comments about it being quieter. So I emulated one of the company SNR measurements:

View attachment 214696

As you see, I am very close to their specs/measurements. Balanced output does not help:
View attachment 214697

You can see the elevated noise level in this measurement and comparison:
View attachment 214698

Our top of the line pre-amps have far lower noise floor in low frequencies.

Distortion is also rather high when we discard noise from it:
View attachment 214700

On the other hand, the RIAA equalization is the best I have ever seen/measured:
View attachment 214699

It is essentially ruler flat!

On subsonic filter, I am measuring its static response. It may do better in practice due to "AI" logic that it say sit has.

Headroom shrinks sharply as frequencies increase which is unfortunate:
View attachment 214701

There is also increased distortion with frequency which is something I don't normally see.

Conclusions
Going into this review, I expected to see excellent measurements across the board. But this did not come to pass as far as noise and distortion. However, RIAA implementation is superbly accurate and this is something you will hear. On the strength of this, I am going to recommend the ifi Zen Phono pre-amp.

-----------
As always, questions, comments, recommendations, etc. are welcome.

Any donations are much appreciated using: https://www.audiosciencereview.com/forum/index.php?threads/how-to-support-audio-science-review.8150/

No measurements for MC? That's where it's really difficult to keep the noise low.

 

[antcollinet]

Yes because each line is for a different frequency.

Thanks - I am of course, a dumbass

 

[nothingman]

Thankful for this test, @amirm ! Agree with others that it’d be helpful to see the other gain settings tested, but I get that doing the full battery would mean quadrupaling the work and that’s a lot.

Question, on the last graph, the headroom test at different frequencies, what sort of minimum performance do we like to see here? Obviously the fact it’s frequency dependent isn’t great, but is there something of a target?

 

[Katji]

? It seems like it might be less expensive than all/most others in the SINAD chart. ?

 

[Risto_H]

A genuine question; given how far they lag behind some other electronics, DACs, amplifiers, etc., is it particularly difficult do design a phono preamp?

You mean the noise performance? The reason for the difference to a typical line level amplifier stage is the input signal level. You need roughly 40 dB of gain for MM and 60 dB for MC and especially the latter is challenging. Note that especially MM input may show quite different results when connected to a real MM cartridge vs. audio analyzer such as AP that is used here. MM cartridges have relatively high impedance with a large inductive component that may lead into much worse real world noise at low frequencies compared to these measurements.

While it is not too difficult to design a decently working RIAA amplifier, there are several challenges if you want to get a better than average result:

• Select the MC and MM input stages to fit the source impedance without sacrificing noise performance
• Select the RIAA EQ topology. There are several different from passive to active circuits and combinations of those.
• Calculate correct component values for the RIAA EQ curve, taking gain stage limitations into account and then select components with tight enough tolerances
• Select gain of all the stages to achieve optimum headroom. RIAA topology also plays a role here.
• Design a power supply that won't let 50 or 60 Hz noise to the system
• Decide how to avoid excessive DC offsets within the amplifier chain and line output. AC decoupling caps, ultra low offset parts, DC servo system or combination of these...
• Design the circuit board to both suppress electromagnetic disturbances and allow the gain elements to work under optimum conditions

 

[tubesguy]

Kind of ugly, is it not?

 

[Curupira]

? It seems like it might be less expensive than all/most others in the SINAD chart. ?

No, Mani 1/2, art pre and pluto are cheaper.

 

[AudioSceptic]

Kind of ugly, is it not?

Fugly. It looks like they had some randomly shaped sheets of metal they needed to use up.

 

[cgallery]

I do like these new frequency-dependent overload measurements a lot.

 

[amirm]

@amirm A question concerning your last THD+N curve. The input generator range goes from approx. 5 to 170 mVrms. My MM cartridge has a spec of 3mV output. So I wonder why this curve is really relevant in this case.

That is to show the headroom in the case of pops and clicks which output much higher voltage.

 

[amirm]

No measurements for MC? That's where it's really difficult to keep the noise low.

It just increases the gain. Everything changes proportionally then. I am on a short fuse to return this unit as it has been here a long time hence the abbreviated measurements.

 

[amirm]

Kind of ugly, is it not?

I think so.

 

[amirm]

Question, on the last graph, the headroom test at different frequencies, what sort of minimum performance do we like to see here? Obviously the fact it’s frequency dependent isn’t great, but is there something of a target?

I like to see no or small difference in overload. The large amounts we see here is not good for the overload characteristics. That said, I agree that I have not posted this kind of measurement enough to get a good handle of the range of the values we get.

 

[Katji]

The entire ZEN range uses that same housing. i prefer box shapes but it made sense for them to differentiate products from all the black boxes. The other iFi products all have that same long rectangular shape...maybe extruded...

 

[JeremyFife]

Aesthetically, I like the shape: quirky and a little bit 50's. It's mostly hidden my setup anyway. I do like that it's not just another box.
White letters on aluminium... that's daft though

 

[TOR]

I would say, this is ifi's trademark color scheme.

Remember the Nano iDSD & micro iDSD Black label. The instructions are printed on the bottom of the unit. Black on black. You can only read that when light bounce off at an angle.

 

[SMen]

It just increases the gain. Everything changes proportionally then. I am on a short fuse to return this unit as it has been here a long time hence the abbreviated measurements.

This is a shame, because it doesn't sound like a proportionate change between the MM and next setting up for low output MM's and high output MC's, making this unit all the more interesting! ... despite it's low price ... and looks.

I think it looks great! Must be the looks . I'm new to this forum but loving the reviews, thank you.

 

[AudioSceptic]

It just increases the gain. Everything changes proportionally then. I am on a short fuse to return this unit as it has been here a long time hence the abbreviated measurements.

Thanks. So the noise would be 20 dB worse (40 dB gain for MM, 60 dB for MC are typical)?

 

[srkbear]

I get the sense that there are some considerable shakeups going on at iFi, based on some redirections evident in their recent flourish of products, and some substantial features they’ve released to existing products (such as the xDSD Gryphon) via firmware upgrades.

Most notable is their deviation from their usual stalwart devotion to Burr Brown chipsets in their most recent Go Bar, for which they opted for a Cirrus Logic DAC chip. I’m hoping that this is in response to their being unable to achieve competitive measurements among their otherwise very well-built and feature-rich DACs. I’m looking forward to seeing how that unit performs on the scope.

They also seem to have begun deemphasizing their historical “clean power” product line (which brought them precariously close to snake oil territory), instead making some notable strides in the general consumer market of late (I hate to say budget because their products are better than what that term suggests).

I say this based on a slew of affordable and accessible desktop and portable options (such as their new Air line and the Go Bar) they have all but flooded the market with in the last year, that offer non-audiophile users access to some pleasingly rendered hi res solutions at aspirational prices.

I admit to some affection for iFi. I’m very fond of my Pro iCAN Signature amp, with its very rock-friendly ASP bass enhancement toggle, extraordinary versatility with inputs/outputs, and choices between solid state vs tube amplification stages (not to mention the 15W of clean, balanced power it achieves). I also find my Gryphon very nifty as a portable. I’m aware that they haven’t fared so well on this site’s measurement standards, but I think their price points are unusually fair, and I’m rooting for them to bring their prestige profile up to the levels they deserve based on their strengths elsewhere.