Discussion in 'Schiit' started by purr1n, May 12, 2021.
Not buying any more Schiit products that have these tiny pots in them. I refuse.
It’s not the size of the pot, it’s how it’s specced. Sorry, could not resist. All joking aside, it would likely add a few dollars to the cost if they requested pots that have better channel matching. May be worth it for a product that can run super sensitive IEMs. More importantly, and only Schiit would know this, it could add a delay in shipping for these custom-specced pots. I’m sure the last thing they want is another product on back order.
Although that low gain means you can use more range on the knob which kinda obviates the issue of channel imbalance at the nether region of the pot.
cannot wait for impressions mine gets here tomorrow to use with my elegia and new thieaudio oracle!
Unfortunately, it doesn't work like that. Can't go to a Digikey or Mouser, and select "slightly more expensive pot channel matched @ a few dollars more" . Even the Alps RK27 Blue Velvet has channel matching issues depending upon the roll of the dice. I've gotten cheap pots where by chance it was "holy crap, this is really good!" The only reason people don't hear channel imbalance in the RK27s is because they have a fancy name "Blue Velvet" so people think they are really good (they aren't) and because no one can hear small differences, 1db still being small. Turning your head 3 degrees while listening to speakers or having the one earpad squished a bit more or offset a few mm will result in more imbalance.
The only way to achieve good channel matching is by hand and with appropriate AP-515/525/555. This is labor intensive, subject to personal judgement as to what constitutes as a good pot, and will result in a lot of pots being thrown into the garbage. That isn't a few dollars more to the cost. That's a lot of dollars more to the cost. There is a reason why milspec pots are $100+ in quantities of 1000. I know because I checked a few years ago for a CS/SBAF group buy.
I'm more than happy to provide a service to install hand-picked pots for $119 for the Magnis. The problem is nobody will take me up on it.
I personally think channel matching on these pots is a slightly overblown issue - out of 15 plus amps that I have tried in the last 7 years, none have been so bad that my central image was affected when listening to mono recordings. Sure, a few had serious balance issues between 7-9 o’clock, but nobody ever listens at these levels.
As for better specced pots, my only source is JDS labs who managed to do this when they revised their Atom a few years ago. I’m assuming they are getting them from the manufacturer directly rather than through distributor. Here’s where they describe a deal they struck with Alps:
One of the many benefits of having a decent passive preamp upstream of my headphone amps is that they all mostly sit at the same volume setting all the time (different settings for different amps tho based on perceived noise and channel imbalance) and all volume control takes place at the preamp. I'm sure that there are potential downsides to this approach like more devices in the signal path etc. (though I am far less sure of their audibility, at least to me) but they are outweighed by the benefits. Ymmv etc...
W are already using custom-spec;d, custom-taper Alps pots on every single one of our products--in the case of IEMagni, a custom-spec'd, higher-tolerance, tighter-matched RK09.
This "tighter matching" still does not obviate some deviations from perfection...hell, I can show you examples of our blindingly expensive, custom-spec'd, tighter matching, RK27114 "blue velvets" that are sometimes 1.5-2dB off in the middle, where the taper splice is.
Hell, I can show you samples of a super custom crazy expensive pot we once played with, paying for their highest tolerance--I mean, this is an insane pot that costs more than a Magni itself--that was off 6dB at -35dB attenuation.
What this shows more than anything is we do what we say--when someone wants a random sample, we give them a random sample. Unchecked, unselected, unaltered. It could just have easily been one with even better matching. Both are inaudible in use.
It also shows why we use relay attenuation when budget permits--because these are essentially perfect.
Thanks Jason for clarifying that - like I said, it's never been an issue for me with any of your amps. I had some DC offset whooshing with the original Magni 3 but not with the Heresy or with the Magni 3+.
Speaking of knobs, I really like the new machined aluminum one that you put on IEMagni.
I have a question loosely related to your entry level combo and very relevant to me: I was thinking of getting an IEMagni/Modi 3+ combo for my home office setup. In the past, I have tried the original Modi 3 at two different occasions and it always picked up some RFI from a local airport radar at high volume levels. Still, it bothered me knowing it was there. The level of this RFI varied depending on the length and types of interconnect cable that I used to connect the Modi 3 to my amplifier as well as the placement of the unit. Nothing else made any difference.
Do you know if the Modi 3+ contains any circuit improvements to minimize this specific type of RFI? I know I am a total outside case but I thought it was worth posing a question. I tried your customer support and they told me that while the Unison module has some filtering, they were not sure if there was anything else that was improved in this respect compared to the launch Modi 3.
Modis, like all of our digital products, are certified as an FCC Class B device. It's more likely you're picking up RFI via cables than via Modi.
Updated my post with initial non-audio thoughts. The potentiometer is not an issue with this amplifier. It wasn't with the Topping L30 either. It's not an issue with all those amps we bought in the 2000's for our IEM's. Would I want a smoother one? Of course, but I'm probably not willing to pay for it with a Magni. My suggestion would be to go the topping route on the volume knob. Increase the diameter and weight of the knob to give a little more luxury feel to the pot. I personally don't like the long skinny knob currently on there, but I acknowledge that it a nice touch with the unique one that they chose.
If you’re suffering from RFI, consider picking up a pair of ICs with Mogami 2534:
you can ask Markertek to make you a pair for a very reasonable price.
For sure, cables contribute massively to it. With some twisted pair interconnects, the RFI issue almost disappears. With others that are shielded for EMI but not twisted, the issue becomes more prominent. However, other DACs like the Bifrost 1 and 2 did not have this issue regardless of the interconnects used, so they must have some additional filtering at these RCA terminals compared to the original Modi 3.
EDIT: Anyway, sorry guys, I don't mean to derail this thread. I am just curious about this kind of stuff.
I might be completely off base here, but the Bifrosts use LPSs and the Modi 3 uses SMPS. Perhaps one is more prone to picking up RFI in your specific environment.
Could be, but it exhibited the same issue regardless whether I used the included power adapter or powered it via PC USB output. It is the internal power circuitry perhaps that exacerbates the issue. I mean, we are comparing $99 device against $699 one, so yeah. Honestly though, the RFI was NOT noticeable during normal use. It only was audible at very high volume levels (95dB+), when nothing was playing.
Impressions thread here: https://www.superbestaudiofriends.org/index.php?threads/schiit-iemagni-discussions-and-impressions.10991/
no other impressions? i happen to love mine a lot. significant improvement from the heresy. obviously less balls that the JOT 2 but 1/4 the price. with a tube pre amp its tearing it up, has more soul than the heresy. but i guess im using 80$ NOS tubes in the pre amp so idk . but isnt that what you want? color it with a pre amp or go straight linear . would love for someone to do A/B using all solid state with heresy/3+ and maybe topping or THX and see whats up @schiit livestream seemed to say that there was more than just the drive unit change? maybe DC protection and other things not sure!
Let the loaners go through their rounds for more impressions. I think in many ways, the IEMatch is anti-climatic. It's "Even More Heresy-er" after all. Now I could see a lot more excitement from the rather jaded SBAF members had it been a IEMagni 3+ using discrete parts. However, that would be a much more difficult design.
I don't remember if the OG Heresy used global feedback. The IEMagni doesn't use global feedback, only local. The x6 OPA1656 used for the output stage are unity gain - that is used as buffers. The sole gain opamp is responsible for the gain at the high-gain setting. At negative gain, there's a voltage divider between the stages. I think Schiit could have gotten better performance, that is measured performance (SINAD) if they used global feedback. Maybe 2-3db SINAD betterer.
The thing is that they they didn't go all-out objective just to take the ASR SINAD crown. Local feedback loops without the use of global feedback sounds better. They stuck to their guns to what sounded better - as if 2-3db SINAD really matters.
1kHz 2Vrms 100k load (most comparable to ASR measurement parameters, except ASR uses 200k-ohm load and balanced inputs on the AP)
Potentiometer channel matching (this specific unit)
Separate names with a comma.