Low Power Piety + Power Hungry CA-1a: All the Small Things

However, I did stumble upon this. (Someone reported that this worked, so I decided to try it). Holy moly, this worked darn good. Piety tiny edition is actually subjectively more ballsy (but murkier) in the lows and wetter in the mids than Pietus. The CA-1a are on the cool and dry side, so Piety is great match. (I'm also using Yggdrasil LiM)

Schiit Yggdrasil LiM -> Nitsch Sound Piety -> TI-1b -> CA-1a


The downside is that we need to crank up the volume. I had no issue getting to moderately high volumes without maxing out the volume with rock tracks that are compressed. With classical tracks with high dynamic range, I had to max out the volume. All this in high gain.

I am messing with you guys? Absolutely not. Sure, some of this is a little bit of recoil from my recent experience with expensive amps that did not meet high expectations**. However, I am dead serious.

THIS. SOUNDS. DARN. GOOD.

I am thinking if I were a year or two into the hobby**, and figured out that I really loved the sound of ribbons, this could be a viable set up. Spend $2500 on the headphone and another $500 or so on Piety and maybe MMB2 (NOS mode) or Modius ES*. The major hang up of course is "lack of power" or is it lack of gain?* Should this even be working? Isn't the Piety a low power amp and the CA-1a power hungry headphone?

Let's find out - scientifically. Now I want to measure the impedance curve of the TI-1b, and run a distortion surface with a similar load. Heck, maybe I should run a distortion surface using the TI-1b and CA-1a as the load.

*With the Modius ES balanced outs, we can do this to get higher output near 4Vrms to the RCA inputs of the Piety. The question is still how much are pushing the Piety? [UPDATE: Scratch Modius E. Piety and CA-1a need a classic R2R sound).

**The fact is, if you are fairly new to the hobby, you should take a transducer first approach because the headphone, speaker, IEM, is going to have the greatest affect upon the sound that you hear. And no, expensive headphones do not necessarily need expensive amps. This is what the organizers of CanJam, and the audio-industrial-new media-shill-complex want you to believe. More audiophiles need to do stuff like this while saving up for a truly game-changer amp.

***There are major hang ups about "lack of power" when we need to go beyond 12 o'clock or near max on the volume knob.

To be continued. Science is fun and about exploration.

 

Here are the impedance measurement results.

TL;DR:

• The impedance using the CA-1a with the T1-1b is approximately what is stated on the tap, 16-ohms and 32-ohms. Nominally measures 19-ohms and 37-ohms respectively, but close enough.
• The impedance rolls off a bit into the lows, with the knee around 100Hz. This means amps with highish output Z may will result slightly (likely not significant) reduce output in the lows.
• The funk bit? The impedance roll off into the lows lessens as we increase the volume.
• Overall impedance curve is fairly flat. Worse case is 25-ohms to 40-ohms (most of it around 37-ohms and no peaks anywhere) with the Piety at 9 o'clock is which nowhere enough to get any kind of volume from the CA-1a with the T1-1b transformer adapter box.
• FWIW, I ran Piety on the 32-ohm tap to lessen the load on the amp and didn't hear any significant reduction in volume.

CA-1a (open pads)
TI-1b 32-ohm tap
Impedance (Volume knob on Piety at 9 o'clock)


CA-1a (open pads)
TI-1b 32-ohm taps
Impedance (Volume knob on Piety at 12 o'clock)


CA-1a (open pads)
TI-1b 16-ohm tap
Impedance (Volume knob on Piety at 12 o'clock)


CA-1a (open pads)
TI-1b 16-ohm tap
Impedance (Volume knob on Piety at 9 o'clock)

 

So this is what I figure. With white-noise, volume knob maxed on Piety, and high gain, we get about 15dBu output. White noise is actually easier for the amp to deal with despite the dBu output is the sum of the noise spectrum, and as such, the levels never aggregate to a voltage where the amp starts to clip. Also, I've noted that current protection comes into play (Piety has one, so don' t be afraid of using it) past 17dBu when it comes to a 1kHz signal.

Thinking 12dBu or 4Vrms would be reasonable max level (oh the irony that we are now measuring at ASR output levels, but remember, this is a severe edge case). Playing back AC/DC Black in Black with the volume maxed measures about 8-11dBu. Also we have another 5db more before current protect kicks in.

Nitsch x Schiit Sound Piety
1kHz 12dBu in 32-ohms
cheat mode off


Yeah, so we can see the Piety is being pushed hard where the odd harmonics are starting to pop up. Looks bad. However, let's turn back the clock and pretend it's 1981 when CD Players started coming out and APx555s didn't exist. 16-bits back then. That meant -96db limit to dynamic range. So this is what we get, and keep in mind that music is still 16-bits.*



Regardless, we are pushing the Piety hard and only have about 5db of headroom before current protect kicks in to prevent the amp from exploding.

*I'm not entirely convinced most of that Hires material from the streamers isn't 16/44 upscaled to 24/96.

 

Nitsch x Schiit Sound Piety
THDN vs Output level
1kHz into 32-ohms
Cheat mode off


Above 18dBu output, we the current protect activates. Note we never get to 1% THDN. 1% is the often used as the spec for power ratings, e.g. this power amp does 125W for 8-ohms will usually mean that the amp will hit 125W with 1% and be able to do that for an hour or X period of time.

1% doesn't sound that bad right? it's only 1%! Here's the scale in dB. 1% THDN translates into -40dB or 40db SINAD. I bring this up because I wanted to point out how ridiculous the SINAD club has become. I wouldn't say it's misinformation, but it's certainly reframing what are acceptable distortion levels and distorting the understanding of distortion.


If anything at least I feel confident that I can use a Modius SE and with some line input transformers to get near double the 2Vrms from the SE (4Vrms or +6db) but still be under 1% THD. In reality the transformer will have losses, so this seems to be just perfect to get more room on the volume control of the Piety for use with the RAAL CA-1a headphone. TBH, I was fine with the volume restrictions, but more room on volume control until max gives us less worry about "lack of power".

Anyway on to distortion surfaces because I want to understand more of what's going on. Instead of 0dBu to -30dBu, we can do 12dBu to 0dBu. It's a more narrow range specific to how we would use the RAAL CA-1a headphone.

 

Ok, typical stuff:

• Piety on high gain
• Assuming edge use case of RAAL CA-1a with TI-1b transformer box - will this work?
• 96kHz bandwidth (because we do care about ultrasonics)
• 32-ohm load
• Adding extra H4 and H5 separate (instead of H4-20), because interesting stuff going on here!

H2


H3


H4


H5


Conclusion:

• We are fine using the Piety like this, powering the CA-1a via the TI-1b on the 32-ohm taps
• The current protect circuit is very conservative
• All bets are off if we decide to go +5db EQ at 50Hz to account for the bass rolloff if using the open pads
• We will very likely be OK running with a hotter output from the DAC (4Vrms), e.g. Modius E with 1:1 XLR->RCA line input transformers, unless the music comprises 50Hz tone at 0dbFS, which is unlikely

 

Yeah, the H2 distortion, which is dominant (again in db which is a ratio) does dip before it goes up. Here's a reminder. Yeah, it's weird. Pretty sure the entire Piety / Pietus line of stuff was a design accident that happened to sound good.


The distortion rise as we go up in level happens more in the odd orders. This is gradually clipping or flattening of the sine wave.

Probably combination of tolerances and parts being pushed harder than typical.

After about an hour of use with the CA-1a:

Common ground at bottom leg of transformer.