Variable Multiwave Update For P5/P10 - Interesting Results

Ok, we figured it out. There’s a couple of things going on here. First, if you look at the graphs on display you’ll note that a considerable amount of time has elapsed between the original reading and the most recent. Months in fact. So, there’s no telling what was going on with the power at the time the first reading were taken.



Second, and this is the big one, the older firmware did not remove the 3d harmonic from the measurement system and the new firmware does. This has nothing to do with the actual output, only the THD measurement system.



So in order to not get people confused with the readings, we purposefully remove the increased 3d harmonic of MultiWave from the measurement. If we didn’t do that, then you’d notice that the THD went up when you went into MultiWave. The THD doesn’t actually go up when you go to MultiWave (in most cases) except where we add in the extended charging time of the 3d harmonic. So recent firmware that’s been shipping for some time, fixes the THD meter to be more accurate of the actual output, sans what we artificially add in if you choose MultiWave. Darqueknight’s original readings were with the older firmware that did not remove the component placed there with MultiWave.

Makes sense.



Thanks, Paul

That’s one mystery solved.

Ok, we figured it out. There's a couple of things going on here. First, if you look at the graphs on display you'll note that a considerable amount of time has elapsed between the original reading and the most recent. Months in fact. So, there's no telling what was going on with the power at the time the first reading were taken.

Second, and this is the big one, the older firmware did not remove the 3d harmonic from the measurement system and the new firmware does. This has nothing to do with the actual output, only the THD measurement system.

So in order to not get people confused with the readings, we purposefully remove the increased 3d harmonic of MultiWave from the measurement. If we didn't do that, then you'd notice that the THD went up when you went into MultiWave. The THD doesn't actually go up when you go to MultiWave (in most cases) except where we add in the extended charging time of the 3d harmonic. So recent firmware that's been shipping for some time, fixes the THD meter to be more accurate of the actual output, sans what we artificially add in if you choose MultiWave. Darqueknight's original readings were with the older firmware that did not remove the component placed there with MultiWave.
Maybe its me not comprehending but again, isn't Darqueknight's observations about the P10 in sine wave mode, not multiwave? If so, wouldn't the measurement oddity be irrelevant?
Darqueknight's original readings were with the older firmware that did not remove the component placed there with MultiWave.


All of the FFT plots shown in the prior post were taken with MultiWave off.

I have taken a new set of measurements going from Firmware version 22, to version 34A, to version 37, and to version 38. All of the measurements in the latest trials were taken with Multiwave off.

The musical selection used for listening evaluation "Antissa", a 21st-Century classical music piece from the "Unearthed" CD by E. S. Posthumus. The piece is well recorded and features a wide and deep sound stage, thunderous bass, soaring highs, vocals, choir, and lush, detailed percussion.

I hope you all don't mind me digressing for a bit of personal history. I don't claim to have "golden ears" or superhuman hearing, but I am very familiar with the sound of my two channel audio system and with the sonic character of my favorite reference recordings. I do not rely on memory alone when doing equipment evaluations. I document everything with spatial maps marking the placement of sound images in three dimensions. I also document the character and quality of the sound images. When I listen, I do not concentrate on trying to "hear a difference". I concentrate on characterizing and documenting all that I hear. Sometimes, with subtle differences, I do not become aware of a sonic difference until I compare notes between listening sessions. I then retest and verify.

I have gone through three iterations if AC regenerators in my two channel system: 1. Power Plant Premier, 2. P5, P10. I use regenerators only to power the two channel system's source components and preamps. The power amps are plugged into the wall.

Since I only use a regenerator to power low wattage source components and preamps, I bought a P5 mainly out of curiosity. I didn't think the P5 would provide a noticeable improvement, and even if it did, it would be so small that the investment wouldn't be warranted. After the first few seconds of listening I realized I was wrong and knew the P5 was a keeper. However, I didn't keep the P5 for long. I became curious about the P10, but surely that brute would be overkill for powering only preamps and source components. Wrong again. After one listen, the P5 was dismissed with haste.


Figure 5. My first PPP.


Figure 6. My first P5.


Figure 7. My first P10.

Getting back to the subject at hand, I conducted a listening trial with the following firmware versions, whose zip files are attached to this post:

Version 22 - "FMC-13" (Released 6/15/11 - the version that came with my P10 when purchased in November of 2011).
Version 34A - "FMC-24" (Released 4/30/13)
Version 37 - "FMC-27" (Released 12/23/13)
Version 38 - "FMC-28" (Released 12/31/13)

The trials began at 1:30 am. Version 37 was already loaded into the P10 from the prior listening evaluation. After listening to version 37 for 15 minutes, version 22 was loaded. The outbound THD meter went from 0.1% to 0.4%. I noticed a loss of definition and detail in the bass and a loss of detail and clarity overall.

After 25 minutes of listening to version 22, version 34A was loaded. The outbound THD meter went from 0.4% to 0.1%. The sound was identical to the previously loaded version 34.

After 15 minutes of listening to version 34A, version 37 was reloaded. There was no apparent change in sonics.

I saw that the filename for the Variable MultiWave update had changed from XXXX.27 version 37) to XXXX.28 (version 38). I decided to give version 38 a try. Version 38 took me back to the same sonics as version 22. After listening to version 38 for 15 minutes, I reloaded version 37 and have been happy since.

There is something going on with firmware version 34A (or some prior update that it includes) that is causing the sound improvements I am hearing. Regrettably, something was done in version 38 that takes away the "new magic"

I did not do any listening or measurements in the trials described immediately above with MultiWave engaged.

The following two sine wave plots show the difference between the power signal from the wall and from the P10.


Figure 8. Power signal from wall outlet that P10 plugs into.


Figure 9. Power signal from P10 outlet (FW ver. 22 - the sine waves for other versions looked the same). MultiWave not engaged.

The following oscilloscope Fast Fourier Transform plots were taken with MultiWave off and were taken within the same ninety minute period. It is much, much easier to see differences in the plots if you copy them to your computer and view them in succession.


Figure 10. FFT plot of power signal from the wall.


Figure 11. FFT plot of power signal from P10 with version 22 firmware.


Figure 12. FFT plot of power signal from P10 with version 34A firmware.


Figure 13. FFT plot of power signal from P10 with version 37 firmware.


Figure 14. FFT plot of power signal from P10 with version 38 firmware.

Going from the wall FFT plot (fig. 10) to the P10 ver. 22 FFT plot (fig. 11), there is a dramatic reduction in noise and harmonic amplitudes, as expected. What I didn't expect was that the 3rd harmonic would be beaten down so low. In prior measurements with ver. 22 firmware, the 3rd harmonic was would be about 3 to 4 dB lower than the wall. Now, with the reloaded version 22 firmware, the 3rd harmonic is 20 dB lower.

Going from the ver. 22 FFT plot (fig.11) to the ver. 34A FFT plot (fig. 12), the overall amplitude and density of the noise is lower with ver. 34A. The amplitude of the 3rd harmonic is reduced about 4 dB in version 34A.

Going from the ver. 34A FFT plot (fig.12) to the ver. 37 FFT plot (fig. 13), the overall amplitude and density of the noise is about the same. The amplitude of the 3rd harmonic is same.

Going from the ver. 37 FFT plot (fig.13) to the ver. 38 FFT plot (fig. 14), there is an increase in overall amplitude and density of the noise. The amplitude of the 3rd harmonic increases by 11 db. The amplitude of the 2nd harmonic also increased by 3 dB. Reinstalling ver. 37 calmed things back down.

I left the oscilloscope plugged into the P10 overnight. When I checked the FFT readings at eight and 12 hours later, the plots were similar to figure 13.

The measurements from the FFT plots are summarized in decibels and in volts in tables 1 and 2 respectively.

Table 1

Table 2

The following sine wave and FFT plots were taken during the initial Multiwave trial. The following seven sine wave plots show what happens to the P10's (and P5's) power signal as Multiwave is engaged with increasing strength from level 1 to level 6.


Figure 15. Output of P10 with Multiwave off.


Figure 16. Output of P10 with Multiwave level 1.


Figure 17. Output of P10 with Multiwave level 2.


Figure 18. Output of P10 with Multiwave level 3.


Figure 19. Output of P10 with Multiwave level 4.


Figure 20. Output of P10 with Multiwave level 5.


Figure 21. Output of P10 with Multiwave level 6.

The level of the third harmonic (180 Hz) is increased, which flattens the sine wave peaks. At Multiwave level 4, the oscilloscope no longer sensed the power signal as a sine wave. It was sensed as a square wave.


Figure 22. The oscilloscope sensed the power signal as a sine wave for Multiwave levels 1-3.


Figure 23. The oscilloscope sensed the power signal as a square wave for Multiwave levels 4-6.

The next seven FFT plots show how the third harmonic is increased in amplitude, which results in a sine wave with flattened peaks.


Figure 24. FFT plot of P10 power signal with Multiwave off.


Figure 25. FFT plot of P10 power signal with Multiwave level 1. The third harmonic is increased 23 dB.


Figure 26. FFT plot of P10 power signal with Multiwave level 2. The third harmonic is increased another 3.5 dB.


Figure 27. FFT plot of P10 power signal with Multiwave level 3. The third harmonic is increased another 4 dB.


Figure 28. FFT plot of P10 power signal with Multiwave level 4. The third harmonic is increased another 2.5 dB.


Figure 29. FFT plot of P10 power signal with Multiwave level 5. The third harmonic is increased another 2 dB.


Figure 30. FFT plot of P10 power signal with Multiwave level 6. The third harmonic is increased another 1.5 dB.

Test Equipment

Tektronix TDS 2012 Oscilloscope

Audio Equipment

Pass Laboratories XP-30 Preamp
Pass Laboratories XP-25 Phono Preamp
Pass Labs X600.5 Monoblock Power Amps (1200 wpc into 4 ohms)
Cary Audio CD 306 Professional Version SACD/HDCD/CD Player
Polk Audio SDA SRS 1.2TL Speakers (Hot Rodded!!!)
Audioquest Sky 72V DBS XLR Interconnects for Amps and SACD player
Audioquest Everest 72V DBS 9 AWG Speaker Cable
Salamander Synergy Triple 30 Audio Credenza
PS Audio PerfectWave P10 AC Regenerator
PS Audio PerfectWave AC-12 Power Cords (8 AWG) For All Two Channel System Equipment.
PS Audio Soloist Premier SE In-Wall Power Conditioners for Power Amps and Power Plant Premier
Dedicated 20 Amp Circuit for each Monoblock Amplifier and for source components and preamplifiers
AI-1 Dreadnought Non-Common Ground Interface with 1000VA Toroidal Transformer for SDA speakers
Hifi Tuning Silver/Gold and Hifi Tuning Supreme audio grade fuses for preamplifiers and source components
Auralex C24 Acoustic Panels
Spatial Computer Black Hole Room Resonance Damping System
PS Audio PowerBase isolation platforms


Figure 31. Two Channel system electronics.


Figure 32. Two Channel system front view.


Figure 31. Two Channel system room view.


Attached below are zip files for firmware versions 22 (FMC-13), 34A (FMC-24), 37 (FMC-27), and 38 (FMC-28):
<fieldset class="gc-fieldset">
							<legend> Attached files </legend> <a href="/wp-content/uploads/attachments/65960=2412-ecc6e0e06b296960855454741c57b3.zip" target="_blank" class="gc-files">/FileUpload/ab/ecc6e0e06b296960855454741c57b3.zip</a> <span class="gc-filesize">(557 KB)</span>Â  <a href="/wp-content/uploads/attachments/65960=2413-bc237b1995c631708811a461b38155.zip" target="_blank" class="gc-files">/FileUpload/3e/bc237b1995c631708811a461b38155.zip</a> <span class="gc-filesize">(552.4 KB)</span>Â  <a href="/wp-content/uploads/attachments/65960=2414-6dab39201f55e0bf8510a665154e37.zip" target="_blank" class="gc-files">/FileUpload/1a/6dab39201f55e0bf8510a665154e37.zip</a> <span class="gc-filesize">(686.9 KB)</span>Â  <a href="/wp-content/uploads/attachments/65960=2415-6c2eb5900b0e95651ff1bfbe28940f.zip" target="_blank" class="gc-files">/FileUpload/5a/6c2eb5900b0e95651ff1bfbe28940f.zip</a> <span class="gc-filesize">(687 KB)</span>Â </fieldset>
1 Like
DarqueKnight said: I saw that the filename for the Variable MultiWave update had changed from XXXX.27 version 37) to XXXX.28 (version 38). I decided to give version 38 a try. Version 38 took me back to the same sonics as version 22. After listening to version 38 for 15 minutes, I reloaded version 37 and have been happy since.


Does anyone have a copy of version 37 they can email me to oddeophile@yahoo.com? Or, post on the forum here? I would like to try it to compare as I have not loaded 38 into my P10's yet. By the time I downloaded the new firmware, version 38 had replaced 37. I do not have any problems with my PWT/D interfering with my P10's as both are behind my rack, well outside the view of the remote control when operating my PWT/D.

Odd

D;oh! just saw Darquenight’s version 37 on the website… sigh.



Odd

D;oh! just saw Darquenight's version 37 on the website.... sigh.

Odd


Unfortunately, filenames of attachments do not show up unless you place your cursor over the attachment. For anyone else interested, the four "blank sheets of paper" at the end of my last post are the zip file attachments for firmware versions 22, 34A, 37, and 38.

DarqueKnight intrigued me with his comment so I backed out and went from 38 to 37.



I had installed 38 after coming back from holiday and missed the earlier releases. Unless I am imagining things I’m sure 37 is audibly superior I was happy with the gains of 38, 37 is the winner.



DarqueKnight thanks. I’m listening to 16/44.1 like I have never heard them before.



Thanks “PS Audio” for this excellent improvement without having to purchase anything :smiley:



Cheers

John

Unless I am imagining things I'm sure 37 is audibly superior I was happy with the gains of 38, 37 is the winner.


Can you describe the improvements you hear?

Hmmm…odd again. All zones on my P10 crashed again in the night. Not had this problem before the new FW but have no idea if it is purely coincidence. I didn’t have to power cycle the P10, I only had to toggle the zones from the front panel.

My immediate impression was that of timing and the way in which instruments seemed to standout and be heard making an awesome musical performance. Tonal qualities of instruments is great I’m loving the sound 38 was good too though I feel 37 has an edge in musical performance. I cannot define any weakness till something better is heard. Bass is huge midrange more detailed and the highs are wonderful metal intranets exceptional. I like the depth to the musical presentation of 37 compared to 38. For me this firmware upgrade has been significant musical improvement and not a minor change. I could not wish for more.



I’m so looking forward to the higher bits rates with WaveStream :smiley:

Is there a change log available describing the difference between 37 and 38?

DarqueKnight, I cannot thank you enough for your work and the clarity of presentation. This is absolutely fascinating.



I am additionally intrigued by your method of listening and note taking. Very robust.



Wonderful job!

Swerags said: Is there a change log available describing the difference between 37 and 38?

My understanding is 38 removed a bug present in 37 where using the remote to change from sine to multiwave switched the multiwave strength to zero.
Swerags said: Is there a change log available describing the difference between 37 and 38?

My understanding is 38 removed a bug present in 37 where using the remote to change from sine to multiwave switched the multiwave strength to zero.


I noticed the Multiwave setting going back to zero when first playing around with MultiWave settings. I thought it was odd, but I just assumed that they didn't want variable multiwave to be engaged from the remote for some reason. Even if I am right next to the P10, I prefer to use the remote or the end of an eraser to interact with the touch screen to avoid smudges and wear on the screen surface.

If all 38 did is remove a remote operation bug, I should not have heard a difference between 37 and 38 just as I did not hear a difference between 34A and 37. I also should not have measured increases in the 2nd (+3 dB) and 3rd (+11 dB) harmonics when I went from 37 to 38.

I am additionally intrigued by your method of listening and note taking. Very robust.


Thank you. I trust my ears and memory up to a point, but usually, there is just too much going on in a stereo sound stage to try to remember it all from one listening session to the next.

Hmm ok, so we might see a 39 soon then :slight_smile:

@dcastle you could try the old trick of downgrading firmware, then upgrading to v38 again. Also, as a previous “sinewave” advocate, I now use multiwave 6 - much better.

@stereophilus - yup. Suspect you are right.

DarqueKnight said:
If all 38 did is remove a remote operation bug, I should not have heard a difference between 37 and 38 just as I did not hear a difference between 34A and 37. I also should not have measured increases in the 2nd (+3 dB) and 3rd (+11 dB) harmonics when I went from 37 to 38.

I fully agree, but the only information we have is 38 was released to resolve this bug.

I believe Dennis handled the programming in house. Perhaps when he is back in, and has a chance to get somewhat caught up, he will comment.