Update On 3rd Version of ADS?

[Johnny]

Actually we have a gorgeous 3D rendering of the chassis as well with a lot of other fantastic steps forward... but I'm not posting anything until I'm holding a near finished prototype in my hands

How about posting what features the new motor controller will have over the standard offering?

[Phoenix Engineering]

Working with AC sync motors is a dry hole IME. Even running them in dual phase is fairly limited in what you can accomplish. Beside the fact that these motors are cheaply built and have a number of mechanical issues, there is only so much you can do by controlling frequency, phase and voltage in a quadrature drive. Compared to 3 phase AC synch motors, it's like taking a shower with your clothes on. I've been running a 3 phase motor and controller for ~2 years and there is no comparison, it is night and day: No vibration, no belt squeal, 100:1 speed range with loads of available torque (even at 78 RPM), 33/45/78 RPM with a single pulley spindle, and speed easily controlled with frequency to 0.0005 Hz resolution with tachometer feedback.

Ask HW; the DD uses a 3 phase motor, although it is servo controlled which is similar to the hunting algorithms of the 70's direct drive tables (albeit with a heavy platter). This is totally unnecessary and can be handled so much more elegantly with the right controller. Others in this industry have discovered what a proper belt drive system is capable of. In time, the rest of the field will follow or get left behind.

[Packgrog]

I've been running a 3 phase motor and controller for ~2 years and there is no comparison, it is night and day: No vibration, no belt squeal, 100:1 speed range with loads of available torque (even at 78 RPM), 33/45/78 RPM with a single pulley spindle, and speed easily controlled with frequency to 0.0005 Hz resolution with tachometer feedback. ... Others in this industry have discovered what a proper belt drive system is capable of. In time, the rest of the field will follow or get left behind.

Or hopefully some 3rd party industrious soul can provide plug & play 3 phase motors with controllers at the same scale to fill that gap, like you all-too-briefly did with the Falcon/Eagle/Roadrunner, for those of us that lack the ability to build such things themselves. And that's not meant as snark, either: I'd *LOVE* to try something like your DIY motor setup, but I simply lack the skill (and necessary equipment/available training) to build it. I suspect that VPI's motor choices may be limited to providers that can meet their scale & cost demands (pure conjecture on my part), and their R&D focus seems to be elsewhere. But they're hardly against 3rd-party mods/upgrades, and I imagine they'd praise someone that provided a nice alternative for their user base.

[Number9]

Others in this industry have discovered what a proper belt drive system is capable of. In time, the rest of the field will follow or get left behind.

Well stated Bill. The three-phase drive is definitely the future. All that remains to happen in order for this to become accepted industry best practice is for Ivor Thiefenbrun to learn of this and then convince himself that it was all his very own brilliant idea and add the requisite number of zeros to the price of the technology.

Everything revealed during the protracted series of listening tests during the development of my Turntable Control Instrument pointed to the fact that the three-phase approach as being inherently superior.

I made a commercial decision to release the synchronous only implementation initially in order to people, even those without much technical prowess, to extract the maximum sonic performance from their existing turntable, as did you.

Fortunately, because we designed the waveform generator board to provide biphase drive for two motors, with precise control of the phase relationships between all four waveforms, it makes developing a three-phase output module, fully compatible with the existing hardware largely a firmware issue.

My plan is to develop a three-phase output module to complement the existing range of interchangeable modules. At present, while we have identified what we believe to be a workable chipset and some candidate motors. there is still a lot to do do, but I hope we can begin field trials in the next couple of months. I'm hoping to keep the price of the three-phase upgrade to around $US300 plus motor. In order the address the value concerns of Linn owners, we also make the same module available for $US3000.

Stay tuned!
Steve

[Stringreen]

What's wrong with the SDS....why would I want to change to an ADS?

[Mr_Putty]

Offer me a kit with full instructions and reasonable price. And I will be all ears!

[Number9]

Hey Mat,
The whole "Wien Bridge" thing is never going to deliver sufficient stability to get the job done. WB oscillators are very sensitive to temperature and will give you less accuracy than the raw mains. We've known this since the early sixties, i can provide you with a fully developed OEM turnkey system that will offer better performance. see 9tci.com
give me acall if you're interested.

steve

[Phoenix Engineering]

Hey Mat,
The whole "Wien Bridge" thing is never going to deliver sufficient stability to get the job done. WB oscillators are very sensitive to temperature and will give you less accuracy than the raw mains.

That would seem to confirm what was posted on http://www.turntablepsu.com/vpi.html:

"Because the sinewave generator is analog based, the frequency drifted by +0.250 Hz after 1 hour of playing time (33.44 RPM) as seen here: http://www.turntablepsu.com/images/ADS1.TXT. Short term stability of the oscillators were good when started at room temp, but as the ADS warmed up, the frequency stability wavered ±0.02 Hz (±0.011 RPM) which is similar to the short term variations of the average wall socket: http://www.turntablepsu.com/images/Wall%20Socket.txt."

The TCI would be a good step in the right direction, but unfortunately, it has digital electronics and VPI has gone "all in" on the analog only approach; it was the raison d'être for the ADS and the claimed reason for why it was better than the SDS which was based on digital electronics. There is a thread on this forum with a large part devoted to the supposed deleterious effects of "digital noise" and why an analog only approach is putatively better. Harry claims there are no digital electronics in his listening room, going so far as removing the remote controls from the room as well. They'd have to eat a heaping helping of crow to do an about-face now.

To correct a couple of errors on your website:

1. Dual 32 bit processors are not required to provide the accuracy you claim (0.0006 RPM). The Phoenix controllers had 0.0003 RPM resolution using a single 8 bit processor. According to your previous post, the N9 controller has an update rate of 6,000 times per second; the Phoenix controllers were 9,320 times per second. While the Eagle and Falcon were single phase, the Condor maintains the same specs with 3 phases and the same (single) 8 bit processor.

2. Likewise a 32 bit DDS is not required, we achieved better accuracy with a lower master clock and 24 bits (DDS frequency resolution is: Clock/2^N, where N is the number of bits; you can increase resolution by increasing N or lowering the Clock frequency, we chose the latter).

3. The TCI N9 controller is not the only one on the market to do 2 phase quadrature drive of an AC synch motor. SOTA's Series VI motor controller also does this with DDS accuracy and user phase adjustment. Speed adjustment is 0.0055 RPM which is the same resolution as the SDS, but with dual phase capability.

On your comparison chart:

4. The Falcon was $379 MSRP not $479; The Falcon RR combo was $614 MSRP not $858.

5. The Falcon uses a 24 bit DDS, not 8 bit (actually it is 28 bit, but we stuff the lower 4 bits with zeros to make it operate as 24 bit).

6. The adjustment resolution with the Falcon/RR combo is 0.0003 RPM not 0.01.

7. The clock accuracy of the Falcon is 30PPM not 100. The Falcon/RR combo is 2.5PPM not 100.

[allvinyl]

What's wrong with the SDS....why would I want to change to an ADS?

I have been using an Aries I with original HRX and Stillpoints upgrades for several years(pic). I bought the Aries with an SDS and was an early adopter of the ADS. My experience tells me that unless you can audition the ADS in your system directly against the SDS before you consider selling the SDS, don't change to the ADS. I regret not having followed my own instinct in this matter. Prior to acquiring the ADS, I ran the SDS in the same rack with my preamp and amp. I will note two issues I have now using the ADS:

1. You may have read about hum issues with the ADS. I have experienced and still live with this issue. VPI provided a MU metal shield I installed internally over the transformer that has done nothing to mitigate the hum. The unit I have hums when powered on and hums loudly through the speakers when play is engaged. I have only been able to banish the play-engaged hum by moving the ADS to a location a good distance from my line stage(pic). Even in this location the powered-on hum is still present. Prior to moving it away from the line stage the play-engaged hum was so loud through the speakers I could not listen to vinyl at all.

2. Speed variation. My basement listening room here in MN is about 8 degrees cooler in winter than the rest of my house. I use a space heater to warm the room when listening or exercising. I have found the ADS changes speed as the room temperature varies. I am constantly adjusting the speed (33 or 45) during listening sessions. I use a Phoenix Roadrunner to monitor RPM so I am able to watch and verify the changes real-time. It's not my imagination.

(I just noticed that I need to update my signature as I am no longer Ops Mgr at Stillpoints.)

John

[Number9]

Hey Mat,
The whole "Wien Bridge" thing is never going to deliver sufficient stability to get the job done. WB oscillators are very sensitive to temperature and will give you less accuracy than the raw mains.

That would seem to confirm what was posted on http://www.turntablepsu.com/vpi.html:

"Because the sinewave generator is analog based, the frequency drifted by +0.250 Hz after 1 hour of playing time (33.44 RPM) as seen here: http://www.turntablepsu.com/images/ADS1.TXT. Short term stability of the oscillators were good when started at room temp, but as the ADS warmed up, the frequency stability wavered ±0.02 Hz (±0.011 RPM) which is similar to the short term variations of the average wall socket: http://www.turntablepsu.com/images/Wall%20Socket.txt."

The TCI would be a good step in the right direction, but unfortunately, it has digital electronics and VPI has gone "all in" on the analog only approach; it was the raison d'être for the ADS and the claimed reason for why it was better than the SDS which was based on digital electronics. There is a thread on this forum with a large part devoted to the supposed deleterious effects of "digital noise" and why an analog only approach is putatively better. Harry claims there are no digital electronics in his listening room, going so far as removing the remote controls from the room as well. They'd have to eat a heaping helping of crow to do an about-face now.

To correct a couple of errors on your website:

1. Dual 32 bit processors are not required to provide the accuracy you claim (0.0006 RPM). The Phoenix controllers had 0.0003 RPM resolution using a single 8 bit processor. According to your previous post, the N9 controller has an update rate of 6,000 times per second; the Phoenix controllers were 9,320 times per second. While the Eagle and Falcon were single phase, the Condor maintains the same specs with 3 phases and the same (single) 8 bit processor.

2. Likewise a 32 bit DDS is not required, we achieved better accuracy with a lower master clock and 24 bits (DDS frequency resolution is: Clock/2^N, where N is the number of bits; you can increase resolution by increasing N or lowering the Clock frequency, we chose the latter).

3. The TCI N9 controller is not the only one on the market to do 2 phase quadrature drive of an AC synch motor. SOTA's Series VI motor controller also does this with DDS accuracy and user phase adjustment. Speed adjustment is 0.0055 RPM which is the same resolution as the SDS, but with dual phase capability.

On your comparison chart:

4. The Falcon was $379 MSRP not $479; The Falcon RR combo was $614 MSRP not $858.

5. The Falcon uses a 24 bit DDS, not 8 bit (actually it is 28 bit, but we stuff the lower 4 bits with zeros to make it operate as 24 bit).

6. The adjustment resolution with the Falcon/RR combo is 0.0003 RPM not 0.01.

7. The clock accuracy of the Falcon is 30PPM not 100. The Falcon/RR combo is 2.5PPM not 100.

Hi Bill,
Thanks for the feedback. I'm an Audio guy, not a programmer, so I had Keith, the project manager and programmer look over your post. Yes, there are some areas where I not clearly understood the ins and outs of the numbers and I will make the necessary changes to the website. Thank you for taking the time. I'll post Keith's comments in full, unedited.

Thus spake Keith:


1. Dual 32 bit processors are not required to provide the accuracy you claim (0.0006 RPM). The Phoenix controllers had 0.0003 RPM resolution using a single 8 bit processor. According to your previous post, the N9 controller has an update rate of 6,000 times per second; the Phoenix controllers were 9,320 times per second. While the Eagle and Falcon were single phase, the Condor maintains the same specs with 3 phases and the same (single) 8 bit processor.

Firstly the controller has one 32 bit controller and one 8 bit controller. The 32 bit controller is the brains for the outfit and is responsible for generating PWM synthesis of the motor drives and also allowing a highly accurate period measuring function for calibration. The 8 bit controller is just there to drive the user interface (LCD, buttons etc). So you blurb is kind of correct: its a 32 bit based generator which has two processors.
What the is this updates per second figure and where did it come from? Is it the PWM rate? If so, the controller updates the PWM value at a rate of 50 KHz or 50,000 times a second.

2. Likewise a 32 bit DDS is not required, we achieved better accuracy with a lower master clock and 24 bits (DDS frequency resolution is: Clock/2^N, where N is the number of bits; you can increase resolution by increasing N or lowering the Clock frequency, we chose the latter).

I have no idea what he is talking about, but the statement is wrong, plain and simple. Accuracy and resolution are two completely different things. In terms of resolution, 32 bits gives a higher resolution than 24 bits. You cannot increase resolution by lowering the clock frequency, as you are not changing the number of steps you can divide the clock frequency period up into. It is fixed by the number of bits you have at your disposal. For 24 bits, the resolution is 2^24 = 16777216 so you can resolve to 1/16777216. For 32 bits you can resolve 1 in 4294967296. Only the number of bits directly relates to resolution. Resolution is literally the smallest fraction that you can resolve or you can divide things up into.
In the first sentence they also mention accuracy. This is directly affected by the accuracy of the reference quantity you are dividing up with your resolution. So you can have the worlds worst reference oscillator, and still claim 24 or 32 bit resolution. In our PWM driven world directly relates to the master clock, which in our case is accurate to less than 5 PPM (from memory).
The are other factors which affect resolution also, including the resolution of any maths being done. In the case of the processor we are using, it is native 32 bits. Yes you can do things with an 8 bit processor, but quite likely for 24 bits of resolution its working hard. Most of the 32 bit controllers functions are happening in hardware, including floating point math. This also includes the period measurement.


See also:
https://www.apgsensors.com/about-us/blo ... tle-things
http://kb.mccdaq.com/KnowledgebaseArticle50043.aspx

3. The TCI N9 controller is not the only one on the market to do 2 phase quadrature drive of an AC synch motor. SOTA's Series VI motor controller also does this with DDS accuracy and user phase adjustment. Speed adjustment is 0.0055 RPM which is the same resolution as the SDS, but with dual phase capability.

Well, I'll take their word for it, but do the other offerings have the capability to drive two, 2 phase motors with adjustable phase locking between them? (i.e. a total of 4 phases all driven with the same resolution and accuracy).

Steve
http://www.9tci.com