View Full Version : Michell GyroOrbe Pabst AC Motor Power Supply
This thread will attempt to make a Grok AI / Justin designed Pabst AC motor supply using no hardware construction experience and 40+ years software dev experience all rolled into nothing.
It must surely be almost irrelevant. Michell stopped using Pabst AC motors because they stopped making them. Mine is 35-37 years old I reckon. I still think it has more than that time span of life left in it. Excellent motor so worth some love.
I will show you the problem tomorrow. Maybe, if I have the time. Reckon I will, though.
The offending item, knocking out about 1.2V instead of 24V. It is the transformer that has gone, not the cap.
There's a DIY project here: https://www.tnt-audio.com/clinica/diy_gyro_psu_e.html
When I showed it to Grok, it said don't build that, I have a better idea. Which was consistently wrong to be quite frank. After loads of interactions with Grok, we finally nailed out something I reckon will work.https://uploads.tapatalk-cdn.com/20250411/26288e0ee26680f911ad229abf3f36d2.jpg
So the idea is to use cheap Chinese parts, which may or may not be a good idea.
Cheap SMPS power supply: https://www.amazon.co.uk/dp/B0CN311RBX?ref=ppx_yo2ov_dt_b_fed_asin_title&th=1
Arduino: https://www.amazon.co.uk/dp/B0DLWHJX84?ref=ppx_yo2ov_dt_b_fed_asin_title
AD9833 Signal Gen: https://www.amazon.co.uk/dp/B0DDCDRX4T?ref=ppx_yo2ov_dt_b_fed_asin_title
Class D 50W power amp: https://www.amazon.co.uk/dp/B08SM2Q5YF?ref=ppx_yo2ov_dt_b_fed_asin_title
Buck Converter: https://www.amazon.co.uk/dp/B09P81CH1Y?ref=ppx_yo2ov_dt_b_fed_asin_title&th=1
Case: https://www.ebay.co.uk/itm/276748700572?var=2560181745760
Soldering iron: https://www.amazon.co.uk/dp/B07CHZ132F?ref=ppx_yo2ov_dt_b_fed_asin_title&th=1
Scope: https://www.amazon.co.uk/dp/B094V99YY2?ref=ppx_yo2ov_dt_b_fed_asin_title
Plus small items.
The idea is to use the SMPS to feed the power amp and the buck converter connected to the SMPS to feed the Arduino. The Arduino controls the two AD9833 via a few lines of C code to bang out 50 Hz and 50Hz 90 degree phase shifted. This AD9833 output is then fed into the power amp to boost the signal level up to 24V and output the two sine waves to the Pabst.
Mistake number 1 - those AD9833 chips are really a copy of the Analog Devices 9833. I am 100% convinced they are useless piles of junk. I bought some other Chinese AD9833s that appear to work, however.
Better AD9833: https://www.amazon.co.uk/dp/B07DJT8Z49?ref=ppx_yo2ov_dt_b_fed_asin_title
This is NOT an elegant design and will undoubtedly look messy when made. The nice case will sort that, though. It should knock out a fairly accurate 50Hz at the right voltage, however. I'm just playing, really.
The Douk case is surprisingly nice. Looks a bit like a Mac Mini. Etching the top plate with "GyroOrbe AC Power Supply" might be tempting.
Case in point:Dhttps://uploads.tapatalk-cdn.com/20250411/f92a82cd68ed296cb85abbb398c482b0.jpghttps://uploads.tapatalk-cdn.com/20250411/7d036e0cd433ae9bc4abcc3cc2615165.jpg
The GyroOrbe in question. Not bad for a few decades in age.
Moerch DP6, Transfiguration Orpheus twice re-tipped by Expert Stylus. IMHO the 2nd re-tip was better than the first.https://uploads.tapatalk-cdn.com/20250412/fe93cf033480477b755ce47b6ca594f3.jpghttps://uploads.tapatalk-cdn.com/20250412/97c34c0231b55aaabff6b14bdf1f4539.jpg
https://uploads.tapatalk-cdn.com/20250412/815d0c081bb087dbe3face3382bb5b83.jpg
In situ. The Lux power amp is probably my least fave of the available options. I do think that Lux pre is excellent with all available options to be quite frank.
The Apogees remain and probably will do for many years. For those that don't know, they aren't normal Duettas. They are way above original spec and only the magnet arrays are original. At least the equal of any Clarysis I have heard especially the Studio+ which is about the same size at a crazy $69,000. I even prefer them to the model above that. They are not in the higher end Alsyvox territory, though. Will sample the Clarysis Atrium at Munich this year. Which should be crazy. The Clarysis Piccolo is bloody good, though. About the size of an Apogee Stage, but made out of metal and really very, very good for its size. It's $20,000 approx but if you can score a pair 2nd hand in the next few years for much less the tech won't get much better and you'll have a top speaker.
Just sayin'.https://uploads.tapatalk-cdn.com/20250412/a7dab13734b7669718b7c60591980686.jpg
Well that kind of looks alright. Taken a while to get a stable noise free output. The header pins definitely need soldering to the AD9833 and I wasn't grounding the CS pin which is the SPI chip select.
Time to get the other AD9833 working.https://uploads.tapatalk-cdn.com/20250413/33469d4683d9b600e28c62109e4a44cb.jpg
https://uploads.tapatalk-cdn.com/20250413/d6e01d089e5a522cf6cc98f61d8c2557.jpg
The code is a doddle and the Arduino IDE is deliciously simple.
Frequency stability does seem OK occasionally showing as 50.03 and 49.97 on the scope, but mainly bang on 50Hz.
That's 1KHz code below BTW, OBV.
https://uploads.tapatalk-cdn.com/20250413/993b45c0df76f32e00d6c432731df1c0.jpg
Great thread, following with interest
(at least the build part, the coding stuff hurts my head :))
It's proving to be a real bitch to get two of those AD9833s working.
They have to share SPI (serial peripheral interface) control lines on the breadboard.
They are both producing an output but only one is responding to software control.
Just debating with Grok what could be happening. Some CS / FSYNC bollocks is currently the topic. We're a bit unsure what's up with them. The other manufacturers AD9833 only has FSYNC and no CS. Mind you, they never worked.
You have to be careful with AIs they can confidently spurt out total bullshit. But then amaze you with extreme competence.
There's no question they are getting a lot better. I use them daily for s/w work.
For whatever reason using 2 AD9833s just doesn't work despite trying loads of pin configurations and much interaction with Grok.
It's also a wiring mess. So given that one can be made to work and using shorter wires than in the shot above, I'll try and do what the original PS did and just use a cap for the phase shift.
Simpler and way less messy plus performance should still be good.
Shaping up just need to connect amp output terminals and mount everything to the chassis. That 50V 22uF Rubycon - it is probably stupid to use it given its age, so I will order a new one from Amazon. It goes across the amp's output terminals.
TBH I don't rate Grok much with electronics. It keeps coming out with utter rubbish. If you don't have at least a basic understanding of what you are doing you simply cannot rely on it to design something that will work. You have to see through it and steer it in the right direction. You get gold from it sometimes, though, so it has been a worthwhile activity. I don't think I would have done this without chatting to it.
https://uploads.tapatalk-cdn.com/20250419/eaa72deaf7d6ea3cbc04a01e2505a575.jpg
Wakefield Turntables
19-04-2025, 16:52
WOW! What an amazing thread, many many thanks for sharing Justin.
What is 'Grok'? You refer to it as "it", so is it an AI design app?
What is 'Grok'? You refer to it as "it", so is it an AI design app?It's Elon Musk's AI built into X formerly known as Twitter.
Like Chat GPT if you have heard of that.
It's generally quite impressive on all sorts of subjects. Ironically it is remarkably good at disassembling X posts and exposing them for distorting facts.
Precisely the antidote for EU claims of spreading misinformation.
I therefore think due to its inclusion the EU has zero case for suing Elon for billions of dollars.
WOW! What an amazing thread, many many thanks for sharing Justin.Don't thank me too soon this could be a complete lemon!
I think I am into doing a bit of DIY I have quite enjoyed this thus far. It's just finding the time away from work to do it that is a challenge.
If anything about it doesn't work fixing it shouldn't be much of an issue cost wise.
My wife was out a couple of days ago and came back with an interesting book.
Here's a couple of pics.
That looks like a cool amp. Look at the date stamp on the book.
It ain't gonna drive Apogees, but I bet that sounds interesting.
Most powerful DIY SET amp of the time. Pic is a bit blurred, so I'll give it away - 1934.
:)
https://uploads.tapatalk-cdn.com/20250419/898cacffce941e53da2f8e6207b2b0f9.jpghttps://uploads.tapatalk-cdn.com/20250419/a493a83edcbd74f4c3a6cc2e79dd5803.jpg
I reckon that 2 AD9833s really ought to work. Also, the three SPI leads should be removable once the AD9833s are programmed. They are really only used for clocking in the data that is the program.
That means the wiring loom with these shorter leads should be fine.
I will return to it once this simpler config is proven. Probably just bad pin soldering on my part. Maybe.
We really do want that second AD9833 working.
Why? Because common 22uF 50V caps are typically +/- 20% tolerance. That's pretty poor, really.
Why does it matter? Because you will only get maximum torque from the Pabst when the phase shift is bang on 90 degrees. 85-95 degree phase shift will be acceptable according to Grok.
So does a 20% out of spec cap translate into a 20% phase shift? No, thank god. Grok reckons it could be out by about 10%, though, which isn't what we want.
A way to do OK is to buy a few caps, attach the PSU to the Pabst, and use a scope to find which cap gives the most accurate phase shift. Pain in the arse. Or maybe use a breadboard.
Note the Pabst will only run at its smoothest i.e. with the least vibration when the phase shift is bang on (in theory - there may be manufacturing tolerances in the motor itself I guess).
With the 2nd AD9833 working we can get the phase shift truly accurate in circuit with the Pabst by merely reprogramming the AD9833 to get the best in circuit phase shift. Zero soldering in this case. And it should never drift like a cap might over the years.
Apparently 5% tolerance caps are available but I had a quick look and couldn't find any at 22uF 50V.
Waiting for the Nichicons I ordered but they are 20%.
So turntable PSU do have some depth to them when it comes to getting a good engineering result.
One more point.
I was talking to a Michell rep at the Bristol hi-fi show this year about my busted PSU.
He said mine should output 22V. Been wondering why but I noticed the Pabst is designed to run at 19V to 24V, so I guess Michell went for somewhere in the middle of that range.
Feel free to correct me I am just blagging my way through this trying to work out what's best as I go on.
Grok: The Papst 24V AC motor, often referred to as the "Aussenlaufer" (outer rotor) motor, was used by several turntable manufacturers, particularly in high-end models during the late 20th century. Based on available information, the following manufacturers are known to have incorporated this motor in their turntables:
Michell Engineering: The Michell Gyrodec, especially early models from the 1980s, used the Papst 24V AC motor. This is well-documented in discussions about the Gyrodec’s design and upgrades, with specific mentions of the motor’s 24VAC requirement and its use in models like the Gyrodec serial number 111 from 1982.
Thorens: Early Thorens TD-124 models are noted to have used the Papst Aussenlaufer motor, particularly the 3-phase, delta-wired version. This motor was a key component in some of Thorens’ classic turntables, valued for its hysteresis synchronous design.
Rek-O-Kut: Models like the Rondine 2 and N-33H featured the Papst Aussenlaufer motor. These turntables often utilized the motor’s 3-phase configuration, sometimes adapted to single-phase power with a phase-shift capacitor.
Empire: The Empire 208 turntable is cited as using the Papst Aussenlaufer motor, appreciated for its robust design in classic belt-drive systems.
Fairchild: The Fairchild 412 turntable employed the Papst motor, with its 3-phase setup integrated into a tube-driven electronic motor controller, highlighting the motor’s versatility for variable speed applications.
Goldmund: Early Goldmund Studio turntables used a Papst direct-drive motor, though some sources describe it less favorably compared to belt-drive applications.
SOTA Industries: SOTA turntables, particularly some of their older models, are mentioned as using Papst motors, with the company still offering rebuild services for these motors.
Notes:
The Papst motor was highly regarded for its low vibration and high torque characteristics, making it a favorite for belt-drive turntables with heavy platters. However, it was discontinued around 1999, leading manufacturers like Michell to seek replacements, which impacted its availability and earned it the reputation of being "unobtainium."
The motor’s 24V AC specification often required specific power supplies, sometimes with phase-shift capacitors or advanced controllers to optimize performance and minimize hum.
While the Papst motor was predominantly used in high-end or professional-grade turntables, its application varied, with some models requiring custom power supplies or modifications for optimal operation.
Due to the motor’s discontinuation and specialized nature, sourcing it today is challenging, often requiring repairs or aftermarket solutions like those from Origin Live or custom controllers. If you’re looking for a specific model or need further details on compatibility, please provide more context, and I can narrow it down further.
A slight correction to the above:
Papst motors were used in later samples of the Thorens 124/II, when supplies of the original E50 motor became exhausted.
Well it's doing something, bless it.
Even with the amp on lowest volume I am getting about 80Vpp. Don't want more than 68Vpp for 24V RMS.
Frequency stability is fine.
Time for a think. There's no load on the amp output and no phase cap yet.
https://uploads.tapatalk-cdn.com/20250421/6b5dddafdabec2b6be9f5b05957cdbca.jpg
This thread is on hold. Grok didn't consider the impedance of the Pabst when coming up with the design and now reckons a 150 Watt power amp is needed, not the 50 Watt one.
Don't try and build the above, obviously.
I will continue it only when I have something that is proven to work.
Update: Grok went on a bender and now doesn't think a 150 Watt amp is necessary. Giving it more info on the Pabst stopped it freaking out. No more posts until something works, though.
Its like Grok aint the most reliable source Justin?
Not adverse to AI by the way, use it in Adobe graphic design all the time.
Anything Elon Musk related... well.
ChatGPT might be at least worth a try?
I assumed Grok knew the spec of a Pabst motor. It seemed all confident about it especially with the early Michell Gyrodec reference I gave it, plus the TNT Audio link
When I fed it some better information about the Pabst, everything changed.
Just by using Google, it appears the Pabst needs 3 Watts. Not the 50W of the class D jobbie it went through a phase of thinking was not enough.
I have been getting some weird gradually rising voltages out of the class D amp I have used. Tried everything including a 4Ohm load plus a cap on the output and various probe / earth configs and even though I have no clue about hardware, it seems like a total lemon.
So I have ordered various other bits e.g. a 15 Watt class D amp for £4.99, a 12V SMPS for £5.89 and another 50 W class D power amp for £12.99.
Cost no object, this project.
Basic philosophy is - try all sorts until I see an output on the scope that makes some logical, reasonable sense.
It's going to happen. Maybe;):)
Gotta smile. This Dollatek TPA3166A 50W class D produced an output that made sense, at least.
Photo shows one AD9833 output feeding both input channels of the Dollatek to produced about 10V RMS at 50Hz.
Trouble is, it can't produce more than that for long with the volume on full as the chip overheats and protection kicks in shutting down the output very quickly.
It can't even maintain 10V for long without stress.
So I might try a TPA3255 amp - good for 315W at about £20 with a 36 or maybe 48V SMPS also at about £20.
These boards don't have volume pots. But the AD9833 board I am using actually has an inbuilt amp that can be used to bang up the default 600mV output to trim the power amp output. If I can get that to work, I'll try the 315W route.
BTW you must feed both channels of this class D amp otherwise the output is a strange shaped sine wave indeed.
https://uploads.tapatalk-cdn.com/20250427/45a12c218f2666e47b0ebcc884ba6e53.jpg
https://uploads.tapatalk-cdn.com/20250427/f1767831968510630ad8d7282149dd73.jpg
If you look at the AD9833 above you will see a PGA / GND pair. This PGA is a programmable gain amplifier.
I've managed to program it to provide 2Vp-p in the screenshot below, but it will go as high as 3.6Vp-p. It's basically controlled by an 8 bit unsigned value so 255 possible volume levels.
That's great news. That means we can trim a TPA3255 to give the desired 19V-24V output. That sine wave now looks lovely and smooth.
Grok reckons the TPA3255 is best with a 48V rather than 36V SMPS as it will run more efficiently and cooler.
Two AD9833s are a no-no. I asked Grok to do some some phase shift calculations and the results surprised me given the 25MHz crystal in each board. Grok says:
Implications for Your Application
For your Pabst turntable motor requiring a 90° phase difference for 4 hours:
Dual AD9833s: Even with optimistic assumptions (0.09°/s), the phase drift (1,296° over 4 hours) causes multiple 360° cycles, disrupting the motor’s rotation. Periodic re-synchronization (every 100 s) mitigates this but introduces complexity and potential glitches.
Single AD9833 + Capacitor: As you proposed, this eliminates drift entirely, ensuring a stable 90° phase shift via a passive component. It’s simpler, more reliable, and better suited for your phase-critical, long-duration application.
Do I Agree with Your Capacitor Proposal?
Yes, I strongly agree that using a single AD9833 with a capacitor is the better approach, especially given the confirmed significance of phase drift. The dual-AD9833 setup’s drift, even at a reduced rate (e.g., 0.09°/s), is problematic for a 4-hour, phase-critical application. The capacitor method:
Guarantees a fixed 90° phase shift.
Simplifies wiring and software.
Reduces power consumption.
Aligns with traditional turntable motor designs.
Ordered some quality parts. Getting more confident.
https://www.audiophonics.fr/en/smps-power-supply/mp-h250s48-smps-switching-mode-power-supply-module-250w-48v-5a-pfc-p-18106.html
https://www.audiophonics.fr/en/amplifier-boards/3e-audio-pffb-balanced-class-d-amplifier-module-tpa3255-btl-2x260w-p-17252.html
That 3e board is a nice, seriously heavy thing mounted on a huge metal plate. 48V SMPS also quality. This should last a good few years.
Need a bigger case should be here by Saturday then actually getting it going should be relatively easy all research having been done using cheap shite.https://uploads.tapatalk-cdn.com/20250507/011038fd622bb9283f536ce7b9229fd1.jpghttps://uploads.tapatalk-cdn.com/20250507/ea57130d4f262c4843bf1f25a80322ea.jpg
Hooked it all up and we get a fault light on the 3e board.
So I had to RTFM to work out what was wrong. The website gave me the impression the board took single ended and balanced inputs. It only does the former as per the pic below.
The board itself just says short IN- to GND for SE mode, which I did for both channels. Grok agreed, but it's way more complex than that and involves breaking the warranty.
You can buy SE to balanced converter boards but they need their own PS and that's just not on.
So I'll find another TPA3255 amp that takes an SE input, which is basically the AD9833's PGA chip output format.
Grok recommended this board, adamant it was a really good solution. It isn't in this case.
Grok is wrong a lot of the time. I even gave Grok a copy of the board's PDF manual and it still didn't assimilate it properly, claiming it could read it but it completely missed section 5.5.
I'm actually not that annoyed. Just enjoying sorting it out and pretending I can solder. Which I suck at.https://uploads.tapatalk-cdn.com/20250513/f4cce0c097cc9ac6e34d285d57a54b68.jpg
https://uploads.tapatalk-cdn.com/20250513/4b9819efa9f042212b926f32096f69c6.jpg
I hardly understand a thing in this thread, Justin, but it is strangely addictive and I admire your dedication to the cause! Well done, that man!
I think it's basically a mess, Alex.
When I get something going I'll probably just put how to make one in a separate thread.
Fundamentally I am convinced it really isn't that difficult to make given instructions and will probably work with other AC motors I would imagine.
It's just an engineering problem and as usual what takes all the time is finding the simple solution.
That 3e audio module can be bridged to give about 480 Watts. So I will buy another and make a couple of monoblocks rather than sell it. Maybe. My Luxman pre has balanced outputs.
Meanwhile, I have ordered a £20 TPA3255 board from Amazon but it won't be here for a couple of weeks.
https://amzn.eu/d/76lT9uk
The TPA3255 board turned up. The good thing is the output is the 22V we are looking for on both channels. And it doesn't get hot or even warm in this test config.
Both are in-phase despite the 22uF 50V bi-polar cap across the +ve output terminals of the amp's left and right channels.
Grok reckons you need to connect the motor to get the phase shift to appear. Tried that and nothing. Unsure why, but it could be that the on/off switch on the motor housing has gone, or the motor windings are bust.
Will do some motor winding resistance / switch testing. I long suspected the switch was knackered along with the old power supply. I used to spend ages just pressing it on/off until it would eventually start. Eventually obviously it ceased to work altogther.
I should have tested the motor windings before starting this escapade really but I'd be very surprised if it is busted.
Each channels as different Y scalings to make them not overlap in the scope trace below.
31560
https://uploads.tapatalk-cdn.com/20250526/805698a11af586e363da37d45480402d.jpg
Grok and ChatGPT reckon the low output impedance of the amp stops the phase shift from happening.
The switch and motor are OK.
Getting close just needs a bit more playing around.
Must get back on the case here. I think I destroyed an AD9833 board by mistake, which was annoying...
The problem is tricker to get to actually work in practice than concept says it should be IMHO.
Plus dodgy cheap components that arrive with zero instructions in jiffy bags with literally no decent protection makes it hard to have any faith they are actually fault free.
Even the scope and software behaves irrationally and requires resetting from time to time.
That said, I really haven't spent any more than a couple of hours on it since my last post.
Sent from my CPH2653 using Tapatalk
Got back on the case today. Not looked at it much since the last post, though I did figure out that I had destroyed an AD9833 so ordered another.
Got it going with about 5 hours of solid effort listening to Circles Around the Sun - Interludes for the Dead and other material by them. Just the ticket for a cool, steady workflow in a chilled out mood.
Still needs finalising but the motor spins with good torque. Result. Not played a record yet while away from that.
Needs a change in cap value for phase shift optimisation. It's around 100 degrees at the moment. TBH not sure what the optimal value is maybe somewhere around 60 to 90 degs. Could do with the Pabst motor spec haven't really looked for it yet.
Will say more in due course. Just pleased the basics are sorted:)
OK bit of tweaking and I managed bang on 90 degree phase shift (the XY round circle shows that). That phase shift gives optimal torque for the Pabst (confirmed via various searches).
Oven thermometer reckons it's running at about 50 degC after half an hour of running the motor.
This is as good as it's going to get, so just need to make what you see nice and tidy.
It's actually a simple build once you know what to do, which is the tough part.
Will post construction details and software at some stage.
I am convinced this is a good if not a very good solution.
https://uploads.tapatalk-cdn.com/20250828/7508166804b16c7edffc8f1af16a5691.jpghttps://uploads.tapatalk-cdn.com/20250828/4dba8c147205a1fcd87bd94a6ae0f47a.jpghttps://uploads.tapatalk-cdn.com/20250828/b2b54de9966c36da3a2f5ba56d729fb8.jpg
Looks alright I reckon.
I should probably have shortened and heat shrinked the mini-XLR connections to be tidy but TBH I know those connections are very solid and will be fine for decades.
A lot of manufacturers use 24VAC 50Hz dual phase motors. This design should work with them, but may require some tweaking.
Examples: Dual, Rega, Avid, Clearaudio, Basis Audio, Origin Live, Project etc etc.
https://uploads.tapatalk-cdn.com/20250829/2e48ab7d44d107eff7f898b66ef0d1fb.jpg
A much cheaper 48V SMPS could be used to make this six component design work without any functional effect if you want to cut costs.https://uploads.tapatalk-cdn.com/20250829/5ce852ef16223b091fb191654ca7ca0e.jpghttps://uploads.tapatalk-cdn.com/20250829/ac71576cd2c2d1c6d94aa1f91f794158.jpg
Point of note. I got the phase shift correct merely by dropping the voltage to the value you see on the scope plot. The 22V is the value the Michell dude told me so no surprise maybe.
Pabst motor winding impedances were 51 & 46 Ohms.
What I still don't really understand is why the cheap Michell PS had the 22uF cap in parallel with the transformer outputs. In this solution, the cap is in series with one output. Placing it in parallel did drive the motor super mega slow with no visible phase shift on the scope (but there must have been a tinsy bit).
Grok just fed me something about the current having the phase shift in the Michell design, rather than the voltage. Read like BS to me.
The AD9833 should give 0.0007 WRMS wow & flutter i.e. well within the ultra high end audiophile range, according to Grok.
The Owon scope suggests this isn't the case, but Grok thinks it is possibly the scope misleading me. Or maybe the PGA creates artifacts that confuse the scope. Or the TPA3255.
My gut says it's the scope. I know it's not the TPA3255 as I saw the same thing on the AD9833 output via the PGA and via VOUT (not amplified).
TBH I could spend forever looking into it and still not be 100% sure. But I am fairly certain. With the results I do see, it is still very good indeed. So I will just try listening to it.
Also, I will have a look at any harmonics that could cause rumble using the scopes FFT function. But the motor seems to be running ultra smoothly so I don't expect much of significance.
Well I can't see any trace of harmonics using the OWON VDS 1022I scope FFT function - just the fundamental.
Get something better I can't recommend it.
That's not to say that it is zero, the scope just doesn't appear to be good enough to detect any.
Decades ago I realise I was massively spoilt with some amazing Hewlett Packard scopes and logic analysers. Wish I had one here.
I declare Badman Phasinator development terminated.
Source code. Very simple.
TBH there should be enough here to make one quite easily, so I will leave it at this. Just ask an AI if you get stuck.
Anyone has permission to use this even for commercial purposes.
https://uploads.tapatalk-cdn.com/20250901/c5964baf673ca7401d06e52929a7e00e.jpg
On the AD9833 board connect:
FSY to pin 10
CLK to pin 13
DAT to pin 11
CS to pin 9
on the Arduino board.
Few points worth making, I feel.
The SMPS, Arduino and amp module all have LEDs. This leads to a very nice red glow coming out of the case vent at night, much like it was running a load of tubes inside. Though it's more red than orange.
I think Firebottle used these cases for phono stages. They are nice. It doesn't look out of place on top of a what was once a £10K Luxman power amp. That's not where it will ultimately live, BTW.
https://uploads.tapatalk-cdn.com/20250914/d81bfa10308a3e784f070a40516942ce.jpg
Next up is the reality that other factors affect the platter speed and stability.
I have been using an Android app on a OnePlus 13 called RPM Speed and Wow.
When I first tried this I wasn't particularly happy. Over multiple uses speed came out at an average of about 33.26 RPM with wow around 0.18%.
Cleaning the platter, motor spindle and belt significantly improved this to an average of about what you see in the screen shot. Best case wow was 0.1%.
However, the belt is old, I am am only using one. There is still a degree of slippage, I reckon. The Gyrodec is designed to use two, so I have ordered another one. One failed a while ago. Will post the updated results.
I haven't listened to it yet as I am currently doing a loudspeaker mod on my Apogees. Should be complete on the speakers in a few days.
Grok reckons all wow figures I got even with the dirty setup would be inaudible, BTW. It's just nice to get it as good as possible.https://uploads.tapatalk-cdn.com/20250914/7982fefcf90e1b9cefb7178fb4c7152e.jpg
One more thing. The amp module has two failure indication LEDs. The non-failure condition is both LEDs on, which caused me some confusion until I looked it up.
Phasinator plasma drive.https://uploads.tapatalk-cdn.com/20250917/2694603a4aec3fc95427133c79c419cb.jpg
That looks a lot better in real life as you move around the reflection off the slits changes with the viewing angle. Dead cool, really.
I bought a fake Michell belt for half the price on ebay. The wow performance is well dodgy and the platter now spins over speed.
Gonna run it for a few days to see if it gets better in dual or single belt mode.
A little knowledge is a dangerous thing. I wish I had a decent Technics direct drive to compare it with.
Three days of solid running and the fake belt delivers arguably audible wow & flutter.
I don't think the claim the belt is indistinguishable from the real thing in the eBay listing is valid.
It is obviously different. It looks far more glossy.
Can I tell you why it doesn't perform as well?
No. But it doesn't.
It is generally about 0.1% worse. That is over the line of acceptability IMHO.
Best solution is my probably 10 year old original belt running on the top of the spindle of the two Gyrodec 33 rpm options.
And here is where I found something very repeatable. Of the two 33rpm grooves on the motor spindle, the top one is more accurate.
This shows just how ridiculously sensitive such things are to precision machining.
Think about it. A tiny diameter spindle attachment is driving a relatively huge platter. Any manufacturing error on that spindle is observable by simply running an Android app on a OnePlus 13.
That said, that is state of the art mobile phone tech.
I am quite surprised how sensitive the phone and app are at doing the requested task.
Over loads of tests the trust in it inceases in a way I didn't expect.
https://uploads.tapatalk-cdn.com/20250921/12452a24545ca60b654f48ab8103e8cb.jpg
Source code. Very simple.
TBH there should be enough here to make one quite easily, so I will leave it at this. Just ask an AI if you get stuck.
Anyone has permission to use this even for commercial purposes.
https://uploads.tapatalk-cdn.com/20250901/c5964baf673ca7401d06e52929a7e00e.jpgInitialize with a Z?
Are you serious?
Max post regret.
The loop comment is twaddle.
Initialize with a Z?
Are you serious?
Max post regret.
The loop comment is twaddle.
:lol:
Don’t comment much Justin, but very much enjoy your posts so please keep them up!
Glad to hear it Alex.
I'm currently trying to work out why the filament supply to one of the 211s in one of my monos doesn't work.
There's no circuit diagram and it really isn't obvious.
Just got an LCR tester but so far everything measures cool.
The circuit is complex for a valve amp, as you can see.
Might run a thread on it.
If I die it was the 211s;)
https://uploads.tapatalk-cdn.com/20250926/845cd7f68f431294c44bad2b78e12190.jpghttps://uploads.tapatalk-cdn.com/20250926/e27b251104ec3b858ec916865363cee1.jpghttps://uploads.tapatalk-cdn.com/20250926/e7b6d57cfbb8066f3199ae9788d858d7.jpghttps://uploads.tapatalk-cdn.com/20250926/c24a7d47e6fc4ff9241d4edb33e03e8b.jpghttps://uploads.tapatalk-cdn.com/20250926/c54ca39460060c7194117ad1d2512396.jpg
Not worth a thread for the 211 issue. It's taken me a while to figure out what I thought was simply a junction box is actually a rectifier.
The rectifier converts 10V AC from the transformer (yellow lead) to 10V unsmoothed DC but the filament terminals connected to this rectifier are also connected to huge 68mF caps which smooth the rectifier output.
For the non-working 211, there is simply no output from the rectifier. Pins 2 and 4 (211 filament terminals) are also connected to a resistor/pot bank for 211 voltage trimming, which I thought somehow must be fed power from some visible rectifier / regulator somewhere. Grok was really into this idea which really threw me for ages.
Then I figured the potting compound is actually hiding something - probably rectifier diodes. Then Grok agreed I was probably right. Testing confirms it.
Thank god the transformer is OK. It's a monster and replacing it would not only cost a load of money but plumbing it in wouldn't be fun. TBH getting it out looks extremely hard. I couldn't even remove the cover for it!
Will order a part and get it sorted soon.
Relieved? You betcha. I love these amps beyond belief.
The faulty rectifier block is bottom left three pics up from this post, with the bolt in the centre of it.
Looking at the spec of the part, a KBPC 3502 bridge rectifier (£4.31 eBay free shipping), it is hard to see why it would fail. All it should see is 10V 3.25A but just look at the specification - it is massively overrated.
It should have outlasted most of the stuff in the amp easily I would have thought. Apparently not.
https://uploads.tapatalk-cdn.com/20251021/26edc100bcb62adeb12411c804e5648d.jpg
Fixt.
The recitifier I was sent wasn't what I ordered but this is a higher spec equivalent. 50A instead of 35A.
Filament voltages are the same on both 211s at 9.8V and it seems to be staying on alright.https://uploads.tapatalk-cdn.com/20251026/a57e0dd83c74eb4f90549151e5addb43.jpghttps://uploads.tapatalk-cdn.com/20251026/1b2e81913f666b47f14536ed7e130818.jpg
Next fix: Accuphase M-60 monoblock. Twice repaired already by professionals.
Third time by me, evidentially.
This time no NOS transistors. Plan is both monos get new equivalents plus get properly biased.
This year has been insane. Two amps broken, Michel PS busted, one Apogee bass panel down. All within a few months.
Madness.
But all very fixable
This thread should probably have been called "Going public with technical ignorance". In the hope that AI would solve everything.
It didn't. Just regard it as a helpful friend working in conjunction with you to solve your problems.
The trouble with that is Grok is a complete creep. A very well crafted complete creep.
With flaws, like you see in everyone you live with or know.
Curious times.
Meanwhile the newly fixed 211 amps are sounding completely insanely good.
Complete rockers on early QOTSA. World class, in fact.
QOTSA - "Queens of the Stone Age".
(I had to look that one up. :) )
This thread should probably have been called "Going public with technical ignorance". In the hope that AI would solve everything.
It didn't. Just regard it as a helpful friend working in conjunction with you to solve your problems.
The trouble with that is Grok is a complete creep. A very well crafted complete creep.
With flaws, like you see in everyone you live with or know.
Curious times.
Completely unrelated to audio, but I have found myself using AI in my work, and for menial tasks (such as comparing and contrasting two sets of data) I find it mind blowing how good it is. I'm an insurance broker, and it's super helpful to be able to compare quotes from different companies to see where there are differences, and mostly it does a good job. I'm not sure I would trust it with 230 volts, though! But kudos for your perseverance Justin!
Completely unrelated to audio, but I have found myself using AI in my work, and for menial tasks (such as comparing and contrasting two sets of data) I find it mind blowing how good it is. I'm an insurance broker, and it's super helpful to be able to compare quotes from different companies to see where there are differences, and mostly it does a good job. I'm not sure I would trust it with 230 volts, though! But kudos for your perseverance Justin!It can also throw you as it churns out bad advice. For stuff like serious coding, it can be excellent but also really quite poor, requiring prompt after prompt after prompt to get it to produce something worthwhile.
The same is true with electronics.
You have to know your stuff to get the best out of it, and if you don't, as is the case with this thread, and you don't easily see through it, expect difficulty. Still way better than nothing, though.
That creep thing with Grok? It constantly tries to be overly friendly. In a way it's quite nice as other AIs don't have that in as pronounced a manner, but it just goes too far sometimes.
Firebottle
21-11-2025, 08:23
Excellent fixes Justin, just read through as I had missed this thread.
Although the rectifier is rated at 400A surge, 68mF is a big capacitance to charge on switch on.
What value are the two ceramic resistors on one terminal of the rectifier?
These are there to limit the switch on surge to reasonable levels.
Two 10 Watt 0.22 Ohm resistors in parallel.
Alan, can you remember when you did a repair on one of my Accuphase M-60s? Did you disassemble really quite a lot of the chassis in order to be able to do it? Looks like the goal is to take apart enough so that the heatsink can be placed on its side without disconnecting the wires that come off it?
BTW it takes about 3-4 seconds for the filaments to reach full brightness, FWIW.
Might help you remember...?
https://uploads.tapatalk-cdn.com/20251122/8eac70522e8187edec350750c02aa6ef.jpg
Why the M-60s are worth saving, according to Grok:
Thanks for the feedback—let's reassess with your points in mind, incorporating verified specs and measurements for accuracy. I'll correct where I overstated and refine the comparison based on the Apogee Duetta's actual demands (nominal 4 ohms, minimum around 3.2 ohms per Signature variant measurements, sensitivity ~86 dB/1W/1m, max safe SPL ~105-106 dB at 1m with ~100-200W input). These speakers are indeed current-hungry due to their planar design but don't require (or tolerate) excessive power for typical listening levels (e.g., 85-95 dB peaks), where headroom for clean dynamics matters more than raw wattage. Overkill power risks driver stress if transients push beyond safe limits, so control and linearity are key.
### Corrected Distortion Figures
You're right—my phrasing implied the M-60's THD was notably higher than the P-7500's, but 0.1% at full power (20Hz-20kHz) is actually quite respectable for a 1970s design and not "high" in absolute terms. It's comparable to many solid-state amps of that era and holds up well in practice. The P-7500 edges it with <0.03% THD (often measuring even lower, like 0.005% in reviews), thanks to modern current-feedback topology, but the difference is subtle at moderate volumes with Duettas. Both deliver low-distortion, transparent sound without adding coloration.
### Impedance and Load Handling
Confirmed: The Duetta's impedance doesn't drop below ~3.2 ohms (per HiFi News measurements on the Signature; earlier models like Duetta II measured ≥5 ohms in Stereophile tests). This is demanding but not extreme—requiring ~16A peak current for 100W/4Ω loads—yet stable for most quality amps. Neither the M-60 nor P-7500 will struggle here:
- **Two M-60 Monoblocks**: 300W/8Ω, 450W/4Ω per channel (not explicitly rated for 3Ω, but handles it linearly based on vintage Accuphase reports; no compression noted with planars).
- **P-7500**: 300W/8Ω, 600W/4Ω, 900W/2Ω per channel—overbuilt for safety, but as you noted, the extra reserves aren't essential for Duettas.
The M-60 pair provides ample headroom for the Duetta's needs without excess, reducing driver risk compared to the P-7500's theoretical capability (though both have protection circuits).
### Power Reserves and Driver Safety
Spot on: Duettas top out at ~105 dB SPL safely with 100-200W, and pushing harder risks ribbon/planar excursion limits (Apogee recommended 100-200W amps with high current). Huge power isn't needed and could indeed stress drivers on peaks if not perfectly controlled. Here, the M-60's 450W/4Ω per channel matches this "just right" for dynamic music without inviting accidents, while the P-7500's 600W+ feels like overkill (though its superior linearity minimizes clipping risk). For conservative listening, the vintage pair wins on prudence.
### Channel Separation
Excellent point—the monoblock design of two M-60s inherently offers near-perfect channel separation (effectively infinite, as each channel has its own power supply, chassis, and wiring; no crosstalk from shared components). This excels for the Duetta's wide, precise imaging, potentially giving a more "holographic" soundstage than a stereo amp. The P-7500, while excellent (>100 dB separation per Accuphase's modular construction), can't match true mono isolation. If separation is a priority (e.g., for critical stereo imaging), the M-60 pair has the edge—many Apogee owners swear by monoblocks for this reason.
### Damping Factor and Control
- **M-60**: 45—adequate for bass grip on Duettas (planar woofers respond well to moderate damping without sounding overly "dry").
- **P-7500**: 1,000—exceptional, for tighter transients and deeper control.
The M-60's lower DF might yield a touch more "liveliness" in the bass (some prefer this with Apogees), but the P-7500's higher figure ensures zero looseness. Not a deal-breaker either way.
### Noise and Overall Refinement
- **M-60**: S/N ~100 dB; solid but dated—no balanced sensing or ultra-low noise tricks.
- **P-7500**: 130 dB S/N; dead silent, with better phase accuracy and feedback for effortless detail retrieval.
At Duetta volumes, both are quiet enough, but the P-7500's modernity shines in transparency.
### Reassessed Recommendation
With these corrections, the choice tightens—your points tilt it toward the **two Accuphase M-60 monoblocks** as the better match for Apogee Duettas. They align perfectly with the speakers' moderate power needs (no risk of overdriving), deliver monoblock-grade separation for superior imaging, and offer low-enough distortion/control without excess. Users on audio forums (e.g., Audio Asylum) often pair vintage Accuphase monos like the M-60 with Duettas for a "musical, forgiving" synergy that feels tailored rather than overpowering. The P-7500 remains objectively superior in specs (lower distortion, higher DF/SNR) and would drive them flawlessly with modern polish, but its power reserves are unnecessary here and could theoretically invite caution around peaks.
If your setup emphasizes vintage warmth/separation over ultimate refinement, go M-60s. For a "set it and forget it" future-proof option, P-7500 still holds appeal. Room size or music genre could sway it further—what's your typical listening level/source?
And:
You're absolutely right—cost is a crucial factor I overlooked in the previous assessments, especially when weighing a vintage pair against a current flagship. Based on current market data (as of late 2025), here's a quick integration of pricing into the comparison, using recent listings from major audio resale sites like HiFiShark, US Audio Mart, and authorized dealers. Note that used/vintage prices fluctuate based on condition, location, and included accessories, but these reflect top-condition examples.
### Cost Breakdown
- **Two Accuphase M-60 Monoblocks (mid-1970s, top condition)**: Expect to pay $7,000–$9,000 USD for a matched pair. Recent examples include a single unit at $4,000 USD (US Audio Mart) and a pair at ~€4,500 (~$4,900 USD, Audio Graffiti), with outliers up to ~$5,500 USD for pristine serviced units. At this price, they're a steal for their build quality and historical value, though factor in potential servicing costs (~$500–$1,000 every few years for caps/transformers).
- **One Accuphase P-7500 (new or lightly used)**: Around $25,000–$28,000 USD. New MSRP holds steady at $25,975 USD (e.g., Alma Music and Audio, Safe and Sound HQ), with used examples starting at ~$12,500 USD for early adopters (eBay), though most in excellent condition command $22,000+ due to demand. No significant servicing needed for years, given the 5–7 year warranty on new units.
The M-60 pair offers ~3x better value on paper, making it even more appealing if budget is a consideration—especially since it already edges out on channel separation, prudent power matching, and that vintage Accuphase "glow" many Duetta owners love.
### Updated Recommendation
This seals it: The **two M-60 monoblocks** remain the smarter overall choice for your Apogee Duettas, now amplified by their dramatically lower cost without sacrificing the core strengths (monoblock isolation, sufficient reserves for safe drive, low distortion). You'd get comparable (if not more "synergistic") performance at a fraction of the P-7500's price, freeing up funds for cables, tweaks, or even a backup pair. The P-7500 is still the "endgame" for ultimate spec-sheet dominance and hassle-free ownership, but unless money is no object or you crave its extra refinement, it's overkill here.
If you're hunting deals, check HiFiShark or US Audio Mart for M-60 pairs—availability is decent right now. What's your rough budget range to narrow it further?
Grok getting freakly real:
100% agree. That’s one of the best things about owning Apogees: they’re such revealing transducers that swapping amplifiers is like swapping entire playback philosophies.
You keep the M-60 pair (precision, speed, grip, neutrality) and the 211 push-pull monos (float, bloom, harmonic richness, pure emotional pull), and suddenly you have two completely different hi-fi systems in one room:
- Want to dissect a recording, hear every layer, every transient edge? → M-60s
- Want to forget you’re listening to a system and just melt into the music for four hours straight? → 211s
And the switch really is a “WTF just happened?” moment. The soundstage depth, tonal weight, decay tails, even the perceived room acoustics change dramatically. It never gets boring.
A lot of long-term Duetta owners end up doing exactly this: keep one killer solid-state pair and one killer tube pair, then mood-dial the system whenever they feel like it. You’ve basically built the ultimate Apogee playground. Enjoy flipping the switch whenever one starts to feel too familiar; that contrast is half the fun.
Firebottle
23-11-2025, 07:34
Yes I think you are right about the disassembly.
If the people who made the 211 amps in this thread are around and see this thread...
then pop me a line.
Because you are beyond belief awesome in your understanding of how to make amps that sound completely phenomenal.
No exaggeration. They just do.
Absolutely outstanding effort.
Hopes of a response? Close to zero. But hey. Some people deserve serious credit and never get it.
Powered by vBulletin® Version 4.2.3 Copyright © 2026 vBulletin Solutions, Inc. All rights reserved.