Sorry, wasn't well explained. Nakamichi used a pair of FETs as source followers to supply DC (and ground) to the *other* polarity's local zener diodes, gates connected to ground. True clean grounds in consumer electronics are made of unicorn horn, so this element at least kept dirty currents away from local ground. Folks enamoured of "shunt" supply regulators need to think about grounds just as much as anybody, maybe more.
Thanks,
Chris
The pure PWM output doesn't bother you? Feed it right into your power MOSFET's and your done (perfect sound forever, finally?)
Yes, that is really strange but i hear nothing over 16kHz. Martina Schoener of Garrard-Loricraft hears much better in the treble then me though. Just today she was complaining about a preamp i build her. With the volume control in a certain position she could hear a crackling noise when i moved the control back and forth. Me with the ear at the tweeter heard ( you guest it ) .... nothing
Hi Scott,
The Nak 1000ZXL machines could really do a good job. Getting rid of the felt pressure pad on step one (no more FM scrape noise). They really did understand analog electronics very well. Some of their CD players used similar methods, and not just the reference ground measures. I still have an OMS-7 (14 bit) that sounds great. I'd love to route the RF signal into a modern chip set and PCM-1702 or better.
Did the Nak analog guys mention anything else of interest? Doing service for many brands, Nakamichi was one of those that I was eager to see new designs from. Most other brands later on were pretty much more of the same inexpensive stuff. Ho-hum.
Hi Chris (Hornbeck),
A lot of the following is for people not familiar with the concepts that Nakamichi used.
Nakamichi also did the same thing with BJTs. Use the ground as a reference and pass the actual operating current to the other supply. To explain this better, a normal regulator sends it's own operating current into the ground it uses as a reference. Nakamichi did two things that were important to this idea. The first was that these regulators were local to the circuit, right at the point of use. They used ground for a reference only, and passed their operating current (with it's fluctuations and input noise) off to the opposite rail before the complimentary regulator. The other related concept was that the absolute voltage was not important. What was important was that the voltage drifted very slowly and did a good job of blocking any noise on the supply. One example of engineering trade-offs made to optimize what was most important.
Now, what noise would there be on the supply rails? After all, these supplies are already regulated. Just as there is a finite impedance in what you call "ground", there is also in the power supply wires and foil. That leads to minute voltage variations due to the demands of anything else sharing that supply (I*R), and also what can make it through the first stage of regulation. These power supplies can also pick up radiated noise along the way inside the chassis. Whatever you want to call this collection of noise, you'd best get rid of it before feeding a high gain stage - like a tape head EQ amp, or maybe the analog stages immediately following a D/A converter. You can make the same argument for the clocks in a CD player, and the I/V converter that processes the signals from the photo-diodes.
There is more than one way to "skin a cat", and this method is pretty powerful if used with the same detail as the rest of the circuits involved.
I would have absolutely loved to have seen a Studer 24 track that utilized Nakamichi designed electronics. That would have been totally amazing! Only a Studer transport would make this worthwhile in my view.
-Chris
The Nak 1000ZXL machines could really do a good job. Getting rid of the felt pressure pad on step one (no more FM scrape noise). They really did understand analog electronics very well. Some of their CD players used similar methods, and not just the reference ground measures. I still have an OMS-7 (14 bit) that sounds great. I'd love to route the RF signal into a modern chip set and PCM-1702 or better.
Did the Nak analog guys mention anything else of interest? Doing service for many brands, Nakamichi was one of those that I was eager to see new designs from. Most other brands later on were pretty much more of the same inexpensive stuff. Ho-hum.
Hi Chris (Hornbeck),
A lot of the following is for people not familiar with the concepts that Nakamichi used.
Nakamichi also did the same thing with BJTs. Use the ground as a reference and pass the actual operating current to the other supply. To explain this better, a normal regulator sends it's own operating current into the ground it uses as a reference. Nakamichi did two things that were important to this idea. The first was that these regulators were local to the circuit, right at the point of use. They used ground for a reference only, and passed their operating current (with it's fluctuations and input noise) off to the opposite rail before the complimentary regulator. The other related concept was that the absolute voltage was not important. What was important was that the voltage drifted very slowly and did a good job of blocking any noise on the supply. One example of engineering trade-offs made to optimize what was most important.
Now, what noise would there be on the supply rails? After all, these supplies are already regulated. Just as there is a finite impedance in what you call "ground", there is also in the power supply wires and foil. That leads to minute voltage variations due to the demands of anything else sharing that supply (I*R), and also what can make it through the first stage of regulation. These power supplies can also pick up radiated noise along the way inside the chassis. Whatever you want to call this collection of noise, you'd best get rid of it before feeding a high gain stage - like a tape head EQ amp, or maybe the analog stages immediately following a D/A converter. You can make the same argument for the clocks in a CD player, and the I/V converter that processes the signals from the photo-diodes.
There is more than one way to "skin a cat", and this method is pretty powerful if used with the same detail as the rest of the circuits involved.
I would have absolutely loved to have seen a Studer 24 track that utilized Nakamichi designed electronics. That would have been totally amazing! Only a Studer transport would make this worthwhile in my view.
-Chris
This i normal.
Pure 50Hz sine wave sounds like "bass" tone. 50Hz triangle, sawtooth or square sound like mid tone, for the reason of harmonic overtones.
The reason why CD sounds worse in the bass is the digital mess it adds to the original sound of the musical instruments. It would play pretty well single sine tone, but it would not play that well natural sound of musical instruments like double bass or cello. This difference of sound in the bass between analog and CD is pretty usual.
One or two.
Enrico Caruso - Recondita armonia - YouTube
(1909, I think)
Rachmaninoff plays Rhapsody on a Theme of Paganini - YouTube
(~1930?)
Rachmaninoff plays Debussy Golliwogg's Cakewalk - YouTube
Willem Mengelberg conducts Mahler Symphony 4; part 1 of 6 - YouTube
(1939)
File:Enrico Caruso, Recondita armonia (Tosca).ogg - Wikimedia Commons
(1909 - some noise reduction)
JUSSI BJÖRLING 1936 - RECONDITA ARMONIA - PUCCINI - YouTube (1936)
Leonard Warren "Prologue" I Pagliacci - YouTube (1940)
Gershwin plays Gershwin: Rhapsody in Blue - YouTube
They used much better heads than others. Playbackhead gap was first 1 um, later they did it with 0,8um. Thus much better frequency response and lower noise floor. The 1000 ZXL did 10 Hz to 25Khz @-20 dB, +/- 0,5dB.
This machine used Toshiba FETs in the first amplifying stage connected to the playback heads. ( also the ZX 9, which sounded much better than the ZX 7 with normal transistors there.)
They used better Bias Circuits and better EQ Circuits with some compensation possibilities for the PB heads. Result was even better frequencylinearity.
The Dolby Chips were almost matched and paired ( with colored dots marked), and the signal had to go trough them, if Dolby was used or not.
And they did much better mechanics with dual capstan( with matched bearings), controlled tapeguidance and controlled torque in the take up wheels and brakewheels.
Also the azimuth control and autoadjustment was unique and gave better frequeny response.
anatech explained the better powersupplies and groundings.
The PB heads with their small gaps and more bandwith were the reason that Nakamichi recorded tapes sounded dull on other tapes, but foreign records. sounded much better on Nakas.
Hi,
Not entirely true.
Long ago I owned a quite old Akai Tape Deck with dual head in glass ferrite, various included adjustments for the recording side and even build in generator for adjusting bias and recording EQ.
The mechanics where of a quality level (including dual capstan and an "aussenlaeufer" [outer rotor] motor of heroic build) that I not seen anything remotely the like of in any of the NAK's I serviced. It was of course completely discrete (IC count of zero). The microphone inputs where excellent too.
I suspect some Stasi P.I.G. swiped it together with my other HiFi Stuff and my motorbikes and my car when I fled east germany (when I was able to get back into east germany almost a year later nothing of what I had left behind could be found...).
This was the most extreme cassette tape machine I ever encountered, at the time I paid around halve a years worth of salary of a skilled worker for it.
Later in London I owned an equally antique Sony single head machine (also glass ferrite heads). It was also zero IC and it too had absolutely heroic build quality and both in terms of sound quality and build quality punched the snot out of all the NAK's I came across (note, I never serviced the Nak Dragon).
I still have a soft spot in my heart for those old machines, but I have zero remaining cassettes and no intention to get started again.
But the engineering that went into these was something else... Beautiful.
Ciao T
Not entirely true.
Long ago I owned a quite old Akai Tape Deck with dual head in glass ferrite, various included adjustments for the recording side and even build in generator for adjusting bias and recording EQ.
The mechanics where of a quality level (including dual capstan and an "aussenlaeufer" [outer rotor] motor of heroic build) that I not seen anything remotely the like of in any of the NAK's I serviced. It was of course completely discrete (IC count of zero). The microphone inputs where excellent too.
I suspect some Stasi P.I.G. swiped it together with my other HiFi Stuff and my motorbikes and my car when I fled east germany (when I was able to get back into east germany almost a year later nothing of what I had left behind could be found...).
This was the most extreme cassette tape machine I ever encountered, at the time I paid around halve a years worth of salary of a skilled worker for it.
Later in London I owned an equally antique Sony single head machine (also glass ferrite heads). It was also zero IC and it too had absolutely heroic build quality and both in terms of sound quality and build quality punched the snot out of all the NAK's I came across (note, I never serviced the Nak Dragon).
I still have a soft spot in my heart for those old machines, but I have zero remaining cassettes and no intention to get started again.
But the engineering that went into these was something else... Beautiful.
Ciao T
Hi Torsten
i dont wanna bash other brands, i simply was a fan of nakas and did the service for them for whole switzerland from 1982- 85, so i had all of them in my hands.
Yep, glass ferrite heads in Akai were a good thing to, they were ultrahard, while the nakaheads were good for a few 1000 hours only, sometimes lesser, depending from tape.
I just meant, the small gap was the real secret togheter with active azimuth correction to get good frequency response.
Most glassferrite heads i have seen in other machines were a sandwich of PB and Rec head togheter and had at 20 Khz almost phase shift of 180 deg or more and could be adjusted just togheter.
The Dragon with PB azimuth control was really a clever idea with the DD- Double Capstan.
But it was lousy made and inside it looked like cooked spaghetti
I also remember the Eumig FL 1000 uP, which sounded pretty good, but always problems with uP and capstan drive.
Revox hat good mechanics with full electronic torque control, heads were not the real thing and the audio circuits i can not comment.
i dont wanna bash other brands, i simply was a fan of nakas and did the service for them for whole switzerland from 1982- 85, so i had all of them in my hands.
Yep, glass ferrite heads in Akai were a good thing to, they were ultrahard, while the nakaheads were good for a few 1000 hours only, sometimes lesser, depending from tape.
I just meant, the small gap was the real secret togheter with active azimuth correction to get good frequency response.
Most glassferrite heads i have seen in other machines were a sandwich of PB and Rec head togheter and had at 20 Khz almost phase shift of 180 deg or more and could be adjusted just togheter.
The Dragon with PB azimuth control was really a clever idea with the DD- Double Capstan.
But it was lousy made and inside it looked like cooked spaghetti
I also remember the Eumig FL 1000 uP, which sounded pretty good, but always problems with uP and capstan drive.
Revox hat good mechanics with full electronic torque control, heads were not the real thing and the audio circuits i can not comment.
Hereby the link to live recording of Louis in Prague, 1965. The file size is 95MB, just one song in 24/96.
Free File Hosting, Online Storage & File Upload with FileServe
Hello Joachim
Was this conversion done inside the Korg MR-1000 onto a CD , could you explain how you did the conversion if you did it differently , and also you playback setup configuration for the comparison.
Regards
Arthur
- Status
- Not open for further replies.