John, i was curious to know the Ampex recorders you had designed some audio parts in. Would-you be nice enough to satisfy my curiosity ?
AD: SSM2141, SSM2142,
or better: TI: DRV134, DRV135
THAT:1606, 1646 ...etc
Why bother ?
AD: SSM2141, SSM2142,
or better: TI: DRV134, DRV135
THAT:1606, 1646 ...etc
Last edited:
The values in Table 6 or Table 8 of the datasheet would be a good starting point- and likely a good ending point.
There's also likely a lot of other candidates in the IC world, I chose the 797 for illustration because I have some and I'm familiar with them.
So, where's that all-important 7th and 9th harmonic distortion data? Lower than -200dB is not good enough?
There's also likely a lot of other candidates in the IC world, I chose the 797 for illustration because I have some and I'm familiar with them.
So, where's that all-important 7th and 9th harmonic distortion data? Lower than -200dB is not good enough?
I worked with the AG-440, MM1000, and the AG500, and 600.
45 years ago, I was testing the AG-440 electronics virtually every day, with an HP wave-analyzer, sometimes for distortion, sometimes for noise.
At the same time I developed the first phase locked capstan servo for the AG-440 back in 1968.
45 years ago, I was testing the AG-440 electronics virtually every day, with an HP wave-analyzer, sometimes for distortion, sometimes for noise.
At the same time I developed the first phase locked capstan servo for the AG-440 back in 1968.
Really, I think you need to review the concept of second and third order intercept.
Well the only thing I get from this is that feedback resistors determine noise performance, not the 797.
Scott must be happy for you to endorse his baby like that!
jan
The Ampex ATR1x0 machines were very special. It used a servo-controlled, direct-drive-capstan tape transport, which allowed the tape speed and tension to be continuously monitored by a servo relay, without capstan and counter-capstan. The capstan was replaced by a free turning weel, in contact with the tape, measuring the speed. The purpose was to reduce wow and flutter.
Working very well, while boring to set the tape: needed adhesive to stick the beginning of the tape, and more free turns of the tape on the receiving weel than habit.
The fast rewind process was incredible fast, affraying, while perfectly monitored; the tape was slowed down in a very controlled and smooth way to the chosen locator's point. A science fiction transport.
They too used magnetic heads made of ferrite, to avoid wear.
This was a VERY bad idea, as the ferrite crumbled in the gap, and, if the head was not wearing with tape friction,some heads were often destroyed very fast and in an non predictable way.
When you had the luck to get good heads, you had what was probably the best analog recording machine ever made.
Working very well, while boring to set the tape: needed adhesive to stick the beginning of the tape, and more free turns of the tape on the receiving weel than habit.
The fast rewind process was incredible fast, affraying, while perfectly monitored; the tape was slowed down in a very controlled and smooth way to the chosen locator's point. A science fiction transport.
They too used magnetic heads made of ferrite, to avoid wear.
This was a VERY bad idea, as the ferrite crumbled in the gap, and, if the head was not wearing with tape friction,some heads were often destroyed very fast and in an non predictable way.
When you had the luck to get good heads, you had what was probably the best analog recording machine ever made.
Last edited:
I don't want to be wrong, but I will stand corrected WHEN you give me something to correct.
As far as the ATR100 is concerned, I preferred Studer transports, and my own electronics. That is what I did for Mobile Fidelity and Wilson Audio.
Some of the systems concepts are very good in the ATR100. I used the record system concept, first shown by Allister Heaslett, the major designer of the ATR100, to get optimum phase response.
As far as the ATR100 is concerned, I preferred Studer transports, and my own electronics. That is what I did for Mobile Fidelity and Wilson Audio.
Some of the systems concepts are very good in the ATR100. I used the record system concept, first shown by Allister Heaslett, the major designer of the ATR100, to get optimum phase response.
Last edited:
SY,
CV=q or C dv/dt + Vdc/dt =i dc/dt is close enough to zero we normally neglect the term so dv = i dt / C or .1 x 1/120 x 1/.1 = 83 mV ripple. (100 mV in the UK.)
So at 3kHz we would expect 833 uV. At 3000 Hz. you don't get 55 dB of reduction you are lucky to get 30 so we are now at 26 uV. Use the AD797 at -90 (Negative rail) and you get your less than a nanovolt or .82nV. Now add the preamp gain of 60 db to see 822 nV even if you run your power amp at 0 db that would put the noise at -131 dBW at 3 kHz.
Now lets run those noise figures for each octave, weight them and add up the results!
Freq.....Harmonic Level...Sensitivity...Regulator...Noise
60...........-6....................-55.............55......1.32069E-07
120...........0....................-30.............55......4.68599E-06
240..........-6....................-15.............52......1.86553E-05
480.........-12......................0.............49......7.42679E-05
960.........-18......................0.............46......5.25777E-05
1920........-24.....................6.............43......7.42679E-05
3840........-30.....................6.............40......5.25777E-05
7680........-36....................-5.............37......1.04906E-05
15360......-42....................-6.............34......6.61914E-06
30720......-48....................-6.............31......4.68599E-06
61440......-54....................-7.............28......2.95666E-06
122880....-60.....................-8.............25......1.86553E-06
................................................................0.000303782 Volts
Probably why most folks don't use the 7915!
Now the other issue is the actual regulator noise will increase that a bit. Or in practice don't oversize you filter capacitor because at some point it is not the ripple and harmonics it is the self noise. Also a calculable number.
Last edited:
a) I don't know what you are talking about within the first paragraph.
b) Fig. 31 shows 75dB of rejection @3KHz, far lower than the 30dB you mentioned.
c) 60dB gain for a line preamp doesn't make any sense.
d) I have no idea what the last paragraph table means and how you used it to prove your point.
You are jealous because you don't found the path to the game's registration.
I won 10 points !
Oh, sorry, so you shifted the discussion from the 7815 here to a 7915, known as having a much worse frequency response. Good to know.
Still, the rest doesn't make much sense. But the 7915 has it's own noise of 175uV in 100KHz BW, much larger than those 26uV coming from that so-called "ripple" @3KHz. Plus that we have something called "decoupling".
You would pull anything out of thin air to make a point, isn't it?
- Status
- Not open for further replies.