Dear all
what about setting the gain of the phono stage to a similar signal level like CD-Players have? The advantage would be a line stage which needs small gain or just only a buffer stage.
How do you handle your input signal of a CDP in a line stage with 20dB gain and a input sensitivity of 500mV at the input - attenuate it down to the level of your phono stages?
what about setting the gain of the phono stage to a similar signal level like CD-Players have? The advantage would be a line stage which needs small gain or just only a buffer stage.
How do you handle your input signal of a CDP in a line stage with 20dB gain and a input sensitivity of 500mV at the input - attenuate it down to the level of your phono stages?
It gives an indication of the asymmetry of the mains waveform?
Output = positive peak voltage - negative peak voltage.
A useful real-world test when fretting about mains DC offset and transformer saturation...
All the same to MIT snobs. I was actually at U of U for grad school- BYU wouldn't have even let me on campus.
John C, yes, I understand that the very lowest noise FETs will be n-channel, but when we're talking about MM, surely there are some commercially available p-channel which are quiet enough? I'm not asking for you to use them in your work, you don't need them, but for my design topology, a p-channel is unfortunately required. If you don't know of any, it's OK to say, "Don't know of any, they're all over 50nV/rtHz and will all have noise figures greater than 3dB with an MM source." If you (or anyone else) do know of any, I'd appreciate hearing about it.
Diagram states 120/48 VCT....without winding polarity indication the schematic is ambiguous.
Eric.
SY, IF you have to use a P channel fet as the source, then a NOISE SELECTED version of the PS geometry, or process 88 by National, is probably your best bet. I would say that you 'lucked out' if you got a very low noise P channel.
Usually, we get about 10 db better performance with a 2SJ74, BUT if you want lowish capacitance (for a P channel) then your selection is OK.
Kirkwood Rough is the expert here.
Usually, we get about 10 db better performance with a 2SJ74, BUT if you want lowish capacitance (for a P channel) then your selection is OK.
Kirkwood Rough is the expert here.
Thanks, John, I'll check those out. The pfets I used in my own preamp were old Siliconix (from a year or two before they disappeared), and given the incorrect pinout, I have no idea of what's actually inside. And "low noise" is relative- you can take quite a bit of en before rising audibly above the Jonson noise of a 5mV MM.
A winner!
Now I am deliberately powering a small transformer from an AC line that does have a DC offset due to unequal loading. I am using a 40 watt light bulb in series with a 1N4004 diode across the same AC line that powers the transformer. I have loaded the transformer to 30% of the rated capability. Will the RMS AC output voltage decline, quickly, slowly or not at all?
Now that the 'side chat' has dimmed momentarily, I would like to point out a few areas of expertise in electronics that overlap, but are not necessarily the same thing.
First, the active device designer. This is the IC or discrete fet, tube, or bipolar designer. This, from an outsider like myself, appears to be more physical processes, some quantum mechanics, and a LOT of 'cookbook' formulas as to how to make a successful specific part. This is not my area at all, and I would think that it would be the domain of Kirkwood Rough and others contributing here. I know from interaction with the principals, that it is NOT always easy or obvious to get a part right.
My area of design knowledge and expertise is circuit topology and parts selection.
To make a successful topology, you have have to know the devices that you are working with, their qualities and their problems. You do NOT need to know the deep internal workings of the devices, just how they behave externally. What makes them noisy, prone to breakdown, and how to get them operate with low distortion. I can usually think of an active device as a 'black box' and use it that way. Physics may be nice to know, but it is not necessary. Just the general engineering rules of how a device behaves.
Device selection is the frosting on the cake, in circuit design. IF you know your active parts, and how they behave EXTERNALLY very well, you can optimize the selection and operation of the active parts in an individual circuit. For example, what is the difference between a low Gm jfet and a high Gm jfet? What about non-linear input capacitance? What is too much? What about noise? Etc. Etc.
Then knowing this, a person can become famous by developing unique circuit configurations that others have not yet looked at. This is what I do for a living.
Now the third group of electronics specialists are the people who build, test, and repair circuit designs, often not their own. This is a whole other area of specialty that is sometimes under-appreciated, but it is very necessary, and worth its weight in gold, when one can rely on someone with the right skills. A degree in engineering or physics is not really necessary, but sometimes useful, but an intuitive understanding of how circuits work is invaluable.
As I said before, these are 3 separate areas of electronics skill, that sometimes overlap, but they are NOT the same skills. So when, even in jest, I am asked to 'design' a part, the request is absurd, any more than I could design an automobile tire.
Alternatively, some people come to me with their broken CD player, thinking that I should be able to fix it. No way, unless it is obvious. I just don't have the skill or experience. Finally, those who lack a specific area of expertise should not presume that they understand everything in this area. This is where someone can really look foolish, trying to tell others about what they really don't understand in the first place. Enough for now.
First, the active device designer. This is the IC or discrete fet, tube, or bipolar designer. This, from an outsider like myself, appears to be more physical processes, some quantum mechanics, and a LOT of 'cookbook' formulas as to how to make a successful specific part. This is not my area at all, and I would think that it would be the domain of Kirkwood Rough and others contributing here. I know from interaction with the principals, that it is NOT always easy or obvious to get a part right.
My area of design knowledge and expertise is circuit topology and parts selection.
To make a successful topology, you have have to know the devices that you are working with, their qualities and their problems. You do NOT need to know the deep internal workings of the devices, just how they behave externally. What makes them noisy, prone to breakdown, and how to get them operate with low distortion. I can usually think of an active device as a 'black box' and use it that way. Physics may be nice to know, but it is not necessary. Just the general engineering rules of how a device behaves.
Device selection is the frosting on the cake, in circuit design. IF you know your active parts, and how they behave EXTERNALLY very well, you can optimize the selection and operation of the active parts in an individual circuit. For example, what is the difference between a low Gm jfet and a high Gm jfet? What about non-linear input capacitance? What is too much? What about noise? Etc. Etc.
Then knowing this, a person can become famous by developing unique circuit configurations that others have not yet looked at. This is what I do for a living.
Now the third group of electronics specialists are the people who build, test, and repair circuit designs, often not their own. This is a whole other area of specialty that is sometimes under-appreciated, but it is very necessary, and worth its weight in gold, when one can rely on someone with the right skills. A degree in engineering or physics is not really necessary, but sometimes useful, but an intuitive understanding of how circuits work is invaluable.
As I said before, these are 3 separate areas of electronics skill, that sometimes overlap, but they are NOT the same skills. So when, even in jest, I am asked to 'design' a part, the request is absurd, any more than I could design an automobile tire.
Alternatively, some people come to me with their broken CD player, thinking that I should be able to fix it. No way, unless it is obvious. I just don't have the skill or experience. Finally, those who lack a specific area of expertise should not presume that they understand everything in this area. This is where someone can really look foolish, trying to tell others about what they really don't understand in the first place. Enough for now.
Last edited:
Very well said.
Good things happen too when circuit topology and parts selection types get drunk with the active device designer types. I have first hand stories of Dallas pizza bar napkin designs arriving in the mail 6 weeks later. Two went on to be PMI/AD big sellers.
The build, test, and repair circuit types get to tell the circuit topology and parts selection types where they screwed up.
Eric.
Good things happen too when circuit topology and parts selection types get drunk with the active device designer types. I have first hand stories of Dallas pizza bar napkin designs arriving in the mail 6 weeks later. Two went on to be PMI/AD big sellers.
The build, test, and repair circuit types get to tell the circuit topology and parts selection types where they screwed up.
Eric.
Not so fast Ed
What is the reason you used two transformers in series/parallel config to get to the pos and neg peak line voltage? Seems to me you introduced an unnecessary inaccuracy by using two transformers that may not be well matched as to exact turns ratios.
We're dealing with presumably small DC voltages. I believe a better solution would be (more accurate) to use a single transformer and a matched pair of diodes?
jan didden
I was wondering the same thing. Wouldn't it be better to avoid transformers altogether with something like this.
The problem with using a transformer is it doesn't pass DC so you'll miss half of what you're trying to measure, unless you just want a measure of asymmetry, ignoring actual DC offset.
Disclaimer: I'm assuming everyone in this thread knows how not to electrocute themselves with a multimeter, but for the record - "Please be careful when dealing with mains or other high voltages".
The problem with using a transformer is it doesn't pass DC so you'll miss half of what you're trying to measure, unless you just want a measure of asymmetry, ignoring actual DC offset.
Disclaimer: I'm assuming everyone in this thread knows how not to electrocute themselves with a multimeter, but for the record - "Please be careful when dealing with mains or other high voltages".
Attachments
Last edited:
I seem to spend a lot of time trying to explain physics to engineers, so they know why the thing they are trying to do does not work. In far too many cases, the basic physics knowledge of the engineer is insufficient to understand the explanation. In the UK, and I suspect elsewhere too, the training of engineers contains far too little physics and maths. Instead, they are just taught to plug numbers into formulas and follow 'engineering rules'. As a result, much of the innovation in engineering is actually done by physicists. (OK, I admit it, I am biased; I am a physicist).john curl said:
This may be true today everywhere ( Physics and Maths fundamentals for Engineering Study)- I am not sure?
I had to study them seriuosly 30 years back in India before studying Engineering. Without Maths and Physics how can anybody understand Engineering. However there are tools avaialble as free to Electronics design that many use them without understanding that the models are not close to real devices or even knowing Ohms law start simulating circuits.
However I am puzzled why many not acknowledging that the Audio Listening experience is personal and unique not necessarily identical to another listner's experience.
kannan
- Status
- Not open for further replies.