Amplifier-corrector RIAA
In a previous article, I talked about the Vega-EP-110-stereo player. We examined its circuitry and began to talk about its restoration. In this article I will tell you about the creation of a new block of the amplifier-corrector with the RIAA characteristic.
As always, I’ll start with a little tour of the topic.
RIAA and phono stages.
What is phono-correto and why is it needed? What is RIAA? Vinyl records are recorded not with a uniform frequency response, but with a rather peculiar one. It has a strong roll-off at low frequencies and quite a lot of +db at high frequencies. This is due to trade-offs between recording quality and the mechanical capabilities of recording systems. For example, low frequencies require a larger groove size, which is why the dimensions of the needle must be larger for normal reading, but then it will begin to read finer irregularities – high-frequency ones – worse. And vice versa. But the golden mean does not give a uniform frequency response. So that all the records were universal and the players too, a standard was adopted. This standard is named after the acronym Recording Industry Association of America (American Recording Industry Association) RIAA. Before the RIAA, there were other standards, but they somehow disappeared, and the RIAA became generally recognized. You can read more about the history on Wikipedia or various thematic resources about vinyl. There are quite interesting and curious cases with the history of vinyl.
Well, I can’t get past the fact that there were state standards in the union, there are vinyl recordings and discs.
https://records.su/pubs/6672864918404634.pdf https://files.stroyinf.ru/Data2/0/4294799/4294799008.pdf
The main thing that the RIAA standard tells us is that the frequency response of the recording has three inflections, which are described by formulas, either depending on the time 3180, 318 and 75 µs, or on the frequency 50.05, 500.5 and 2122.1 Hz.
First knee = 3180 µs, sets bass boost frequency = 50.05 Hz to reduce low frequency noise.
The second inflection = 318 µs, sets the frequency of the section of the low-frequency and mid-frequency regions = 500.5 Hz.
The third inflection = 75 μs, sets the frequency of the section of the mid-frequency and high-frequency regions = 2122.1 Hz.
Here is a picture from the Internet, what frequency response the records have and what an amplifier should have to restore it:
So. We took a record and with a pick-up needle we get music from it, but it is very unlike a familiar performer. To restore the frequency response, bringing it to a normal, even one, we need to make a corrector-amplifier that has an inverse recording frequency response. That is, reflect it vertically. It seems that the task is not difficult. Moreover, there are formulas for this.
Creating a phono stage.
In a previous article, it was told about the restoration of the device. I didn’t immediately like the types of capacitors there, and the difference between some ratings and the circuit hinted at a not very conscientious approach to the assembly of that time. Since I still had to change some rotten (in the truest sense of the word) electronic components, being a person not very far from electronics, I decided to do everything again. And even a little better.
Generally speaking, in one of the editions of my favorite art of electronics there is just an analysis of such an amplifier with the following characteristic:
To begin with, I formed in my head what exactly I want to improve and what I want to have as a result. Then I decided which blocks of Vega would remain untouched for now. And after such a mental compilation of the TK, he began to work.
Power unit.
It was decided to leave the transformer as it is, but change the voltage stabilizer. The decision was made to slightly better filter the voltage and slightly increase the load capacity. Since it is made in a single copy for yourself, you can not look at the price much and not optimize for mass production. I drew a diagram of a stabilized power supply. And even for growth, not with zener diodes, as it was before, but with linear stabilizers. You never know what kind of cascade I want to put there in the future. Maybe there are not enough resistors in front of the zener diodes …
After that, the old power board was measured and, in its dimensions, on its seat, developed a new one.
For those who do not develop boards or are far in the past, let me remind you that now a two-sided board costs the same as a single-sided board.
And when ordering them in China, you also don’t have to pay for the preparation of production.
And now the board is already in place. Alas, this is the only photo left.
Amplifier – corrector.
The scheme seems to be the most classic. The first stage is the corrector-amplifier itself. But the second one is far from often seen in such players. Many are limited to one stage and save one op amp. When one electronic unit is being made, without any plans for a series, I can afford to spend money on 2 DIP8 cases instead of one. Here I wanted to have a margin of amplification. That’s the problem. The chosen op amp, quite popular for audio circuits, has a unity gain bandwidth of up to 10 MHz. It would seem that turn the gain to madness, but remember that with the gain, the band will fall.
I can spin the maximum at 800 without leaving the sound band with a flat response. The level from the pickup needle is 3.5mV at best. That is, I will get 2.78V at the output. What about phase margin? And if I want a higher level at the output or the load will be with a certain capacity, well, there are all sorts of wires, or you never know what … I don’t want to get a generator instead of an amplifier. As a result, I think that the second cascade is not just a whim, but a well-founded decision. Next, I throw all sorts of components onto the circuit: R1, R2 (by the way, this is a universal place where it is convenient to put the capacitor), C33, C34, R15, R16, so that there would be an opportunity to play around with different options during debugging.
As a result, a completely universal scheme turned out.
The corrective circuit values are taken from the classics from Horowitz and Hill, they were slightly changed during the tuning process.
Next, the native board was measured. And a new one is drawn along its contour.
Nothing special either. Two layers and output elements. Why withdrawals? Don’t know. We have it all at home and it’s all of fairly good quality. Wima capacitors alone are worth something.
Here are the pictures in 3D.
But such a fee was sent to me from heaven.
Setting.
In general, the setup process seemed to me the most exciting of all this entertainment.
How can I check that I am accurately restoring the characteristic? There are several ways.
It is not necessary to have a frequency response meter in order to measure the frequency response. You can make a test plan. Connect the generator to the input of the amplifier. Set the desired value of the output signal amplitude on a conventional tunable generator. And change the frequency of the signal on it in steps. Watch the amplitude of the output signal at the output of the device under study and mark each step in Excel. After that, just build a graph there. For visual correspondence, the scale should be made logarithmic. Better yet, build a reference using the RIAA curve function and apply the measured value to it.
And there are such devices as a frequency response meter. It simplifies the job and saves time. It is especially pleasing when you select some components and you do not need to run through all the points with a generator to measure new results.
Our civilization develops according to an exponential (it seems) function. And what I could only dream of in the radio club, turning the knobs of the generator, now I have it in the attachment to the laptop (there is a generator, and an oscilloscope, and a frequency response, and a spectrum analyzer, and something else, all in one). And working with markers really simplifies the work.
As a result, the block acquired the following configuration:
R11, R12, C33, C34 not installed.
C21, C22 became 1nF, and C31, C32 became 3.9nF.
But the configured board is in its workplace.
The amplifier block itself is placed inside a metal shield inside the case.
Then there is still a little work left in appearance, and if there is time to redo the speed control unit. I don’t like him very much. And now you can already enjoy the music.
