Unfortunately the layout, the PCB is lost on one of my hard disks, but never mind. Mk2 is the final version that several test setups paved the way for, my copies of the documents show 12 / 2017 as the timestamp. Here is a quick drawing of the entire RIAA equalizer without the simple 18V power supply.
Photos to follow. Have fun with the design. I myself used an SSM2134 from PMI, you can do without the Ckomp = typically 22pF with a clear conscience.
tbc
HBt.
Photos to follow. Have fun with the design. I myself used an SSM2134 from PMI, you can do without the Ckomp = typically 22pF with a clear conscience.
tbc
HBt.
Attachments
Here are a few quickly found pictures of the EXakt Mk2 project.
There is no noise, no distortion, no hum and it fits practically any MC cartridge. The accuracy of the equalization is astonishing, which is exactly what the name implies.
There is no noise, no distortion, no hum and it fits practically any MC cartridge. The accuracy of the equalization is astonishing, which is exactly what the name implies.
Attachments
If a good knight dares to follow in Haudegen's footsteps, I recommend the reproduction of this unusual (simple concept) EQ and will be happy to answer any questions that may arise.
HBt.audio
HBt.audio
24 hours later, and still no positive or the obligatory first negative voices. EXakt MK2 even overshadows Bob's circuit, which is currently being discussed again 😉.
kindly,
HBt.
kindly,
HBt.
OH NO! You are using BC550C for MC input, but think about the noise! It will create so much noise that will feed back to the pickup resonating ultrasonically exiting the pickup that that will cut up your precious records to plastic spirals and catch fire and burn your house down!
But seriously I thought transistors with lower base resistance would be better for MC preamps although they will typically have lower hfe than BC550C, What are your thoughts about the input transistor? The BC337-40 used for power filtering have lower rbb than BC550C and still decent hfe.
But seriously I thought transistors with lower base resistance would be better for MC preamps although they will typically have lower hfe than BC550C, What are your thoughts about the input transistor? The BC337-40 used for power filtering have lower rbb than BC550C and still decent hfe.
😎
Just for a moment, I'm in a hurry.
I can hardly believe it myself to this day, but at some point during the development process a simple and absolutely rebuild-proof circuit with available standard components was to be found. Originally, of course, I also tested ZTX xxx as well as various parallel circuits and other gimmicks. Everything actually worked perfectly in the simulation, but not at all in practice. Frustrated, I changed the method: brain plus pencil and paper, as well as a standard pocket calculator (for convenience), which one don't even need.
So yes, of course I completely agree with you, but it works perfectly well with the BC5xxC types - if you pay attention to two points (and I won't tell what they are, because they are more or less self-evident).
Amazing!
(The BJTs in my personal circuit are strictly selected.)
greetings,
HBt.
Just for a moment, I'm in a hurry.
I can hardly believe it myself to this day, but at some point during the development process a simple and absolutely rebuild-proof circuit with available standard components was to be found. Originally, of course, I also tested ZTX xxx as well as various parallel circuits and other gimmicks. Everything actually worked perfectly in the simulation, but not at all in practice. Frustrated, I changed the method: brain plus pencil and paper, as well as a standard pocket calculator (for convenience), which one don't even need.
So yes, of course I completely agree with you, but it works perfectly well with the BC5xxC types - if you pay attention to two points (and I won't tell what they are, because they are more or less self-evident).
Amazing!
(The BJTs in my personal circuit are strictly selected.)
greetings,
HBt.
@MartinX
The thoughts behind the selection:
The leitmotif simply was the question of "why should an LF small-signal standard all-world transistor not be suitable for its actual task?"
However, I have to admit that later in the process I specifically looked for original BCxxCs from Siemens or Philips, because goods from unknown manufacturers tended to have all kinds of conspicuous noise behavior - I'll just put it that way without opening a new can of worms.
That is absolutely right.
And one of the reasons why I shied away from publishing this circuit variant. But the circuit and its dimensions have been working flawlessly for eight years now.
A circuit without frills!
HBt.
I know that this circuit is very difficult to swallow. It is neither a pearl numbered 3, nor an ultra sweet salad. This RIAA EQ is simply nothing more than a well-known standard circuit with a few little tricks thrown in.
As for low noise and finding the most suitable bipolar input transistor, I can think of absolutely nothing clever.
Everything is out in the open. Of course we can prefer other BJTs, also other topologies, a very simple differential amplifier with for example two MJE371 and one BC414 may fulfill its primary task in the entire circuit just as well. Ultimately, it's all about the whole ensemble, the package, the finished product. Cost issues also play a role - and the small equalizing preamplifier does not have to do more than its actual task.
We can even do without a spectacular power supply design ... Only the mains transformer presents us with a challenge! The simplest solution is also the cheapest, outsource it and place it far away ... or use a tiny switching power supply, with other negative consequences.
Of course I don't recommend a switching power supply, hopefully that goes without saying.
HBt.
Psst
Have I been able to answer most of your (@MartinX) questions about choosing the BC550C? The secret lies in the operating point, the topology and the selection, as well as in the alignment (the required key data of the entire EQ).
As for low noise and finding the most suitable bipolar input transistor, I can think of absolutely nothing clever.
Everything is out in the open. Of course we can prefer other BJTs, also other topologies, a very simple differential amplifier with for example two MJE371 and one BC414 may fulfill its primary task in the entire circuit just as well. Ultimately, it's all about the whole ensemble, the package, the finished product. Cost issues also play a role - and the small equalizing preamplifier does not have to do more than its actual task.
We can even do without a spectacular power supply design ... Only the mains transformer presents us with a challenge! The simplest solution is also the cheapest, outsource it and place it far away ... or use a tiny switching power supply, with other negative consequences.
Of course I don't recommend a switching power supply, hopefully that goes without saying.
HBt.
Psst
Have I been able to answer most of your (@MartinX) questions about choosing the BC550C? The secret lies in the operating point, the topology and the selection, as well as in the alignment (the required key data of the entire EQ).
I have previously made a surface mount version of the Preconsonant RIAA stage and i have been thinking of making a MC variant by simply adding a pre stage, when i saw your version here i thought that might be a good idea because the transistors used in the pre stage are the same as in the Preconsonant, I use BC850C/BC860C from Nexperia. As you know there have never been a schematic posted in this forum that did not need modification 😉 so I immediately started to think about using another input transistor to possibly improve noise performance. In another thread someone pointed out 2SD2226K that i supposed to have low rbb and also high hfe so that in my mind could directly replace BC550C.
I simulated your pre circuit, gain is about 17 and that would be perfect in font of Preconsonant that has kind of low gain unmodified, the input transistor runs at about 550µA higher than the usual 100µA often seen in MM stages.
Here is a picture of my Preconsonant build, 1% NP0 ceramic capacitors used in riaa correction, I also added a TL431 based shunt regulator. Thanks for the food for thought HBt!
I simulated your pre circuit, gain is about 17 and that would be perfect in font of Preconsonant that has kind of low gain unmodified, the input transistor runs at about 550µA higher than the usual 100µA often seen in MM stages.
Here is a picture of my Preconsonant build, 1% NP0 ceramic capacitors used in riaa correction, I also added a TL431 based shunt regulator. Thanks for the food for thought HBt!
Gain should be (R8/R9 +1) approx +19dB over the thumb ...
But I could set up a new simulation, now I'm a bit confused.
To be honest, the following must be said about this very simple linear amplifier:
it
Take the skeleton, so to speak, or better formulated the construction plan and select more suitable BJTs. I wanted to guarantee reliable results. The result is really something to be seen and heard.
H(jOmega)
Further notes:
CoreSupplyVoltage
Vcc < +5Vdc
THD1k
always < 1m% under fullscale, max_swing
The upper cut-off frequency and the corresponding roll-off corrects the precisely defined fourth time constant of the negative feedback equalization network.
It is not a lowestnoise design, unfortunately that is a fact.
HBt.
it
- does not perform miracles
- fulfills its task according to its physical limits!
Take the skeleton, so to speak, or better formulated the construction plan and select more suitable BJTs. I wanted to guarantee reliable results. The result is really something to be seen and heard.
H(jOmega)
Further notes:
CoreSupplyVoltage
Vcc < +5Vdc
THD1k
always < 1m% under fullscale, max_swing
The upper cut-off frequency and the corresponding roll-off corrects the precisely defined fourth time constant of the negative feedback equalization network.
It is not a lowestnoise design, unfortunately that is a fact.
HBt.
Attachments
The parallel-regulator is a very clever decision on your part, and I like the consonant. I think @catd once praised this elektor equalizer very highly.
A significant improvement in SNR can already be achieved, by using Q1=2N2484 & Q2=BC177B & Q3=BC109C. If we can believe MC12 and trust the used models?!
Lift R10 down to 2k7.
#
This basic circuit simply has great potential - the entire EQ circuit is a blueprint. Metal film resistors with 1% tolerance and FKPs or MKPs with 2.5% tolerance already guarantee success.
I strongly recommend the replica, with subsequent comparisons (preferably double-blind, of course) to the well-known road sweepers (with the high-end coating) ..!
EXakt Mk2 is a workhorse.
@pfarrell Patrick,
are you on board ? - there is still a comparison to be made 😉.
greetings,
HBt.
Lift R10 down to 2k7.
#
This basic circuit simply has great potential - the entire EQ circuit is a blueprint. Metal film resistors with 1% tolerance and FKPs or MKPs with 2.5% tolerance already guarantee success.
I strongly recommend the replica, with subsequent comparisons (preferably double-blind, of course) to the well-known road sweepers (with the high-end coating) ..!
EXakt Mk2 is a workhorse.
@pfarrell Patrick,
are you on board ? - there is still a comparison to be made 😉.
greetings,
HBt.
... Hi, I'm always on board with building anything...current ever-expanding skillset not-withstanding....I don't have the immediate ability to fab your circuit from a schematic beyond a breadboard—which wouldn't be a fair comparison in my experience—
If you have gerbers then I would get some PCBs made, figure out a PSU setup, build it and report back ASAP. (I would need some lessons in how to test/match BC critters...)
If you have gerbers then I would get some PCBs made, figure out a PSU setup, build it and report back ASAP. (I would need some lessons in how to test/match BC critters...)