Still without casing but I couldn't resist having a listen. This riaa preamplifier is considerably "firmer", better defined, than my previous. I believe this is due to the input stage driving the riaa network. My preamplifier following is rumored to have a pretty low impedance, maybe as low as 4kohm, and that can explain that I subjectively think I'm lacking in treble. My previous riaa has an opamp output that probably has a lower impedance output. I will try verify or denounce this by inserting a mockup unity gain opamp buffer between the stages. With earth of the turntable connected there is virtually no hum that I can trace to OREAD, even with the casing not completed. I got absorbed in the music instead of building and went through a handfull lp sides before bedtime. I'm very pleased already 🙂
Last edited:
As the dad of the OREAD (the namesake is "Chestnutspread"), I'm delighted that your first impression is already positive.
Treble should not be missing under any circumstances.
Treble should not be missing under any circumstances.
I really should buy myself a test record. There are so many factors that can influence my perception. I shouldn't speculate. But what You are saying is that even if the input impedance of my "pre amplifier" is around 4 to 5 kohm that should not influence treble reproduction?
As for C1, in the circuit diagram I've built from it is 4,7 uF and I'm fully satisfied with the amount of low-end and lack of noise. But again, that still is my subjective opinion from just listening.
You ask a very good question,
which is causally related to the negative feedback and the frequency-dependent output resistance of the five transistor OP.
However, it seems to me that the question about your MM system, i.e. what terminating impedance it would like to see, is more important at the moment (practical & quick cause research). Too large a line capacitance could be the cause of missing treble. However, we older people no longer consciously hear treble at all, we have to admit that to ourselves.
Perhaps you are missing the upper mids, that would also be conceivable. If the terminating is not right, the upper mids can also end up in a small bathtub.
An additional impedance converter between the OREAD and a preamplifier whose input resistance is less than or equal to 10kOhm will solve any loading problem (of OREAD's outputside) - it won't do any harm. The EQ itself complies extremely well with RIAA-compliant equalization and is extremely accurate over the entire audio frequency range.
On my test turntable, the capacitance of the connecting cables alone provides the capacitive termination. This point should not be underestimated!
well,
greater (or equal) then 0,47µF is absolutly fine.
which is causally related to the negative feedback and the frequency-dependent output resistance of the five transistor OP.
However, it seems to me that the question about your MM system, i.e. what terminating impedance it would like to see, is more important at the moment (practical & quick cause research). Too large a line capacitance could be the cause of missing treble. However, we older people no longer consciously hear treble at all, we have to admit that to ourselves.
Perhaps you are missing the upper mids, that would also be conceivable. If the terminating is not right, the upper mids can also end up in a small bathtub.
An additional impedance converter between the OREAD and a preamplifier whose input resistance is less than or equal to 10kOhm will solve any loading problem (of OREAD's outputside) - it won't do any harm. The EQ itself complies extremely well with RIAA-compliant equalization and is extremely accurate over the entire audio frequency range.
On my test turntable, the capacitance of the connecting cables alone provides the capacitive termination. This point should not be underestimated!
😉
well,
greater (or equal) then 0,47µF is absolutly fine.
If your PREAMP actually has such an extremely low input impedance, then a buffer is recommended and would even be absolutely necessary.
Which preamplifier is it?
This means:
that
OREAD's PCB could also conquer the American continent 🙂. Or catd can instruct the Asian manufacturer deliver a certain quantity directly to a trustworthy address in the USA - doesn't something like that also work?
Last edited:
OREAD frequency response vs. output load resistance.
RIAA feedback network tweaked for loads around 18k
RIAA feedback network tweaked for loads around 18k
Last edited:
I'm spoiled by all service You supply. I should be able to plot these graphs myself if I wasn't that lazy. I've have to admit I've been listening to records instead. My direct reaction is the impedance can't be 4kohm because of the well defined bass I get.
The "preamplifier" is the defunct Najda project here at diyaudio. The technical specification of the analog input is unknown. From the forum I've only read speculations on the impedance being quite low.
I measured the capacitance of my consumer grade interconnect to be 540 pF
I have adjusted (increased) the level of the amplifier driving the tweeters which in turn are only experimentally set up and would benefit of being adjusted directionally (they are pointing a bit off my listening position).
I haven't measured capacitance of the interconnect from the tonearm to OREAD. It's complicated to disconnect it inside the record player.
If you have an automatic LCR meter, you could proceed as follows:
Carefully remove one of the connection wires from the system (using a suitable tool) or remove the entire system with the headshell from the TP4 or whatever /Ortofon ...
Without the system,
you can now connect the LCR meter directly to the end of the line, i.e. to the cinch plug or xlr, and determine the total resulting parallel capacitance.
Cp should definitely be less than 150pf.
Oh, thats very high 🙁.
Carefully remove one of the connection wires from the system (using a suitable tool) or remove the entire system with the headshell from the TP4 or whatever /Ortofon ...
Without the system,
you can now connect the LCR meter directly to the end of the line, i.e. to the cinch plug or xlr, and determine the total resulting parallel capacitance.
Cp should definitely be less than 150pf.
Oh, thats very high 🙁.
About the Najda project:
Totally unknown to me, but apparently a DSP-based concept, so first and foremost an AD converter and input multiplexer.
If this is the case, I would strongly recommend a buffer (impedance converter) between the EQ (the sources) and the multiplexer (input), from painful experience with DAQ systems from NI, this is really essential.
Totally unknown to me, but apparently a DSP-based concept, so first and foremost an AD converter and input multiplexer.
If this is the case, I would strongly recommend a buffer (impedance converter) between the EQ (the sources) and the multiplexer (input), from painful experience with DAQ systems from NI, this is really essential.
OREAD noise
of the input transistor(s) depending on the value of the input capacitor
Since I have a bunch of unpolarized 4.7µF film capacitors, I use them on the OREAD boards as the input capacitors (C2).
But there is already enough space for a 10µ WIMA MKS2-50 film capacitor, too.
For higher values than 10µ you will have to use electrolytic capacitors.
of the input transistor(s) depending on the value of the input capacitor
Since I have a bunch of unpolarized 4.7µF film capacitors, I use them on the OREAD boards as the input capacitors (C2).
But there is already enough space for a 10µ WIMA MKS2-50 film capacitor, too.
For higher values than 10µ you will have to use electrolytic capacitors.
Last edited:
@catd First I got a little surprised. But Your graph shows frequency dependent noise level plotted for different coupling capacitors. I'm trying to make sense of the information. The MM signal should in the region of mV's and the y-axis is in uV, it's a factor 1000, and subsonic to that. Doesn't that make it negligible? Of course a 22uF as @hbtaudio uses is better. The resonance frequency of tonearms is ofter in the region 6-9 Hz, isn't it, but I couldn't say If that has any influence. Speaker systems usually roll off quite steeply from at least 30 Hz, usually higher up. I would argue it isn't audible?
To be honest, I didn't think about this dependency before reading post #324 by Mark Tillotson.
So I simulated it in LTspice and got the result shown.
Let's wait and see if someone else says something else that sheds another light on this. 🙂
So I simulated it in LTspice and got the result shown.
me too
Let's wait and see if someone else says something else that sheds another light on this. 🙂
Last edited:
short and simplified, quick answer
I find it very difficult to comment on Mark's quick interjection and insight.
Firstly,
I have not written down the value >=470nF entirely without any background and specialist knowledge (and intent).
Secondly,
we can't just throw anything into the pot without possibly causing damage - causing confusion that leads to false conclusions (i.e. assumptions).
Basically,
one must know and distinguish that the processes at a PN junction (of a semiconductor) are pure physical effects, for our investigated frequency range below 100Hz special disturbances, just other than ... are.
Don't panic, the peak is always between 3 and 5 Hz.
What exactly is audible here is something the dedicated self-builder likes to try out:
he listens to OREAD
a) with 470nF MKP
and
b) with 47µV/100Vdc (in each case and compare),
the tonearm with cartridge is lowered into the groove, but the turntable drive must be completely de-energized, i.e. switched off.
Now we use headphones to listen deeply and for a very long time to the entire noise package generated.
#
-jX = 1/(2*PI*f*C)
R = const.
We distinguish between reactances (X) and active resistances (R) and the complex interaction of both candidates as impedance (Z).
Noise is a very, very special subject area, regardless of which type of noise you want to look at in detail. It makes no sense whatsoever to determine which X is present at f in terms of magnitude.
greetings,
HBt.
I find it very difficult to comment on Mark's quick interjection and insight.
Firstly,
I have not written down the value >=470nF entirely without any background and specialist knowledge (and intent).
Secondly,
we can't just throw anything into the pot without possibly causing damage - causing confusion that leads to false conclusions (i.e. assumptions).
Basically,
one must know and distinguish that the processes at a PN junction (of a semiconductor) are pure physical effects, for our investigated frequency range below 100Hz special disturbances, just other than ... are.
Don't panic, the peak is always between 3 and 5 Hz.
What exactly is audible here is something the dedicated self-builder likes to try out:
he listens to OREAD
a) with 470nF MKP
and
b) with 47µV/100Vdc (in each case and compare),
the tonearm with cartridge is lowered into the groove, but the turntable drive must be completely de-energized, i.e. switched off.
Now we use headphones to listen deeply and for a very long time to the entire noise package generated.
#
-jX = 1/(2*PI*f*C)
R = const.
We distinguish between reactances (X) and active resistances (R) and the complex interaction of both candidates as impedance (Z).
Noise is a very, very special subject area, regardless of which type of noise you want to look at in detail. It makes no sense whatsoever to determine which X is present at f in terms of magnitude.
greetings,
HBt.