Very nice work @curryman and very clever idea to install dipswitches on the bottom.
I'm still trying to understand what to do to improve low freq response.
Measured the RIAA caps and i have small deviations like 3400-3430pf instead of 3300 amd 1050-1065pf instead of 1000pf. So all my Riaa caps are sligtly higher in capacity.
Also output cap is 950nf-960nf. I have a microcontroller based precision capacimeter (Roman Black project) and i believe the measurements are very accurate.
What is your opinion fellows about where to look and what are components responsible for low freq response ?
I really like how the voices are projected and the 3D of the sound, a bit more bass is all that i need.
Adrian
I'm still trying to understand what to do to improve low freq response.
Measured the RIAA caps and i have small deviations like 3400-3430pf instead of 3300 amd 1050-1065pf instead of 1000pf. So all my Riaa caps are sligtly higher in capacity.
Also output cap is 950nf-960nf. I have a microcontroller based precision capacimeter (Roman Black project) and i believe the measurements are very accurate.
What is your opinion fellows about where to look and what are components responsible for low freq response ?
I really like how the voices are projected and the 3D of the sound, a bit more bass is all that i need.
Adrian
40V output voltage
I finished mine a few days ago.
My first impression is absolutely positive. I believe this RIAA will be a keeper.
I'm a bit concerned about my output voltage, measured at SENSEV and SENSEGND.
I'm not able to get 40.0V exactly with various combinations of RP2 and RP3 sensing resistors.
Right now I'm a 41.1V with 25K3 and 1K, but I've also been at 39.2V with 19K1 and 845R, and at 40,9V with 19K1 and 768R.
The Vf of my regulator LED is unknown, however according to specs max. Vf is 2V. However, I believe it's around 1.6V.
To feed the regulator I use a 48V/500mA POE injector wall-mart. I measure 48.7V.
Opamp is THS4031C and regulator is IRF510.
By the way, I use BC550B transistors with legs crossed, but will probably buy 2SC2547 soon.
Enclosure used to be an old network disk, and the PCB fits in just nicely.
How important is it to be at exactly 40V for this circuit?
I'm reluctant to continue on my experiments with different resistor values, fearing I will damage the PCB traces.
Thanks,
Jan
I finished mine a few days ago.
My first impression is absolutely positive. I believe this RIAA will be a keeper.
I'm a bit concerned about my output voltage, measured at SENSEV and SENSEGND.
I'm not able to get 40.0V exactly with various combinations of RP2 and RP3 sensing resistors.
Right now I'm a 41.1V with 25K3 and 1K, but I've also been at 39.2V with 19K1 and 845R, and at 40,9V with 19K1 and 768R.
The Vf of my regulator LED is unknown, however according to specs max. Vf is 2V. However, I believe it's around 1.6V.
To feed the regulator I use a 48V/500mA POE injector wall-mart. I measure 48.7V.
Opamp is THS4031C and regulator is IRF510.
By the way, I use BC550B transistors with legs crossed, but will probably buy 2SC2547 soon.
Enclosure used to be an old network disk, and the PCB fits in just nicely.
How important is it to be at exactly 40V for this circuit?
I'm reluctant to continue on my experiments with different resistor values, fearing I will damage the PCB traces.
Thanks,
Jan
Attachments
Sound differences?
Not wanting to be controversial like some audio personalities, as John Curl or even Stan Curtis, but I also noticed a certain "dryness" in bass on this project, even though the measured response is impeccable. But I had a possible clue to explain the phenomenon.
In my prototype I use transistors BCY58 with +B=30V, to not damage them (by the way, do not worry about small variations in the +B, this product tolerates well variations). When I measured the distortion at 1kHz and 10kHz, the THD values were expected. But at 77Hz for example (bass), I measured somewhere around 0.12% (or 0.22%, I do not remember, I am with measurements here, are at home), with some harmonics (in this case the harmonic profile is perfectly expected/normal) . Perhaps this explains something. (((but my measurements have a lot of 60Hz noise due to my all-unbalanced and not shielded gig))). Maybe by +B/trasistor choices results in lower OL and since bass uses high gain...
Different transistors with different beta and different Early voltage (and +B values, maybe) probably result in different values and may probably different signature sound.
Currently my prototype is with a friend (very happy with the sound), since I'm a fan of valves and I use a pre with tubes, which sounds less dry in bass despite not measure big difference in bass THD (but different harmonic profile) or measured response. Alas, probably I put this High Octane pre in my second best in my personal preferences, ie, the sound of this pre worth it!
Not wanting to be controversial like some audio personalities, as John Curl or even Stan Curtis, but I also noticed a certain "dryness" in bass on this project, even though the measured response is impeccable. But I had a possible clue to explain the phenomenon.
In my prototype I use transistors BCY58 with +B=30V, to not damage them (by the way, do not worry about small variations in the +B, this product tolerates well variations). When I measured the distortion at 1kHz and 10kHz, the THD values were expected. But at 77Hz for example (bass), I measured somewhere around 0.12% (or 0.22%, I do not remember, I am with measurements here, are at home), with some harmonics (in this case the harmonic profile is perfectly expected/normal) . Perhaps this explains something. (((but my measurements have a lot of 60Hz noise due to my all-unbalanced and not shielded gig))). Maybe by +B/trasistor choices results in lower OL and since bass uses high gain...
Different transistors with different beta and different Early voltage (and +B values, maybe) probably result in different values and may probably different signature sound.
Currently my prototype is with a friend (very happy with the sound), since I'm a fan of valves and I use a pre with tubes, which sounds less dry in bass despite not measure big difference in bass THD (but different harmonic profile) or measured response. Alas, probably I put this High Octane pre in my second best in my personal preferences, ie, the sound of this pre worth it!
That then means you don't know if the differences are from RIAA accuracy. It may well be that the one with the new caps is less accurate, who knows.
Jan
The apparent bass dryness might be fixable with a swap to the polarity of the output follower.
At present the output stage isn't output current invariant - in that because the load is in parallel with the CCS to 0V there's a varying current drawn from the supply. But changing Q3 to a PNP (2SA1085 is the complement) and moving the CCS to be fed from the positive supply fixes this - then the output EF is in effect a shunt and the CCS to the +ve rail ensures no variations on the rail.
I'm not sure if the shift in DC potential at the output by 1.2V as a result of the gender swap has any adverse effect - perhaps a quick simulation would answer that.
At present the output stage isn't output current invariant - in that because the load is in parallel with the CCS to 0V there's a varying current drawn from the supply. But changing Q3 to a PNP (2SA1085 is the complement) and moving the CCS to be fed from the positive supply fixes this - then the output EF is in effect a shunt and the CCS to the +ve rail ensures no variations on the rail.
I'm not sure if the shift in DC potential at the output by 1.2V as a result of the gender swap has any adverse effect - perhaps a quick simulation would answer that.
This may help those who wish to modify their preamp.
http://www.diyaudio.com/forums/blog...stor-ltspice-sim-phonostage-3-transistors.asc
I'm planning to try the 2Sc2547 and the BC337-40 in the first stage of the preamp as I always found the BC550C a bit noisy.
Sharif.
http://www.diyaudio.com/forums/blog...stor-ltspice-sim-phonostage-3-transistors.asc
I'm planning to try the 2Sc2547 and the BC337-40 in the first stage of the preamp as I always found the BC550C a bit noisy.
Sharif.
Last edited:
Jan,
Really, in practice they are even better than most signal transistors in this parameter: I've measured myself gain of 500 to ~750 in my BCY58 specimens and I used/selected the higher gain ones. Leakage current are OK, in nA range.
Abraxalito,
Good idea! Like Power Follower from Ciufolli, another guy to use the zero PSU current variation, and various other examples exists.
I made some amps in that style and always liked the sound, and my friends are impressed with.
Discrete design: More on 2-transistor shunt-feedback amplifiers
Maybe the THD becomes a little worse like Self example (compared with original) but this values/change are very low and worth changing if bulder wants it.
Really, in practice they are even better than most signal transistors in this parameter: I've measured myself gain of 500 to ~750 in my BCY58 specimens and I used/selected the higher gain ones. Leakage current are OK, in nA range.
Abraxalito,
Good idea! Like Power Follower from Ciufolli, another guy to use the zero PSU current variation, and various other examples exists.
I made some amps in that style and always liked the sound, and my friends are impressed with.
Discrete design: More on 2-transistor shunt-feedback amplifiers
Maybe the THD becomes a little worse like Self example (compared with original) but this values/change are very low and worth changing if bulder wants it.
Last edited:
But I thought you said above that you thought the lower gain would be responsible for lower bass? I don't understand now.
Jan
Me too... 
oops (my bad), to clear thins:
Maybe gain (beta) in some cases, "Early voltage" in others (ops, I forget the correct term to deviations to pure pentode action = partial internal resistance), or some other not-verified combination
So beta is out of question in my case I think
And all of this only to change the THD level/character (and perceived bass quality, perhaps), and response is impeccable unless one use very very low beta defective transistors
oops (my bad), to clear thins:Maybe gain (beta) in some cases, "Early voltage" in others (ops, I forget the correct term to deviations to pure pentode action = partial internal resistance), or some other not-verified combination
So beta is out of question in my case I think
And all of this only to change the THD level/character (and perceived bass quality, perhaps), and response is impeccable unless one use very very low beta defective transistors
Last edited:
An externally hosted image should be here but it was not working when we last tested it.
On the right shows the sim eq network for the original High Octane Riaa Preamp.
On the left it shows a modified sim eq network.
The BC550 and the BC337 did not change the response.
R8 was 720K now 700K
R10 was 2K7 now 820R
I've not tried measuring the response of my built preamp.
Sharif.
Since Sharif has posted the .asc on his blog, I quickly checked the PSRR. Rather a disaster at LF - injection via R2 is doing the damage. Its easy to fix up though - 100R and 470uF (RC filtering of the supply to R2) gives >20dB better PSRR @ 40Hz. I would hope this small tweak would fix up your apparent bass lightness. If not then swapping the output device polarity would be worth trying.
Hey there,
it's been a long time since my last visit to this thread as it's extremely busy lately.
Many thanks to all for your great works and discussions!
I see already the first mods show up - very interesting!
Many thanks also to those that like the sound of this little circuit, I'm flattered on all accounts
it's been a long time since my last visit to this thread as it's extremely busy lately.
Many thanks to all for your great works and discussions!
I see already the first mods show up - very interesting!
Many thanks also to those that like the sound of this little circuit, I'm flattered on all accounts