In terms of linearity, no. In terms of noise, yes- D3a being one of them.
http://www.diyaudio.com/forums/tubes-valves/164920-guess-tube.html
At 8.5VRMS out, -70dB second, nothing higher visible above -100dB
Hi,
Thank you.
I don't mind at all 0.1% THD, not even 1%, as long as there are no measureable higher harmonics, say above the fourth.
We saw here different view, probably based on actual measurements
One thing is that I don't care at all about small amount of 2nd harmonic distortion.
Second is that differential configuration, when the two sections are matched closely, cancel even harmonics distortion,
Indeed, it is so. I have taken all priorities into consideration. So far, I don't see why and how the circuit I have in mind doesn't meet well all requirements and priorities. All else being equal, I prefer that there will be no measureable harmonic distortion above say the fourth.
So far, I don't see how different architecture may be better for my priorities, namely low noise, low microphonics, enough gain, good immunity to RFI, flat wide-band, frequency response, ample headroom and instantaneous recovery from overloads.
Thank you.
Thank you.
I don't mind at all 0.1% THD, not even 1%, as long as there are no measureable higher harmonics, say above the fourth.
We saw here different view, probably based on actual measurements
One thing is that I don't care at all about small amount of 2nd harmonic distortion.
Second is that differential configuration, when the two sections are matched closely, cancel even harmonics distortion,
Indeed, it is so. I have taken all priorities into consideration. So far, I don't see why and how the circuit I have in mind doesn't meet well all requirements and priorities. All else being equal, I prefer that there will be no measureable harmonic distortion above say the fourth.
So far, I don't see how different architecture may be better for my priorities, namely low noise, low microphonics, enough gain, good immunity to RFI, flat wide-band, frequency response, ample headroom and instantaneous recovery from overloads.
Thank you.
Yes there are. Try the 6SF5GT or the 6SQ7GT. Same behavior down to very low output and better and better at higher level. In a simple common cathode amp you can get 20-30% more output for the same distortion for any given power supply. With higher load than ususal (using CCS) even better.
At the same time less problems with biasing on average.
Forget distiortion at very low levels. You can get low distortion without using special valves. In the end you need to look at the circuit as a whole. I don't think A. Wright wanted high distortion from his circuits. Actually all his circuits are quite well behaved in this department.
From my experience the D3a can be easily anything but low microphonic device and well selected standard hi-mu triodes specifically designed for that are usually better! Sometimes as in my case with the 6SF5GT it is even better than low noise ECC83's. It might seem strange but the full glass solution and special filaments do not guarantee the best result by default. This is also true for power devices. I still have to find a quiter power device than my 45's! Could this be something related to its "magic sound"? Maybe....
As a rule of thumb, with high gm devices there are more variations from lot to lot and device to device, more potential issues with osciallations.
I cannot help you more than this as I have found my solution for phono preamp and valves have been ruled out!😀
You can have it if you want as this is a finished product.
Last edited:
I am not a fan of interstage transformers. As a general rule I regard transformers as being much less like ideal components than capacitors. Transformers need a reasonably well-defined impedance on both sides, which cascode may not provide.Joshua_G said:
In the end it is not a big difference in practice and you get basically a bit lower input capacitance of a triode connected D3a with higher gain.
In a phono preamp input the relevance of input capcitance can be anything. Too high or beneficial (i.e. useful) when looking for the best input impedance of the amp to match a specific cartridge.
D3a also gives you dramatically lower noise and lower plate resistance. Ditto triode-strapped C3g (at least plate resistance- I don't have noise data, but the transconductance is nice and high). The high plate resistance and low transconductance of ECC83 and 6SF5 make design quite a bit trickier, but you get the same problem with cascodes (which is why I ended up abandoning that approach).
You have to take those numbers with a pillar of salt. They are generally useful at RF, but often wildly off for audio. One major issue is the corner frequency of the noise spectrum. See, for example, Vogel's article in Linear Audio vol. 4. The bottom line of his paper was that there's a lot of variance between tube types and from sample to sample of the same tube, none of which correlated well to datasheet numbers. ENR is a good starting point, but isn't even close to the final word.
Hello ,
you can reach very low distortion with the D3a in a RIAA circuit :
Kind regards ,
Alexander .
you can reach very low distortion with the D3a in a RIAA circuit :
An externally hosted image should be here but it was not working when we last tested it.
Kind regards ,
Alexander .
If you look hard enough you can always get measurable components at any order. Some of them may be just noise; some of them may be FFT windowing artifacts. -100dB at a few volts out should be inconsequential for the first stage of a phone preamp. The danger of straining at gnats in electronic design is that you may end up swallowing a camel!
- Status
- Not open for further replies.