|
Audio Asylum Thread Printer Get a view of an entire thread on one page |
For Sale Ads |
32.213.224.132
In Reply to: RE: Everyone said it could not be done??? posted by TomWh on November 10, 2014 at 07:18:06
If your RIAA filter relies on the plate impedance of the tube for correctness, it will never be quite correct, and it will change with the age of the tube. Obviously there are lots of tube phono stages that ignore this and do just fine, so whether it is an issue worth worrying about or not is up to you.
Ways around it include using pentodes for gain (where the plate load will dominate the output impedance), using a cathode/source/emitter follower before the filter for increased consistency, including a series resistor which will dominate the impedance, etc. DHTRob's solution doesn't strike me as ideal.
Follow Ups:
The influence of the first stage output impedance is reduced by its proportion to the RIAA filter impedance. This is as true for pentodes as for triodes, and is of roughly the same magnitude.
First off, I am talking of passive RIAA networks between two stages, and driven by a low impedance, loaded by a high one.
Typically, let's assume a first stage of 10k plate impedance, an RIAA network of 100k impedance, and the second stage with a 1.0meg grid resistor. Drift in the 10k plate resistance is 10% of the 100k input impedance of the network, so its effect is reduced by a factor of 10. A typical pentode operates as a pentode might have a plate resistance of 1.0meg, which is 10 times that of the RIAA network, so its drift is again reduced by a factor of 10. Six of one, a half-dozen of the other.
I have used both approaches with success.
Hey Paul
How is he getting away with interstages on both tubes. I was under the impression that the choke (inductance load) varies changing the riaa curve.
How much does the resistance change and how much does that effect the curve? I like the natural way interstages handle music and have them any where I can put them. Most everyone in the past have said it can not be done on the riaa correction.
So I am wondering how far off the curve that a circuit like dht Rob's would go?
Thanks Tom
If you will share a link to the circuit, I will comment. Until then, and in general, if the inductance is large enough, or the driver plate impedance is low enough, it will not affect the RIAA within the audio band.
As a further comment, there are a few nutcases (I admire that!) who cling to the very old LC constant-resistance filters for equalization. I've never heard one myself, but the elegance is very attractive to me - and they are especially well-suited to transformer coupling.
Paul
Hope this link works http://www.dhtrob.com/ He used 5842/5842 which is about as good as it gets for gain vs low impedance also did one with 5842 /d3a combo. The choke is rated in ohms and I am use to seeing it in impedance hys.
If you have the time it would be great to hear your insights in a design like this.
Thanks Tom
Hmmm... not enough information about the transformers' inductance, but they appear to be rated 5K:5K which is an appropriate load for the 5842. I only looked at the 5842 version, since I am more familiar with that tube. Assuming the transformers have decent bass response, the RIAA will not change that. Looks pretty reasonable to me.
The RIAA itself appears to not take into account the tube's plate resistance. Since the unbypassed 110 ohm cathode resistor adds about 4500 ohms to the 2000 ohms internal impedance, that would be a significant problem - I estimate errors of 4dB or less. I have not read the rest of the site, so maybe there's something I missed. I am also assuming that the DC resistance of the transformer windings is negligible relative to the 6500 ohm source impedance.
I do understand the desire to keep the RIAA network impedance as low as possible, and the unbypassed resistor does stabilize the effective plate impedance. I also suspect the 10K resistance might be chosen to optimize the high-frequency behavior of the interstage transformer, so you probably don't want to change that without a lot of further study.
Since I have a spreadsheet at hand, I will offer what I think is a more accurate RIAA network that retains the 10K resistor:
Change 1K45 to 2K37
Change 0.22uF to 0.13uF (0.1uF plus 0.033uF in parallel)
Change (68nF+6nF8 in parallel) to 47nF
This should give errors of less than 0.2dB.
Earlier this year I built a dual mono phono stage for myself, where the two gain stages are inductively plate choke loaded and RIAA EQ is performed split-passive. RIAA curve tracking is very accurate: I calculated the initial component values and fine-tuned using my AP. You need very good plate chokes and shielding around them. As Paul said above, you need to use high inductance chokes and/or low-medium impedance tubes.
Edits: 11/10/14
Donald
What chokes/tubes did you use? Also could you have used interstages and avoided the coupling capacitors?
If you do not mind letting everyone see your design I would love to see a schematic. Also any information on how it sounds in your system would be great.
Thanks Tom
I use an EC86 for the input tube and 6J5 for the second stage, with both loaded by Audio Note UK plate chokes. Traditional 1:1 interstages could not be used because they don't have high enough primary inductance. I capacitor couple the output of each tube, which then goes through a RC EQ stage.
I would say the sound is engaging, vivid, dynamic, and powerful. The better the components, the greater the layering in the recordings.
Post a Followup:
FAQ |
Post a Message! |
Forgot Password? |
|
||||||||||||||
|
This post is made possible by the generous support of people like you and our sponsors: