diyAudio

diyAudio (http://www.diyaudio.com/forums/)
-   Tubes / Valves (http://www.diyaudio.com/forums/tubes-valves/)
-   -   Need help on RIAA (http://www.diyaudio.com/forums/tubes-valves/213911-need-help-riaa.html)

Andersen 5th June 2012 08:28 AM

Need help on RIAA
 
I plan to build the phono stage pre-amp based on the below schematic:
http://i1235.photobucket.com/albums/...agepre-amp.jpg

I use this link to calculate the values of RIAA components by choosing the value of R1 (i.e. R17 in the schematic) at 68K in order to get the values of the other RIAA components close to standard values:
http://i1235.photobucket.com/albums/...ACalculate.jpg

I also note from the link that the driving tube V1 should have a low enough output impedance relative to R1 and that the following tube V2 should have an input impedance of at least 50 times R1 to prevent loading.

So my question is if my chosen value of R1 of 68K is appropriate in this application or which value of R1 I should choose for this application.

Thanks and regards
Andersen

revintage 5th June 2012 09:35 AM

Are you aware the input stage is a cascode? If so you should know Zout is in the ballpark of R7. So your values are way off. This means, if you go for R17=68k you should use 68k+22k=90k as R1 in the link.

Still you will only get ballpark figures. You must use a rev-RIAA (Hagerman) to get it right.

Andersen 5th June 2012 04:08 PM

Many thanks, Lars for your response. But I don't really think that R1=68K+22K as I did check some schematics with this calculators and the results are accurate. Anyway, my concern is how high or low the value of R17 should be in this specific case of 6922 cascode with Ra of 22K?

Thanks and regards

kevinkr 5th June 2012 04:12 PM

You should pay attention to Lars, he is absolutely correct.

If you want to use values close to those based on your original understanding change R17 to 47K, this should put you in the ballpark allowing for the cascode source impedance to be slightly lower than the plate load resistor of 22K. (The cascode will look like an rp of up to several hundred K ohms or so in parallel with the 22K plate resistor.) You have to take this into account in the selection of R17. I've designed a number of cascodes as well as conventional RIAA stages with this RIAA network. (First analyzed in detail by Stanley Lipshitz, University of Waterloo some 3 or more decades ago)

An inverse RIAA network is not strictly necessary, but you do need to know what the response is supposed to be at any given frequency and have a good way to precisely measure it. Expect to tweak values, particularly R17 to get it close. Don't neglect the miller capacitance in the second stage either if you want accurate HF response.

Note that the assumption made in all of these calculators is that V1 has a source impedance of 0 ohms, (usually an op-amp) your cascode does not and its source impedance is effectively part of R17.

Andersen 5th June 2012 04:25 PM

Noted with thanks. Can you all please kindly help me determine the values of the RIAA components: R17, R23, C7, C9 to complete the schematic?

Thanks and regards
Andersen

kevinkr 5th June 2012 04:30 PM

Quote:

Originally Posted by Andersen (Post 3049088)
Noted with thanks. Can you all please kindly help me determine the values of the RIAA components: R17, R23, C7, C9 to complete the schematic?

Thanks and regards
Andersen

See the above post. Try 47K for R17 with the existing values if you have them - otherwise I would recommend slightly higher effective resistor value for R17 in order to reduce the effect of variations in cascode source impedance due to aging (reduction of transconductance) in the cascode stage.

I'd aim for something in the 110K range total so say R17 = 88.7K + ~21K (cascode) for the basis of the calculations. This should be better for linearity in the cascode as well.

jan.didden 5th June 2012 04:33 PM

You can take R25/R26 in account by assuming they are in parallel with R17, and that R17 is, as said, the sum of the 'real' R17 + (Ra // Rp).
Then just use the online calculator.
And be prepared to tweak.

jan didden

kevinkr 5th June 2012 04:36 PM

Quote:

Originally Posted by janneman (Post 3049098)
You can take R25/R26 in account by assuming they are in parallel with R17, and that R17 is, as said, the sum of the 'real' R17 + (Ra // Rp).
Then just use the online calculator.
And be prepared to tweak.

jan didden

Or place R25/R26 before R17 as I do which gets rid of the implicit voltage divider in the network. (I was remiss for not mentioning this sooner.) Make sure the sum of R25/R26 and R17 does not exceed the maximum recommended grid circuit resistance of the 6922. Should be safe with 301K for R25/R26. C1 should be much larger to assure good response down to 20Hz.. I'd recommend at least 0.1uF.. (roughly -3dB @ 5Hz, -1dB @ 10Hz)

Andersen 5th June 2012 04:45 PM

Many thanks, Kevin. I will try your 110K recommendation. So in order to get the values of other RIAA components I will apply R1 at 110K in the calculator, right?

kevinkr 5th June 2012 04:56 PM

Quote:

Originally Posted by Andersen (Post 3049115)
Many thanks, Kevin. I will try your 110K recommendation. So in order to get the values of other RIAA components I will apply R1 at 110K in the calculator, right?

Yes, that's correct. Just remember to move R25/25 after C1 and before R17. Be prepared to do some tweaking to compensate for component tolerances and small variances from predicted circuit performance.

Usually what I do is go down to the next closest standard 1% resistor value and pad it up to the target value. (Then adjust by measurement if necessary) Try to purchase the caps from someone who can measure and select parts that closely match your target values. If you done really well usually all you will need to tweak is the value of R17.

Do not forget to take into account the approximate miller capacitance of the second stage as it is significant. Say it is 50pF then you would want a value that is 50pF less than the calculated value. Paralleling a few selected caps to get to the right overall value is a time honored technique. If you can get them polystyrene foil caps are excellent for EQ.

Consider infra-red or red leds for cathode bias (now really fixed bias) as opposed to RC based bias, good performing electrolytics at these low signal voltages can be hard to find.


All times are GMT. The time now is 03:54 AM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright 1999-2014 diyAudio


Content Relevant URLs by vBSEO 3.3.2