Author Topic: Tube phono preamps  (Read 10053 times)

Offline aquarat

  • Trade Count: (0)
  • AVForums Member
  • **
  • Posts: 84
  • ...and his trilogy of philosophical blockbusters.
Re: Tube phono preamps
« Reply #15 on: November 01, 2015, 05:37:07 PM »
I'd be interested in a valve-based phono preamp. I currently use a Toad-Audio unit sourced from the Turntable Guy (Tokai), which isn't at all bad.
"You know it's ready when the cat starts meowing."

Offline iondb

  • Trade Count: (+1)
  • AVForums Veteran
  • ***
  • Posts: 772
Re: Tube phono preamps
« Reply #16 on: November 02, 2015, 12:04:14 PM »
I will also be interested, but I am on the budget side of things.
Currently using a Project phono box which does MM duty. I already have a Karel Mars amp, so if the chassis could look similar, it will be great. Maybe same depth and then width as needed to fit everything. If it can service MM and MC for options further down the line if one decide to upgrade one's cart from MM to MC.  :roll:

Offline Ampdog

  • Trade Count: (+1)
  • AVForums Veteran
  • *****
  • Posts: 6,819
Re: Tube phono preamps
« Reply #17 on: November 03, 2015, 08:39:48 PM »
Fredep,

Very useful list that Millet one, BUT:

A cascade topology is mentioned - that is not the same as cascode. In cascode two valves are used in series (one on top of the other as it were).  The anode voltage on the bottom one (going to the cathode of the top valve) is somewhere in the 60 - 70s, maintained by setting the top triode G1 at some 60 - 70Vdc. The top valve anode operates at some 150 - 200V, in series with the usual load resistor to B+.

Cascode triode connections exhibit pentode gains and characteristics. The figures given in that list rather suggest one valve with a CCS as anode load. The last entry for a 6SN7 with gain of some 100 would indeed suggest a cascode. (Eg. ECC88 triodes in cascode give a gain of some 150+ if memory serves.)
Audio must be the only branch of engineering where lack of basics' knowledge is considered a superior form of wisdom. (Anon)

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • ***
  • Posts: 956
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: Tube phono preamps
« Reply #18 on: December 03, 2015, 07:49:54 PM »
I've built one RCA and two RJM type preamps. The RCA does the job well, but needs a high impedance input; higher than most amps are nowadays. The RJM is great, but the gain is a bit low. So I settled on the TCJ remake of the RCA.



Here is some photos of the build:






Input




The Aikido Cathode followers



RIAA filter section



Power supply




OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline ghostinthemachine

  • Commercial Member
  • Trade Count: (+13)
  • *****
  • Posts: 5,163
Re: Tube phono preamps
« Reply #19 on: December 03, 2015, 08:03:41 PM »
Clever enclosure!
Inactive

Offline handsome

  • Trade Count: (+1)
  • AVForums Veteran
  • ***
  • Posts: 843
Re: Tube phono preamps
« Reply #20 on: December 04, 2015, 06:58:21 AM »
I tweaked the RCA phono stage myself, slapped a buffer on the output, re-diddled the eq network and used a semi regulated PSU for the front end. It is a very pleasant phono stage - i prefer it to my ARC SP9 - but it has way to much input capacitance. High gm tubes are definitely the way to go but again it's the input capacitance that lets them down - unless of course you are using an MC cartridge.. As ampdog mentioned the cascode is the logical engineering solution but now one has to babysit a low noise input that has almost zero PSRR........

Offline handsome

  • Trade Count: (+1)
  • AVForums Veteran
  • ***
  • Posts: 843
Re: Tube phono preamps
« Reply #21 on: December 04, 2015, 07:06:05 AM »
Also the high impedance of the RCA requires a suitably high impedance eq network for both linearity and stability of eq characteristics and that gaily plays hob with the low noise requirements.......i really enjoy designing these stages so much conflict one has to deal with; brings out the (latent) Bruce-Lee-versus-ten-baddies in me!

Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • ***
  • Posts: 956
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: Tube phono preamps
« Reply #22 on: December 04, 2015, 09:13:28 AM »
I listened to some LP's last night, and I am very pleased with the sound :groovy: The EQ is a bit out though. All the singers sound timid and there are too mucht tops. Looks like the road to flat response is a bit longer than I thought :sd:
OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline handsome

  • Trade Count: (+1)
  • AVForums Veteran
  • ***
  • Posts: 843
Re: Tube phono preamps
« Reply #23 on: December 04, 2015, 09:42:41 AM »
I think Broskie's eq values are ballpark....You need to build an inverse eq network to measure accurately. use 0.1% or better resistors (you dont need many) and 1% caps........wish someone would give me the other Audio Precision they are not using anymore  :thinking:

Offline tangmonster

  • Trade Count: (-1)
  • AVForums Veteran
  • *****
  • Posts: 1,752
Re: Tube phono preamps
« Reply #24 on: December 04, 2015, 09:49:41 AM »
after recently spending a fair amount of time checking my d3A phono preamp RIAA curves I chose the route of chickening out and using online calculators and comparing them and using a function generator to measure real world.

http://mh-audio.nl/CalculateRIAA.asp

http://www.marignoni.it/ElettronicaHiFi/UTL/RIAA/Calcolo%20rete%20RIAA%20passivait.xls



My understanding is dat you will have to parralel your 100k and your 200k resistor so your Zin is 66k ohms (and tehcnically that 1Meg :) )

That gives 9.7k where you have 33k

and 33nF where you have 2 x 4.7nf in parralel

And then 11nF where you have 3.3nF

Please correct me if my understanding is wrong.

I built this design a few years back:

http://www.audiodesignguide.com/Phono/pre-phono4.html

Sadly it was lacking in the bass department and it soon became clear that the 50ohm with 270uF in parralel was attenuating low frequencies too much . I have since changed to IR led cathode bias for that d3a design.

Remeasuring with signal generator and comparing voltage gains at frequencies from 20hz and up showed i was very close , at least within 1 dbv or so currently.
« Last Edit: December 04, 2015, 09:51:43 AM by tangmonster »

Offline tangmonster

  • Trade Count: (-1)
  • AVForums Veteran
  • *****
  • Posts: 1,752
Re: Tube phono preamps
« Reply #25 on: December 04, 2015, 10:03:21 AM »
i think i need to correct myself there , shouldn't do this on a friday morning it seems.

You need to add roughly 62k for 12ax7 in parralel with 100k to the 200k in your calculations.

so 35k

3.1nf

and 9nf all checks out.


Offline Mars

  • Commercial Member
  • Trade Count: (+1)
  • ***
  • Posts: 956
  • “The path is the goal.” Mahatma Ghandi
    • Mars Tube Audio
Re: Tube phono preamps
« Reply #26 on: December 04, 2015, 12:53:40 PM »
Thanks tangmonster

That D3A design looks very interesting. I need to order some tubes!

Back to my existing design: You recommend me paralleling the 200k with a 100k? Q.e.d.

I used a 10nF cap in stead of the two paralleled 4n7's. Looks like I need to return there and get a 9nF rather.

The 33k is correct?

I'll check the online calculators.
OEM manufacture of Tube Audio
Power and Integrated Amplifiers
Phono Pre-amplifiers
DIY Kits
DIY Workshops

Offline tangmonster

  • Trade Count: (-1)
  • AVForums Veteran
  • *****
  • Posts: 1,752
Re: Tube phono preamps
« Reply #27 on: December 04, 2015, 01:08:38 PM »
this calculator give your 12ax7 output impedance as 38.46kohm

https://www.ampbooks.com/mobile/amplifier-calculators/output-impedance/calculator/


You need to add that 38.46kohm to the 200k when calculating your RIAA circuit

So.

I get

sticking to 200k as you have it.

34606ohm (where you have 33k , you might ad 1.5k to be more precise?)

3.1nF (where you have 3.3nF which is as good as it gets)
9.1nF (where you have 10nF, i'm sure you are going to be close enough with 10nf)

I messed around with my d3a's riaa but mostly the cathode bias to get a good riaa.

In the end i only trusted using an oscilloscope and function generator and scrolling through the frequency. But after i fixed the tubes cathode bias problems i was VERY close to perfect riaa using the calculators.

My understanding is for your circuit you need to make R1 = 238000 and do the calculation since you have a 200k resistor in your circuit and the tube output impedance is 38k.

http://mh-audio.nl/CalculateRIAA.asp

« Last Edit: December 04, 2015, 01:21:13 PM by tangmonster »

Offline handsome

  • Trade Count: (+1)
  • AVForums Veteran
  • ***
  • Posts: 843
Re: Tube phono preamps
« Reply #28 on: December 04, 2015, 01:58:12 PM »
the output impedance of the first triode will add resistance to the 200k. the input capacitance of the second triode will add capacitance to the 3.3nF capacitor. thus you cannot calculate your networks without considering these values. also as the valves age these values change......... It is very hard to measure riaa accuracy accurately as the riaa curve covers a 40dB range which then requires either a very (very) accurate meter. An inverse eq network is the best but then you need to be aware of proper loading and termination in order to maintain its accuracy. the fancier measuring equipment can (usually digitally) generate such inverse responses thus making checking a doddle.

Offline tangmonster

  • Trade Count: (-1)
  • AVForums Veteran
  • *****
  • Posts: 1,752
Re: Tube phono preamps
« Reply #29 on: December 04, 2015, 02:19:20 PM »
It is very hard to measure riaa accuracy accurately as the riaa curve covers a 40dB

not that tricky.
Just change your function generator amplitude over the range. And measure with an oscilloscope Dual trace is easier or you can use single channel and go back and fourth to measure difference between function generator and output.

I start at 1khz.

Then check the preamp gain by converting the Volt difference between input and output and converting to dbv

remember each double in voltage difference is 6dbv difference.

example:

1 khz sine wave input 1mv peak = output at 0.1V peak = 100 times amplification = 40db  gain. (useless info is that I measured my d3a based one at aroun 38db, which is I suppose good enough)

So for 20hz to 20khz you want to see 60db gain to 20db gain from your input.

For that 1khz baseline you want your 20hz to be around 20db higher or at least 19db.
So for the same 1mv peak and 20hz you want to see close to 1V peak at the output.  40db gain + another 20db at 20hz = 60db = 1000 times gain.


20khz is different.

You are going to want to see only 20db gain. ( 40 db (at 1khz) - 20db)

so for a 20khz 0.1V function generator peak input you are going to want to see 1V peaks on the output.


If you prefer only testing down to 50hz your are going to want to see a 17db or 7X gain from your 1khz measurement.

I'm hoping i didn't rape math too much in above post.....


P.S.

nice list of attenuation curves vs frequency

http://www.beigebag.com/case_riaa_1.htm


Gain/Loss vs db calculator:

http://www.sengpielaudio.com/calculator-gainloss.htm
« Last Edit: December 04, 2015, 02:24:54 PM by tangmonster »