Tube DIY Asylum

Do It Yourself (DIY) paradise for tube and SET project builders.

Return to Tube DIY Asylum


Message Sort: Post Order or Asylum Reverse Threaded

RIAA equalization circuit in this one

193.219.3.76

Posted on October 7, 2010 at 02:22:49
Bartas
Audiophile

Posts: 69
Location: Vilnius
Joined: June 5, 2008
Hi,
I wonder if any of you could educate me how to calculate if RIAA is correct in this circuit.



This is a Hagerman Cornet 2 based phono stage but with major modifications. The output buffer tube is 6SN7, it is fed from choke input PS, gain stages operate at lower voltage. Gain stage operating points are done according Jim's latest thoughts on Cornet modifications.
Though I am not sure about the RIAA part and do not have the knowledge to check it.

 

Hide full thread outline!
    ...
Circuit is wrong, posted on October 7, 2010 at 06:55:36
Slick
Audiophile

Posts: 435
Location: Maryland
Joined: April 10, 2002
Please don't attempt to build that circuit as drawn. The series resistors in front of each section of the RIAA components are missing. It can't work properly.

As far as the calculations go:

If the first (missing) series resistor is called R1
The first shunt is called R2
The first shunt cap is C1

Then for the first stage: R1 = 9 x R2; R2 x C1 = 318uS; (R1 + R2) x C1 =
3180uS

The output impedance of the first stage is subtracted from R1. The grid leak of the second stage is in parallel with R1 + the 1st stage O/P impedance.

For the second stage: same part nomenclature, same comments about drive and driven impedances: R1 (missing in your schematic) x C1 = 75uS. The posted circuit does show an "R2" in the second stage, the 3.18uS corner that Allen Wright adds to stop the HF rolloff. If that's your plan, then R2 x C1 in the second stage = 3.18uS.

First select and match the caps as close as you can. The better you do, the better the result. Then choose theoretical resistors based on the actual cap values. Choose resistors ~5k too small, except second stage R2). Insert series 10k trim pots at each resistor (except the 3.18uS resistor, for which the theoretical value is probably close enough).

Then buy or build the Hagerman inverse network. Feed a signal generator's square wave input (maybe 1kHz) into the inverse network and the network into the phono stage. Measure the output and adjust the pots for the best square wave.

Some people stop right there, but if you're really anal (and who amongst us isn't? We're audiophiles!) then sweep the audio range (I stop at 100kHz)with sine waves, and adjust for the closest to flat match you can get. This part can be really time consuming and frustrating.

Finally, carefully remove the trimpots and substitute fixed resistors for the trimmers. Precision and quiet are really important here.

Good luck, it's a lot of work, but rewarding. when you're done, you really want no more than 0.2dB variation. Even that's audible. You must do the manual trimming. Simulations and mathematically determining the precise part values never works. Too many variances. I wish it would work, would save a lot of time.

One other thought, elimination of the current source from the tail of the follower (in the original Hagerman) is a bad idea. If you don't want to use the Hagerman version, look up some of Pimm's designs. The simple versions using a cascoded pair of depletion mode MOSFETs work very well.

Then again, you could add Stuart Yaniger's servo output stage (A Heretical Line Stage) in place of the existing cathode follower (by all means use the 6SN7) and eliminate that nasty big output cap in favor of a 0.01uF cap at the input of the servo follower. DC coupled at the output, works great and drives damn near anything.

Stuart

Smoke is on......here come the burners!

 

RE: Circuit is wrong, posted on October 7, 2010 at 08:08:54
Bartas
Audiophile

Posts: 69
Location: Vilnius
Joined: June 5, 2008
This is interesting.
This part of the circuit (gain stages with RIAA) is taken directly from Jim's sakura (cornet 3) except for the coupling RC between stages which is a leftover from cornet 2. And there are no "R1" in it. And it is no mistake. I think Jim did the most accurate calculations and compensated the R1 by the right combination of plate, cathode, coupling and tube resistance. This is why his circuit is much dependent on tube rolling.



And it is playing music right now with definite improvement regarding original cornet 2. Though one can not define which change added what to sound as it was all done at once.

Are there any more opinions on CCS on follower?
I somehow am a bit superstitious about using active components anywhere near signal path.

And yes the nasty output cap will go. This is planned but not yet scheduled :)

 

RE: Circuit is wrong, posted on October 7, 2010 at 14:22:15
Tre'
Industry Professional

Posts: 17260
Location: So. Cal.
Joined: February 9, 2002
I have a different question about that circuit.

How can that 12ax7, with a output impedance of about 30K ohms, hope to drive 6.8K? (at some higher frequencies the cap is a dead short leaving the 6.8k to ground)

The phono preamp circuit I build will drive the RIAA filter with one hand tied behind it's back. I think that's key to good sound.

What I mean is, the filter is designed to shape the frequency response the way we want it shaped. If the tube is "gasping" because it just can't drive the load.....then the frequency response will be wrong and the distortion will be high.

You want at least a ratio of 1:10, output impedance vs. load impedance, to insure a nice horizontal load line (= low distortion) for the triode.

Tre'
Have Fun and Enjoy the Music
"Still Working the Problem"

 

RE: Circuit is wrong, posted on October 7, 2010 at 08:46:31
Slick
Audiophile

Posts: 435
Location: Maryland
Joined: April 10, 2002
The only explanation would be depending on the (ever changing) stage output impedance as the equivalent of R1).

Given your experience with tube rolling, that must be what he did.

Many successful designers seem to shoot for a R1 around 10x the driving stage's plate resistance for immunity to aging and tube rolling effects.

With a 12AX7, that means BIG resistors. And that means big noise.

I like high gm, low plate resistance tubes for those reasons and others.

Stuart

Smoke is on......here come the burners!

 

RE: RIAA equalization circuit in this one, posted on October 7, 2010 at 05:42:06
Stephen R
Audiophile

Posts: 1427
Joined: January 11, 2002
If you don't have measuring equipment, then sim it using free LTSPICE. All info needeed to do this on Dave's site in the link below.

I've designed loads of phono stages this way.

Jim tends to design so that tube rolling has quite an effect on the end result so a sim can show you how this happens with minor changes in tube characteristics if you're that interested.

cheers,

Stephen

 

RE: RIAA equalization circuit in this one, posted on October 7, 2010 at 05:47:10
Bartas
Audiophile

Posts: 69
Location: Vilnius
Joined: June 5, 2008
Thanks Stephen, will look into that.

 

You can build an inverse RIAA passive network to use at the input and adjust accordingly, posted on October 7, 2010 at 04:51:56
There is also, I thought, a RIAA calculator on Jim Hagermann's web site. TUBE CAD and AUDIO DEVICES by Broskie has a calculator but there should be one on the web. Ray

 

Calculations are very difficult..., posted on October 7, 2010 at 03:16:45
Allen Wright
Manufacturer

Posts: 5652
Location: Schaffhausen
Joined: January 31, 2002
...and I wouldn't waste the time on it.

By far the best solution is to buy Jim's Inverse RIAA kit for $50 or so and MEASURE the RIAA response. Then you will KNOW the answer, not just a hopeful calculation.

Regards, Allen

 

RE: Calculations are very difficult..., posted on October 7, 2010 at 05:30:10
Bartas
Audiophile

Posts: 69
Location: Vilnius
Joined: June 5, 2008
What do you think of PC scope/generator. Would it be any good with iRIAA?

 

Difficult even with a scope and generator, posted on October 7, 2010 at 12:01:30
richardl
Audiophile

Posts: 3555
Joined: September 5, 2002
measuring precisely is rough. The inverse RIAA surely helps in this regard. All the same height is far easier than trying to measure changing values off of an Osc. screen, a quick sweep will tell you most of what you want to know with the iRIAA.

Richard

 

RE: Calculations are very difficult..., posted on October 7, 2010 at 03:28:22
Bartas
Audiophile

Posts: 69
Location: Vilnius
Joined: June 5, 2008
I was just looking into Jim's iRIAA.
It is one fine instrument, but I do not have a scope or signal generator.
Will have to think of something...

 

RE: Calculations are very difficult..., posted on October 7, 2010 at 10:27:30
dave slagle
Manufacturer

Posts: 5429
Location: NYC
Joined: April 27, 2001
you could burn spot tones with the inverse RIAA curve built in and use a DVM at the output to check.

I would first burn the tones at the same levels and measure them directly so you have a reference to the accuracy of your meter wrt frequency then you can correct for your meter when you measure the "iriaa" tones.

Entering them all into excel will plot the results nicely.

dave

 

Page processed in 0.035 seconds.