My 12AX7 zero negative feedback preamp, with measurement

[alexw88]

All of my previous projects are Opamp based pre and power amps. Now I have tried tubes and I really love it! I am a tube newbie so I hope I can learn more.....

My 12AX7 zero negative feedback pre-amp : http://www.shine7.com/audio/12ax7_pre.htm

Here are some photos and measurements from the above page :

[SY]

Very nice construction, and the distortion spectrum looks excellent. What is the output level where it was taken?

I like those prototype boards and terminals. Where did you get them?

[alexw88]

Thanks, the output is at 1V p-p.

I have got the prototype board and the terminals at a local DIY components shop in Hong Kong.

[dhaen]

I agree: nice construction. Also good philosophy IMO

[GlidingDutchman]

Bravo! Very nice indeed!

I envy your skill!

GD

[CaliforniaBob]

love your site!

Bob

[zigzagflux]

That's freaking gorgeous !

[daanve]

SY,

if you ever make it to HongKong you have to visit the store where I saw those solder stands.
Bush Electronic Supply, in an alley behind a pretty big street in Hong Kong central or the Centre.
You enter the alley, then walk up some steps and you enter the shop. Amazing, the contrast.

http://hktubeaudio.homestead.com/files/Central_tube.htm


Daan

[Majestic]

Very nice looking job, Alex!

(I need to get myself some of those prototype boards too!)

[SY]

Daan, I still have to get back to Holland before I can start thinking about travel for pleasure!

[kstagger]

similar PCB is available on E-Scam
http://cgi.ebay.com/MULTI-PURPOSE-TU...QQcmdZViewItem

not associated with the seller.

[rdf]

Gorgeous work alexw88, it suggests a professional's hand. That harmonic graph is a little strange though, specifically the 20 dB drop in noise floor above the fundamental. Any idea what that's about? RMAA can be a little flakey at times.

Incidentally, if you're not afraid of solid state, replacing the front end plate load and cathode follower cathode load with something like an IXYS CCS will probably drop those spikes much lower. And thanks for the board lead!

[abzug]

It looks very nice. Don't put a case on it--maybe a mesh cage or plexiglas would work better showing off these internals.

However,

There's NFB inside the 12AX7

Also, I see electrolytics in the signal path

[EC8010]

Originally posted by daanve
Bush Electronic Supply, in an alley behind a pretty big street in Hong Kong central or the Centre.

An excellent shop! Sadly, Mr Chan (who previously ran it) died three years ago. Further details:

Real electronics/audio shops

Nice construction, by the way.

[Salas]

Is the gain ok? Or is it excessive? What's the gain of your main amp and the SPL/2.83V sesitivity of your speakers?

[alexw88]

Thanks everyone for the comments

Originally posted by rdf
That harmonic graph is a little strange though, specifically the 20 dB drop in noise floor above the fundamental. Any idea what that's about?

Incidentally, if you're not afraid of solid state, replacing the front end plate load and cathode follower cathode load with something like an IXYS CCS will probably drop those spikes much lower.

I *guess* the 20dB noise floor at lower frequency is caused by the regulator capacitors charging and discharging pattern. I tried removing one tube (to lower current draw) and do the measurement again, that 20dB noise started to drop at 1kHz instead of the current 1.5 kHz. I think if I increase the supply capacitance, I can make the noise lower at lower frequency.

What is IXYS CCS? I am new to tube design.....

Originally posted by abzug
There's NFB inside the 12AX7

Also, I see electrolytics in the signal path

Oh, so I can't claim it is a zero negative feedback amp...
The Jensen electrolytics are for the B+/12.6V supply filtering. The Blackgate NX is to bypass the cathode resistor. And this one need to be large, so I have no other choice.

Originally posted by salas
Is the gain ok? Or is it excessive? What's the gain of your main amp and the SPL/2.83V sesitivity of your speakers?

I think the gain is more than enough, however it is ok in my system as my speaker is 86dB and I have just changed the gain of my power amp by re-wiring its input transformer (LL1545A) to a 2:1 ratio.

[rman]

Hello.

First I would like to echo the others in saying beutiful job on that preamp.

I think you can justify saying no negitive feedback. While triodes
by their natture do have some built in negative feedback, you haven't added any extra feedback, and you have fully bypassed the cathode resistor to minimise the feedback as much as you could. As a bonus, bypassing also lowers the output mpeadeance
of the stage.

Have you considered battery or LED bias? This way no bypassing is needed and, (if I am wrong someone please correct me) there might be even less feedback. This circuit is definatly going to give you the classic tube sound.

Cheers!
Rolf

[jackinnj]

It ain't pretty -- but I breadboarded AlexW's line stage to check it out -- for one thing using 680R + 470uF on the cathode of the input results in a bit lower THD% than an LED -- THD measures about 0.14% using 80kHz bandwidth -- and I think that this is high because the device isn't in a more usual aluminum habitat. The third harmonic is down about 44dB from the second (better than Alex shows -- measured in linear mode on my HP3577). I used WIMA polypropylene's and carbon resistors and my regulated Heath Supply as that is what I had on hand, and I plugged in a rather ancient Dynaco 12AX7 since that was the first one I reached for:

[alexw88]

I have never researched on LED / Battery bias yet, will do some homework on it.

Jackinnj, thanks for trying it! When you mention the 3rd is 44dB below 2nd harmonics, were you talking about LED bias or the RC bias?

Thanks,

Alex

[jackinnj]

RC Bias -- 680R and a 470 uF Nicchicon -- getting near bed time her in NJ so I will look at the LED tomorrow.

[alexw88]

Thanks

BTW, the THD% in my measurement shown above is 0.035%.

[Majestic]

Originally posted by kstagger
similar PCB is available on E-Scam
http://cgi.ebay.com/MULTI-PURPOSE-TU...QQcmdZViewItem

ooh, thankies.

Any ideas about the gold turrets too?

[abzug]

CCS means constant current source. IXYS is a manufacturer of solid state devices.

Bypassing the cathode degeneration resistor means you are putting an electrolytic capacitor in the signal path--it's just as much in the signal path as a DC blocking capacitor on the output would be.
Whatever criticisms can be made about global negative feedback, one cannot possibly argue that the very local one provided by an unbypassed cathode degeneration resistor can have negative sonic contribution, let alone one worse than the insertion of an electrolytic.

You should try an unbypassed resistor. If there's not sufficient gain, you can decrease its value; there are other means for adjusting the DC operating point. You might like the sound of the capacitor better, but you might not and end up preferring the linearized tube and lack of coloration from dielectric absorption and electrolytic nonlinearities.

[alexw88]

Originally posted by abzug
You should try an unbypassed resistor.

Thanks for the suggestion I will give it a try, and there are LED and battery bias for me to try as well. That should keep me busy

[rdf]

As long as you don't take all this chatter the wrong way, it's lovely work and no doubt sounds good as it stands too.