John Curl's Blowtorch preamplifier

Status
Not open for further replies.
Jump to Latest
A

alansawyer

Member
Joined 2008
audiolinear.com
You may like to try running trace route (tracert) from a command prompt to find out where you are getting blocked:

C:>tracert www.synaesthesia.ca

I get:



Tracing route to www.synaesthesia.ca [66.6.63.32]
over a maximum of 30 hops:

1 <1 ms <1 ms <1 ms server.sawyer.local [192.168.2.4]
2 <1 ms <1 ms <1 ms 192.168.1.1
3 21 ms 21 ms 20 ms 62.241.160.185
4 25 ms 19 ms 20 ms 62.241.167.133
5 19 ms 21 ms 20 ms po2-104.cr05.hx2.bb.gxn.net [62.72.139.25]
6 21 ms 21 ms 21 ms vl3952.cr05.tn5.bb.gxn.net [62.72.137.9]
7 91 ms 91 ms 91 ms rd1ny.ny.shawcable.net [195.66.224.213]
8 91 ms 90 ms 91 ms rc1hu-pos8-0.ny.shawcable.net [66.163.74.22]
9 104 ms 104 ms 104 ms rc1sh-pos4-0-0.mt.shawcable.net [66.163.76.53]
10 104 ms 105 ms 104 ms ra2sh-tge11-1.mt.shawcable.net [66.163.66.74]
11 104 ms 103 ms 104 ms rx0sh-hydro-one-telecom.mt.bigpipeinc.com [66.24
4.255.38]
12 139 ms 123 ms 123 ms 142.46.128.14
13 109 ms 109 ms 108 ms tol-gsr.telecomottawa.net [142.46.130.10]
14 142 ms 126 ms 127 ms Momentus-gw.telecomottawa.net [142.46.196.50]
15 * * * Request timed out.

Becaus I can see the pages OK, the final "request timed out" most likely just means the target server doesn't accept "ping" requests for security reasons.
 
syn08

syn08

Banned
Joined 2005
synaesthesia.ca
courage said:
Nope, tried three different browsers but still no access. I did visit syn8 website a while ago without any problem, so I guess something has changed.

Thanks for the help Pavel ;)

Franklin
Click to expand...

:censored: I don't know what's happening. Could you please run

tracert www.synaesthesia.ca

and post the results?

Guys, I don't mind hosting the schematic(s) on your web sites, provided that you credit properly and occasionally follow up and update with the latest changes. There might be a few minor changes over the next weeks.

I'll post later today the full schematic, including the power supply. BTW, I gave up the Jung regulators. They are not really required.
 
Joshua_G

Joshua_G

Member
Joined 2008
gomeh.com/
janneman said:



Joshua,

This test feeds the 60Hz test signal in *series* with an 8 ohms resistor into the amp output. It is a specific protocol to measure IIM and compare different amps. As such, it doesn't have a direct relation to a real speaker.

If you want I can send you some literature on this, I think I have something in my library.

Edit: The attached is the way Jean Hiraga did this test. It looks like I was wrong, he does include either an 8 ohms load or speaker load, and feeds the test signal through 250 ohms. In a perfect amp you would only see the 1kHz at the output, not the 50 (or 60) Hz external signal.
But this may be a slightly different test than Bob alluded to.

Jan Didden
Click to expand...


Using an actual loudspeaker makes a difference.
It seems that measurements that are to be relevant to real-life conditions would better be done on a real loudspeaker, or on a working emulation of one.
 
jan.didden

jan.didden

AX tech editor
Joined 2002
linearaudio.net
Paid Member
Joshua_G said:



Using an actual loudspeaker makes a difference.
It seems that measurements that are to be relevant to real-life conditions would better be done on a real loudspeaker, or on a working emulation of one.
Click to expand...


Of course, in general, you have a good point. OTOH, we should design amps that would be 'blind' to whatever speaker we would attach. If I would design an amp that changes performance or character depending on the speaker used, I would not be very proud of it!
So, I agree that in developing and testing we should take care of it, and in my view a comprehensive worst case speaker model should suffice. If we then buy an amp that *does* change character with the speaker, I'd bring that amp back for a refund!

Jan Didden
 
S

Sigurd Ruschkow

Member
Joined 2004
Good work, Syn08!

I guess you are now in the 80dB 20-20k, flat, MC-RIAA club :)

What kind of eq total input noise 20-20k do you get incl RIAA transfer function?

BTW,
your matched JFETs, what noise did you select for for the SJ74 and SK170?

I am glad to read that you will make it signal cap less! and also as dual mono! For me, that is the only way.

Problem with BF862 is that there is no P-ch compl part.




Sigurd


syn08 said:

With the attached MC input stage I got 0.3nV/rtHz using regular K170BL/J74BL with Idss=11+/-0.5mA and 0.25nV/rtHz with selected (for noise) devices. It is also rock solid stable.

The complete MC uses a few more high voltage opamps (OPA552), has servo (OPA445, hence no film or electrolytic caps in the signal path, except for decouplings and the output optional 10uF polypropylene). 28dB headroom (+/-24V supplies, +/-16V for the input stage), 64dB gain, 0.01% distortion mostly 2nd harmonic, RIAA better than +/-0.1dB.

All the above measured and consistently confirmed. BTW, it's a dual mono construction hosted (preliminary) in a single 6 x 6 x 2 inch 0.1" aluminum case with two magnetically shielded 25VA toroids on board. 108dB channel separation, if somebody cares about.

It's part of HPS 3.0 soon to be finished and posted in the HPS thread. Don't worry, it sounds amazing even with that much feedback.

I'm not planning to get into business and sell this stuff so your old SCP-2 record is safe from a commercial perspective :D

Next, down to under 0.2nV/rtHz with BF862, SMD parts (including inductors), etc... Hard to find good SMD parts (in particular resistors), but not impossible.
Click to expand...
 
G

GK

Disabled Account
Joined 2006
syn08 said:
BTW, I gave up the Jung regulators. They are not really required.
Click to expand...


Well duh.

Way back when you asked about getting around the PSRR problem of single common source input stage, I suggested the discrete differential input + opamp topology.

Saves having to find complementary input devices too.

Cheers,
Glen
 

Attachments

  • lownoise.jpg
    lownoise.jpg
    31.4 KB · Views: 545
syn08

syn08

Banned
Joined 2005
synaesthesia.ca
G.Kleinschmidt said:



Well duh.

Way back when you asked about getting around the PSRR problem of single common source input stage, I suggested the discrete differential input + opamp topology.

Saves having to find complementary input devices too.

Cheers,
Glen
Click to expand...

Sorry Glen, but that's a pretty poor design for ultra low noise. A differential stage has 3dB higher noise than a single ended (and HPS is single ended!). The CCS is happily injecting common mode noise in the input stage. Maybe not a big problem with bipolar superpairs (haven't tried, I don't think bipolars are the best way for MC) but certainly a problem with JFETs (even 2SK389); even the smallest mismatch will show the noise at the output. Designing an ultra low noise CCS should be treated as a separate project...

BTW, the spec in the schematic you posted is :bs:. Only the 100ohm feedback resistor has a noise contribution of 1.3nV/rtHz. Do you think I'm using 1ohm 0.5% feedback resistors just for fun?

Jung regulators are now replaced with LM317/LM337 and parallel post regulators as shown. Works as a charm, PSRR is measured at about 106dB for the positive supply and 100dB for the negative supply. The negative regulator has a slight noise peak at around 10KHz, certainly nothing to worry about.
 
B

Bob Cordell

Member
Joined 2006
cordellaudio.com
janneman said:



Joshua,

This test feeds the 60Hz test signal in *series* with an 8 ohms resistor into the amp output. It is a specific protocol to measure IIM and compare different amps. As such, it doesn't have a direct relation to a real speaker.

If you want I can send you some literature on this, I think I have something in my library.

Edit: The attached is the way Jean Hiraga did this test. It looks like I was wrong, he does include either an 8 ohms load or speaker load, and feeds the test signal through 250 ohms. In a perfect amp you would only see the 1kHz at the output, not the 50 (or 60) Hz external signal.
But this may be a slightly different test than Bob alluded to.

Jan Didden
Click to expand...


Hi Jan,

Interesting, although the added value of the L-C is questionable because the back-feeding source is at a single frequency.

Somewhat more generally, it should be said that a perfect amplifier with finite output impedance and no IIM will show no evidence of 50 Hz (60Hz) modulation of the higher frequency signal sent through the amplifier under test in the forward direction.

Cheers,
Bob
 
B

Bob Cordell

Member
Joined 2006
cordellaudio.com
Joshua_G said:



Using an actual loudspeaker makes a difference.
It seems that measurements that are to be relevant to real-life conditions would better be done on a real loudspeaker, or on a working emulation of one.
Click to expand...


Hi Joshua,

What you are saying is generally true, but just not really for this particular test.

It is more generally true for tests that usually just use a resistive load in a conventional test. This of course fits most tests out there. The IIM test, in my opinion, is just not one of them.

Cheers,
Bob
 
B

Bob Cordell

Member
Joined 2006
cordellaudio.com
janneman said:



Of course, in general, you have a good point. OTOH, we should design amps that would be 'blind' to whatever speaker we would attach. If I would design an amp that changes performance or character depending on the speaker used, I would not be very proud of it!
So, I agree that in developing and testing we should take care of it, and in my view a comprehensive worst case speaker model should suffice. If we then buy an amp that *does* change character with the speaker, I'd bring that amp back for a refund!

Jan Didden
Click to expand...


Be careful, Jan, you're talking about most tube amps :).

Cheers,
Bob
 
jan.didden

jan.didden

AX tech editor
Joined 2002
linearaudio.net
Paid Member
Bob Cordell said:



Be careful, Jan, you're talking about most tube amps :).

Cheers,
Bob
Click to expand...

I know. I still want my money back ;)
It's my personal preference only for amps that are as neutral, transparent and universally applicable as possible. I have no intention to force it on anybody else. I think. Whatever floats anybody's boat!

Jan Didden
 
jan.didden

jan.didden

AX tech editor
Joined 2002
linearaudio.net
Paid Member
Bob Cordell said:



Hi Jan,

Interesting, although the added value of the L-C is questionable because the back-feeding source is at a single frequency.

Somewhat more generally, it should be said that a perfect amplifier with finite output impedance and no IIM will show no evidence of 50 Hz (60Hz) modulation of the higher frequency signal sent through the amplifier under test in the forward direction.

Cheers,
Bob
Click to expand...

Indeed. If I understand what you mean, I would still find the 50 or 60Hz at the amp output, but no intermodulation products.

Jan Didden
 
Status
Not open for further replies.
Top Bottom