John Curl's Blowtorch preamplifier part II

Status
Not open for further replies.
slew rate limit, and may distort if you approach this limit. So what? Good design will be aware of this. Signal levels are low, so quite 'fast' signals have small slew rates.

A large pop contains no information so nothing is lost in any case. The slew rate issues were probably more relevant in the 5532 days, there are numerous choices now that recover in 100's of nsec which makes the issue moot.
 
Well everyone, at least we have had some technical input generated (I think).
Let me clarify a term: When I say 'mis-track' I mean temporarily leaving the groove due to acceleration forces generated by the recorded velocity on the record. This can be done with a test record that continues to increase in level with a 1KHz tone, for example, or you can do it with certain passages of music. With some phono cartridges and some recordings, you can't do it at all. With most phono cartridges and certain recordings, you can do it every time, on certain passages. This is shown in my 1978 IEEE paper: 'Omitted Factors In Audio Design'. Yes, the stylus appears to leave the groove and crash back down, putting out a high level dV/dt pulse, along with the rest of the test signal. Do the test yourself with a B&K test record and a storage scope, if you don't believe me.
PMA, you are, of course, correct in the recommendation of the 2SK369, IF it is as good as a 2SK147, BUT is it? Have you noise tested it for 1/f? Different process means different 1/f. Trust me, I had to return 100's of 2sk147's when they changed the process, 20 years ago. Perhaps this should be checked out.
For some reason, my costs are criticized, yet they are MY costs. If your costs are lower for the SAME item, I am all for learning how to save a few dollars. The transformer costs I cited were for the samples purchased by us for this project.
Also, OF COURSE, it is not COMPLETE INTEGRATION, but the effect is the same with fast pulses. This is quibbling, that obscures the instruction offered. (enough for now)
 
On closer inspection, I suggest that anyone seriously attempting to understand what I do for a living, should make Werner's circuit. It is as close to my approach as anything that I can offer. It is buildable with available parts, and it manifests virtually everything that I have stated so far about 2 stage phono design.
The only real departure from the Vendetta and other more exotic designs is the lack of a complementary pp aspect, that reduces open loop distortion.
 
If you DO make this circuit, be sure to use a AUDIO QUALITY coupling cap and for your EQ caps. I strongly recommend Rel RT caps, because they work for most of us pros. I use them, Charlie uses them, Audio Research has been known to use them, etc. etc. You can't go wrong with a Rel RT. Don't let ANYONE tell you to use something of lessor quality, or even a different material. If you want to know what I do is all about, don't compromise on the passive parts.
 
$400-$440 for the transformers we chose.

1. What transformers are they?
2. If I remember correctly, the Vendetta SCP has SS MC input, without a transformer. Why did you choose transformer this time? Is it because transformer input sounds better, or because of obsolete JFETs?

Edit:
Oops, now I see you already answered about the reason for using input transformer.
 
Last edited:
Depends. If you want the valves tone you go for my Valve Itch pre. Of course my 3XK363 input Low MC for 103R shows -105dB at 20Hz and -145dB at 20kHz noise bed when the Itch shows -80dB & -120dB with SUT. One slams and analyses better, the other is silkier. Noise bed differences and resolution impact grow pro Jfet as the speakers move from 90dB sens mark upwards never the less. Its also the SUT cost issue. And then again they are both open loop & passive filtering, you may prefer closed loop with OPAMPS in the consecutive stages and active Riaa for more sense of low THD and control tone.

So, if I get what you say, there is no definite winner here, some may prefer trafo input, while others may prefer JFET input. Is that right?

Also, thank you but no, thank you – no opamps for me, nowhere in the signal path.
 
They changed to ion implantation and destroyed many 1000's of parts, around 1990 We caught them with a Quan-Tech, and returned the REALLY BAD ones, greater than 10nV/rt Hz at 10 Hz. Sorted out the others, using them for second and 3rd stages, instead of input stages.

You must not read Nelson's forum. It looks like the SK170's still can have problems (top plot). This one is 3nV at 10Hz, luckily. The lower plots are various SK369's and SK372's all better than 2nV at 10Hz. -87dB = 1nV
 

Attachments

  • noise1.JPG
    noise1.JPG
    87.4 KB · Views: 286
Why should I read Nelson's form? To clarify this, I am an expert in my own right, and I don't want to 'steal' anything from Nelson, consciously or unconsciously. If I contributed to many threads here, then there might be a good reason for my constant moderation. I can be brutally honest, it comes naturally, and not tempered by 'prep school' training that many have had here, to use diplomacy, rather than direct honesty, which protects tender feelings, with the exception of when one is actually aware of what they are saying, below the polite phrases. I am continually insulted on this forum, both politely and impolitely. Apparently, as I am a known figure, it is OK for many to take cheap shots, usually poorly thought through, such as SY's pricing of input transformers, on this thread recently. Then, I have to respond, double check my figures, and then patiently wait while an overworked moderator finally gets to my input and clears it, for affecting any tender feelings of you grown up guys out there. And so it goes.
 
Last edited:
Status
Not open for further replies.