An Attempt to Quantify Semiconductor Differences

Discuss x0x construction and related issues

Moderators: altitude, adafruit_support_bill, adafruit, phono, hamburgers

Please be positive and constructive with your questions and comments.
User avatar
altitude
 
Posts: 995
Joined: Wed May 11, 2005 5:17 pm

Re: An Attempt to Quantify Semiconductor Differences

Post by altitude »

This is an x0x?

Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am

Re: An Attempt to Quantify Semiconductor Differences

Post by Brassteacher »

antto wrote:is that recorded thru a microphone or something?! :shock:
the aiff doesn't play from box.net
you might wanna check if you can record audio at higher sampling rates.. 96KHz ..
this is needed sometimes, especially for looking closer at the envelopes and accent (on sonogram)
i even record some stuff at 192K ;]
Obvious thing #2 detected! Yes, a Sony powered mic going into a Sony mini-disc recorder. I'm working on the noise problem. The mini-disc recorder is of course battery powered, but in their infinite wisdom, Sony put it in a metal shell that is grounded to the chassis. You don't dare hold it in your hand unless you want a LOUD 60Hz hum. Higher sampling rate? Ok, Audacity it is from now on then! (I didn't have it downloaded when I made this recording).

I found out the file wouldn't play back through the site, and if I'm going to go 96K or 192K, either AIFF or .wav may be the way to go. I'm really skeptical of something that calls itself "lossless compression" 8)
Altitude wrote:This is an x0x?
Excellent! Obvious thing #1 detected. You are correct, it is not a x0x. It is a Korg Monotron.

Since I've spilled the beans on the Monotron, is it obvious from the recording that it is a sawtooth wave? And by this I mean does it show up if you're looking at the waveform, I'm aware it's quite obvious by listening. Also, is it obvious that it's a different sawtooth from the x0x or the TB-303? Another question, is it detectable that the filter is 12dB as opposed to the 18/24dB filter in the xox/303?

Thanks for the responses. I'm finishing the complete build of the x0x, since I figured no one ever sits and just listens to their VCO!

User avatar
antto
 
Posts: 1636
Joined: Thu Apr 15, 2010 3:21 pm

Re: An Attempt to Quantify Semiconductor Differences

Post by antto »

i DO listen to the VCO
you know i had to record raw VCO sound in order to make a better approximitation of the saw->square shaper for my soft synth
i also needed the filter raw output to get clearer sonogram picture for looking at the envelopes and accents

Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am

Re: An Attempt to Quantify Semiconductor Differences

Post by Brassteacher »

antto wrote:i DO listen to the VCO
you know i had to record raw VCO sound in order to make a better approximitation of the saw->square shaper for my soft synth
i also needed the filter raw output to get clearer sonogram picture for looking at the envelopes and accents
Ok then, I'll make a few short recordings right off the VCO output when swapping transistors in the VCO. I'll go ahead and save the trouble of recording part of the testing: Putting a 2SC2259 in the VCO in place of the 2SC1583 makes NO difference at all! So, that is definitely a valid substitute part. I will add though, after swapping and running it through the scope, I measured the Hfe of the C2259, and it happened to be in the low 600's. The Hfe of the C1583 I had plugged in was in the 630's. The biggest difference between C2259 and C1583 is the range of Hfe. The Hfe range of C2259 is 200 above C1583, both on the bottom and the top, which leaves a LOT of potential overlap. Other than that they are basically the same transistor.

I think the next recording, the first real one anyway :) , I'll do with 2SK30A-O and 2SK30A-Y vs. J201 in both positions. With the square wave, there was a quite visible difference on the scope when swapping those out. J201 was a noticeably "cleaner" square wave, flatter across the top and with a faster rise time. Amplitude was also increased slightly. Pics and recordings when I get the chance.

Oh yeah, recordings off the speakers with the mic, recordings by plugging headphone output directly into the computer, or both?

User avatar
altitude
 
Posts: 995
Joined: Wed May 11, 2005 5:17 pm

Re: An Attempt to Quantify Semiconductor Differences

Post by altitude »

I think that you should also be looking at several different 2k30s as well, the J201 is a much tighter spec'd part and I would expect considerably more variation in the older parts.

Dry direct recordings only please, the room may sound nice but I dont think that's what you are shooting for..

Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am

Re: An Attempt to Quantify Semiconductor Differences

Post by Brassteacher »

Altitude wrote:I think that you should also be looking at several different 2k30s as well, the J201 is a much tighter spec'd part and I would expect considerably more variation in the older parts.
Ah, don't worry, I have O, Y, R, and GR varieties to play with as well!
Altitude wrote:Dry direct recordings only please, the room may sound nice but I dont think that's what you are shooting for..
Done.

Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am

VCO Recordings - 2SK30A vs. J201

Post by Brassteacher »

Finally had a chance to run a few real tests. Let me know if these recordings and screen shots do not have enough detail to make any real determinations. I have only listened to these through my cheap headphones, not on my main system, so I cannot make much comment on any audible differences yet.

Recording equipment: PowerMac G5 Dual 2.0, Audacity 1.2.6, x0x VCO fed directly into line input via a 2µf polypropylene capacitor.
Recording info: 44.1KHz, 32-bit floating point, .aif format. Due to the age of the G5, 44.1KHz is the upper limit on sampling rate. The polarity of the recordings has been inverted. Why? For some reason, either the ADC in the Mac, or Audacity itself inverted the polarity of the signal, and waveforms were originally displayed upside-down. They played back that way also. I'll burn these to a CD and run them through the scope to see if the wave polarity comes out correctly that way or not. If you don't think the signal polarity matters, I'll explain in another thread, but I would imagine synthesizer folks would be more savvy of this than your typical audiophile type.

I'll post the one most people will be interested in first, the square wave. File names should make which sample is which obvious. Each recording is 7 to 8 seconds long. First, the screen shots:
SK30 SQ.jpg
SK30 SQ.jpg (109.55 KiB) Viewed 2776 times
J201 SQ.jpg
J201 SQ.jpg (111.61 KiB) Viewed 2776 times
And the sounds:
2SK30A Square Wave
J201 Square Wave

If these turn out to be successful at being able to see and hear differences, I'll post the Sawtooth files, which are already in my folder on box.net. Also, I have a few other JFETs to test in place also.
Last edited by Brassteacher on Wed Dec 15, 2010 5:07 pm, edited 1 time in total.

User avatar
antto
 
Posts: 1636
Joined: Thu Apr 15, 2010 3:21 pm

Re: An Attempt to Quantify Semiconductor Differences

Post by antto »

erm, the URLs you posted.. they are relative to your account, thus they don't work
use the "Get Link to this file" or when you preview the file check in the top-right corner for the link itself (in a small text field)
the link should be something like box.net/shared/23bh9dd <- pretty short :wink:

Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am

Re: An Attempt to Quantify Semiconductor Differences

Post by Brassteacher »

antto wrote:erm, the URLs you posted.. they are relative to your account, thus they don't work
use the "Get Link to this file" or when you preview the file check in the top-right corner for the link itself (in a small text field)
the link should be something like box.net/shared/23bh9dd <- pretty short :wink:
I guess that's what happens when I do things in a hurry. Thanks, antto. The links are working correctly now...

User avatar
antto
 
Posts: 1636
Joined: Thu Apr 15, 2010 3:21 pm

Re: An Attempt to Quantify Semiconductor Differences

Post by antto »

now.. i can't load em in my sound edittor, maybe you used some newer aiff compression :roll:

Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am

Re: An Attempt to Quantify Semiconductor Differences

Post by Brassteacher »

antto wrote:now.. i can't load em in my sound edittor, maybe you used some newer aiff compression :roll:
Hmm, aiff, like .wav isn't compressed. The were saved in that format from Audacity. But, it is trivial to go back and convert them to .wav. I'll do that in a little while and re-upload. Let me know if the .wav files work, and I'll save future recordings in that format as well. I probably should have done .wav to start with.

EDIT: .wav files now up, everyone let me know how well they work/don't work, or if the detail isn't good enough. I may have to find a way to come up with an M-Audio PCI card, or an M-Audio Firewire interface. Too much $$ at the moment (been spending it all on transistors!), perhaps I can take a survey among my friends to see who has one I can borrow.

Links to the square wave .wav files:
SK30A SQ Wave
J201 SQ Wave

Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am

.WAV files are up!

Post by Brassteacher »

Bump. I didn't know that editing a post wouldn't show that it had been updated by appearing as an unread post.

User avatar
altitude
 
Posts: 995
Joined: Wed May 11, 2005 5:17 pm

Re: An Attempt to Quantify Semiconductor Differences

Post by altitude »

so they look and sound pretty darn close to me. What is perceived difference? I have a J201 in mine right now and I cant find a reason to go back to the original part..

Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am

Re: An Attempt to Quantify Semiconductor Differences

Post by Brassteacher »

Altitude wrote:so they look and sound pretty darn close to me. What is perceived difference? I have a J201 in mine right now and I cant find a reason to go back to the original part..
I haven't listened to them on my high-quality system yet, didn't want to influence any one else's opinions. If you think current file quality is acceptable, I'm going to record these again, but with a slow frequency sweep to see how much the square changes with frequency between the two JFETs. Any real differences should definitely show up then.

User avatar
altitude
 
Posts: 995
Joined: Wed May 11, 2005 5:17 pm

Re: An Attempt to Quantify Semiconductor Differences

Post by altitude »

Out of curiosity, is there any reason that 1n4148 diodes were chosen? The original Rohm parts are still in production (1SS133)

Locked
Please be positive and constructive with your questions and comments.

Return to “Making x0x”