USE UPS NEXT DAY AIR FOR ANY USA ORDER BEFORE 11AM ET *TODAY* 12/19/2014 TO GET YOUR PACKAGE IN TIME FOR XMAS - PLEASE SEE OUR SHIPPING DEADLINE NOTICE FOR MORE DETAILS!

An Attempt to Quantify Semiconductor Differences
Moderators: Altitude, adafruit_support_bill, adafruit, phono, mome rath

Re: An Attempt to Quantify Semiconductor Differences

by Brassteacher on Sun Dec 19, 2010 9:43 pm

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


I wasn't aware the 1SS133 was still available. Then again, I didn't really look, as I literally have a few hundred 1N4148 in the workshop and I was following the x0xb0x build manual.
Don Taylor, a.k.a. Brassteacher
"Jack of all trades, master of none, proficient at even fewer..."
Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am
Location: Kennesaw, Georgia, USA

Re: An Attempt to Quantify Semiconductor Differences

by altitude on Sun Dec 19, 2010 9:47 pm


altitude
 
Posts: 978
Joined: Wed May 11, 2005 5:17 pm

Re: An Attempt to Quantify Semiconductor Differences

by Brassteacher on Sun Dec 19, 2010 10:57 pm



I don't know man, those are FOUR CENTS a piece! I mean, a set of 25 will cost a whole EIGHTY CENTS! That's half a chocolate bar, I have to save up for that, I have my priorities you know. Besides, they only have 9,346 of the things, by the time I scrape enough change out from under the floormats in my car, they'll be out of them already... :mrgreen:

Ok, seriously, I wonder if it is enough to make a difference? Some differences in specs are not really a factor, but the 0.2V difference in voltage drop is significant, as well as the 2pf difference in capacitance, especially since one is double the other. However, reverse recovery time is the same.

Looking at the VCO, and without doing a lot of complex calculations (Kind of in a lazy mood right now), it appears that with D30 and D31 any differences would be irrelevant, and with D25 being located in a current source, I'm not sure how much, if any, the effect of the voltage drop and capacitance would be.

Unless you count D24, there are no diodes in the filter, D24 would affect accent more than anything else. Given its location, I wonder if the difference in voltage drop would be noticeable. It may be, perhaps you or bcbox could take a look and help figure out if it's work testing.

D24 of course brings us to the VCA and the EG. If there is a difference, I think it would show up the most here, mostly because of the difference in voltage drop. Can you or someone else help make a determination on whether or not this would be worthy of testing? I just may be. I don't know if the 0.2V difference is something antto can model in his software synth or not, but it may be interesting if he can.

If enough people chime in thinking it may make a noticeable difference, I'll at least go ahead and unsolder and socket D24 and all the diodes in the VCA and EG. By the time I finish some more of these other tests, perhaps I will have scraped enough loose change out of the car to order the 1SS133 diodes :mrgreen:
Don Taylor, a.k.a. Brassteacher
"Jack of all trades, master of none, proficient at even fewer..."
Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am
Location: Kennesaw, Georgia, USA

Re: An Attempt to Quantify Semiconductor Differences

by altitude on Sun Dec 19, 2010 11:15 pm

well i just ordered 250 of them and will be building 5 b0xes in the next couple of weeks, so we will see

altitude
 
Posts: 978
Joined: Wed May 11, 2005 5:17 pm

Re: An Attempt to Quantify Semiconductor Differences

by antto on Mon Dec 20, 2010 4:30 am

D24 has something to do with the accent going to the filter cutoff?
sure!
the x0x accents behave a bit different than on the 303..
We are here too: irc.freenode.net >>> #x0xb0x
..:: c0nb0x v1.00 ::.. http://forums.adafruit.com/viewtopic.php?f=7&t=29282 (new control app - win/osx)
VCO Tuning: http://forums.adafruit.com/viewtopic.php?f=12&t=24759 (do it the EASY way)
User avatar
antto
 
Posts: 1395
Joined: Thu Apr 15, 2010 3:21 pm
Location: 127.0.0.1

Re: An Attempt to Quantify Semiconductor Differences

by guest on Mon Dec 20, 2010 5:35 am

i gave a quick look at the v-i curves
on the datasheets
and they have similar forward voltage drops
although the rohm datasheet
doesnt show much below .1mA
and this is where they are usually run
guest
 
Posts: 3155
Joined: Fri Feb 17, 2006 5:35 am

Re: An Attempt to Quantify Semiconductor Differences

by Brassteacher on Mon Dec 20, 2010 3:14 pm

guest wrote:i gave a quick look at the v-i curves
on the datasheets
and they have similar forward voltage drops
although the rohm datasheet
doesnt show much below .1mA
and this is where they are usually run


I saw that also, and was confused a bit because it seemed to be contrary to the specs published above the curves. I actually read the 1SS133 data sheet three times just to make sure I wasn't seeing things. I guess I won't know until I try it, but Altitude will probably get to that stage of things before I will.

Hmm, I wonder if maybe that dead Juno has some of those diodes, just thought of that...
Don Taylor, a.k.a. Brassteacher
"Jack of all trades, master of none, proficient at even fewer..."
Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am
Location: Kennesaw, Georgia, USA

Re: An Attempt to Quantify Semiconductor Differences

by guest on Mon Dec 20, 2010 7:58 pm

they spec them at different forward currents
so the Vf varies a bit

i think there are a couple of different types of diodes
used in the 303
and if they have different forward voltages
then there could be some sonic differences
especially in the envelope generators
guest
 
Posts: 3155
Joined: Fri Feb 17, 2006 5:35 am

Re: An Attempt to Quantify Semiconductor Differences

by Brassteacher on Mon Dec 20, 2010 9:19 pm

guest wrote:they spec them at different forward currents
so the Vf varies a bit

i think there are a couple of different types of diodes
used in the 303
and if they have different forward voltages
then there could be some sonic differences
especially in the envelope generators


I agree. I've heard many times that no two 303's sound exactly alike :)
Don Taylor, a.k.a. Brassteacher
"Jack of all trades, master of none, proficient at even fewer..."
Brassteacher
 
Posts: 248
Joined: Fri Oct 08, 2010 4:49 am
Location: Kennesaw, Georgia, USA

Re: An Attempt to Quantify Semiconductor Differences

by altitude on Thu Jan 13, 2011 12:12 am

ok, so I have the first couple boxes done based on the orginal Rohm diodes and one thing I notice right off the bat is I am not hearing the mystery click nearly as much. I wont have the VCA offset pots in till tomorrow but with the volume cranked up, I really dont hear anything other than the DC offset thump and free running click, the one sharp one I am used to seems to be gone.. For reference, the parts I am using for this run are:

Sliver face 945s
"vinatage 733s" for all locals except Q8,9,10 which use modern Fairchild ones. The modern parts seem to have the biggest spread of beta so I can get the best ones sorting those
beta matched PNP and NPN parts in the current mirror
5 and 3% poly caps (mixed Xicon and Panasonic brand)
Elna electrolytcs

There is a lot different from the last run I did so I guess what is in play is speculation right now but I will have all 5 boxen don this week so I should be able to draw a consensus off that. For some reason, my boxes are getting popular so a fifth run looks likely and I have recently discovered that Arrow carries and stocks pretty much the whole line of Panasonic film caps which covers everything in the BOM so I plan to fully switch to those.

altitude
 
Posts: 978
Joined: Wed May 11, 2005 5:17 pm

Re: An Attempt to Quantify Semiconductor Differences

by bcbox on Thu Jan 13, 2011 3:09 am

Here's a blind test for you guys.
These two samples contain 4 bars each. Can you hear any differences of any part of each sample?
test sample a
test sample b
bcbox
 
Posts: 587
Joined: Thu May 25, 2006 1:30 am

Re: An Attempt to Quantify Semiconductor Differences

by decoder23 on Thu Jan 13, 2011 9:42 am

wow, huge difference. sample a sounds much smoother, not so sharp and scratchy as sample b :o
decoder23
 
Posts: 96
Joined: Wed Dec 15, 2010 10:17 am
Location: outer space

Re: An Attempt to Quantify Semiconductor Differences

by antto on Thu Jan 13, 2011 2:15 pm

smooth scratchy.. :shock:
sample B has the envmod increased alot, sample A doesn't.. that's all
bcbox: what was this about?
We are here too: irc.freenode.net >>> #x0xb0x
..:: c0nb0x v1.00 ::.. http://forums.adafruit.com/viewtopic.php?f=7&t=29282 (new control app - win/osx)
VCO Tuning: http://forums.adafruit.com/viewtopic.php?f=12&t=24759 (do it the EASY way)
User avatar
antto
 
Posts: 1395
Joined: Thu Apr 15, 2010 3:21 pm
Location: 127.0.0.1

Re: An Attempt to Quantify Semiconductor Differences

by bcbox on Thu Jan 13, 2011 2:42 pm

sorry guys, I didn't explain very well.
I meant within each sample can you hear any differences. Basically each sample has 4 bars, do they all sound the same?
The two samples themselves are just different settings.
bcbox
 
Posts: 587
Joined: Thu May 25, 2006 1:30 am

Re: An Attempt to Quantify Semiconductor Differences

by Sokkan on Thu Jan 13, 2011 4:07 pm

Oh, totally different!

On bar 2 and 4 you have an old panasonic cap mounted backwards in C29, my guess is that it is between 17 and 17 and a half years old...

;)


No, I could not here any apparent differences between bars within the two examples.
Sokkan
 
Posts: 240
Joined: Sat Sep 01, 2007 12:56 pm
Location: Stockholm