Component C29 is backwards on all TB-303's... (more inside)

x0x0x0x0x0x

Moderators: altitude, adafruit_support_bill, adafruit, phono, hamburgers

Please be positive and constructive with your questions and comments.
guest
 
Posts: 3155
Joined: Fri Feb 17, 2006 5:35 am

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by guest »

and if anyone is interested
i wrote up a more thorough report
http://wiki.openmusiclabs.com/wiki/Electrolytic

User avatar
resonator
 
Posts: 102
Joined: Fri Dec 08, 2006 10:17 pm

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by resonator »

Just reversed C29 and it really made a difference. Thanks to everyone here for all the info.
This is perfect timing as well since I'm working on some more acid tracks. 8)

elmacaco
 
Posts: 31
Joined: Tue Mar 07, 2006 3:43 am

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by elmacaco »

Nice one resonator! I told you that BANNED is amazing! Thanks to bc box for sharing this significant discovery about the tb 303. It is amazing to me that an error makes an unintentional improvement in sound and character, possibly against conventional engineering wisdom. This may be something that can be exploited in other synths effectively. Seems like a simple mod to try. So before I look at the schemes, does anyone care to explain the role this cap plays a bit more? And do other filter topologies have a capacitor performing this role? I wonder if any other synths have this error? Someone should tell the 303 designer : ).

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

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by antto »

seriously.. why didn't you record a before/after pattern?

User avatar
phono
 
Posts: 1502
Joined: Wed May 02, 2007 4:01 pm

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by phono »

elmacaco wrote:Nice one resonator! I told you that BANNED is amazing! Thanks to bc box for sharing this significant discovery about the tb 303. It is amazing to me that an error makes an unintentional improvement in sound and character,
i wouldn't say its an improvement, since if the capacitor was in the other way that would be "the sound"

elmacaco
 
Posts: 31
Joined: Tue Mar 07, 2006 3:43 am

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by elmacaco »

So do you like it as the schematic? There are before and after patterns recorded here by bc. I know my xox sound really well, and the reversed cap sounds like it causes changes to how the filter responds to the envelope and fast series of notes. The way my xox was it had a more typical synth response without that reactionary variation that the 303 had when I compared them. Maybe I'm wrong, but that is what my ears tell me. Also the accent sounds more right to me now.

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

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by antto »

all i hear is:
bla bla bla sounds more 303 now bla bla bla to my ears bla bla bla i'm almost sure about this bla bla bla bla bla but there is a 0.1% chance i could be totally wrong bla bla bla bla and i actually talk without doing any kind of measurements bla bla bla bla bla and the difference in the bcbox samples might be all due to the different resonance levels bla bla bla bla whut resonance? bla bla bla bla bla but i am not trained to really tell de difference between a different resonance level and different accent/envelope curves bla bla bla bla bla but on the other hand my ears are very good i can hear sounds like "fat" "clean" "thin" "warm" "cold" bla bla bla bla bla "purple" "orange" "red" "violet" "dark-green" bla bla bla bla bla bla "50 dolars falling on the carpet" "55 dolars falling on carpet" bla bla bla bla bla and i once heard my cat speaking to me bla bla bla bla i wasn't drunk bla bla bla bla bla C29 bla bla bla bla...
okay, i'm really sorry for posting this :?
i hope at least 1 person understands what i meant..
do a propper test and spends some time.. before you try to make anyone believe there is a difference
once again, the bcbox sample is only different in the resonance level
even IF there is any difference in the accent or envelope behaviour (which i doubt) it cannot be clearly observed due to the pattern used, which isn't the best pattern for doing such tests as i already said

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

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by antto »

i mean... prove that there is an actuall difference other than the difference in resonance level (which is expected and can be compensated)
i expected any kind of difference from change in the accent envelope (gimmick) accumulation response or resonance warble, or even filter tuning change and more
there was NONE
you are saying there IS a difference and it sounds like it's sooooo obvious you instantly hear it
at the same time i can't hear it, i can't even SEE it in my sonogram/oscillogram
there are only 3 possible cases:

1) your x0xb0x is not like mine and there really IS a difference?!?!?
2) you are hearing whispers in the wind and your x0xb0x isn't too different than mine
3) a little bit of both :mrgreen:

User avatar
resonator
 
Posts: 102
Joined: Fri Dec 08, 2006 10:17 pm

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by resonator »

I would've recorded examples, but honestly I was too excited after talking to elmacaco and reading this thread. So I had my x0x open and closed before I could give it a thought.

I'm also wondering about the difference combined with swapping R1 on the power supply with a 200 ohm resistor (which I had already done prior). That resistor did make a subtle difference, but after reversing C29, it's sounds much looser, elastic, and has more of that 303 bite than ever before.

Thanks again bcbox... and thanks again elmacaco for calling me about the news. :D

User avatar
tim stinchcombe
 
Posts: 11
Joined: Wed Dec 01, 2010 6:25 pm

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by tim stinchcombe »

OK, so here is what I have tried.

I pulled C29 off my TBX-303 (the Analogue Solutions 'clone attempt'), and mounted it on a double-pole switch so that I could swap its polarity at will. Having done so I hear nothing like the difference in bcbox's samples. The measurements taken by guest suggest that if there is a difference, it is likely to be due to the bias level of the diff amp being changed by the (now very) leaky reversed capacitor. Running one of bcbox's samples ('saw no. 3') in the only software I have available for such things, 'Audacity' (OK it's a freebie, but useful nonetheless), it is clear that the difference we hear should be easily measurable - the waveform with the cap reversed looks to have a significantly reduced amplitude for the high-frequency resonance, and the overall amplitude itself seems to be down by a good few dB, perhaps 3 or 4.

So I measured the voltage across my C29, and it is around 3.318V for the correct polarity, and a mere 5mV down at around 3.313V when reversed. With such a small difference, I was not surprised I couldn't hear any difference. (And for those paying attention, the TBX-303 has a 5V reference voltage, and not the TB-303's 5.33V...).

Armed with this information and guessing that guest's measurements could be approximated to by adding a resistor in parallel with the (ideal) capacitor C29, so that it would 'leak' a considerable amount of DC current when reversed, I moved to a simulation - how much voltage would C29 have to pull the R116/R112/R109 bias chain down by in order for the amplifier's gain to be lowered by a significant amount? The diff amp gain is normally around x100 (40dB) in its passband (the 100n caps coupling to the ladder give a significant highpass affect, with a cut-off at around 95Hz or so...); I've attached the circuit used, showing the resistor added, R29p, in parallel to C29 (no attempt to use the right transistors, or number the other components as per TB-303 - I'm only after relative effects...), and the plots obtained when sweeping the value of R29p from 1kohm to 1Megohm. The red trace at the right shows the typical 40dB value; as R29p decreases and we move left, it needs to be in the 10-20k region to show a decent 3 to 6dB drop that might account for the loss of resonance when the cap is reversed, but look at where the (green) voltage line is - a good volt or so lower than where it would be with the correct polarity. If (and that is a big if!) this is what is causing the difference in the sound, it should be easily measurable on a unit exhibiting such a difference in sound - a before and after voltage reading would be great, but the voltage is apparently so far from what it should be, a reading taken on a machine purportedly exhibiting the aural difference after reversing it will probably suffice!

Whether this is the cause or not, I think I would expect the differences due to reversing a capacitor in this way to vary enormously from machine to machine, as there will likely be a huge variation due to different cap makes, age of machine, how often it is used etc., and thus simply turning the cap around in a x0xb0x and miraculously expecting it to sound 'more like a TB-303' could be a forlorn hope!
guest wrote:and if anyone is interested
i wrote up a more thorough report
http://wiki.openmusiclabs.com/wiki/Electrolytic
Thanks for that, I did find it interesting - what you have measured broadly fits into other work I have seen in the past whereby electrolytic capacitor behaviour can be modelled by various combinations of resistors, capacitors and several diodes, and it is nice to know that some of this may be directly applicable in the current context!

Tim
Attachments
Change of gain and voltage on sweeping C29 parallel resistance
Change of gain and voltage on sweeping C29 parallel resistance
c29_resistance_sweeps.gif (8.43 KiB) Viewed 3744 times
Circuit used for simulation
Circuit used for simulation
c29_sim_circuit.gif (4.88 KiB) Viewed 3744 times

User avatar
tim stinchcombe
 
Posts: 11
Joined: Wed Dec 01, 2010 6:25 pm

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by tim stinchcombe »

Some more simulation output, this time transient analysis runs (i.e. normal time-based plots). I'm hoping that the simulation circuit files are still good in terms of component values etc., as they were generated up to two years ago, and they are a little artificial in that I haven't made dynamic envelope voltages (but since that generally only affects the cut-off frequency, I don't think it will be a major shortcoming) nor any 'accents' etc. I feed a 5-volt amplitude, 80Hz sawtooth into the filter, the main plots are for the filter voltage out (so again, no simulation of the VCA etc.). (Resonance at max; cut-off frequency at max; envelope voltage a constant 10V, but env pot at 75%.)

All this, with the standard value of 1uF for C29 gives some nice 4kHz or so resonance, the red trace; dropping C29 to one tenth its value, i.e. 100nF (which seems to be unrealistically low even when the real cap is reversed), and there is literally no difference in the plots (the reasons become apparent in a moment) so I have not included it; adding a 100k resistor in parallel to C29 (which we know will alter the diff amp gain), indeed drops the ringing quite considerably, the blue trace (the vertical bars, '||' to be read 'in parallel with').

In the second set of plots we see how much the audio signal affects the voltage at C29: the red trace is with the standard 1uF value, and we see that the impact of the signal is tiny, tiny, tiny, only changing the voltage by about 3 microvolts (uV); lowering the cap value does increase this voltage swing, but it is now still only less than 10uV; adding the 100k resistor in parallel with the 1uF value (blue trace) again shows little amplitude variation, but the DC level has shifted down by 200mV or so, hence the separate axis to display it.

So once again, I feel that if the reduced bias voltage is the cause, I reckon it should be measurable, so it is over to someone (the OP, hopefully...?) who has a unit that sounds different and who has the means to take some reasonably accurate voltage readings at C29!

Tim
Attachments
Filter transient analysis output
Filter transient analysis output
c29_filter_out_transient.gif (12.84 KiB) Viewed 3736 times
Voltages at C29
Voltages at C29
c29_V_transient.gif (11.42 KiB) Viewed 3736 times

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

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by altitude »

I'm not hearing a huge difference either. I suspect it is localized only to certain settings and I'll let someone else A/B that

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

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by Brassteacher »

Altitude wrote:I'm not hearing a huge difference either. I suspect it is localized only to certain settings and I'll let someone else A/B that
I've just parted out a piece of audio gear about the same age or slightly older than a TB-303, a Pioneer integrated amp. There are a few 1uf 50V caps the same make as in some of the pics bcbox posted, and I'm seriously thinking about repeating these tests with a more or less "authentic" cap!

By the way: Some of you that may still be looking for rare parts, your search may be as short as a trip to your garage or attic. It seems that about the same time the TB-303 and similar Roland gear was current, Pioneer was a HUGE fan of the 2SA798 and the 2SC2291 transistors! It would definitely be worth a peek into such an item if you happen to have one lying around gathering dust or propping up a bookshelf. Most likely to be found in Pioneer's mid-line to nicer preamps, integrated amps, and receivers of that era. I've found these transistors in at least two different models. Happy hunting! 8)

textile
 
Posts: 221
Joined: Mon Nov 23, 2009 9:34 pm

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by textile »

All old gear from JAPAN is just asking for it. 8)

Thanks for this info Brian! A discovery as obvious as this is amazing that nobody has noticed ( or told anybody )

There were other mistakes / typos in the schematic.
If it is meant to be well then do it.

2sc1815 and 2sc1015 , this is good news as I have been considering to use these Japanese components.
:arrow: Load all these alternatives up in to the Alternative parts page.

Great to have you on board Mr Stinchcombe!! I appreciate your papers! :D

I understand antto..
man if you put you finger in your ear and stand on your head... you will hear your cat tell you how to modify your kitchen stove to make contact with a near by galaxy.. or at least cook some tuna.

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

Re: Component C29 is backwards on all TB-303's... (more inside)

Post by antto »

i'm sorry, i get very angry when someone talks about the 303 like he really knows what he's talking about
about 2 years ago when i started gathering information about the 303.. there was so much info on all kinds of websites
and by following the details i figured something is not right.. they were wrong? very wrong!? extremely wrong!? complete bullsh*t!
starting from the wikipedia article (which afterwards got a little fixed now)
there were things like "The oscillator has two waveforms: Sawtooth and Triangle wave" sometimes Sine wave
lots of weird things said about the slide (which actually is very simple)
"the accented notes modify/increase the filter resonance and vca"
"the square wave is produced by inverting every other cycle of the sawtooth" :shock:
i wasn't so familiar with analog gear back then (not that i am very familiar now but..)
i had no real clue how this thing works and i wanted to mimick the sound of it, so i searched and searched and watched videos, listened to recordings..

now i know more about the 303 (and i have a x0xb0x)
but i had to waste my time with all the bullsh*t written everywhere (and i believed it at first)
in fact, there is very little bits of true told about this synth

so now when someone writes stuff about the 303 that sounds like he is the 303 guru and knows what he's talking about but he's just interpreting another BANNED-article he's read before.. i just get very angry.. that's me :roll:
EDIT: i mean.. one should be responsible for the words he writes on forums or web articles, other people will be reading it eventually, some might follow it

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

Return to “General x0xing”