TR808 BD modulation...

From circuitbending to homebrew stompboxes & synths, keep the DIY spirit alive!

Moderators: luketeaford, Joe., lisa, Kent

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

TR808 BD modulation...

Post by nologin » Wed Oct 23, 2019 4:14 am

Hi, i have a friend who bought an 808, he tried it quickly and got a little scam.
He realized that later, that kick isn't ok, it did not get along with other sounds playing at the same time ... :ripbanana:

So initially, the kick had a problem as if the decay modulated, when it was at most, he replaced the pot decay and it has not changed anything. Then he left 808 for me to watch.

So i started by replacing components on the decay part, no difference ... Then persuaded that it came from the analog part, i replaced all the capacitors. No difference ... Then all the transistors. Then all the resistances.
The bug is less obvious, but if we listen well, we still hear a little, i recorded:

https://www.cjoint.com/doc/19_10/IJxiVn ... 08-bug.wav

the bug is very light, you must listen to hear it...

To a certain extent, we hear more beug, decay thoroughly, full tone and full accent, but still it seems to me always present, it's just that some settings of the BD optimize its harmful presence .
Out of curiosity i compared with my yocto clone, the TR808 has a more pronounced attack, detail that i noticed.
I did some tests, the scope of the trig kick, has some little variations. But i do not think that small trig voltage irregularities could pollute the sound of kick.
I tried a few things, without really knowing, i disconnected trig from kick and snare and reversed it. If i sequence the kick with the trig of the snare, the kick is still not good, no difference. Does anyone have an idea to propose to me?

Sebo
Common Wiggler
Posts: 62
Joined: Thu Oct 13, 2011 6:17 pm
Location: Argentina

Post by Sebo » Fri Oct 25, 2019 4:19 am

Are the PSU voltages clean? I mean no ripple or noise... may be some voltage variations causes the modulation.
Sebo
----------------------------------
http://www.cosaquitosenglobo.com/

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Fri Oct 25, 2019 1:53 pm

Yes the power supply is OK, this is the first thing i checked. By cons, only BD jokes, all other instruments work perfectly.

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Wed Oct 30, 2019 5:48 am

I checked pin 17 of the power supply, which is the BU point. When the TR808 runs, the voltage oscillates from 4.2VDC to 4.5VDC. I ask if this voltage variation would not be too high. :hmm:

Sebo
Common Wiggler
Posts: 62
Joined: Thu Oct 13, 2011 6:17 pm
Location: Argentina

Post by Sebo » Wed Oct 30, 2019 6:17 am

I think the variation is a bit high, but that voltage is for the memory part of the circuit, I think, so don't think is going to affect the analog part... but if you have this variation in a part of the PSU may be you have a problem in it...
Sebo
----------------------------------
http://www.cosaquitosenglobo.com/

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Wed Oct 30, 2019 7:31 am

Yes i think this allows a reserve of charge for the memory, when the batteries are out of order. I compared with my TR606, BU point oscillates too, but much less, that's why i have a doubt.
On the other hand + 15 / -15 / + 5 are perfectly stable.

User avatar
SMS303
Tha BABELFISH Wiggler...
Posts: 903
Joined: Wed Jan 14, 2009 9:08 pm
Location: EU-NL, Rotterdam

Post by SMS303 » Wed Oct 30, 2019 10:26 pm

My old 808 had a kind of distorted BD caused by a bad trace on the pcb. Maybe that's the source?!?!? :despair:
------------> Jaime Murray does Capua.... ------------------------------------------>

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Thu Oct 31, 2019 4:02 am

I thought about that for now, i checked the via visually with a magnifying glass, i'll check with the tester. But this is not distortion the problem, I think it looks like variations of attack, it means that when the decay potentiometer is more than half.

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Tue Nov 05, 2019 6:47 am

I was wrong it is not the pitch of the bass drum that is affected, but the accent.
I also noticed that the bass drum is very bory when when the rotary switch "INSTRUMENT SELECT" is pointed to Accent. But it's a bit boring, pointed BD. And bug even less rotary switch pointed to the other instruments.

I made other recording, where it's very bugz, we hear that kick drum are stronger with randomly, there are variations while the pattern is "total accent":

https://www.cjoint.com/c/IKfkBIzObYv

Besides, i photographed the recording, we see that there are higher peaks and it's random. (sorry i could not capture pics the computer's, it's an old XP dedicated to audio)

Image

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Wed Nov 06, 2019 3:34 am

Nothing to do with my problem.
I found a funny detail on the service manual, the accent in of all instrument are named trig. . For example, on this image capture, i indicated in green the accent input and in red the trig input.

Image

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Tue Nov 12, 2019 7:19 am

I continue my search for the beug. Last night i went to a friend who has a better scope than mine. So i sequenced 1/2/5/13 steps for BD, SD and accent. We see that the trig of the kick and snare, seem to be OK:

Image

By cons my ears were not wrong, we see that at R165, the input of common accent, there is variation, the signal should be the same and it is not:

Image

By cons what is strange only the BD is affected, the other instruments have no variations, while the TR808 accent is common to all instruments. I must concentrate my research on the ACC module. :*:

User avatar
guest
Super Deluxe Wiggler
Posts: 4328
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Tue Nov 12, 2019 1:08 pm

so, that shorter hieght pulse is only when BD is accented? otherwise it is tall like the rest?
openmusiclabs.com

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Wed Nov 13, 2019 7:26 am

Yes if i program only accented notes, there are short and high pulses, whereas there should be only high, i think. And i hear the abnormal behavior.

If i turn the potentiometer accent down, there are only short pulses and we do not hear the abnormal behavior.

User avatar
guest
Super Deluxe Wiggler
Posts: 4328
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Wed Nov 13, 2019 12:25 pm

i would suggest remving R152 and seeing if the pulses work as they should (you wont hear the BD, though). there is probably a short or bad component right where the TRIG signal enters the BD circuit.
openmusiclabs.com

Sebo
Common Wiggler
Posts: 62
Joined: Thu Oct 13, 2011 6:17 pm
Location: Argentina

Post by Sebo » Thu Nov 14, 2019 5:21 am

No, the accent should be present all time.
The transistors at the input of all drum circuits are a AND gate, so to trigger the sound it should be present both trigger and accent, the non accented notes have an accent of 5V, and the accented notes have an accent of something beteween 5V and 15V (depends on the setting of the accent knob), but always have to be a signal in the accent input.
Sebo
----------------------------------
http://www.cosaquitosenglobo.com/

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Sun Nov 17, 2019 9:33 am

Thank you for your answers, i have not desoldered R152, i will try soon. But i did a quick test, i sequenced the accent alone and probe it, without any other inst. With the ACC knob set to the lowest, we can see that everything is normal:

Image

There is 4.6V in a regular way, I put the "measure" option with my scope.




But when Accent potentiometer thoroughly, nothing goes well. I have 4.6V to 13.2V randomly. :omg: While i should get 15V. :doh:

Image

User avatar
guest
Super Deluxe Wiggler
Posts: 4328
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Sun Nov 17, 2019 12:45 pm

is this with only the BD active, and all BD hits accented? it would be good to set the BD to hit once a second, or something slow like that, and see how often the accent trigger appears on your scope. is it just once a second, and in time with the BD sounding? it might be that the small pulses you are seeing are due to the line being reused elsewhere in the circuit when voices are not being triggered.
openmusiclabs.com

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Sun Nov 17, 2019 2:47 pm

The last measurements that i did (pictures with rigol scope), it was without sounds, only the accent was sequenced and we can see that there are pulses variations.

User avatar
guest
Super Deluxe Wiggler
Posts: 4328
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Sun Nov 17, 2019 4:24 pm

yes, i think it would be good to do with with sounds. the digital line from the microcontroller that controls the accent, also is used to read and write to the memory, so there will be pulses on that line which have nothing to do with accents. so whats important is the level of that line only when a note is being triggered. to do this, use a second probe into your scope, and take a look at the BD control signal (the one going to R152), and set a second probe to the accent signal. trigger on the BD signal, so youre only looking at the accent when the BD goes high.
openmusiclabs.com

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Thu Nov 21, 2019 9:37 am

OK, thanks. One of the first things i did, was replace the μpd444C. But that did not change anything.

I did another test, probed the accent at R155, the yellow wave; probed the BD trig at R152, the blue wave.

I sequenced the 1/4/9 /13th step of the kick and the same for the accent. (pre scale 3) With the accent knob turned all the way.

I first tried with a very fast BPM. And what surprised me the variations of height of waveforms is regular, we see that 3 short and 1 high succeed one another at infinity. While normally i should have only highs, because i programmed "total accent":

Image

Then i did the same but with the slow BPM, I notice sometimes that there is a time lag between the BD trig and accent. The BD trig seems to me rather irregular, i do not know if normal, next time i will try with trig snare trig to see if i have something different:

Image

User avatar
guest
Super Deluxe Wiggler
Posts: 4328
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Thu Nov 21, 2019 6:07 pm

it probably activates the trigger on each step, regardless if there is an instrument present at that step. it needs both the instrument line to be held high and the trigger high to make a note. so it just holds the instrument line low for those other triggers. since the level is consistent, i dont think the issue is with the trigger level.

when you first started this thread, i listened to the sound sample you put up, and i could not hear the variation you were referring to. is this variation not present in other TR808 you have used?
openmusiclabs.com

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Fri Nov 22, 2019 2:52 am

Yes i see what you mean, it is not obvious sound recording, but in reality with a loud speaker, it really gets along. The 808 had no problem before, curently where's problem, with a rhythmic having a lot of instrument that does not get along, by cons a rhythm with 1 snare and 1 kick that makes it completely scrappy... I compared audio with my yocto V1 actually the kick is OKt on my yocto V1 and it does not comply with the 808. I go compared to the oscilloscope on my yocto V1. I do not have an other original 808 to compare.

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Fri Nov 22, 2019 3:47 am

Aaahhhhhhhhhhhh :santacruz: :santacruz: :doh:

Indeed i think you're right, the problem seems to me that the fruit of my ears too attentive! I explain: BPM really fast, we hear significantly this modulation.
I compared with yocto we also hear it with the decay thoroughly, and total accent. But less, because the yocto kick seems to be less dynamic. I think it's the variations of the envelope that i hear with the decay all the way down and total accent.
In fact, these TR808 had other problems when my friend left me, i repair it but my perfectionist side made me imagine things because of the living side of the machine.
:bang:

User avatar
guest
Super Deluxe Wiggler
Posts: 4328
Joined: Mon Aug 19, 2013 11:49 am

Post by guest » Fri Nov 22, 2019 2:23 pm

well, at least you can now consider it "fixed"!
openmusiclabs.com

User avatar
nologin
Common Wiggler
Posts: 201
Joined: Thu Jan 21, 2016 2:53 am
Location: France

Post by nologin » Mon Nov 25, 2019 5:46 pm

Yes.

I looked at the service manual, there is this drawing the place where probed the accent called common trig, i marked it with red. We see on the left side of the drawing, there is schematically small and high waves.The waveform my display is a few different because it's not an old analog oscilloscope that i use:

Image

I did not probe the 2nd point that i marked green. Because now I know that accent is OK


By curiosity, i compared with yocto.
And i can see that the probing accent is similar, i compared not accented it is 4,4V and accented I have 13,4V and it is clear that the yellow wave is similar than TR808.

Image

But i see that with yocto, The trig to R152. I have a quite different trig, which is regular and it is 5.4V.

And finally everything is not "fixed", i tried last night, SUPRISE! there is another problem, if i make patterns, they are not all patterns at the same BPM and if i try to synchronize with a TR606, the 808 is desync. :ripbanana:

Post Reply

Return to “Music Tech DIY”