Yes I made up four of the CGS gate to trigger circuits ages ago thinking they would be handy one day. I should measure the output and see how wide it is. Anyhow, it has stopped that crazy trigger with an echo on the falling edge (I presume). I can see myself needing more of them in future.
Joined: Mar 20, 2012 Posts: 2179 Location: Chicago
Audio files: 16
Posted: Tue Sep 11, 2012 6:56 am Post subject:
Quick question, probably off topic, might be a good idea to start another topic...
How does one approach a stripboard layout? Perfboard or protoboard stuff is pretty straightforward, but determining how to lay out to the strips is something that I haven't been able to wrap my head around.
I use a free software called DIY Layout Creator https://code.google.com/p/diy-layout-creator/. It can be a tedious task, but I usually start by studying the schematics and then following the signal flow from some point by placing components vertically across the strips. You want to try lay it out in a way that reduces the need to cut traces - after all, the strips are the main benefit of those boards. In my case it usually involves a lot of optimisation, i.e. moving components around to save as much space as possible...
Good question. What I used to do is hand draw with pen, loads of holes between lines on a note pad to represent the stripboard tracks. I'd then draw the components in pencil so I could easily erase them.
Then I drew up stripboard paper! I could print it out and draw in pencil on it.
I then went one step further. I drew up components in Illustrator. What I do now is I work directly in Illustrator. I import the schematic first onto a layer. I arrange my stock components by copying and pasting. At first I work roughly. I use a red X to represent the cuts. I colour the + GROUND and - in red, green and blue so I know what I'm doing. It's handy to start with an IC and work outwards from that. I just follow the schematic. Then it's a matter of shuffling things around to make the footprint small. It's an intellectual challenge, like sudoku, and very relaxing. Once the rough looks ok, I make the board. If it doesn't work, I try figure out where I went wrong. If it looks like it is going to take too long figuring it out, I just try a different layout. When I get something which works, I then draw it up again and make it look pretty. I've got quite a lot of stripboard layouts here I have drawn but never posted. Mainly MFOS stuff and CGS...
Here's the stripboard paper anyway. If you have a vector based programme like Illustrator, I can post the .ai file with all the resistors, jumpers, caps, IC's you'll ever need
EDIT: yeah it's good to NOT cut the traces. That's part of the challenge... of the gentle art of stripboarding. It's not so easy when you have transistors all over the place.
I like LochMaster for doing stripboard layouts. It has a pretty good selection of usable components and lots of functionality specific to stripboarding. Uncle Krunkus seems to be the master of the medium (IMO), as it was his work that got me started with it. Just my 2c.
Unfortunately Lochmaster is PC only and I currently don't have a multiboot installed
DIY LC is pretty basic and buggy, but it does the job as well...
I've tried Illustrator for stripboarding before, but I was having trouble with properly aligning elements to the grid. I also found scaling components a bit of a chore. How do you deal with these issues? Are you using 'Symbols'? If you don't mind, I'd love to take a look at your file as well ...
It's awesome to see how everything is working out for you guys. I just started putting together my stripboards. Looking to the future of my poor mans 808 I'm thinking about setting it up in a 2U rack and controlling it with this.....
Joined: Mar 20, 2012 Posts: 2179 Location: Chicago
Audio files: 16
Posted: Tue Sep 11, 2012 9:42 pm Post subject:
Some more goodies from the Service Manual...
TRIGGER PULSE
"Common Trig with pulse width longer or shorter than 1ms will be a cause of deteriorative voices."
I was talking about how the trigger from the CPU into the common trigger circuit was likely to be 5V, but that's not the case. It's 5V out from the CPU, but then it goes through a transistor inverter and a monostable multivibrator, the purpose of which is to level shift to 15V and widen the pulse from 10us to 1ms. It is a negative going pulse because it's inverted.
(I am still pretty sure that the "gate" inputs to each voice are supposed to be 5V positive pulses)
So the input to Q31 - Q34 which combines the trigger with the accent level is 15V. And again, we have the "1ms" requirement mentioned.
It's interesting that you had to decrease the two Cowbell oscillator resistors; the original units had 390K and 330K respectively instead of 150K and 100K. I'm guessing that differences between the original 4584 hex inverter and the 40106 may have something to do with it, but....
I'm going to try and take this from a different angle. I'm going to set up my arduino to generate the timing pulse and two instrument gates--one for the voice I will test and one for accent. Once I can do that under pot control, I'll add in the accent circuit to generate a common trigger, and then see how those behave with the voices as I build them. This will let me filter out any complications due to using a cheapo sequencer (in my case a 555 timer, in minus' a baby8) and make sure I have the supposedly optimal trigger conditions.
Of course, I'm supposed to get a module from someone to repair here in the next couple of days, so that will likely take my attention away from this for a bit too....
Interesting research there. It does sound like the trigger is important here. I'd be interested to see what results you get with the arduino. I have no experience with them. Maybe you can sort out the triggering issue AND come up with a sequencer for us.
The cow bell resistors had to be changed because as they were I was right at the end of the trim pot rotation and still slightly off the required frequency.
With the Illustrator file, I'll tidy up a master copy and put everything on their correct layers. I should make it snap to grid. At the moment I just align things by eye. Using symbols would be a much better option. I should look into that. I don't really scale things. I have resistors which span across 2, 3, 4, 5 ,6 etc holes. Same with jumpers etc. I simply copy and paste. Using a proper programme to do this would be a better way to go. Might be a lot quicker. I don't have a PC so Lochmaster isn't an option. I don't like the look of the graphics on these things. It's as if they have gone to no effort at all. I guess I sacrifice functionality for aesthetics. Oh well...
Joined: Mar 20, 2012 Posts: 2179 Location: Chicago
Audio files: 16
Posted: Wed Sep 12, 2012 10:18 pm Post subject:
Arduino is generating a beautiful (and easy) 1ms wide pulse, with an input pot to vary the rate. Rate appears to match original spec, within tolerance....oh. For the "TEMPO CLOCK" which is not likely to be the same as the output clock.range is 8ms to 65ms which is .... not going to be the output trigger rate And I realize if I'm going to use the original Accent circuit I will have to invert it as well.
I'm impressed by your research and scientific methodology elmegil. This shall be a good indication how the circuit will perform under optimum conditions. Excellent work!
Joined: Feb 19, 2011 Posts: 73 Location: Newcastle Upon Tyne
Audio files: 4
Posted: Thu Sep 13, 2012 4:40 pm Post subject:
Alright all
Great thread by the way, I've been following your progress. I was working on 808 clones but in PCB
format a fair few months ago now. I had to put it on hold for one reason or another and i'm now in the process
of moving house so all my stuff is in storage, no building for me.
Anyway i'll get down to brass tax. I didn't get any building done while working on this project but i did do a
fair amount of bread boarding and experimenting with possible mod's. in particular the 808 Clap circuit.
I saw eric archers clap circuit here http://ericarcher.net/devices/clap-raca/ and realized that
they were further possible modifications. unfortunately they was no schematic, so i set about trying to find
Eric's group timing mod.
With the help of RichardC64 i was able to finalize the modifications. The schematic is below.
In the 808 clap schematic R333 is replace with a 10k pot to ground
REVERB-DECAY
Thanks to RichardC64 for pointing this out
here is a quote from him
Quote:
Although creating more envelopes would be difficult, it should be relatively simple to change the decay(s)
of any or all of them. The caps involved are C139, C143 and C144. (808 designators.) Increase any of these and add a resistor and pot in parallel, and there you go.
I used a 0.47uf cap and 1M pot
GROUP-TIMING
The group timing Mod consists of two controls.
R350 and C140 control the trigger burst duration -
R356 and C142 control the rate at which the triggers happen within the time frame set by the duration control.
So by adding pots and limiting resistors for R350 & R356 allows you to adjust the comparators
trigger/timing mechanize.
C140 & C142 values allow you set the range of the control pots
depending on what you want from this circuit will in the end, guide the value for C140 and C142.
By choosing large values you make this thing act like a Amplitude Modulation LFO.
For a demonstration the sound files called "trigger burst demo" below have silly values for C140/C142 (2.2uF).
I still need to fine tune the cap values i'm going to use, but the values stated in the schematic are not a bad starting point.
elmegil awesome work on the trigger input research. I'm wanting to use the arduino as midi to trigger so your findings are really helpful. Keep up the good work man.
minus great layouts.
clap-Standard-Values.mp3
Description:
stock 808 clap sample
Filesize:
179.59 KB
Duration:
0:11
Viewed:
2918 Time(s)
clap-Standard-Values-Filter.mp3
Description:
Filter mod sample
Filesize:
145.3 KB
Duration:
0:09
Viewed:
2860 Time(s)
Burst-Trigger-Demo-increase-Dur-Rate.mp3
Description:
I increased the (Group Timing) duration control first then I modulate the rate control.
Joined: Mar 20, 2012 Posts: 2179 Location: Chicago
Audio files: 16
Posted: Thu Sep 13, 2012 6:37 pm Post subject:
Thanks for the kind words. That looks like some good research on the clap too.
Hm. That doesn't sound quite right
I now have a real project that arrived today that ought to pay some money, so I'm really seriously torn between further work on the arduino reproduction of the triggers and that thing. We shall see.... I will get back to it, but patience if I don't continue at the breakneck pace I was going.
Joined: Mar 20, 2012 Posts: 2179 Location: Chicago
Audio files: 16
Posted: Thu Sep 13, 2012 10:37 pm Post subject:
Of course I couldn't keep away I did spend some time checking the other project out, but I really need to tear it down, I think. So I came back to this.
I have the Arduino generating the BD and AC signals in some semblance of the original, including running the output through 1) a level shifter/inverter and 2) the Accent + Trigger sum circuitry (from my previous post).
I'm using pin 13 as the accent "gate", pin 2 as the BD gate, and pin 9 as the clock/trigger out. The only thing not on that previous circuit is that to make pin 9 do the right thing, I run it through a voltage divider (22k/10k/ground) into an NPN transistor, whose collector is pulled up to +15V through a 33K. That goes into the "TRIGGER" point on that other circuit. The Accent gate goes into "ACCENT" and the output is the common trigger.
When there's no accent gate, I get a reasonable drum sound, and when there is I get a louder drum sound, how much louder depending on the ACCENTLEVEL pot.
BTW I'm varying the speed with a pot wired as a voltage divider, being read by the Arduino on Analog in A0.
Here's the code:
Code:
int Trig = 9;
int BDGate = 2;
int ACGate = 13;
int analogPin = 0; // potentiometer connected to analog pin 0
int val = 0; // variable to store the read value
int cycleTime = 57; // max cycle time of 65ms - 8ms constant factor
int scaleFactor = 18; // divide by this to scale cycle time from 0 to 57
And attached is an MP3 of me varying the speed of the full accent sound from fastest to slowest and back again.
This was all complicated by the fact that something is wrong with my input on my macbook, and I can't take the mono signal off the tip any longer. I've double checked things I've seen cause this before, and so far nothing is resolving it . But at least I was able to use some cable tricks to record it on the second stereo channel....
Joined: Mar 20, 2012 Posts: 2179 Location: Chicago
Audio files: 16
Posted: Thu Sep 13, 2012 10:50 pm Post subject:
P.S. Once I have some more time and get another voice or two (probably snare next, and then the cymbals etc from the top of this thread), I think I will use dip switches into the Arduino to program semi-fixed patterns (read at start only) and see how that sounds. I'm also considering whether or not I want to throw the trigger circuits onto a proto-shield so I can get them off my breadboard....
This is great work! I wish I was capable of understanding all that code, but it's a bit beyond me. Maybe my 'neck-beard' son can help me.
Nice sound clip there too. Would you say you are finding it is better with the correct triggering method? Mine does sound similar. How about increasing the decay? Does the weird bounce Boom BOOM thing happen? I was thinking of ways around that. Like maybe somehow discharging that decay cap instantly the next time it is triggered. Or flip flopping between auxiliary caps somehow... Maybe then we could have the BOOOOM if we want at a rapid tempo. I've been sending my output through a cheap distortion pedal. Getting a sound which I do believe kids these days refer to as 'hardstyle'
Joined: Mar 20, 2012 Posts: 2179 Location: Chicago
Audio files: 16
Posted: Fri Sep 14, 2012 5:24 am Post subject:
Understanding the code isn't strictly necessary, you can get an Arduino and just load it, as along as you pay attention to which output pin is which it should work.
I tried switching it back and forth between different triggering methods, but using the Arduino as the trigger source in all cases, and honestly I couldn't hear much difference. Then again, I was not listening on headphones until late in the process so there could have been something subtle...
Even with the Decay all the way up, I didn't notice as much of the variation from one step to the next this way. I think that may have to do with the 1ms width being pretty exact. It ought to be easy enough to change that width and see what happens (which might give a clue for the people posting about why they were getting it so bad--perhaps something in the monostable was getting old and extending the pulse width....)
I did found I got better boom running it through the Mac's Compression, but when I tried to run the results off to MP3 I couldn't hear any difference. Could be because I had "normalize output" selected when saving the MP3, I'm not sure.
Joined: Mar 20, 2012 Posts: 2179 Location: Chicago
Audio files: 16
Posted: Fri Sep 14, 2012 5:37 am Post subject:
Revised schematic. TRIGGER-1MSPULSE is 5V, 1ms wide, positive going pulse, as output by the arudio sketch above. ACCENTGATE is a 5V gate signal, as generated by the code it will end up being the same width, which is just fine, but it could be wider.
Screen shot 2012-09-14 at Sep 14 7.35.15AM.png
Description:
Filesize:
14.97 KB
Viewed:
1421 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
Joined: Mar 20, 2012 Posts: 2179 Location: Chicago
Audio files: 16
Posted: Sat Sep 15, 2012 7:32 am Post subject:
Buit this on a proto-shield yesterday, realized after finishing it I need to address a couple other concerns :/
Also, in testing, now the click seems a lot louder to me as well.
The click is generated by the upper left pair of transistors in the BD ckt, as I understand it, so I'm going to breadboard again and see what effect increasing the 8.2k resistor, changing the cap between them, or changing the type of transistor might have on things.
Posted: Tue Sep 18, 2012 3:10 pm Post subject:
Cow bell not working
Hello minus, I stripboarded the sound source and work perfect.
I stripboarded the cow bell and doesn't work, please can you check your layout?
Transistors T1 and T2 are in the correct place?
Thanks.
Hi. Sorry to hear the cow bell isn't working for you so far. Yes I can confirm that T1 and T2 are correct. I've uploaded my rough I did before I redrew the posted cow bell diagrams. It will give you an idea of the power rails etc. Red is the positive, blue the negative and green is ground. The orange rail was for my own reference so I wouldn't get confused with the two 100K resistors coming off those diodes. The red X's are cut marks. I worked off this diagram, then I opened a copy of it and simply redrew it neater over the top. Check that you have made all the cut marks in the correct places. Check that your components are in the right holes. If you look at the schematic from the service manual I posted on page 1 of this thread, you can compare it one step at a time with the stripboard layout. It's a fairly small circuit.
I assume you are feeding it from the sound source correctly from the... oh... wait a sec. I had the inputs incorrectly labelled! They are meant to be cow bell input, not cymbal input. I'm not sure this will be your issue though, but make sure you are taking wires from the noise source cow bell outputs to the cow bell board at the correct points. I have just updated the original diagrams on page one of this thread. SORRY!
Check your power to the board... check the sound source to the cow bell board ... check your cut marks are all done and in the right places... check components are all in place... check components are of the right value and orientated correctly... inspect the bottom of the board for solder bridges, cold solder joints, cut marks which haven't cut through the stripboard copper...
Let us know how you get on.
EDIT: Check that you have the accent tied to the +V when you test it. Leaving it unconnected will result in no sound.
TR-808 cow bell rough stripboard 1.jpg
Description:
Filesize:
868.96 KB
Viewed:
1387 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
Posted: Thu Sep 20, 2012 1:17 am Post subject:
COW BELL!
Ok minus, big thanks for your explanation! Fantastic layout.
I found a difference between booth layouts, I marked in yellow the jumper at strip 16/17, which one it´s correct?
COWBELLrt.jpg
Description:
Filesize:
1019.98 KB
Viewed:
1584 Time(s)
This image has been reduced to fit the page. Click on it to enlarge.
Alfredo that is excellent spotting! You should be working for Roland! That jumper should be going to the + rail as seen in the rough version. You need to extend the jumper one hole down. I have no doubt this is the cause of the problem. I shall now update the artwork and repost. Thank you so much for spotting this. I try to make these images correct but an error can slip through. I'll check all the others and see if I have gone wrong with them. I do check these things by ticking off the parts with a pink line on a separate layer. Somehow this slipped through. Please post again when you have the circuit working for you.
You cannot post new topics in this forum You cannot reply to topics in this forum You cannot edit your posts in this forum You cannot delete your posts in this forum You cannot vote in polls in this forum You cannot attach files in this forum You can download files in this forum
Forum index » DIY Hardware and Software
FAQ
Calendar
Search
Memberlist
Usergroups
Register
Profile
Log in to check your private messages
Log in
Chat Room
Once the rough looks ok, I make the board. If it doesn't work, I try figure out where I went wrong. If it looks like it is going to take too long figuring it out, I just try a different layout. When I get something which works, I then draw it up again and make it look pretty. I've got quite a lot of stripboard layouts here I have drawn but never posted. Mainly MFOS stuff and CGS...
...
They are meant to be cow bell input, not cymbal input. I'm not sure this will be your issue though, but make sure you are taking wires from the noise source cow bell outputs to the cow bell board at the correct points. I have just updated the original diagrams on page one of this thread. SORRY!
You should be working for Roland!