unmuxed-one pot...tons of jitter - but works with traktor

[djiceman]

Hi

Newly built box, running bootloader 1.2 and mios 1.9c with midibox64 2.4

I know many are going to say its a connection or wiring problem.

But ive checked the connections and wiring 4 times with a beep meter.

All unused pins are grounded.

When i plug in the box to the pc, its quite, i get only one upload request(normal).\

Once i turn the pot, then the jitter starts, in midi-ox in the port activity window, input ports 1-16 flash at random but very fast.

How can this be, if unused pins are ground and i checked the ground with the beep meter, it beeps when i touch the pins and ground on the other end of the board.

Im only using one pot to debug this.

When the pot is in the middle and i use the beep meter on its center leg and on a grounded pin, it wont beep.

If i put the pot value to min, the meter beeps. So wiring checks out.

Just a reminder, I dont know if this is normal, but the jitter only starts once i turn the pot.

So if i put turn on the box, it comes on as i get the one upload request, then it sits quite. Once I turn the pot, its endless jitter, all different midi events.

So wiring/connection fault or something else?

Just abit more info,when i connect the ain module and ground the proper pins and so on, same as above, only, the jitter comes from potwhich i turn. So if i turn only one pot, the jitter comes from there, if i continue to turn other pots, the jitter comes from thos aswell. This is according to the port activity monitor in midi-ox. The jitter is only reverb depth.

Ive tried the box with fruity loops. The jitters cause some bad play back. But if i turn the box off and on while im in fruity loops, it works. I set the pot to control the volume. Works perfect.

Then I tried traktor dj studio. The box works fine, no problems and I dont need to off/on the box.

Even when I use the learn function, more than once, it all seems fine.

So i removed the entire ain module and im using one pot to check if its a problem with the ain or core module. Seems like a core problem.

The software only works with the ain module in, but when i use the unmuxed one pot with the software, i get jitter problems.

Offcourse, when i removed the ain module and connected one pot directly to the core, i ground the proper pins.

[henrygr]

When you have the core unmuxed, I assume you set the program to beahave as such- ie set mode to unmuxed.

So, when you have done that, set the number of pots to one, That way, the PIC will only accept signal from that pin. It doesn't matter if the other 7 are ground or not, as the PIC is not looking for signal from them. So set it to one, test it, set to two, test them and so on......

Then post the results. That is how I trouble shoot, so personally would be better able to read the problem through that route of troubleshooting.

Mark.

[djiceman]

When you say set the program? do mean edit the midibox 2.4?

or something else?

[Wild_Weasel]

Yes he means the MB64 application.

You need to recompile it afterwards.

regards

WW

[henrygr]

Ahh. You have the full app, of course.

If you are set up to compile in C, follow this link and adapt as previously mentioned

http://ucapps.de/mios_c_send_ain.html

and here

http://ucapps.de/mios_c_send_mapped.html.

If not, then a workaround would be such. I know you have done most of it, but bare with me...

Ground everythhing bar te first pin. Try your pot. If it jitters then do the following.

Run three leads from the Vs, Vd and the AIN.

Take a resistor of a value about 7.5k, and connect it between the Vs and the AIN, There should be no jitter, as your AIN is esentially grounded, as though you had turned a pot three quarter ways around away from where the Vs was connected. Take a second resistor of a value about 100R. Use this to connect the Vd to the AIN, while the Vs is still connected. The value in MidiOx should change to about 126 in Midi terms, and when you break the contact, the value should return to zero, WITHOUT jitter. If it helps, use the Vs of J2, and the Vd of J4. If this works, then your pot is dodgy, and if there is still jitter, check C6 for proper connection.

Hope this helps,

Mark.

[djiceman]

ok im going to try this just now.

But in the mean time, yesterday i removed the one pot and used the ainx board which has 8 pots on it.

Jitter only comes from the pot which i turn. Once i turn the pot for the first time after turning on the box, it starts jitting on that pot.

Any explanations there?

Im going to try the resistor testing method you described now and will post back shortly.

[henrygr]

Were you holding the pot in your hand?

[djiceman]

i did a few times, but its lying on a wood top.......i just turn it and then leave it.....

[djiceman]

just did your resistor method test.....

results: My pots is 100% fine......

I get jitter from the resistor test...the jitter only starts when i touch the 100kO leg to the ain.

What can I conclude and do to fix this?

I checked c6 (330nf), its connections are proper.

[henrygr]

I think it's a fair conclusion to say that there is something loose on the core side of things. I mean, with the resistors, we have given a voltage that is exact, leaving no room for mechanical error, and still the jitter comes. Jitter generally only kicks of after a voltage return of some sort has been sent to the AIN (this is a broad generalisation, but concise enough for this examle).

If you are getting jitter, then there is leakage, and that has to be somewhere in the soldering or in the capacitors. Check your caps with a multimeter set to, say, 20k resistance. If there is nothing but 100% resistance then the cap is faulty (can be a bit tricky checking them on the board, as the circuit may completes at another point that will give you a fals reading). Check tha solidity of all soldering. If it turns out to be none of the above, at least you're two steps closer to the resolve ie figuring out what it is not!!

Mark.

[djiceman]

removing components is hard, very hard to get rid of the solder with the sucker.

But i have to try.

In the mean time, my mb64 application is in asm, so im downloading mplab. Going to try that first.

I dont really have a problem, my main intention is to use the box with traktor and it works fine, just with fruity loops and with midi-ox, i get jitter problems.

Need to get more board to build the dinx module.

What do you guys recommend as materials for the case for this midibox? something light and easy to work with.

[djiceman]

i tried the rebuilding of the software with unmuxed and only 1 pot.

I get jitter....guess its a hardware problem, im thinking to build another core.......

because i checked the core out and everything seems in order.....

[smashtv]

removing components is hard, very hard to get rid of the solder with the sucker.

Before you pull parts lets look toward your power supply.....can you give us the specifics on what you are using?  (a noisy supply will cause jitter)

I dont really have a problem, my main intention is to use the box with traktor and it works fine, just with fruity loops and with midi-ox, i get jitter problems.

It sounds like traktor is averaging the value(s) on the fly to avoid issues like this, while FL and midi-ox are not......Even though it works now if the jitter gets any worse* you might see jumpy values with traktor also.

*"jitter gets any worse" like if you use the box in a different/higher noise environment, plug it into a dirty/noisy source of power, etc.

Best

Smash

[djiceman]

my psu is a "deluxe universal ac adapter"

Model: mw96

input: 240v 50Hz

output: 1.5, 3, 4.5, 6, 7.5, 9, 12 (theres a switch on it that you can select the output)

current: 500mA 9W

Im using a 9v output.

This is a wall adapter.

I can put up pics of this and my box if you guys want. But the midibox isnt in a case yet. Though the pots and the ainx module are on some ply wood.

[djiceman]

in regards to fruity loops, the only time i get a problem is when i press the play button(on the software using the mouse) to play the sounds, then it like jitters and the midi light(in the software) flashes very fast.

If i turn off the box, fruity loops plays normal, while already playing, i can turn on the box and itll work fine.

But if I press stop/play again, then i got to off/on the midibox.

Thats basically my problem with fruity loops. Other than that, it works fine.

[henrygr]

The PSU sounds fine, althouh SmashTV was right, I should have pointed you toward that earlier. SamshTV also makes a very valid point in that if you don't nail this now, it will get worse. Before you remove any componants, set up he resistors as I pointd out earlier, and then jiggle bits on the core (not very scientific, I know). If there is a loose connection, and you somehow find it, the jitter may stop momentarily.

Also check your voltage across the board with reference to the following

http://www.ucapps.de/howtodebug/mbhp_core_extract_measuring_vdd.gif.

Actualy, do the above in reverse order....

[djiceman]

i just did the vdd check as in the gif file.

though my meter reads 4v everywhere. Im not sure if thats cos im using a 24 euro meter.

Even on the output of the 7805, it reads 4v and the inlet of the 7805 reads 10v, though my 7805 never gets hot, not even warm.

I then put the meter on j1 from the adapter, it reads 12v, though the adapter is set at 9v output.

So i set the adapter to 12v output and i put the meter at j1, it reads 17v.

So is there a circuit fault or is it my meter. Because with a 12v input, the 7805 should put out a nice 5v but the meter reads 4v.

My meter does not display decimal values and maybe it is also just rounding off the value to a lower value?

So my voltage coud be 4.8 but the meter shows me 4v. Just a theory.

[henrygr]

I would imagine if it was rounding off 4.8v it would go to 5v. You have a major problem here in terms of functionality.

So is there a circuit fault or is it my meter. Because with a 12v input, the 7805 should put out a nice 5v but the meter reads 4v.

Directly check your meter if you want to see if it is that.

But, irrepestive, even with 12v at 500ma, you should be getting 5 to 5.4 volts at most of the locations. Lets remove that PIC till we get to the bottom of this as a safety measure.

The drop in voltage would explain the random data. Check the joints on IC3 (7805). If its nor even getting warm, then it sounds that it may not be connected properly. Also check the joints on C4, , C5, C6, T1 and BR1.

The best way to do this is, on the printed side of the PCB, trace the printed circuit from the base of the pin you are testing to the next available bare area (if you are using SmashTV. Otherwise touching the coppper would do fine). Put one end of the multimeter at this area, the other to the pin on the far side of the board (this may be difficult if you have soldered the components close to the board) and check for resistance. For example, to check the middle pin of TI, check for resistance between the pin itself, and the joint of R4 where it meets the board closest to IC3. Make sense?

Anyway, you'll figure it out. Another tip is a soldering one. If you find that a joint is not right, rub the solder that is already there with a bit of flux. (I use plumbers flux, which has a smooth ceamy consistency like Oil of Ulay, but I wouldn't recommend rubbing it on the face, though admiteddly, considered giveing it to my mother-in-law for christmas). This has the effect of thinning out the solder as you heat it, and lets is set better around any contact that is there.

Keep me posted.

Mark.

[djiceman]

i did solder the components close to the board. Im not using smashtv, im in south africa and the price was too high for me, so i just created boards using the layouts provided.

I changed the 7805, i seem to have ordered an extra one. The voltage is still 4v.

Interesting, at j1, its 12v but at the output of the rectifier, its 11v.

Think maybe its cheap components?

the 7805 is by WST and is a lm7805cz.

The shop didt have stock of ST components and the 78xx components from ST have a different size.

Can i use a 7806 in this circuit?

[djiceman]

i used a 7806 by ST in another circuit. It measures 5v output.....but the power circuit design for that is the same as the one in the midibox.

Ok, i will try the 7806

[smashtv]

i just did the vdd check as in the gif file.

though my meter reads 4v everywhere. Im not sure if thats cos im using a 24 euro meter.

Even on the output of the 7805, it reads 4v and the inlet of the 7805 reads 10v, though my 7805 never gets hot, not even warm.

This is bad, if you can't trust your meter not much point in testing anything....  :-\

I then put the meter on j1 from the adapter, it reads 12v, though the adapter is set at 9v output.

So i set the adapter to 12v output and i put the meter at j1, it reads 17v.

This is not too unusual.  Most of these adapters are not very well regulated, and will output way more or less than printed spec in relation to load....Current draw goes up, voltage goes down and your meter pulls almost no load.

(I use plumbers flux, which has a smooth ceamy consistency like Oil of Ulay, but I wouldn't recommend rubbing it on the face, though admiteddly, considered giveing it to my mother-in-law for christmas)

hehe.... ;)

Another thing to keep in mind, some PC MIDI interfaces tie the PC's ground to the center pin on thier MIDI input......This can cause jitter (ripple from the PC's switching supply leaking across the ground).  To eliminate this possibilty disconnect the center pin of the CORE MIDI out (J12).  The MIDI out will still work fine without this.

Can i use a 7806 in this circuit?

I would strongly advise against this.....If your meter is lying to you 6 volts will kill parts.....

Best

Smash

[djiceman]

im using j11 to gameport.

I just tried the 7806 which gives a 5v output according to my meter which cost 200 bucks locally...thats alot for me....

I did the voltage check with the meter, it shows 5v in the needed places.

Hooked the pic back in and did the resistor test, i still get jitter.

Im thinking, i will just build another core but use the same pic. I will leave out the diode and the optocoupler since i use j11 to gameport.

Btw, the optocoupler is not on the board. I wasted money buying it in the first place.

I need a new soldering iron, the tip on my is eaten away. So no more work tonight.

Until i build the new core, ill just use my 8 port midi box with traktor. hopefully I get parts and new soldering iron by tomorrow.