Weird noise with 3.2 kit

Antwort erstellen


Diese Frage dient dazu, das automatisierte Versenden von Formularen durch Spam-Bots zu verhindern.
Smileys
:D :) ;) :( :o :shock: :? 8-) :lol: :x :P :oops: :cry: :evil: :twisted: :roll: :!: :?: :idea: :arrow: :| :mrgreen: :geek: :ugeek:

BBCode ist eingeschaltet
[img] ist eingeschaltet
[url] ist eingeschaltet
Smileys sind eingeschaltet

Die letzten Beiträge des Themas
   

Ansicht erweitern Die letzten Beiträge des Themas: Weird noise with 3.2 kit

Re: Weird noise with 3.2 kit

von bovist » 16. Jan 2013, 00:57

edwinvp hat geschrieben:Hi,

Perhaps anyone knows about the following?
Just assembled a HOAX 3.2 kit. Upon applying power, the C test chord can be heard as described in the manual.

However, the notes sound distorted. Some notes more than others.
Please see viewtopic.php?f=3&t=15 for this topic.

Re: Weird noise with 3.2 kit

von edwinvp » 29. Aug 2012, 20:31

Hello Carsten,

Welcome back, hope you enjoyed your holiday!

Who said anything about shorting the stereo channels in the plug?
I use stereo plugs, not the mono ones.

And yes it's stupid to have connected the headphone directly to a line level output.
Took some time before that quarter dropped, I should've known better.
I guess I was bit tired after soldering all those parts ;-)

Oh and it sounds logical that the analog scanning is shut-down after receiving MIDI messages.
Otherwise there would be two captains on the bridge.
You probably found this out while implementing it, right? ;-)
(This would happen to me)
4. There is an HOAX issue with Yamaha Tyros "MIDI Active Sensing" (may also occur on other master keyboards). May lead to missing or hanging notes when using only one channel, as reported by a customer.
Don't have one, but interesting... What does Active Sensing do?
5. Use a heatsink on U21 if external 9 to 12V wall wart is used. Heatsink (aluminum plate) is big enough if you don't burn your fingers when touching.
I'd suggest to keep the input voltage lower than that anyway (below 9V). Regulator supplied is of the low drop type.
Combined with a stable power supply the board will be okay (but don't bypass the regulator of course).

Regards,

--
Edwin

Re: Weird noise with 3.2 kit

von bovist » 28. Aug 2012, 14:18

Hi folks,

sorry for late answer due to vacation. To lighten things up, here are some gentle words from the designer ;-)

1. F*CKN' NEVER EVER bridge Leslie stereo or Organ/Bass channels directly (e.g. in the 1/4" plug) to obtain a MONO signal. It will produce distortions. I really could SMASH MY HEAD TO THE WALL (ok, I've done that already): Having used four 2k2 output protection Rs instead of those DAMNED 470R, and everything would be fine. Use separate channels instead. Leslie mono output is best if only left channel is used, so leave roght channel open.

2. Do NOT EVEN DARE to use a F*CKING headphone on the outputs. It will produce distortions for the same reason.

3. My NI B4 MIDI implementation should be complete for drawbar, percussion, vibrato/chorus, volume, leslie control and other basic settings, so complete MIDI remote control should be possible. However, MIDI control may be too slow if you play "Flight of the Bumblebee" while working the drawbars simultaneously.

3a. If HOAX receives a valid MIDI controller message (exept volume/expression control), it shuts off scanning ALL of it's own analog and digital control inputs (exeption: volume pedal). To re-enable the analog handling, select and de-select one of the four common presets. So you may use either MIDI or direct analog drawbars/potentiometers/switches.

3b. If MIDI expression/volume control is received, direct expression pedal (input jack) is disabled. It is re-enabled by simply using it which overrides the MIDI expression value.

4. There is an HOAX issue with Yamaha Tyros "MIDI Active Sensing" (may also occur on other master keyboards). May lead to missing or hanging notes when using only one channel, as reported by a customer.

5. Use a heatsink on U21 if external 9 to 12V wall wart is used. Heatsink (aluminum plate) is big enough if you don't burn your fingers when touching.

6. Zener Diode on HOAX board is for short-time surge protection only, not for regulating. If you go over 6.8V input voltage with some amps for several seconds (U21 disabled by jumper JP12), the Zener will emit it's magic smoke. And as you know, once the magic smoke escaped, it will refuse to work. With some bad luck, the atmel controller and the ADCs will also blow off their magic smoke.

Re: Weird noise with 3.2 kit

von happyfreddy » 15. Aug 2012, 23:18

hi
I know what You mean talking about using the HOAX via MIDI. The difference is the point of view.
When I talk about HOAX then the HOAX board is implemented in a real organ iE there are two manuals
each with 61 keys, a 25 key pedal and of course two drawbarsets for each manual. To make different sound
when playing You have to variate the drawbarpositions as You like or switch between the two sets. When You remember
great organists like Jon Lord, Brian Auger or Jimmy Smith ( and many others ) they often variate the drawbars with pressed
keys. Not even one drawbar, they use full hand to variate the complete drawbars. If You want to imitate a sound like WahWah
You have to regulate 4` and 2 2/3´ together up and down while playing a pointed cluster on keys.
I´m sure You will hear in this case a difference between real drawbars connected or using a midicontroller. The Midicontroller will work
but not so quick as real drawbars.
Okay perhaps extremly example but I use the Hoax as that it was planned : implemented in a real organ and that´s the point of view
when I talk about the hoax.
happyfreddy

Re: Weird noise with 3.2 kit

von edwinvp » 15. Aug 2012, 22:38

happyfreddy hat geschrieben:
edwinvp hat geschrieben: I mean with the adjustments only the levels between HOAX and amp . Output level from Hoax is line level and some clipping will be the result of the position of the gain pot at the amp.
Volume setting in real time via pot is more comfortable than any setting by programming parameters. If You want more volume in lesliechannel You only have to turn the pot nothing else.
Full playback and full control from a sequenzer even in MIDI Mode will cause a LATENZ in the system. Full control means that each Drawbar each Pot or switch must be acitavted and regulated by a controller PERMANENTLY ! plus the NOTE ON/OFF commands. Remember the Baudrate of 31250 Bd and what you can transmit per second. The Latenz only causes in the MIDI Commands not in generating the sound. Therefore it is decided : no MIDI Control of the drawbars. If You want to regulate the Drawbars live You have to regulate them manually at real drawbars.
Now what if you have a fine to very good MIDI controller that already has all the pots and sliders you want?
Would you still attach all those extra PCBs?

Granted, there will be some latency, as with all digital systems. But for the HOAX, I read it will be down to only a few [ms] or so.
So that's still very good (if true) and barely noticeable while playing. If noticeable at all.
I bet most people will start to feel the latency effects only after about roughly 8-10 [ms].

BTW: I could actually control the drawbars over MIDI yesterday, so that's apparently already built into the HOAX.
That's when I wondered if the "other" settings couldn't be controlled as well. Why one setting but not the other?
Seems random.

Realistically, I reckon it has to do with the original purpose in mind: converting/upgrading an existing Hammond organ.
Obviously, MIDI was just a distraction to the designer. But for me, it's presently the only way to get sound from the board!

Now about your MIDI explanation: I think the picture painted by you about it is a bit too dark.
In any case, if I were you I'd immediately buy another MIDI controller if yours is indeed working like you just described.

Constantly using MIDI-bandwidth to update something with the same controller value over and over? Doesn't sound good!
Even my simple E-MU MIDI-controller only sends updates when the pot (or slider) is actually changing.
If any pot is left alone, no MIDI traffic is generated for that.
Sounds quite what you'd expect, right?

It also means you are safe as long as you are not trying to adjust hundred pots/sliders *at the same time*.
But who actually does that?

For fun, I watched the MIDI messages as they came in from my digi piano.
Also this device only sends pedal (controller) messages when you vary the foot pressure on the pedal
At any other moment, MIDI stays quiet (apart from some sync. messages that are always transmitting).

The only problem I can think of now is what would happen if the A/D-converter of the pot would not give a monotone reading.
But something with noise instead: i.e.: 33 33 35 33 33 34 35 35 33 33 33 34
Then the computer would think the value is almost constantly changing and indeed waste bandwidth like you describe.
Horror!

But hey, when even a cheap E-MU controller seems to be able to get this right, there's still hope :)

And when there's a (configurable) choice between either hardware scanned controls or indirect MIDI controls, everybody is happy.

Regards,

--
Edwin

Re: Weird noise with 3.2 kit

von happyfreddy » 14. Aug 2012, 21:20

edwinvp hat geschrieben: From your words, it sounds like some adjustments can only be made with potentiometers?
A nice feature here would be if the source for certain controls could be selected. As with the scan cores.
Then you could configure if - for instance the leslie volume setting - would come from a pot -or- a MIDI controller change message.

This would allow for full playback and full control from a MIDI-sequencer. The Xilinx FPGA probably has the values in its registers right now anyway.
So, this 'routing' should be trivial. However, the Atmel MPU might also be involved with the MIDI-processing. That could make things complicated again.
Edwin
hi
I mean with the adjustments only the levels between HOAX and amp . Output level from Hoax is line level and some clipping will be the result of the position of the gain pot at the amp.
Volume setting in real time via pot is more comfortable than any setting by programming parameters. If You want more volume in lesliechannel You only have to turn the pot nothing else.
Full playback and full control from a sequenzer even in MIDI Mode will cause a LATENZ in the system. Full control means that each Drawbar each Pot or switch must be acitavted and regulated by a controller PERMANENTLY ! plus the NOTE ON/OFF commands. Remember the Baudrate of 31250 Bd and what you can transmit per second. The Latenz only causes in the MIDI Commands not in generating the sound. Therefore it is decided : no MIDI Control of the drawbars. If You want to regulate the Drawbars live You have to regulate them manually at real drawbars.
MIDI IN is an option for the HOAX, best results even with direct connected manuals like FATAR.
All what is possible in MIDI section is documented in capture MIDI.
happyfreddy

Re: Weird noise with 3.2 kit

von edwinvp » 14. Aug 2012, 20:19

happyfreddy hat geschrieben:hi
The diodes anyway if 1N4004 ode 1N4007 doesn´t matter, its only for security if changed polarity of DC supply, In conclusion with this fact
the input Voltage must be 8,7Volts DC MINIMUM ( 3 Volts for Regulator 7805 correct operating and 0,7V for 1N4004 or equivalent ).
The output level of the HOAX 3 is LINE OUT signal. If You want to drive a headphone You must first connect a headphone amlifier.
There are two OUTPUT connectors DIRECT or via LESLIE. The leslie channel has a separate volume pot on Plug 9 Pin 11 (Regulator Pin on plug 9 /11 and GND and Vref at the ends same as drawbars). You can measure the Vref direct at the drawbars , it must be +3V3 to GND. Drawbar pots have the value of 2K2 to 10K but use only one value for all, never mixed.
There is no need to change any level via MIDI or so. DON´T TRY ANYTHING WITHOUT KNOWING WHAT YOU´RE DOING !! ( specially RANDOM Changes in try and Error matter , marked in RED in DOC Hoax 3 Syntax !!) The problem You have is a uncorrect connection to Your Amplifier. The output of HOAX is low Impedance so there is any problem with th input impedance of Your Amp. Use a volumepedal as marked and reduce first the Input Level of Your Amp.
happyfreddy
Be gentle on the red font and the caps... I promise that I will take proper care for my new instrument! ;-)

And you are completely right: the exact diode type doesn't matter.
But what might matter for future builders is that they do not place the zener instead of the general purpose rectifier..
That was the real point I was trying to make. It only needs a few extra hints on the top diagram I guess.

It sounds like I have an additional soldering job ahead of me, in order to get the sliders and volume knobs at my disposal ;-)

From your words, it sounds like some adjustments can only be made with potentiometers?
A nice feature here would be if the source for certain controls could be selected. As with the scan cores.
Then you could configure if - for instance the leslie volume setting - would come from a pot -or- a MIDI controller change message.

This would allow for full playback and full control from a MIDI-sequencer. The Xilinx FPGA probably has the values in its registers right now anyway.
So, this 'routing' should be trivial. However, the Atmel MPU might also be involved with the MIDI-processing. That could make things complicated again.

Meanwhile the organ now plays via an amplifier.
Think I will get myself the Yamaha FC-7 expr. pedal soon...

--
Edwin

Re: Weird noise with 3.2 kit

von happyfreddy » 13. Aug 2012, 20:16

hi
The diodes anyway if 1N4004 ode 1N4007 doesn´t matter, its only for security if changed polarity of DC supply, In conclusion with this fact
the input Voltage must be 8,7Volts DC MINIMUM ( 3 Volts for Regulator 7805 correct operating and 0,7V for 1N4004 or equivalent ).
The output level of the HOAX 3 is LINE OUT signal. If You want to drive a headphone You must first connect a headphone amlifier.
There are two OUTPUT connectors DIRECT or via LESLIE. The leslie channel has a separate volume pot on Plug 9 Pin 11 (Regulator Pin on plug 9 /11 and GND and Vref at the ends same as drawbars). You can measure the Vref direct at the drawbars , it must be +3V3 to GND. Drawbar pots have the value of 2K2 to 10K but use only one value for all, never mixed.
There is no need to change any level via MIDI or so. DON´T TRY ANYTHING WITHOUT KNOWING WHAT YOU´RE DOING !! ( specially RANDOM Changes in try and Error matter , marked in RED in DOC Hoax 3 Syntax !!) The problem You have is a uncorrect connection to Your Amplifier. The output of HOAX is low Impedance so there is any problem with th input impedance of Your Amp. Use a volumepedal as marked and reduce first the Input Level of Your Amp.
happyfreddy

Re: Weird noise with 3.2 kit

von Gast » 13. Aug 2012, 20:01

Kermit hat geschrieben:Hi Edwin,
So now I´m just waiting for Carsten to get back from holiday, so I can order the licensekeys :D

Claus
Makes me wonder what those undocumented resistors are for... ;-)

Disclaimer: just joking!

--
Edwin

Re: Weird noise with 3.2 kit

von Gast » 13. Aug 2012, 19:40

edwinvp hat geschrieben:Hey there and thanks for all replies so far, it now works far better!

Now there's a new problem: playing two or more notes simultaneously still clips, but in an entirely different way.
One note mostly doesn't clip.
Lacking further investigation, I'm thinking that it now clips to the 16-bits digital min/max values?
Perhaps the 'volume' is all up to eleven, because I lack any knobs, additional boards?

I'm hoping to find a MIDI parameter to put the volume down. Since I have no extra drawbars, pots...
MIDI controller CC 7 seems to do the trick for me...

Nach oben