Strange Noise

[SLP]

Hi,

I've finished one SID and one Core of my dual-SID, I've loaded the SID testtone-app on my master-Core, switched it on and after 2 seconds of a (nearly) pure triangle-wave, I heard another strange sound.

As it's very difficult to describe this kind of question, I've attached 2 pics and an audio-file.

(I know it's pretty useless to zip a wma-file, but wma-files are not allowed).

Parts of the audio-wave:

1 ? no idea. quite loud noise though it's switched off.

2 switch-on-click

3 either the releasing of the power switch or the Relais which switches the 12V-line. (black box on the                                  picture)

4 here begins the pure triangle-wave

5 here starts the additional strange noise

6 triangle + noise

7 the noise stops. triangle

8 the noise starts again

9 switch-off-click

My "Class 2 Power Supply" also offers a GND-connection. should I use this one instead of the normal Vd ?

thx,

matthias

sidnoise01.zip

Unbenannt.JPG

sidnoise01.zip

DSCN06380001_1.JPG

[dj3nk]

do you have 6581 or 8580 sids ? what powersuplly do you use ? the original c64 one in the "optimized version" ?

greetz

[SLP]

oh, I forgot to mention the most important things...

the 2 cores run the V1.74685 firmware.

Master: connected to 8580, no CS

Slave:  connected to 6581. not connected to master at the moment (CAN)

I'm using the black powersupply you can see at the picture.

I've just put the Master PIC into the slave-core. the 6581 works pretty well. (except for that "normal" noise)

I'll try to connect the 8581 to the slave board tomorrow to see if it's the core that doesn't work.

[stryd_one]

What kinda PSU is that?

[SLP]

it's a "Skynet DMH-1202-A" "class 2 power supply".

I've bought it at pollin.de.

it offers

+5V (1A)

+12V (0,2A)

-12V (0,2A)

and 2 other lines. I don't know how you call them in english.

there is a normal "minus" (Vs) wire and another wire that is connected directly to ground (earth, green/yellow cable).

I think that the powersupply is just the reason for the 50Hz-buzz and the usual (white? pink?) noise.

I'll try to solve this problem in the next step.

The problem I want to solve now is this kinda disturbing second signal. It doesn't occur at the 6581.

any ideas?

matthias

[SLP]

ok. problem solved

One definition of an expert is "a person who has made all the possible mistakes in a narrow field".

hmm.. there aren't very many mistakes left that I haven't done  ;D

I just didn't push one multiplexer entirely into the socket.

New problem - same headline:

If I upload the SID-application and start the MB, I hear some really weird sound.

I'll record it later.

I tried every single Core-SID combination, and tried with both the PIC18F452 and PIC18F4685. It's always the same.

After booting up, i shouldn't hear any sound, right? There's no BankStick connected.

thx,

matthias

[Sasha]

I suppose what you hear is the noise that comes from the powering the circuit, similar to what you heard when powe up the amp that have no delayed speaker output relay. I think it is normal.

[SLP]

I do always hear that noise, not only after powering on..

noise2.zip

noise2.zip

[Sasha]

I hear same noise in the background after I turn the volume to the max, but seams signal to noise ratio is still 90 and something dB, as it should be. I`m probably not right person to answer it, but I think that noise is normal. MAybe it could be improved using some better PSU.

[SLP]

I hear same noise in the background

background? 

everything you can hear shouldn't be there.

the 50Hz buzz and the "usual"-SID-and-PSU-noise isn't the problem.

I turn it on and the noise starts. If I send any note-on messages nothing happens.

[SLP]

No ideas?  ???

Is it likely that this problem is caused by the PSU?

[TK.]

Last week I've heard exactly the same noise on a MBSID - the reason was a bad soldering joint on the C1 cap (connected to pin 1/2 of the SID).

So, it makes sense to check the C1/C2 cap connections

Best Regards, Thorsten.

[SLP]

Unfortunately that's not the reason  :(

[SLP]

I tried something different (see attachment).

result: No noise  :D

now there are 2 possibilities:

- The noise was just because of the bad PSU

or

- I've just killed something (I'll try the testone-app tomorrow)

...something inside of me sais that Murphy was right...  ::)

[Altitude]

Hmm, Interesting power supply you have there.. I wonder if the quatro has better ripple current performance

I tried something different (see attachment).

result: No noise  :D

now there are 2 possibilities:

- The noise was just because of the bad PSU

or

- I've just killed something (I'll try the testone-app tomorrow)

...something inside of me sais that Murphy was right...  ::)

[TheAncientOne]

Wow! Are you are going MIL spec then? I'm just guessing by the environmental tests. High temp, vibration, etc.

On a serious note, it might indicate a power supply  problem. My first ever project from 'Practical Electronics' was a 3 transistor (germanium!) amplifier. It made all kinds of weird 'motor boat' noises until I shunted the dry battery I was using with an electrolytic. My first lesson in power supply stability.

Looking at the noise in 'Audacity', and 'blowing it up', I notice a background 10mS rough half sine wave, plus the HF noise. You're coming out of regulation, unless you're pulling more than (say) 150mA, so I'd try a series inductor in the 12V make sure you have local decoupling capacitors on all your boards. You should have both a small electrolytic with a ceramic in parallel, (Electrolytic don't work too well at high - like PIC clock - frequencies). All in all my vote is for the PSU as a source of the noise. From the description, I'd say your supply was a switch mode. I prefer a linear for most audio kit. There are usable switch modes, (I've just bought an old Akai 3000 sampler that uses one), but they are 'instrumentation grade' and expensive.

Try loading the supply with resistors - suggest 4.7Ohm on the 5V and 150 Ohm on the 12 V rails, (get big ones 10 watt for the 4.7 and 2 watt for the 150's), and then put a scope across the power rails, on AC coupling, an see what you get. I use a test like this often, before trusting the supply with the expensive stuff. I've just done something similar with both the Oakley 3030 I've just completed, (welll all bar 1 LED), I put the regulator chip in before any others and set and tested the power supplies before inserting the expensive stuff. With the AVRx board I've just started, I built the PSU section first, out of the normal order, both to make sure it worked, and as a first test of my surface mount skills. With surface mount a failure is usually a lot more expensive....

Perhaps as an overall hint to MIDIboxers in general. A couple of bench power supplies, either second hand, or kits to save money, are a really worthwhile investment, second only to a test meter. I say a couple because you are going to need at least a 5V and the analogue power.

[stryd_one]

Wow! Are you are going MIL spec then?

Argh don't do it unless your funds are limitless hahahahah... I was gonna go that route but trust me, commercial will do (especially seeing as even the military don't bother with MS any more, it's all commercial rated) and even then it's a total mission... Let me know if you're gonna do it and need specs, I have a few hundred meg in a collection here.

[SLP]

it might indicate a power supply  problem.

Don't you think the stabilizing caps on the Core and SID-Module are enough?

I've downloaded my noise-record again and must say that the record-level was too low. it's much louder.

There seems so be an other problem... ::) (This thread is never going to end  :P )

When I power only the Core, something seems to oscillate. There are some sounds that are generated and then led through the ground-wire to the SID and then to the audio out and to the amp.

When I power the SID on now, there is a 1KHz triangle-wave (?-quite the same as the testtone-app, though the SID-app is on the PIC) and the additional noise from the Core.

When I power just the SID, there's a quite loud noise.

...I'll make a video and put it on youtube  ::)

edit:

I won't put a video on youtube..... modem.... :P

[SLP]

[depressed mood]

ok...

I've completely disconnected the SID-module from the Core and removed every IC from the SID-PCB.

If I power it on, I hear the strange noise. If I take the connector which is used to connect the Cables for the SID-Core-connection and move it closer to the PIC, the noise gets louder.

It seems that the PIC sends some electromagnetic waves which are recieved by the SID-PCB and forwarded to the amp.

Now I take the input-wire of my amp (TDA 7052,btw.) and connect it to every pin of the PIC, I can hear the loudest noise at pin#:

6,19,28 (noise#1) and

22,21 (noise#2)

edit:

Testtone application works well, btw.

[SLP]

here you go:

http://www.youtube.com/watch?v=rqWhOjqHhq8

Crappy noise recorded with crappy Digital camera mic.  :P

[TK.]

It sounds like digital noise, couldn't this be a corrupted patch?

You are writing that the testtone application works well - does it mean that you don't hear such noise in background anymore?

Then I would say, that it is very likeley that you missed to solder the two 1k Pull-Ups for the SC/SD line of the BankStick?

Best Regards, Thorsten.

[TK.]

Second assumption: I guess that you are not using the audio input of the SID - is it connected to ground? (this is recommented)

Best Regards, Thorsten.

[SLP]

Second assumption: I guess that you are not using the audio input of the SID - is it connected to ground?

yes. the 2 pins are shorted.

the strange thing is that even though the SID-chip isn't plugged into the socket the noise occours.

I'll try to fix the bankStick-issue now... and I pray that this was the error  ::)

[SLP]

I've put a 1K (2*2K in paralell) resistor between SC and +5V and another one between SD and +5V.

still noise.  :'(

btw. the noise starts when the display shows the MOIS version.

[TK.]

Once the MIOS screen pops up, the application starts to access the SID via the 74HC595 shift registers. If you would connect an amp to one of the 74HC595 pins, I'm sure that you would hear such digital noise.

Are you sure, that there isn't a "hidden" connection between the 74HC595 and the transistor amp stage (compare with http://www.ucapps.de/mbhp/mbhp_sid_v3.pdf Schematic)

Best Regards, Thorsten.