noisy cricket problems

compuwade · January 12, 2010, 09:58:37 PM

Quote from: dano12 on January 12, 2010, 06:54:59 PM
Quote from: compuwade on January 12, 2010, 03:47:12 AM
I know this is going to sound very very strange. However, the Noisy cricket was my very first project and I had the exact same problems with it that you are having using the radio shack board. I tried many different IC's, different caps, resistors and even rebuilt it on a new RS board. All resulting with the same issues. When troubleshooting it for some reason I decided to connect the 220uf cap across pins 3 and 6 of the IC and this fixed the problem. I even emailed Dano from Beavis Audio and he said that it should do absolutly nothing. But it did fix the circuit. I was wondering, if you haven't figured out the problem, could you humor me and connect a 220uf cap across pins 3 and 6 and see if that clears up the problem. As I said, I know it sounds strange and shouldn't do anything but I'm curious as to why mine worked.
Thanks,
Wade

I remember your email!

What you suggested, in bridging pins 3 and 6, was effectively connecting connecting another 220uf cap from V+ to ground.

Looking at the schematic, I was pretty sure that would not fix anything:

But since this worked for you it means that you have added an additional 220uf cap to ground for power supply filtering.

This is a particularly vexing issue as I've not seen LM386 circuits ever calling for more than 100uf between V+ and ground. If I recall correctly, both the little gem and ruby from runoffgroove call for 100uf caps.

If you don't mind, a few questions: Did you build yours on a radioshack board as per my project pdf? Also, did you use a wall wart or battery?

Thanks for any help you can provide.

Hey there! I did use the instruction per the PDF you have on your site. I could never get the project to work without the 220uf across pins 3 and 6. I used a wallwart for that project. However, later on I used your layout PCB for the project and it worked with no issues at all. I no longer have the original project I built, but after looking at the PDF again I realize exactly what I was doing with the cap. I never tried using a small cap to see if I could accomplish the same thing. That's actually pretty interesting now that I think of it. It may have worked with a 1uf cap instead.

-Wade

El Heisenberg · January 13, 2010, 03:08:15 PM

When you are putting a cap across pin 3 and 6 you are basically just putting the cap from pin 6 to ground. This shouldnt do anything.

Ive built a ton of 386 amps and sometimes they have weird problems, and im not surprised that something totally weird fixed it. These things are so frustrating!

puretube · January 13, 2010, 04:58:55 PM

Physical layout/wiring...

markrocknroll · January 13, 2010, 09:25:58 PM

Quote from: compuwade on January 12, 2010, 03:47:12 AM

I know this is going to sound very very strange. However, the Noisy cricket was my very first project and I had the exact same problems with it that you are having using the radio shack board. I tried many different IC's, different caps, resistors and even rebuilt it on a new RS board. All resulting with the same issues. When troubleshooting it for some reason I decided to connect the 220uf cap across pins 3 and 6 of the IC and this fixed the problem. I even emailed Dano from Beavis Audio and he said that it should do absolutly nothing. But it did fix the circuit. I was wondering, if you haven't figured out the problem, could you humor me and connect a 220uf cap across pins 3 and 6 and see if that clears up the problem. As I said, I know it sounds strange and shouldn't do anything but I'm curious as to why mine worked.

Thanks,
Wade

Well I got a 220uf cap and went to work tonight. First powered it up to verify it was still working as before and couldn't get a sound out of it. Checked out the guitar cable, nothing. Changed the IC, nothing. Retouched all solder joints and got it working again (with the same problems). Must have giggled something loose last night taking voltage measurement. Noticed that I had wired the grit switch power straight to the battery. Reran it to the board and tested it. Still working with same problems. Then I soldered the 220uf cap across pins 3 and 6. It was if little fairies sprinkled f'n pixie dust on the amp; it started working. No sputtering with the tone pot above off. No noise and distortion. No hum/howl with the grit switch off. Everything worked like it should. I can get clean sounds and great distortion. Grit switch gives it more bite. Sounds great on battery or wallwart. THANK YOU Wade!

compuwade · January 13, 2010, 09:49:07 PM

WOW!! I'm glad that worked for you....However, thanking me is really not warrented since this is more like a bandaid than a real fix. Because we still don't know what the heck the deal is and why that's working. Of course sometimes it's best not to ask why and just accept the fact.

Anyway...for what it's worth...you're welcome.

-Wade

markrocknroll · January 13, 2010, 10:04:29 PM

Wade

You said a couple posts up that after relooking at the PDF you realized what you were doing with the cap. What were you doing? Why did you try it?

Boprikov · January 13, 2010, 11:01:59 PM

Hello! I think it may be important to get this additional 220uF cap assembled physically near the chip. Unwanted oscillation occurs easier when there is even the slightest resistance between power supply filter cap(s) and the chip.
Power supply filtering is important even with battery power. Maybe it is the physical layout of the circuit, that makes some 386 amps easier to get working. Just as Puretube said...

compuwade · January 13, 2010, 11:15:06 PM

Quote from: markrocknroll on January 13, 2010, 10:04:29 PM
Wade

You said a couple posts up that after relooking at the PDF you realized what you were doing with the cap. What were you doing? Why did you try it?

LOL, well I didn't say I knew what I was doing at the time, just that I know kinda what it does now after looking at it. Basically, besides filtering any noise created by AC in the power supply, it also charges the cap with power giving the IC a little extra power boost. Having the cap very close to the power input of the IC ensures that the IC gets the required amount of power without any voltage drop created by wires. At least this is how I understand it. Someone else could correct me if I'm wrong or explain it a little better I'm sure.
Why I tried it....Well, thats kinda beyond me. I think I tried it because the original cap ( C1 on the schematic ) seemed a little suspect to me after looking at it, so I simply just added another cap between the two pins because it was easier than desoldering C1. But when I eventually replaced C1 with a new cap, it didn't work. So I just added the cap between 3 and 6, boxed it up, and forgot about it. So other than the IC requiring a little more power than the layout provides, I can't see any other reason why it would work.

El Heisenberg · January 13, 2010, 11:42:06 PM

this is cool.

compuwade · January 14, 2010, 02:28:03 AM

Quote from: El Heisenberg on January 13, 2010, 11:42:06 PM
this is cool.

Isn't it though

markrocknroll · January 14, 2010, 09:59:57 AM

Since I built it in a cigar box and not a stomp box enclosure the wires were longer. Also, I installed C1 on the AC adapter jack which I had some distance from the board (put the input jack next to the board). These factors caused noise, oscillation, voltage drop and required the cap close to the IC input to boost the power and therefore work properly? Is this how it works? If that's the case, since I was building a cigar box amp, I don't think I could have ever gotten this thing to work without this suggestion and I would have built some other amp in it.

compuwade · January 14, 2010, 12:22:45 PM

Quote from: markrocknroll on January 14, 2010, 09:59:57 AM
Since I built it in a cigar box and not a stomp box enclosure the wires were longer. Also, I installed C1 on the AC adapter jack which I had some distance from the board (put the input jack next to the board). These factors caused noise, oscillation, voltage drop and required the cap close to the IC input to boost the power and therefore work properly? Is this how it works? If that's the case, since I was building a cigar box amp, I don't think I could have ever gotten this thing to work without this suggestion and I would have built some other amp in it.

Well, thats how I understand it. After building a few chip amps, I learned that having the power input filter close to the IC helps prevent voltage sag. The futher away the filter is, the larger the posibility for voltage sag at higher volumes. In many LM3886 designs I've seen the filter soldered right to the chip. Probably not nessesary with a good PCB layout, but doesn't hurt one bit.

puretube · January 14, 2010, 01:48:33 PM

Oscillations...
(parasitic)

Brymus · January 14, 2010, 04:04:50 PM

I think it even says on the data sheet ,that the filter cap needs to be X amount of millimeters from the IC pin or closer for it to function properly.
Or maybe that was a different IC data sheet...
But it apparently applies here too.

El Heisenberg · January 14, 2010, 04:13:03 PM

I think beavis's diagram shows the filter cap connected directly to the dc jack. Maybe an update is in order

markrocknroll · January 14, 2010, 05:29:54 PM

Quote from: El Heisenberg on January 14, 2010, 04:13:03 PM
I think beavis's diagram shows the filter cap connected directly to the dc jack. Maybe an update is in order

It does and that's why it was there. I really like the amp now. Sounds great and was a blast to build. I'm thankful to Dano for his design. Troubleshooting is all part of the fun for me.

El Heisenberg · January 14, 2010, 09:34:01 PM

Next try a tda2003/2 design. Or you could try bridging two lm386s with all the controls from the noisy cricket. I used a dual gang pot for the gain control. I added a little bass cut control after the FET in my amp, so i had treble and bass controls. It was just a .0022 cap and a 1meg pot.

the bridged lm386 design in the datasheet is also really good for a lunchbox iphone speaker system.

compuwade · January 14, 2010, 10:07:25 PM

Quote from: El Heisenberg on January 14, 2010, 09:34:01 PM
Next try a tda2003/2 design. Or you could try bridging two lm386s with all the controls from the noisy cricket. I used a dual gang pot for the gain control. I added a little bass cut control after the FET in my amp, so i had treble and bass controls. It was just a .0022 cap and a 1meg pot.

the bridged lm386 design in the datasheet is also really good for a lunchbox iphone speaker system.

Or if you want to build your own custom amp you can do what I did and go pickup a nice 100w a side car stereo at the thrift store and pull the amps out of it. You might have to search for the datasheet, but if you find yourself a popular brand it'll probably have a couple TDAxxxx's in it. Which datasheets are easy to find for. Then you can build a stompbox of your choice and use it as a preamp and end up with a really nice sounding amp.

tuckster · January 17, 2010, 07:47:55 AM

Maybe someone can give me a hint regarding my cricket problem:
I used the pcb version. It works. But only the volume pot "works".
I do hear my guitar signal and I'm able to change the volume but the gain and tone pot doesn't make any difference to the sound.
The sound is very muffled, to much bass in it.
The grit switch works like a bright switch, no grit but a brighter sound.
I want to use the cricket as a headphone amp. The jack is soldered like the one on the last picture but without a speaker http://www.generalguitargadgets.com/pdf/ggg_ruby.pdf?phpMyAdmin=4a28f86a515b7883e7bc35a68d4e7b6d
With the audioprobe I can hear a tone on the pot pins.

dano12 · January 17, 2010, 04:39:19 PM

Quote from: El Heisenberg on January 14, 2010, 09:34:01 PM
Next try a tda2003/2 design. Or you could try bridging two lm386s with all the controls from the noisy cricket. I used a dual gang pot for the gain control. I added a little bass cut control after the FET in my amp, so i had treble and bass controls. It was just a .0022 cap and a 1meg pot.

the bridged lm386 design in the datasheet is also really good for a lunchbox iphone speaker system.

Hi Ed, one of the things that has kept me from moving the Noisy Cricket to one of the TDA designs is supply voltage. The idea of running of a single 9 volt battery simplifies the design and lots of people who build them like that convenience. I think the next Noisy Cricket version (Mark IV) will be a marginally improved 386 design.

However, it would be interesting to take some of the TDA1013 designs I've tried and work them up into a "Power Cricket." Minimal supply voltage is 10-12 volts, so the 9 volt battery would have to be replaced by a stack of AAs or somesuch. I'll start breadboard this tonight and see what I can get out of it.

News:

noisy cricket problems