Print

Please login or register.
1 Hour 1 Day 1 Week 1 Month Forever
Login with username, password and session length

[J M Fahey]

It's not a dummy load but part of a "Zobel" stabilizing network.
8.2 ohms will be fine.
Also 1W isn't needed; 1/2 (and even 1/4W) will be fine.
And no, you can't use a speaker there, it needs to be pure resistive.

[rocklander]

ha ha.. so result correct, but assumption wrong... thanks very much 

[rocklander]

alright.. so pretty much a copy and paste from FSB (am a member there too) but thought it may be useful here, and also someone here may be able to help too..

I built this using the original (reversed) file, so turned my ICs inside out ( ) and got nothing, but the lm386 got warm

so I scored some more lm386 and started again with my own layout... built it and got nothing.. no sound at all..

I made the preamp and power amp modular.. easier to troubleshoot, and if I want I can swap the power amp out for something bigger maybe

the lm386 in the power amp are both getting pretty hot.. not melting yet, but I've not left it on very long cos I think they will if I leave it too long..

so, I disconnected the power amp and put the preamp into my ruby. I got a farty sound and only on full gain..
audio trace got it to the first opamp with good signal as far as pin6
as of pin 7 it was the awful farty sound.. kind of like a badly biased fet.. so I pulled out the LEDs and boom.. works awesome..
so a) do I need the LEDs in there (remembering I have zero theory knowledge so if they're vital, then .. um.. :blackeye )
and b) why would the LEDs cause an issue?
I noted that whilst they're both 3mm red LED, one is physically slightly larger than the other (maybe 0.5mm? but still nowhere near a 5mm LED).

thanks in advance

oh.. and layout and PCB are attached if anyone wants it, but bear in mind the power amp is as yet unverified and the preamp sounds like poop with the LEDs I had in there

[phatt]

Hello Rocklander,

Note on Leds,
They clamp the the AC swing (the sound you hear) so the amp will be louder and clearer without them.
Maybe this is confusing the outcome?

Hold up the LEDS to strong light and you can see the two ends inside.

Note the small bit is positive and the bigger bit is Negative.

These need to be mounted *opposite or reversed to each other* otherwise it will only clamp one side of the AC swing.

It's debatable as to whether such a small Amp even needs diode clippers or not as being so small in the wattage department it will be easy to get distortion without diodes at a low volume.

As to the heating of the Power stage you might have messed up the PCB,,, which is easy to do.

As a wise teck once said;
Turn the schematic up side down and start again  LOL
This will (by design) force your brain to look at at it all from a different perspective and you may find you made a mistake.

Having done exactly that I've learned that it is indeed true. Winky.

Phil.

[rocklander]

thanks Phil.. pretty sure I oriented the LEDs correctly as the both light up, and I also tested them with a 9V battery (and a 3K9 resistor) to get the polarity correct.


the power section seems pretty simple, but you're right in that I may well have goofed.. think I'll grab the datasheet and ensure I've got things within spec.

[phatt]

A quick glance at the poweramp section tells me you have Chip A and B reversed.
But I might be mistaken?
Whether that will mess up the result I don't know? Better minds will tell more I'm sure
Phil.

[rocklander]

hmm.. not sure, but should that matter? they're both the same chip.
and since the 47n and 10r are in series, does it matter which chip each goes to?

[J M Fahey]

Both amps are "reversed" so they end up being out of phase.
Notice that the speaker does not have one end grounded but goes from output to output.
It ends up being a bridged amplifier.
The implementation is horrible, but happens to work on LM386 (I guess nowhere else)