LM380 guitar amp has a pulsating sound at high volume level only, how can I fix?

[Jasonmatthew911]

Here is exactly what I have  in the supply right now, starting from where the power goes into the pre-amp:

... then goes into 4580 IC and continues to power section, starting with .1uf cap to GND, 10 ohm resistor w/ ferrite in parallel, 2200uf electrolytic to GND, then goes directly into LM380....

That is backwards! The power supply should always go to the current-hungry power stage first, then progress on to the preamp. This is almost certainly the source of your problems.

Ok, I tried the 470 ohm resistor on the power supply rail between the pre-amp stage and going into the power amp stage, the other leg of the resistor going into the (+) of the 10uf cap, and the (-) to GND...

You connected a 470 ohm between V+ and ground?? Why would you do that?? PRR's instruction was pretty clear; replace the 10R with 470R. But, per my comment above, I think you need to reverse the other of your PSU first.

No I had done what PRR told me of replacing the 10 ohm with 470, but that didn't fix it...This 470 resistor and 10 uf between V+ and GND was the LPF I was trying out that was recommended by Gurner, based on the 4 transistor amp link he posted, but I probably didn't do it right anyways....I keep thinking that you're probably right about feeding the power stage first so it gets all the power consumption it needs and then feeds the 4580 which doesn't need much at all...I hope this is it, thanks for bringing this up...I'll let you know what happens, if nothing, I'll post schematic.

[Jasonmatthew911]

Here is exactly what I have  in the supply right now, starting from where the power goes into the pre-amp:

... then goes into 4580 IC and continues to power section, starting with .1uf cap to GND, 10 ohm resistor w/ ferrite in parallel, 2200uf electrolytic to GND, then goes directly into LM380....

That is backwards! The power supply should always go to the current-hungry power stage first, then progress on to the preamp. This is almost certainly the source of your problems.

Ok, I tried the 470 ohm resistor on the power supply rail between the pre-amp stage and going into the power amp stage, the other leg of the resistor going into the (+) of the 10uf cap, and the (-) to GND...

You connected a 470 ohm between V+ and ground?? Why would you do that?? PRR's instruction was pretty clear; replace the 10R with 470R. But, per my comment above, I think you need to reverse the other of your PSU first.

Ok I tried feeding the power stage first, but unfortunately that wasn't it either....I'll try and get the schematic up as soon as I draw a clear understandable version of the one I have...Someone mentioned that I should try star Grounding, but is that really necessary with this little amp?....I am using 2 stripboards for this and the GND's all end up connected to the 6 center pins of the LM380, where the heat sink is also soldered...So if I was to star GND my project it would have to be sending my GND rows on the stripboards with cables grounding all to a star GND point on my chasis???...I'm not completely understanding how it would make a difference, or I would probably have to start all over and re-design my stripboard layout...My jacks are switchcraft, which I can't really isolate from the Chasis either....Can you tell me what I would have to add to my V+ to maybe lower the supply voltage, so that a lower voltage and lower consumption is used by the LM380 and 4580? 

[Jasonmatthew911]

Ok, below are my schematics for my LM380 guitar amp with 4580 Pre-amp section...After trying a few things, I'm still getting motorboating when my Volume and Gain levels are turned all the way up...If anyone notices anything weird or out of the ordinary that might be causing the problem in my schematics, please let me know...Hopefully you DIY experts can help me out better by checking the schematics below...Thanks again...

[Jasonmatthew911]

Is it alright that I used the non-inverting inputs on both the 4580 and LM380 in this case?....Can anyone tell me if the 10K resistor going to GND on pin 6 of  the LM380 shouldn't be there?

[merlinb]

Is that an inductor in the power supply, in parallel with a 10R resistor? Would I be right in thinking you have drawn the PSU from right to left? (So the "9-18V" is where the power comes in?)

[Jasonmatthew911]

Is that an inductor in the power supply, in parallel with a 10R resistor? Would I be right in thinking you have drawn the PSU from right to left? (So the "9-18V" is where the power comes in?)

It's actually a small ferrite bead...Yes, the power comes in where it says 9-18V...

[merlinb]

I'm still a bit confused by your schematic. It shows pin 14 of the LM380 being connected to "+18V", which is the raw power supply with no extra filtering?
It shows pin 8 of the 4580 being connected to "+9-18V", which is also where the power comes in, no filtering?
So the only stage getting filtered power is the FET which goes to "+V"?

[Jasonmatthew911]

I'm still a bit confused by your schematic. It shows pin 14 of the LM380 being connected to "+18V", which is the raw power supply with no extra filtering?
It shows pin 8 of the 4580 being connected to "+9-18V", which is also where the power comes in, no filtering?
So the only stage getting filtered power is the FET which goes to "+V"?

My mistake, they're all going to +V and through the filtering components...I just tend to draw my schematics that way, sorry for the confusion...

[thedefog]

Ok I tried feeding the power stage first, but unfortunately that wasn't it either....I'll try and get the schematic up as soon as I draw a clear understandable version of the one I have...Someone mentioned that I should try star Grounding, but is that really necessary with this little amp?....I am using 2 stripboards for this and the GND's all end up connected to the 6 center pins of the LM380, where the heat sink is also soldered...So if I was to star GND my project it would have to be sending my GND rows on the stripboards with cables grounding all to a star GND point on my chasis???...I'm not completely understanding how it would make a difference, or I would probably have to start all over and re-design my stripboard layout...My jacks are switchcraft, which I can't really isolate from the Chasis either....Can you tell me what I would have to add to my V+ to maybe lower the supply voltage, so that a lower voltage and lower consumption is used by the LM380 and 4580? 

Your grounding sounds fine the way you described it, I was just asking in case you had a loop somewhere and to see what your shielding looked like. Parasitic oscillations are a pain to debug.

[blackcorvo]

Try adding a 4k7 resistor connected from LM380's pin 2 to ground.
These "pc speaker" chips usually have this resistor there, to keep their input impedace low, so they can match propetly with the low impedance from the sound output of pcs.

Without that resistor, the chip's input "floats", and that's what makes it "motorboat".

[merlinb]

Try adding a 4k7 resistor connected from LM380's pin 2 to ground.
Without that resistor, the chip's input "floats", and that's what makes it "motorboat".

The chip has 100k pull down resistors "inside", so it doesn't float, but it will be interesting to see if it makes any difference.


Jason, did you actually mention *what* speaker you're driving with this circuit? You said your PSU is rated for 1A, but maybe you're trying to suck more than that? For example, 5W into 8ohms would be over 1A of peak current...

[Gurner]

We know the problem...ie when the PSU to the pre & power are independent - it's cool, when they're joined it motorboats.

Motorboating is +ve feedback from the output stage's large current draw, getting into the rail by way of extra ripple...which in turn bleeds into the preamp stage, which in turn amplifies the error, which in turn gets amplified by the output stage, which in turn causes more current draw & ripple on the rail - rinse repeat.

you have a preamp there with potentially 500x worth of gain (which is a ludicrous amount ... just 1mV of output stage induced supply ripple getting into your input stage  = 500mV at the output of the preamp...this 500MV will then be amplified by the output stage)

My plan of attack would be...

Reduce the gain  - ie take the value of that 500k pot down to say 100k (or conversely increase the value of that 1k resistor connected to pin 2 to say 5k)

Place a very large cap on pin 14 (as close as you can get it) to ground - I'm thinking here  the largest cap you have (you can reduce it's value later)

Place another large cap to ground as close to pin 8 as you can (again, the largest you can get your hands on - you can always take it down in value once you've proved the point)

slap a resistor ....something like 1k (you can always reduce it later) feeding the supply to pin 8 of the preamp IC (this in conjunction with the cap on pin 8 is effectively a low pass filter and helps get rid of any power amp induced ripple)

Consider placing an electrolytic cap at the junction of the two 10k resistors & 1M resistor to ground (on your preamp - I do wish you'd numbered the components...it makes giving suggestions so much easier!)

As an side ...that 2,200uf to your speaker looks excessive ...that allows frequencies down to 9Hz to your speaker (and might be resulting in some low frequency issues on your rail) ....you might want to limit that - a 330uf cap will take this threshold up to about 60Hz (the lowest frequency on a std guitar is 82.4Hz therefore there'll be no impact - a 220uf might be even better but the threshold then is 90hz, so might rob you of a ittle bit of bass on your open low E)

If you wish to nail it completey....put a regulator (say a 78L12) between your 18V and your preamp power pin (+ the additional supporting caps the regulator requires)

[merlinb]

We know the problem...ie when the PSU to the pre & power are independent - it's cool, when they're joined it motorboats.

In a previous post Jason said

What about the power supply going to both stages at the same time, in parallel?...Cuz I had tried them in parallel and the motorboating persists

So it doesn't sound like a pure gain problem to me...

[Gurner]

We know the problem...ie when the PSU to the pre & power are independent - it's cool, when they're joined it motorboats.

In a previous post Jason said

What about the power supply going to both stages at the same time, in parallel?...Cuz I had tried them in parallel and the motorboating persists

So it doesn't sound like a pure gain problem to me...

i think that could just be a terminology thing, a pay more heed to this...

...I also know that my preamp section and power section can work perfectly together because I had already tried them together but feeding off of separate power supplies, but both @ 18V....Now I put them together to feed off of 1 power supply, so the issue is in the power section for sure....Can anyone help me?...Let me know if you need more info about the circuit.

If I'm reading it right, when they're independent...no motorboating, when the pre & power source from the same rail...motorboating.

[merlinb]

...I also know that my preamp section and power section can work perfectly together because I had already tried them together but feeding off of separate power supplies, but both @ 18V....Now I put them together to feed off of 1 power supply, so the issue is in the power section for sure....Can anyone help me?...Let me know if you need more info about the circuit.

If I'm reading it right, when they're independent...no motorboating, when the pre & power source from the same rail...motorboating.

Ah I missed that. Sounds like his 18V 1A PSU doesn't have a big enough reservoir to do both jobs.

[Jasonmatthew911]

I will try what Gurner said as well, but...
Do none of you think it has anything to do with my Grounding connections, as R.G. wrote in my other thread of "How to isolate pre-amp & power amp stages on same supply rail?" about "Star Grounding"...Originally I didn't think that grounding would be an issue with a low power amp, but maybe I was wrong since I have 2 stages running on 18V....

You think 1A isn't enough?....It does say that the LM380 has a peak current of 1.3A, plus all the other current needed, I might be needing more???

[blackcorvo]

Try adding a 4k7 resistor connected from LM380's pin 2 to ground.
Without that resistor, the chip's input "floats", and that's what makes it "motorboat".

The chip has 100k pull down resistors "inside", so it doesn't float, but it will be interesting to see if it makes any difference.


Jason, did you actually mention *what* speaker you're driving with this circuit? You said your PSU is rated for 1A, but maybe you're trying to suck more than that? For example, 5W into 8ohms would be over 1A of peak current...

I had this same problem before, and my power supply was a 12v, 7A sealed battery. The only way I found to stop the motorboating is to add this 4.7k resistor there. I remember seeing this resistor being used in a pc speaker system using an 8-pin tda2282 stereo chip. both inputs had this resistor there.

[PRR]

[PRR]

> Do none of you think it has anything to do with my Grounding connections

Nope.

[Jasonmatthew911]

FINALLY, Problem solved...I ended up doing a change Blackcorvo and Gurner mentioned...I added the 4.7K resistor from pin 2 of LM380 to GND and lowered the 2,200uf speaker cap to 560uf, cuz thats what I had, and one of these 2 things or both eliminated the motorboating...I'm starting to think that it was probably more the changing of the big 2,200uf cap at the output, it sounds good now, but even still feels like it has a little too much low end maybe, so I will try lowering  it again to the recommended 330uf instead, just cuz with the 560uf cap on the output at full volume and full gain it adds a minor low note oscillation when the guitar isn't being played, but it's barely noticeable...So I figured that bringing that cap down to the recommended 330uf will probably get rid of that and some low end...

I may also try what PRR pointed out, of adding the 10uf electrolytic where the (2) 10K resistors are at the input of my 4580, just to see if this helps improve my signal as well....Unless PRR doesn't think this will make much of a difference anymore....If you think it will still help contribute to a cleaner signal, then I will do it for sure...Let me know PRR, and thanks for helping...

I'm surprised at all the noise and motorboating that went away by just doing this minor change(s)...Special thanks to all of you, especially Gurner and Blackcorvo for putting me on the right track....

I forgot to mention that I also ended up switching the LM380 to an LM384 when I did these 2 changes, but I doubt that made any difference, I just thought it would handle higher voltage better, although LM380 takes up to 22V....I never took into consideration that the 2,200uf cap at the output could add so much low end that would cause it to motorboat...Luckily I didn't have to do all the other stuff Gurner mentioned, it was enough with changing the big output cap and adding the resistor in the input of the LM384...

PRR, let me know if you think I should still add that 10uf cap?

Thanks again guys...