Ground and V- on an LM3886 opamp

[antknee]

Hi,

I am trying to get an LM3886 opamp working on a breadboard, I thought I understood the pins but realised I don't! It isn't working, I'm having a problem understanding the difference between ground and V-, so I don't know where the pins are supposed to go.

I've attached the datasheet. Essentially I want to build the circuit in figure 1 on the first page.

I'll be using a signal generator as the input. Vin+ is pin 10, Vin- is pin 9. So i connect the red lead of the singal generator to the Vin+ input, it goes through a variable resistor and 1Kohm resistor to get there. The black lead of the signal generator goes to where exactly? Ground or V-? I'm sure they are different as |Ground| - |V-| is meant to be over 9V from the datasheet notes.

Thanks,

Antknee.

 

[Nigel Goodwin]

It's NOT an opamp, it's an audio power amplifier, use the circuit out of the application sheet.

Layout for these kinds of chips is usually critical, a breadboard is likely not to work - usually they have a recommened PCB layout in the datasheet, which should be used.

 

[Hero999]

It requires a bipolar power supply: -V and 0V are not the same thing. If you must use a single supple then you should double the total voltage and connect all parts requiring a DC connection to 0V to half the supply voltage.

As Nigel said, don't use bread board and build the circuit on the datasheet.

You might be able to get away with stripboard if you're careful but certainly not breadboard.

 

[antknee]

Hi Nigel / Hero,

Ah yes it is an audio power amp and not an opamp. I will check the datasheet for what it says about PCB layout. I started out soldering it to a stripboard but having never used one before found it hard to work with, so I took it off again.

What is a bipolar power supply? I was going to use a single 24V DC supply, I do have some other power adapters, a 12V DC for example, so I could use that also. I can follow the schematic from a component point of view and I have them all, I'm struggling to know how I power the amp and which pins.

I'll essentially be doing this: Signal generator ----> LM3886 ----> Audio speaker

Regards,

Antknee.

 

[Hero999]

What is a bipolar power supply?

A spilt supply with a positive raile, 0V in the middle and -V rail.

It can consist of two batteries connected in series with the middle taped to form 0V or a centre tapped transformer connected a rectifier with capacitors connecting to the tap.

I was going to use a single 24V DC supply, I do have some other power adapters, a 12V DC for example, so I could use that also. I can follow the schematic from a component point of view and I have them all, I'm struggling to know how I power the amp and which pins.

That will not work, you need a 24V bipolar supply or a 48V single supply to power this amplifier.

 

[antknee]

Hi Hero,

I see. A bipolar supply isn't something I've ever thought about, I'm a novice to electronics, it makes sense though. I dont have a bipolar supply so I will use batteries for now, i'm really just testing for the moment so there is no need to buy a specific supply.

So how do I tap 6 batteries (each at 9V) ?

I'm thinking

+ - + - + - Tap - + - + -+

Pin 9 (Vin-) looks like it is also connected to the same earth. Does that really go to the bipolar source 0v? it seems an unlikely place, but I don't really know...

Regards,

Antknee.

 

[Hero999]

No, it goes to -V, the ground/earth symbol goes to 0V and the +V goes to +V.

**broken link removed**
**broken link removed**

Unfortunately this amplifier won't run off a 9V battery - the voltage and current requirements are too high.

 

[antknee]

Thanks for the diagrams they were very helpful and I have got it working.

The signal changes tone as expected but still could sound better, I might have missed something. I'll check later.

The next task is to remove the speaker and attach a transformer, to this I'll attach a piezo and see if I can drive it. The transformer is 3.2ohms to 1.2Kohms which is impedance matched, more or less, I think. It is only 0.2W though, so under powered and I'm not sure if it will like the frequencies I wish it to work at - 50KHz to 250KHz.

Regards,

Antknee.

 

[antknee]

One thing that has confused me is that I'm only using one lead from the signal generator, going to Vin+. If I connect the other lead of the signal generator to ground or V- the tone from the speaker disappears, I wasn't expecting that.

 

[Nigel Goodwin]

Both the single and dual supply circuits show a volume control Rin, the signal connects to the top, and the screening connects to the bottom (ground).

I'm somewhat dubious about your hoped frequency response though?.

 

[Hero999]

The other signal generator lead should be connected to 0V, not -V.

Do you have the 0V rail connected to safety earth/ground?

If so, it sounds like you're shorting the negative rail!

It sounds like you're doing this the wrong way. You should use a lower powered amplifier and a transformer with a primary impedance of 8R if you don't want much power.

 

[antknee]

Hi Nigel,

Ah yes, that is it! Why didn't I think of that!? Thanks.

The LM3886 was chosen because it has a bandwidth gain product of 2MHz minimum, works at 8MHz. The gain I'll be looking for is 5 (or 10 at most) so the chip won't be the problem. The transformer probably... I chose it primarily because it is impedance matched to the chip, 1Kohm input and 4 or 8 ohm output. It would be hard to find the right transformer, I'd have to build it myself and I'm not keen, nor do I know how exactly, so i'm hopeful but not optimistic with this transformer.

 

[Hero999]

I think you've got that backwards, you need an 8R primary and a 1k secondary.

You can get pre-made transformers with a 1k primary and 8R secondary which could be reversed but I doubt they'll work at 50kHz.

50kHz to 250kHz sounds like very wide bandwidth for an ultrasonic transducer, are you sure it's up to it?

This isn't the normal way of doing it, you either use a squarewave which is converted to a sine by a transducer with a sharp resonant peak or use a class C amplfier to drive the transducer. The trouble with this, is it doesn't work at such a wide bandwidth.

 

[antknee]

Hi Hero,

I don't have the O V rail connected to anything else no...shorting the negative rail...hmm... I can but smile! Must check the batteries haven't drained!

The problem I have is that I'm going to build the piezo myself, this means I need circuitry which is flexible with regard to power consumption, voltage and frequency. Ideally I'd like to get away with using an opamp, but the current output from those is very low. To this end I have some opamps to have a look at and get working too. I wouldn't argue with you that a solution between these, i.e. a low power audio amp is the right way forward. I'd need one that had enough bandwidth though.

I was under the impression I could simply switch the primary and secondary transformer leads, I have 1.2K primary and 3.2 secondary. Does that change the frequency at which they are happy? I have a feeling the transformer I have won't be happy at the frequencies I wish. We'll see. At that point I start work on a) buying in the transformer parts from coilcraft and b) start wiring up the opamps for a look.

Regards,

Antknee.

Edit: Batteries are ok. Phew.

 

[Hero999]

How are you going to build your own piezo transducer?

You need to buy a block of piezoelectric material, then you have to cut it into a suitable sized wafer and mount it to a suitable diaphragm. To be honest I wouldn't know where to start but do you? If so I'd love to know how you're going to make it. Building the transducer is more of a project than the circuit.

 

[antknee]

I have built the circuit in figure 1 of the datasheet and it does work, in that I get a signal across the speaker more or less the signal I have put in from the signal generator. The waveform looks a bit ugly tho, rather than a sine wave it is more square.

Oddly the variable resistor doesn't change the volume, that is up high no matter what. If I connect the earth of the variable resistor to the earth rail I get no signal. Also if I connect the negative lead of the signal generator to earth rail the signal also disappears. So i don't have amplification at all at the moment. The variable resistor may be faulty, I did have a few sparks off it!

I would like to buy the equipment to machine pzt myself but it wouldn't be practical. So I will buy it in, this means using an industrial company, there is usually a minimum order value of around £300 ($400), for that you'll get around 100 pieces. Same again with the metal you'll be mounting the pzt on, although it'll be cheaper. Glue the two together with a conducting epoxy resin. Test what you have and damp down the frequency with weights if necessary, based upon those experiments reorder pzt/mounting plates. Re-iterate until you have what you want. Rapid prototype the seal/gasket and any other parts. I'll have to buy in bulk and accept the waste. This takes months of spare time and a fair bit of cash. So yes building the transducer is more of a project than the circuitry, but one won't work without the other.