Op-Amp issues on breadboard

[tpaairman]

Branching off my thread about parallel MOSFETS, which I've decided to just go with a single set (psuh-pull), I'm running into some issues trying to design this thing.

First though, let me apologize for being a pain with this one. I know that I can go buy a kit and just do it that way, but i'm trying to learn the design side of things with this, and if I can get it to actually work by Christmas, I can build it for my Son for his bass guitar.

Below is my schematic 9sketched out). Overall, I got a tone through form my function generator, things weren't quite right, and I'm getting a parasitic oscillation. A nice sine wave that looks to be at 20KHz, according to my oscilloscope. My power supply has + and - voltage, and I have it set to 10 volts +/-. When I vary the voltage, the frequency shifts slightly. I pulled the MOSFETs from the circuit, so it's just the op-amp (and the feedback resistor moved to the op-amp output instead of the MOSFETs), and no change. In the mean time, I set the gain of the op-amp to 100, and figured let's try connecting my guitar to it, but no sound comes through. But if I connect my function generator (and have it set almost as low as it goes) I get a tone, thought it's more of a buzz, but I figure at that gain, it's clipping, so no surprise there.

So my two questions are why am I not getting anything when the guitar is connected? And second, while I understand it's on a breadboard which can lead to stray crap, nothing changes with the oscillation when I tap any of the parts. You'd think some movement would have an effect.

[tpaairman]

One other quick note, I know that for the final amp, I will need a pre-amp stage, and there will be a tone control. Right now, I'm just trying to get this part up and running, then go from there.

[nickb]

It important for you to realize that there is likely to be quite a gap between the the circuit given, which is just a textbook functional illustration, and one that is practical, i.e. is electrically and thermally stable , doesn't overheat and doesn't produce unacceptable distortions. The circuit you have there will likely suffer from all of these perils to some degree or another.

So, since you are interested in how to make it work let me run through the problem areas that I see, there may well be more.

#1. You have a known stability issue, even without the MOSFETs. My first instinct is to add 100nf decoupling capacitors as close a possible from each power pin of the opamp to 0V. I suggest you set the gain to 10 to give you about 1.2Vpk signal in for full output. The opamp will not have enough current drive capability to drive a speaker, you'll get about 0.0001W out.

#2. The bias network ( the LEDS ) is going to be a problem. Either the voltage across the LEDS to going to be so high that the MOSFETS bake or so low that you'll have crossover distortion. The high gain of the opamp will help to hide this but not solve it. To make matters worse they don't compensate for the negative temperature co-efficients of the MOSFET ( yes, negative, they only go positive at high currents, excepting lateral MOSFETS). Therefore you need a NTC thermistor thermally coupled to the FETS along with properly designed bias circuit to make it thermally stable. To make it even worse, MOSFETs have a huge trans-conductance meaning a small change in Vgs will give a massive change in current. To tame this add a 1 ohm resistor between each source and the output (again, not such a problem with lateral MOSFETS).

#3. RF stability. Add a 1k resistor between the gate and drive close to the MOSFET.

#4. More stability issues. The opamp will not see a purely resistive load and this can lead to stability issues. Same for the MOSFETS. You may need a zobel network and a small inductor in parallel with a resistor to tame the reactive load that is an actual speaker.

#5. You'll need some bulk capacitance on the drains of the FETS to ground. Suggest 1000uF on each.

#6. There is risk of the inevitable signal that will remain on the power rails getting into the opamp and causing yet more stability issue since it has a finite PSRR (power supply rejection ratio).

#7. You'll need a decent heatsink for the FETs.

#8. After having done all the above there is still a very high chance that the wretched thing will still hoot like a banshee at one frequency or another. You'll have to get into loop stabilization techniques to tame it. See all the little extra caps and res's in real designs? That's why they are there.

#9. There is nothing to protect the MOSFET from short on the output

#10. There is nothing to protect the speaker from a MOSFET failure.


The above are some of the reasons that you don't see the circuit you show in real amps. You might get it to work but the performance will be poor.

[tpaairman]

I knew what I had was a starting point but dang, that’s a list. I’ll see how far I can get with it though.

[J M Fahey]

I suggest you *start* with a basic but complete amplifier, build it successfully, enjoy, experiment, and afterwards go on with more complex ones.

Start with a chipamp, TDA2030 and LM1875 are still in production, include an Op Amp front end, and real world power transistors, in fact they can be considered a Power Op Amp and follow same basic rules, within certain limits: open loop gain is not *that* high and often can´t be less than 10X to 20X, the reason being they are optimized for typical Audio use.

They are also properly biased, thermally compensated and to boot have a reasonable short circuit protection.
All in a compact 5 pin package and for less than $2.

You´ll still find a good layout, first Protoboard them and then make PCB which will be a necessary skill.

Then build a dedicated supply, etc. ... plus adequately heatsinking the chipamp, kludge some kind of chassis, cabinet, add a Preamp, make a front panel ... you have weeks of fun ahead before going to the next step.

After success and experimenting, maybe we can suggest a "next step" discrete amplifier.

[Dave H]

After success and experimenting, maybe we can suggest a "next step" discrete amplifier.

Like one of yours for example?

[J M Fahey]

Oh, the old workhorse
Some 8000 of them with 2N3055 plus some 4000 in this TIP14x incarnation happily running around.
As simple as can be, very reliable if well heatsinked and *easy* to repair, important when it´s in a small town 2000km away from Buenos Aires and only the local TV repair Tech is available.
And parts are as common as dirt, always locally available ... if he doesn´t already have them in his parts bins.

Not only the schematic is simple, pulling 4 screws removes the full back panel and power amp + supply board are exposed, both sides accessible, for easy testing and replacing.

You can literally repair it in 10 minutes, and I value that 1000 times more than any suppossed improvements fancier amps can offer.

But I was thinking more along Tpaairman desires, I guess after he´s fully explored chipamps in all possible ways, his next step might be either a larger more complex one for higher, live Show level power, maybe LM3886 or TDA7294 based, or a boosted TO220 chipamp, somewhat similar to the first schematic he posted here.

Going step by step is fine in my book and the best learning tool.

[Zozobra]

This a somewhat timely thread as I'm musing trying out some chipamp stuff, probably something with a TDA7294 (or maybe two bridged as I'm an idiot that likes loud things )

[R.G.]

This a somewhat timely thread as I'm musing trying out some chipamp stuff, probably something with a TDA7294 (or maybe two bridged as I'm an idiot that likes loud things )

I highly recommend that you don't go down the TDA7294 path. I've played with that chip and its relatives, and they can eventually be coaxed into working with a whole lot of care and tinkering. On the other hand, the LM3886 is remarkably easy to get running well and reliably.

@tpaairman:
Every one of nickb's comments are correct, and each is a hole deep enough to spend a LOT of time on puttying over. I'd add to them the idea that most beginner amp builders get very ,very deep into the hole of making a good amplifier circuit, and then run into the issue that the power supply and enclosure are incredibly harder to do than the fiddly little circuit of the power amp.

It's good to step back and think about objectives every now and then. What is the objective here? If it's a hand made Christmas present, you will have a tough time getting that circuit running reliably by Christmas unless you're just lucky - there are too many potholes. If the objective is to just d@mbit get that circuit to work, and as a side effect have a Christmas present, it might be well to split those two objectives, get an easier to accomplish Christmas present ready, and go grind out that circuit at your leisure.

[J M Fahey]

I hadn´t noticed the time constraint issue, doubly so it including "Christmas" , "present" and "family" added goals.

Then start now with a kit LM3886, period, so you remove that stumbling block out of your way, (it includes sourcing power supply and transformer), decide within 3 or 4 days tops about chassis and cabinet, you will also have to design and make some kind of front panel graphics (plotted self adhesive vynil works fine, but even that will require a couple days), possibly cutting, assembling, sanding/rounding and Tolexing the cabinet, and then spend a week at least developing a Bass Preamp.
And it "should" be ready at least 1 week before Christmas so you can troubleshoot and debug it.

>>>you are very short of time for your goal<<<<<

It is just "doable" but your path must be ruler straight, any deviation along side roads will simply avoid your reaching it on time.

[tpaairman]

If I don’t make it by Christmas then he gets a tent for camping with the scouts and his birthday is a couple months later.

In the mean time I’ve been looking over the kits and narrowing it down.

And then there’s the amp project sitting on the breadboard. So I got to tinkering with it, figured the easiest place to start was the cap across the power supply rails. There just so happened to be a .01 uF disc cap laying on the bench. I put it between the + and - pins at the chip and it solved the misbehavior. Nice straight line on the scope with no signal. Good trace with a signal.

I have a LM386 for a pre amp. I started with the basic gain of 20 version, with the speaker just connected to it. Hooked up my guitar and it’s low volume (due to the low gain) but it works. I tied it then to the op-amp/MOSFET setup and it’s distorted as hell. In fact, it seems to be the same issue. But I’ve done all I can for the night. When I’m wondering why my scope won’t show a trace, then realize I connected the Chanel two probe instead of the one probe, it’s time for bed. They are clearly color coded.

So I believe I will order the kit, keep working on this, and either way, I should be good to go.

[tpaairman]

One other thing I forgot to mention. I did change the gain of the op-amp to 10. I will try lowering that, maybe even back to unity and see what I get.

[Zozobra]

I highly recommend that you don't go down the TDA7294 path. I've played with that chip and its relatives, and they can eventually be coaxed into working with a whole lot of care and tinkering. On the other hand, the LM3886 is remarkably easy to get running well and reliably.

Ah damn, I was gonna snag a kit off ebay to tinker with Are they a dead cert waste of time?

I was mainly interested in these as they're used in the Orange CR120. They're pretty well regarded and importantly they can deliver a good chunk of power into 8 and 16 ohm loads.

[R.G.]

It is not dead certain that they're a waste of time. They can be made to work, else they would not be used in commercial stuff. But I had a lot of problems with them and in my mind I dropped them into the "temperamental ICs" bin. My preference for the LM3886 is based largely on the fact that my tinkering with the LM3886 was largely trouble free.

The 7294 family can work with much larger power supply voltages than the LM3886. This is attractive for running higher power amps simply. The problem is that you need to run two, three, or four of them in parallel as added output buffers to spread the waste heat out to be able to keep them cool reliably IIRC. There is also a power supply sequencing issue somehow. It's been a while since I worked with them.

The LM3886 also has to use multiple chips to get to high power. On its own, an LM3886 will reliably do 30 to 50W into 8 or 4 ohms. They can be hooked up for higher powers to get up to 250-300W by hooking them in bridge parallel. This is also a pain. However, something that works really well is to make a BIG +/-30V power supply and run one LM3886 per speaker. The speakers are effectively acoustically paralleled, so you get 30W per LM3886, up to as many watts as you need, without any tinkering with the circuit to electrically parallel them. You can hardwire parallel them with some more tinkering. Ebay has many PCBs and kits for LM3886 power amps and power supplies in the few-dollars range.

And as always, the dollars go into the power supply. Expect to pay bet,ween $20 and $100 for the power transformer. The filter caps are not cheap,either. The electronic complexities are in the amplifier circuits, but the money is in the power supply.

[J M Fahey]

Yes, believe it or not the amplifier itself (semiconductors, passives and PCB) is the *cheapest* part.

The bulk $$$ (as well as the physical bulk and the weight bulk) are in power supply, *transformer* , *chassis* , and more or less on cabinet/case, depending on how fancy you get.

And we have not even mentioned the speaker and speaker cabinet.

There is no loss of prestige or bad karma for buying an LM3886 kit: both the typical EBay supplier and yourself will use the same Factory datasheet suggested circuit, and he´ll supposedly use a properly designed and very well made (in China) PCB.

You will have to pre designPCB on a napkin, learn to use a CAD package, do the layout and design and then have it made in China anyway ... buying the kit is exactly the same thing, only you saved a couple steps.

"In principle", the magic is not in the Power amp (we are talking SS here) which just needs to be a robust workhorse, capable of carrying your signal to a speaker, but in the preamp, EQ, etc.
.
A good speaker and cabinet is another full job by itself.