Transistors

Yup. THS3062, practically a unique opamp, blinding fast with +-15 supplies. But if it's amplifying a sine wave at, say, 20 volts p-p out, and the frequency goes up to roughly 12 MHz, it crashes, pulls tons of power, gets red hot, and phase inverts! If you drop the input signal way down, it recovers!

We've been pestering TI support about this for about 6 weeks now. All they say is that they don't have an engineer available to look into the problem. Sounds bad.

John

Yikes! Is it noisy?

John

On a sunny day (Sat, 17 Nov 2007 09:15:55 -0800) it happened John Larkin wrote in :

Well, John!!!! the datasheet specifies input common mode range as +-13.9V MAX. Page 3 of ths3062.pdf. It is a video amp :-)

Yes, magnetostriction-generated, but not too bad.

On a sunny day (Sat, 17 Nov 2007 18:40:25 GMT) it happened Jan Panteltje wrote in :

Oops, you said out, I thought you said in. What is your load? Output voltage is only specified at 1k and 150 Ohm.

We're running these at gains in the 3 to 5 range, and it does this working inverting or non-inverting, loaded and unloaded, so it's not a common-mode issue. And at 12 MHz, we're not even slewing a volt per nanosecond yet.

It crashes and phase inverts in either config! It made a nasty blister on me poor finger, it did! But the effect doesn't seem to be thermal, in that freezing it hard doesn't affect the frequency trip threshold much.

You'd think that TI would find something this exotic to be interesting.

John

That's 27.8V p-p, and that's the guaranteed minimum CM range, not the maximum. Allowable input range is +/-Vs where Vs can be as high as

It's not clear what Bad Thing happens if you exceed the Absolute Maximum Vid of +/-3V though.

Best regards, Spehro Pefhany

On a sunny day (Sat, 17 Nov 2007 11:06:59 -0800) it happened John Larkin wrote in :

Ideally, differential input voltage is always zero, regardless of the feedback network. Reality might be different due to saturation, slew-rate limiting, offset voltage, phase shift, gain at the input less than infinity and so on.

Best regards, Spehro Pefhany

On a sunny day (Sat, 17 Nov 2007 14:57:52 -0500) it happened Spehro Pefhany wrote in :

Exactly. An input step could well for a short moment exceed that spec. I was thinking that it is a pity they do no longer provide schematics of the circuits. Would be nice to see what is in there, and then you can get a better idea. As this is a video amp (in my view) I think 1.4V pp input would be safe LOL. We also need to look at John's circuit.

I note the 2nd Harmonic Distortion rises rapidly around 10-20MHz. Probably an inadequately stabilized multi-loop that is amplitude dependent.

...Jim Thompson

Assuming NPN, and 0.7V junction Vf:

Vb approx. 5.7V Ve approx 5V Ic is negative

This is a current-mode opamp, so in the inverting config, with the + input grounded, the - input looks like a low impedance to ground, even open-loop... spec'd 71 ohms typ. So there's no way the input signal is going to pull the - input very far off ground.

Given that this amp is flat to .1 dB to 100 MHz, and can slew 7 v/ns, the closed-loop diff input voltage should be tiny in either config at

We've seen the same effect on final-product boards and separate test breadboards, with +-12 supplies. I suspect the amp is broke.

I'd love to see a schematic of the guts of this thing.

The datasheet on the TI web site, dated 2002, still has DRAFT ONLY splattered on every page.

John

On a sunny day (Sat, 17 Nov 2007 12:58:20 -0800) it happened John Larkin wrote in :

What is you load? 75 Ohm (video like)? If 75 Ohm you get 133mApp for 10Vpp With 50 Ohm you get 200mApp, and that exceeds Iout guaranteed.

You can argue that if you have a 50 Ohm resistor on the output, and a 50 Ohm cable, that then it sees 100 Ohm. The datasheet says it can source and sink about 60 mA in 50 Ohm, than makes 6Vpp. It also says the swing is at least +-3.5 V in 150 Ohm. So, if you have 50 Ohm, and 50 Ohm cable, then that makes 100 Ohm load, and expect less then 7Vpp max output swing. capability (page 6 pdf). Do I see this right? I maintain it is just a video amp cable driver with a bit more then 2 x 1.4Vpp output capability.

On a sunny day (Sat, 17 Nov 2007 12:58:20 -0800) it happened John Larkin wrote in :

PS, small correction, that was for +5 and -5v supply, I see the output is about 12 V in 100 Ohm for +-15V supply. When I calculate power for that case I get to more then 1 W.

There are several packages, but I see things like 50 C/W thermal resistance. So, yes, you may burn your finger, 25 + >50 = 80 C. They are talking about powerpad layout guidelines too on page 19 pdf,for SOIC-8.

It is possible the output flips over because of some internal current or temperatue / power limiting, measure device temp, calculate dissipation for whatever load you have, at the supply voltage you have, and the output level you have.

If temp rises to 85C then it already draws 10mA by itself, 300mW,

I sort of agree with Joerg that often it is betetr to make your own amp with discrete. But you likely already have the boards.

I will stop on this now, maybe we can design a better one.

Avalanche breakdown on the Vbe side, then. Vcb would seem to be reverse biased, but not at breakdown levels. Regardless of whether the bulk of emitter current comes via base or collector, the base is going to be jacked up about 6V or 7V above Ve, so within 2V of -10V -- probably very close to -10V, in fact, if there is any Ic at all (and I think there would be.)

I have heard about a 'Miller's avalanche multiplication coefficient' that is supposed to operate on Ic in this mode, but I frankly don't know anything about it in practice. I'm just a hobbyist, though, without so much as a single hour of schooling on the subject, so no surprise on that point. Ignorance noted, I'd tend to guess that there is significant Ic here and that the result of this is that Ib is tiny and negative, with a Vb close to -10V, leaving Ve around 3-4V. This means an Ie of about -600uA to -800uA with most of that showing up as Ic. I think there is also a case where if Vce is large enough, Ic goes without much limit... but at -5V I think this is well below that dangerous threshold. Anyway, my guess is that most of Ie will be seen as Ic, with Ib being small. So with the 1k, Vb won't be a lot different than -10V -- maybe on the order of (3 to 10)uA*1k away from it (guessing that 'beta' may still apply to a degree.)

I would like to hear your analysis of it, though. This is a novel application I haven't played with. I've only heard it helps for some kinds of faster pulse circuits and that's about it.

Jon

MAX.

It does the same thing, loaded or not loaded. The worst-case design load is 200 ohms to ground, 50 mA max.

output

TI doesn't say that. The datasheet specs and the example waveforms clearly show big swings. They even run it near both rails, 25 volts p-p out, to demonstrate its overload recovery... see fig 43.

It's not load that breaks it, it's frequency.

We're going to try the single, THS3061, and the older THS3001, and see if they have the same bug. National has a couple of amps that have high bandwidth and slew at +-12 or +-15 volts, but they're not as fast and don't swing as far.

John

I think it's likely another matter, one I've struggled with in high-frequency high-voltage-amplitude push-pull amplifiers, namely getting the pullup devices completely off before the pulldown devices turn on. And visa-versa.

Actually, it's generally not possible, or even desirable, to completely achieve this goal, but one struggles hard to avoid having too much rail-rail current under these extreme conditions, because various local parts become overheated. I fear too that many high-frequency high- slew-rate amplifier designs don't pay enough attention to this issue and may experience failures as a result.

Fred Abse snipped-for-privacy@cerebrumconfus.it posted to sci.electronics.design:

I figure ib ~= ie ~=1.6 mA & 0.6 v Vbe.

Vb then is about 8.4 V and Ve 7.8 V. Ic is well below uA.