[amb]

The PPAv2's output transistors are rated for continuous 1.5A (BD139/140) or 4A (MJE243/253). There are 2.2 ohm emitter resistors that would help limit the current in a short circuit. There is also current limiting in the PSU (The LM317 regulator for example will current-limit and shut down if the draw is over 1.5A). As long as you don't keep the short circuit condition persist over a long period of time while the music is playing loud, then there should be little risk of damage to the circuit. If you're really paranoid, add a fuse at the DC power input.

 

[jamont]

Quote:

Originally Posted by Syzygies Just wondering, before ordering. This comes with a barrel connector, right? Anyone know an alternate vendor you'd rather have me support?

I'd get it from Allied rather than Newark. Newark really sucks. And yes it has a barrel connector.

 

[tangent]

Quote:

Originally Posted by Syzygies This comes with a barrel connector, right?

Yes.

Quote:

Anyone know an alternate vendor you'd rather have me support?

Allied Electronics.

Quote:

Originally Posted by amb There is also current limiting in the PSU

Yes, though I wonder if that is enough. There is no current limiting between the rail cap bank and the transistors. When that big bank discharges suddenly, what is the peak current? I suppose it depends on the output resistors.

 

[Syzygies]

Quote:

Originally Posted by amb There are 2.2 ohm emitter resistors that would help limit the current in a short circuit. There is also current limiting in the PSU (The LM317 regulator for example will current-limit and shut down if the draw is over 1.5A).

Thanks, all.

Shorts are most likely during travel. I've seen NiMH AA cells put out 6A in a short, though I'd guess that even my 1000 mAh AAA cells are incapable of 2.5A.

Those 2.2 ohm resistors are only spec'd for 1/4 watt, they'd certainly see more than that in a short.

For that matter, I just reread various of Tangent's voltage notes. I'm not coming away with a clear idea of the maximum watts my PPA is capable of putting into, say, a 10 ohm dummy load. I might have to measure this. To be safe, I'll extrapolate first from a bigger dummy load.

People are going to use 1/8" plugs, perhaps in "we're just kidding ourselves" 1/4" adapters. It would make sense to have a conservative recommendation here, for anyone who wants to add output resistance.

 

[tangent]

Quote:

I've seen NiMH AA cells put out 6A in a short

They should be able to do better than that. I've seen 9V alkalines put out that kind of flash current. Surely a big cell like that can do better? Just dividing voltage by nominal impedance, I get 35A for Sanyo AAA NiMHs.

Quote:

Those 2.2 ohm resistors are only spec'd for 1/4 watt

It's not yet covered in the docs, but you're supposed to use 2W resistors on the output, for precisely this reason. We allowed such large resistors there because we heard stories of people with PPL diamond buffer modules turning their 1/4W output resistors into fuses.

Quote:

I'm not coming away with a clear idea of the maximum watts my PPA is capable of putting into, say, a 10 ohm dummy load.

No, there is indeed no way to get that number from what I've written so far. In fact, I think it's probably a complex problem, having to do with the hFEs of your transistors, supply voltage, and other such variables. You could use SPICE to simulate it.

I did an informal test a few weeks ago and got many watts into a 32 ohm load. I don't have the data on me right now. It was "more than enough", so I haven't bothered to memorize it.

 

[Syzygies]

Quote:

Originally Posted by tangent Just dividing voltage by nominal impedance, I get 35A for Sanyo AAA NiMHs.

Thanks. My shorts were never intentional, just misguided late-night experiments with some circuitry in the way.

While I wouldn't bet on your rule of thumb staying linear in a short circuit, it makes it clear that for purposes of this discussion, an NiMH battery pack might as well be viewed as capable of infinite current. Something else will give first.

I'm upping the wattage of my 1/8" jack output resistors, to stay commensurate with the 2 watt, 2.2 ohm output resistors on board the PPAv2.

As a rough rule of thumb, I'd say that additional R ohm output resistors should be rated for 4/R watts, e.g. 1/2 watt, 10 ohm output resistors can handle the same scenarios envisioned by the 2 watt, 2.2 ohm on board resistors.

 

[steinchen]

still a few caps missing, but my PPA v2 is up and running

two questions left for me atm:
1) across which resistor do I have to measure the buffer bias ? R24 ? What is the default bias setting ? 50mA ?
2) do I have to connect BBR with S2R for bass boost ? What does BBR stand for ? BBR = bass boost right (channel) ?

 

[tangent]

Quote:

Originally Posted by Syzygies While I wouldn't bet on your rule of thumb staying linear in a short circuit, it makes it clear that for purposes of this discussion, an NiMH battery pack might as well be viewed as capable of infinite current. Something else will give first.

Sure. Also keep in mind, my rough-and-ready 35A estimate is based on the pack impedance, and that is further reduced by the rail cap bank's parallel impedance. So while I'm not sure what the actual number is, I can't see that I'm greatly overstating the potential.

I think you're right, though, that it's effectively infinite. To get 8A (the max peak the MJE243/253 allows) through 2.2 ohms, you need to put 17.6V across it. That just can't happen in a typical PPA. Ohm's Law poops out on you first. Now, that doesn't put us fully in the safe zone, because between that and the 4A continuous rating there's a danger zone. But I'm optimistic that the supply voltage and discharge current will be dropping at this time, so the ability to sustain more than 4A for a prolonged time should be difficult.

The less studly BD139/140s would be easier to kill. In fact, I think it wouldn't be all that hard to kill them with 2.2 ohms of output resistance.

It's worth pointing out that the resistor can easily die before the output transistor, even with 2W resistors. If you're somehow putting 4A through them, the wattage is 35.2W. Fortunately, film resistors tend to fail open; they're hard-to-replace fuses, in effect.

During the PPAv2 development, we went through several design changes to the buffer, and at several points I had 10 ohms on the output. This will raise your distortion a bit but it will ameliorate these problems greatly. And because it's inside the feedback loop, it doesn't hurt as much as putting resistors inline with the output wires. That's why the distortion doesn't go up much.

Quote:

As a rough rule of thumb, I'd say that additional R ohm output resistors should be rated for 4/R watts, e.g. 1/2 watt, 10 ohm output resistors can handle the same scenarios envisioned by the 2 watt, 2.2 ohm on board resistors.

Careful there. We didn't pick 2W through any special wisdom. It's just the largest value that fit in the space we had.

 

[morsel]

Hi steinchen,

Measure output bias across the output resistors R24 or R34.
Set bias to 30mA, or 10mA if using a battery board.
Connect BBR with S2R for bass boost.

 

[Syzygies]

Quote:

Originally Posted by tangent We didn't pick 2W through any special wisdom. It's just the largest value that fit in the space we had.

It looks to me like PercyAudio's "MILLS Non-Inductive Wire Wound Resistors" MRA-5 (5 watts) $2.15 each in a range of nearby values, fits the spot. I'm about to go out, I'll check again. This resistor is highly regarded, if one believes their catalog copy!

 

[steinchen]

raised the bias to 30mA (= 66mV across R24 2.2Ohms), now Q24 / Q34 are getting warm but not hot, sound improved considerably

it's 2 a.m. round here, I'll wire BBR tomorrow

DC offset: -0.4mV / 5mV (didn't match trannies out of lazyness)

I suggest adding a setup-section to the PPA v2 pages (OpAmp bias, buffer bias, bass boost wiring, ... )

I substituted Q3 with BF245C since I'm using AD843 which are drawing 13mA quiescent current. Using 2n5486 would require hand picking for IDDS > 13mA and pn4392 would have a significant lower noise isolation due to it's hig IDSS. Maybe you can add some words about that, too.

first impression: nice clean sound, not that warm and rich like M³, rather analytical, but that has to be verified the next days with AB tests. Building cost should be much lower without monolithic buffers, a big step forward.

PPA v2 is definitely a great amp. Thanks and regards to Morsel, Tangent and PPL

 

[jamont]

What should be considered in choosing the output transistors? Are the BD139/140 and MJE243/253 equally good in this circuit?

 

[stadams]

Quote:

Originally Posted by jamont What should be considered in choosing the output transistors? Are the BD139/140 and MJE243/253 equally good in this circuit?

jamont,

ppl addressed this earlier in post #16 I believe. He gives a good explanation between there differences and possible effects on the circuit.

Later,

 

[jamont]

Quote:

Originally Posted by stadams ppl addressed this earlier in post #16 I believe. He gives a good explanation between there differences and possible effects on the circuit.

I'm going blind, thanks!

 

[Syzygies]

Quote:

Originally Posted by Syzygies It looks to me like PercyAudio's "MILLS Non-Inductive Wire Wound Resistors" MRA-5 (5 watts) $2.15 each in a range of nearby values, fits the spot.

Ok, this resistor is a bit long to fit cleanly, but there's room on the board for its body:

Their other boutique resistors, the $7.50 "BE YAMAMURA Non-Inductive Wirewound Resistors", include a 2 watt resistor that makes only a slightly better fit:

Then, back at the working class lunch counter, there's Mouser Xicon 2W 2.2OHMS 5% Small Metal Oxide Resistors, $0.19 each.

The perhaps better Vishay Dale 2W 2.2ohm, Mouser 71-CCF02-J-2.2K, is non-stocked.

Perhaps most promising is Vishay/BC components at Digi-Key, RES 2.2 OHM METAL FILM 2W 5% for $0.32 each. These have an L2 body dimension of 12mm, a tight but feasible fit with a board hole spacing of 0.5".

I'm willing to buy 60 or 120 of these and do some matching, drop sets of 6 in the mail to people who PM me, but 5%, yikes, how close do we want to get?