Hi!
I'm having some trouble resolving an issue on a Phase linear 400 series 2 but with what seems to be the internals of a series 1 (differential input pair instead of an opamp for one). It has been fully restored by kind of the phase expert here in France and worked fine and sounded great until the owner had a wiring problem to his speakers and the positive rail fuse popped. With the fuse back in place all seemed ok at first but actually there is an audible (and very visible on scope) distortion at any power level starting at about 1kHz (nothing visible under that frequency). It's present in the same amount on both channels. I have the schematic but there are no voltages on it so it's quite hard to check. There's no offset and bias seems ok (about 37mA).
Any suggestions would be much appreciated
I'm having some trouble resolving an issue on a Phase linear 400 series 2 but with what seems to be the internals of a series 1 (differential input pair instead of an opamp for one). It has been fully restored by kind of the phase expert here in France and worked fine and sounded great until the owner had a wiring problem to his speakers and the positive rail fuse popped. With the fuse back in place all seemed ok at first but actually there is an audible (and very visible on scope) distortion at any power level starting at about 1kHz (nothing visible under that frequency). It's present in the same amount on both channels. I have the schematic but there are no voltages on it so it's quite hard to check. There's no offset and bias seems ok (about 37mA).
Any suggestions would be much appreciated
37 mA of output stage bias? That’s actually too high, and could even be causing oscillation. The old quasi comp versions are supposed to run fully class B. My series 2 is *very unhappy* at what’s a normal bias for any other amp. The distortion increases as you bring up the bias, and only decreases again once the bias gets so high that it tries to run away. Sounds like hell on piano when mis-adjusted. Factory setting is 350 mV across the base-emitter of one of the output transistors (same as the old Crowns, for same reason).
actually I have 0mV across almost all the emitter resistors except for 3 on one channel where I get 5mV on each of them on one side.
I have had the NPN predriver transistor (40327, I think) go short circuit from base to emitter. That upsets the output stage bias drastically and drops the current gain resulting in distortion. There is then no “correct” setting. This failure mode CAN happen if the positive rail fuse blows and you continue to try to run the amp that way (it will try, and sound like hell and clip real early). It can also go open circuit, but that would clip the positive half of the waveform completely OFF. When that transistor has died, I use MJE340 as replacement. Check it for sure, as well as everything in the current limiter circuits. They’re not very “reliable” as built.
From my last tests (I will try again tomorrow and check those mentioned parts carefully) it got less bad the higher I pushed the volume. I got up to 80W without clipping, I didn't push much more because my resistors are only 100W.
OK so I removed and tested all the transistors on the main board and they all test ok (the 2 current limiter transistors had an extra reading ICE0 that wasn't there on other parts, this is with a simple component tester). Distortion measurement at 20kHz gives me about 0.4% at very very low volumes, .08 at high volumes so within spec (around 80W) and worst is at about 0.2W where I get 2.6% distortion. I should probably add that it's only visible on the positive half of the cycle and is present on both channel at the same amount.
Last edited:
That looks just like the distortion profile I get when the output stage is over-biased. That is, if it try to run it like every PA amp I own, with 3 to 5 mA per output transistor. But it normally happens on both half cycles. Happening only on the + side suggests that the bias is mismatched between the upper and lower halves. How can this happen? One possibility comes to mind - a leaky bootstrap capacitor. The - terminal is NOT connected to the output. It goes to the base of the output transistor. If a few uA or tens of uA leaks, it will add a bit of bias to the upper output transistors. The lower driver will simply absorb the excess so no extra is generated in the lower outputs. This bias mismatch will cause distortion, and that cap DOES go bad. Pull the cap, and if the problem disappears get a really *good* one so it never happens again. Why did they connect the cap there instead of the output? Because it has lower crossover distortion that way when everything is running *normally*. I have tried both ways.
Be glad it’s not a series 2 if you’re seeing a leaky bootstrap cap. The same problem causes LATCH UP in that op amp input stage. The mechanism is quite complicated to explain, but if you’re getting DC to the speaker for no apparent reason on a S2 look there FIRST. Took me weeks to track down. Because it just doesn’t make any sense. If the cap is connected in the traditional manner, all these problems disappear - BUT the distortion under 5 watts goes up.
Be glad it’s not a series 2 if you’re seeing a leaky bootstrap cap. The same problem causes LATCH UP in that op amp input stage. The mechanism is quite complicated to explain, but if you’re getting DC to the speaker for no apparent reason on a S2 look there FIRST. Took me weeks to track down. Because it just doesn’t make any sense. If the cap is connected in the traditional manner, all these problems disappear - BUT the distortion under 5 watts goes up.
The current limiter ICEO is normal on some versions. The original has germanium current limiter transistors. At some point they switched to silicon and redesigned that portion of the circuit. You could copy the circuit from the S2 if it turns out to be that. Very well could be.
I disconnected D9 and 10 diodes to eliminate the current limiter circuit as the manual for the S2 suggests and it didn't change the distortion. And yes they are germanium.
By the bootstrap cap I suppose you mean C11 on the schematic of the S1 going from the bas of the positive half output transistors to the diode string for the bias circuit ?
I tested this with a simple component tester, it's a elna silmic II of 47µF (measured 51) with .5% leakage (I compared it to a low leakage brand new KL series cap and that was at .4%) and .3ohm esr.
At low power the distortion is only visible on the positive half cycle (looks like wind is blowing the top of the wave a little to the left) but at higher power it kind of looks like crossover distortion visible around null point but only on the falling portion of the waveform, rising looks clean.
By the bootstrap cap I suppose you mean C11 on the schematic of the S1 going from the bas of the positive half output transistors to the diode string for the bias circuit ?
I tested this with a simple component tester, it's a elna silmic II of 47µF (measured 51) with .5% leakage (I compared it to a low leakage brand new KL series cap and that was at .4%) and .3ohm esr.
At low power the distortion is only visible on the positive half cycle (looks like wind is blowing the top of the wave a little to the left) but at higher power it kind of looks like crossover distortion visible around null point but only on the falling portion of the waveform, rising looks clean.
It doesn't seem to be the bootstrap cap, I removed it and the distortion was still there. I measured voltage drop across the emitter resistors with some power output and the negative half transistors are conducting more (15mv across emitter resistors of the positive half and 21mv across those of the negative half). It's the same thing on both channels. Also there is about 400mv DC on one channel.
It's most definitely a bias problem : when I turn bias up it gets way better, I can see the distortion going away on the the scope. It does go down a lot on the distortion analyser but not all the way (that is with bias pot maxed out which only gives me 500mV across the 10ohm driver emitter resistor).