Garbage switch

[stephensank]

There is a good bit wrong, but at least each is in a distinct area and neither seems to affect the other. There are 4 limiting resistors and thermal fuses. Only one is bad. Bingo. One trouble cleared. I have replaced the 1 ohm resistor and its accompanying output transistor. Since the output transistors are all connected to common base feeds, common emitter or collector "rails,"I I also replaced the corresponding transistor. I hope to be able to say BINGO another trouble cleared. Right now my quanddy is wbat has caused the resistors to cook! They seem to be adequately sized for the load, but barring hivh rail voltages, I don't get it! The transistors Q230 &I Q231, Q235 &I Q236 all I check ok. I do plan to replace them anyway. Any thoughts anyone?

I'm looking at the PA5 service manual, which I happen to have on my home pc here, my PA7 manual being at my shop, but the component numbering & etc. is very close.

Unless the relay was replaced with an improper one with wrong coil impedance, the R151/251 burning is mysterious. As for R243 & 244 going, that pretty much has to have been caused by the channel going into ultrasonic oscillation, which is pretty much always caused by cold solder joints on the predriver & etc. transistors, i.e, ones that are on individual heatsinks, where the thermal & sink-weight stresses crack the joints. Also is the only thing I can figure causing an output emitter resistor to open.

 

[stephensank]

What do folks think about trying a Music Fidelity buffer? I have never caught the glass bug, but what the heck? Does it do anything I can't do with just a LP filter? Do I just need to mask the top of bad CD's because the new stuff is letting it through, or is it just plain harmonic distortion that the higher BW analog happily passes along? I have not checked if the schematic is available. Sometimes current production is pretty controlled.

Too many projects to take on shortening the Nak cabinet. At least for now. Holding off on the Pa5, modifying a 535 or building an F5. I still have a project waiting to compare about 10 CCS implementations based on what I learned here and several weeks of Spice modeling. Finding fast fets is the issue. I want to know more about Linn, Music Fidelity, Arcam, Naim and other higher end smaller amps. I have never had an amp with regulated rails. Are they really that much better? I need to build a decent crossover and a couple power supplies. Regulators seem harder than is obvious.

The musical fidelity units like that have pretty darned shabby build quality, but not a terrible thing to add to a system as a temporary bandaid for poor cd's.

I think regulated rails for the output stage of an amp are a VERY bad idea, unless there are VERY large filter(energy storage) caps AFTER the regulators, so that the dynamic transients aren't strangled. I have personally never heard an amp with regulated output rails sound remotely as good & as dynamic as any decent unreg'd amp.

 

[tvrgeek]

Thanks. Cheap try. I'll see how my cheap bidding goes. There seem to usually be several around so no need to get excited. Their build can't be much worse than the Chinese ripoff copies.

Interesting on regulation. Linn, Levenson etc. I was thinking just the IPS and VAS. Modulating the low power rails sure does show up in the Spice sim. Unless yo use a booster power supply, you do reduce the swing of the VAS, so you do limit the output just a tad. On the other hand, just an RC or maybe LC filter between output rails and VAS rails may be just as good. I have found old circuline florescent ballasts to be quite handy as cheap power supply chokes for wall wart filtering. Only building tells for sure.

 

[rogerbent]

Stephen,
You mentioned "two resistors" that needed to be changed in the PA7 protection circuit. Which ones are they and what values are recommended, please? I ahve done everything except change the bath water in the Left-channel of a PA7, but cannot get it to go out of protect. The bias is fine and the speaker output rail voltage is near zero. I have changed almost all the transistors in the protection circuit and the pre-drivers all to no avail.

One other thing, the external two-transistor board on the PA7 I'm working on has designations Q241 and Q242 with three resistors R266, R267, and R268. R266 is 390K and R267 & R268 are 100K each. The schematic I have - and all I have been able to download - depict an off-board Q238 and Q239, but these numbers are on the main board of the unit I'm nursing, not on an external one. Do you have any insight on this?