Phase Linear 400/500/700 Repair Procedures


SAFETY WARNING!

The repair of these amplifiers require working with LEATHAL VOLTAGE LEVELS THAT CAN AND WILL KILL. The voltage levels in the PL400 excede 140 Volts across the powersupply rails. The Dual 500 and PL700 is higher still. If you are not comfortable working with high voltages and have no technical experiance in doing so, DO NOT ATTEMPT TO REPAIR THESE AMPLIFIERS!

In each step of the repair, you must discharge the powersupply caps to protect yourself and to prevent damage to the amp while you are working on it. Use a 10 watt 10-25 ohm resistor to discharge the two powersupply rails together before you put your hands inside the amp. DO NOT SHORT THEM WITH A SCREWDRIVER!



FAILURE MODES:

This procedure is intended in particular to the PL400/700/Dual 500 amplifiers, but the basic procedure applies to any discrete amplifier design using bi-polar output devices with a linear/brute force type powersupply.

Repair of a broken PL400/PL700/Dual 500 amp starts with a very carefull inspection of all semiconductors. Under normal failure modes of the amp, failure consists of 2 or more output devices, the drivers, one or both of the pre-drivers and several resistors associated with the related devices. The PL700 and Dual 500 will often take out many of the diodes and the current limiters right back to the input pair. Always check every semi-conductor device for shorts/open/leakage! Always check all resistors associated with any device that failed. Check the BIAS (and OFFSET in early amps) pots for noise/erractic operation.



RE-ASSEMBLY:

DO NOT install any output devices during the initial bring up of the amp. These will go in after the initial bring up phase. DO install the final driver devices (the RCA T0-3 cased devices at the bottom of the heatsink). All of the final outputs must be the same type device. Do NOT mix Delco/Fairchild/Motorola or NEC devices on any one channel! If you have any of the DELCO (steel cased PL909 or PL910) devices, it is STRONGLY recommended that they ALL be replaced with a suitable device such as the Motorola 15024 series devices (use 15024 NPN and 15025 PNP pairs on the very later version of the amp and the Dual 500 that used the fully complimentary output design, for the more common 400/700 amps quasi-complimentary output design, use the 15024 NPN device and any amp using the Delco PL909/PL910 devices or the Fairchild devices). The NEC 2SD555 (NPN) is an excellent device and nearly indestructable but it is pricy and becoming hard to find as it was discontinued several years ago. It's PNP compliment is the 2SB600.

Note:

If you must use the DELCO devices, they MUST be matched for beta. The color and code markings indicate family and beta maching on these devices. When using the Delco or Fairchild parts do not mix devices of different color/part-number/gain types. They WILL fail. With the Delco/Fairchild devices, the current sharing should be matched within 10% if at all possible. With the Motorola or NEC devices, it is almost of no concern. The Motorola/NEC devices are similar in performance and can tolorate a 20% mismatch with no noticable ill effects. I have under test/repair on the bench, run just 2 of the NEC devices in a PL400 up to full power for several hours with no trouble at all. The RCA drivers are just fine and should be left alone. Try to use only the original devices elsewhere in the amp for rest of the circuit. Use only flame proof resistors for any that have failed. Check ALL of the diodes on the board as they are the most common failure on the amp if anything is wrong besides the outputs.


BRINGING IT UP:

The PL400 should be serviced as a pair of amps and both channels checked out together. The PL700/500 can be serviced one channel at a time.

You must use a signal generator with a means to control the output level and frequency, a dual channel oscilliscope, a 120 volt variac with voltmeter and AC current meter prefered, and some high power 250 watt+ 8 ohm and 4 ohm load resistors as a minimum to startup the amp. Failure to follow this procedure will cost you a lot of money. I have no trouble spending your money, but you might.

Replace the two power rail fuses (PL400 uses only 2 fuses for both channels) with small value fuses (about 2-3 amps).

Remove the output devices if you have not done so already. This phase of the test is to ensure that the amp is actualy operational while not allowing it to destroy your costly output transistors if something is wrong (grin). Typical worst case failure this way is a resistor or at most a few smaller transistors if you don't follow this procedure (BIG GRIN). With the amp connected to the variact (turned off down to ZERO and SWITCHED OFF!), the signal generator connected to the inputs with an output level of about 100 mv at 1KHz., and NO LOAD connected to the output terminals. Connect the scope probes to the output terminals with the scope set to about 10v per division and the sweep so that you can see several cycles of the sinewave clearly. Apply power SLOWLY to the the amp while watching the output on the scope. At about 30 volts into the amp, the output will suddenly jump slightly and then start to produce a sinewave on the scope. If this does not happen, shut the amp down and check things. If you see the jump on the scope and you see no output, check your setup of the generator/scope. If all is looking well at this stage, you should see a sinewave that is either "normal" looking or shows symetrical clipping. After adjusting the output level to just give symetrical clipping (adjust the scope as needed to give a good view on the display) you should be able to raise the AC input some to about 50 volts and see a clean sinewave again. Check the offset voltage on the outputs. It should be under 50 mv. If you can not get a clean symetrical sinewave, power the amp down and start looking for things broken. Most often it is an overlooked resistor if all of the semi-conductors are good. If things are looking good, the next step is to raise the AC input more and again repeat the above checks for symetry of the signal. The generator input should be able to be raised higher to the point of symetrical clipping at each step. If at anytime the amp fails to clip symetricly, shut it down. Failure to do so will cause you smoked parts and money! If you can take the amp all the way up to a full line voltage in (120 VAC) and the amp will produce a clean symetrical signal up to the point of clipping, you are pretty much done with this phase. The next step (2nd step) is to install only one pair of the output devices (one on the + rail side, one on the - rail side) and repeat the above steps. If all looks fine, repeat once more (the 3rd step) but with the load resistors connected. As the AC input is raised and the amp "turns on" you should be able to actually drive some power into the load at this time. Check for non-symetrical clipping!!!! If you see that signal looks good with no imbalanced clipping, then you are home free at this point. Raise the AC input levels slowly and watch the scope. If all is fine, you should be able to run the amp up to full power and get a hundred watts of power out of the amp before the fuses pop. Check the BIAS as described below. Once the amp is operational and checked out, you can start installing the rest of the outputs. BE CAREFULL. Make sure you power the amp down and disconnect it from the AC lines and SHORT OUT THE POWERSUPPLY CAPS before you poke your fingers into the amp. Do one final step through of the above procedures with all of the output devices install.



BIAS ADJUSTMENT:

The amp should draw about 25 ma across the emitter resistors on the OUTPUTS (.22 ohm) after the amp has reached operating temperature (running the amp at 50-60% power into 8 ohms then a 5 minute cool down period). The offset on the output should be under 40-50 mv or less. If it is not, replace the input differential pair on the channel that is not meeting specs.


MORE NOTES:

I HATE SILLIPADS. I do not care what the manufacture of these things say, they suck!
If you really want these beasts to stay together, replace ALL of the rubber silipad insulators on the outputs with mica insulators. I have run tests on these by placing thermal probes on the underside of the outputs at the heatsink, and another probe on the topside on the transistor and have seen 40-50 degree differences between the rubber vs mica. Mica is the ONLY type I use. Also the rubber pads puncture thru if not torqued just right (torgue to 6-7 inch pounds EVENLY - NO MORE!). The mica should also be torqued same way and a small amount of Wakefield compound (zinc oxide) on both sides of the insulator. Do not use excessive compond. A proper amount is such that it just starts to show a very small trace of compound squeeze out around the device as it is tightened down. Do not use any compound on the rubber pads. They will deteriorate and fail.



MODIFICATIONS:

Also remove the current limiter devices (Q8/Q9 in the 400 schematic)). These are 80% of the cause for failure of the amp. They tend to latchup during high power operation at high frequencies due to slow recovery times of the limiter circuit. There are other reasons to remove them anyway. The one that has the most affect on the sound is the fact that the amp clips way before it needs to causing the outputs to run hotter than desired as well. The original intent was to "protect" the outputs from excessive power disapation, but the VI limiter clipping actually causes them to disapate even more power then if they were to just run to the rails. Also change the power supply bridge rectifier to a 40 amp/600V device. Under heavy loads it will fail. The amp is capable of much more power then rated. The PL400 should see about 265 wpc into 8 ohms, about 450 watts into 4 ohms, and just short of 900 watts into 2 ohms. The PL700 amp is capable of much more power then rated. You should see about 465 wpc into 8 ohms, about 900 watts into 4 ohms, and about 1450 into 2 ohms.


FUSES and Other Good Things:

I use 10 amp AGX fuses and have on several occasion popped them on very heavy transients. Put fans on the outputs. I use 5" 120V boxer fans wired in series (this will cause them to run at half speed and QUIET) or 240V units in paralell and place them between the heatsink fins with a couple of "L" brackets to hold them in place, although I have also used double sided sticky tape for installation in the home where the amp does not travel around anyplace. I have used more than 30 of these amps with these mods on the road under very heavy usage doing sound re-enforment duty for rock-bands/clubs with ZERO failure over a 12 yr period. My amps at home have run for more than 24 yrs with no failures and are making 900+ wpc into 2 ohms. The load that they see in my system currently is under 3.2 ohms (4 NHT 1259 subs in 2.8 cu/ft enclosures each).