* Recently In The Surgery | The Amp Doctor
 

PG MS2125 anniversary

Decided to restore this once we’d sorted the baseplate. About 1/2 way through, main rail caps upgraded, one bank of 12v caps removed. Once these are replaced with uprated parts the preamp gets the works.

PG 2

 

Preamp mostly upgraded, just waiting on some Holco resistors. Orange * parts uprated already, black X parts still to be done.The preamp is the weakest link of the amp and upgrading it should yield significant improvements.

PG

More Phoenix Gold ZX Modifications

ZX had been in for cap upgrades a year back, then changed hands. New owner wants SQ tweaks which includes high bias. Previously we added individual heatsinks to the predrivers for this but it’s a fiddly job.

 

 

Skills learned in 4 years of Metalwork at High school have come in handy. Piece of ally bar cut and tapped to allow fastening to end drivers. Green goo is thermal gap filler which transfers heat and holds the bar in place.

 

Ally bar soaks up heat from predrivers and it’s then transferred to the cooling tunnel by thermal gap filler. Middle predrivers bonded to the bar and have thermal compound to ensure good heat transfer. This keeps them cool and the bias stable.

 

Usual underbiased distortion waveform that we see with ZX amps as they arrive with us. If the bias is turned up they can go into thermal runaway as the predrivers get warm, so the factory sets it low.

 

 

After optimisation, and the bias is now completely stable. That’s less than a tenth of the previous figure.

 

 

 

If I could get these anodised it would look totally stock. Even as it is, it doesn’t seem out of place. This is now better than new SQ wise, and should give many years of listening pleasure.

 

 

Common fault on 1000/1′s sorted – May 2013

Insides of the venerable 1000/1 showing the main cap bank, the 8 shiny topped black bits. These are ‘SamYoung’ branded parts and after 5 plus years they are well past their best. Symptom is cutting out whenever the amp is cranked.

 

 

 

Original cap (black), bulging out the top like a Shoebury girl. New higher grade Cornell Dubilier part (blue) has better specs and shouldn’t need replacing again.

 

 

 

New parts fitted and bonded to the board. Typical power on one of these with the high grade cap bank is around 1150 – 1170 watts into 4, 3, 2 or 1.5 ohms at a current draw of 140 or so amps. We provide a 1 year fully comp warranty on these amps as the build quality is high.

 

 

Rocky T10001

Amp had a dead output, on these they are soldered 4 at a time onto a heat spreader. It’s possible to heat the spreader up and replace the outputs, as I’ve done on 30001′s before.

 

 

But it’s a lot easier to replace the old bank with a new factory fresh one. Bought a big box of Rockford parts from the previous service company and in there were several of the correct strips.

 

 

 Old strip removed and holes are in the process of being sucked clean of solder and leads. Main rail cap has a bulge like a teenagers shorts so will be replaced.

 

 

 

 

Mounting surface cleaned up, new thermal compound spread on it ready to fit the new module.

 

 

Module in place along with a brand new CDE cap, and another module replaced the same way. 1000w amp makes 1479 watts at 1 ohm. That’s the difference in buying a decent brand versus a cheap one. This amp will get a 12 month warranty as I know it will go the distance.

 

 

D5000 common faults

I’ve grown to really like working on these amps. They are well designed and reliable when repaired properly. There are a few common faults that come up as shown below, but overall I feel it’s a well designed product.

Fried output coil, one of 7 strands has shorted and got hot enough to fry the insulation. This needs to be unwound and completely rewound with new wire. It’s a bit of a tedious job but satisfying when the amp works again.

 

Thermal pads on bottom plate, vibration has caused the coil from the previous pic to short onto the base. We add Kapton tape and new pads so it won’t short out in future.

 

 

Pic of a cap that’s been vibrated off the board. Yellow marking is from arcing where the broken lead was making then not making contact and arcing and is vapourised metal.

 

 

New caps vs old, we can even uprate them with new EPCOS parts.

 

 

Input socket vibrated to death, missing 5 of it’s 8 connections. 4 Shiny Neutrik sockets fastened to the front plate and wired with flexible leads sorts this out very nicely.

 

 

 

This endplate is from an SPLD D5 but you get the idea. I believe D5000 amps are definitely worth repairing so much so that we are happy to offer a 3 month fully comprehensive warranty on any we sort out.

 

 

Soundstream Continuum Modifications

Bias mods to a Continuum, absolutely necessary if the bias is to be set into Class A. The sensor now measures the output device temperature, not the heatsink, and therefore responds around 200 x faster to changes. Without this mod the outputs overheat.

 

Pile of switches from two 705′s and a Continuum. These were very low quality and I change them as a matter of course as they will fail at some point, it’s just a matter of when. Shafts get replaced with metal parts that fit properly.

 

Crappy OE switch on the left, quality branded part on the right. Price differential is around 8 times which is probably why SS went for the el cheapo option.

 

 

Main board now has new switches, uprated rail caps, upgraded preamp chips, bias adjusters, new control shafts, gain mods so it can be used with horns. More to follow next week.

 

The casing has a sheet of Polyamide / kapton to insulate the power devices but let the heat into the heatsink. When we increase the bias on the front channels we need a thermally better interface, so the Kapton sheet is removed where the 4 channel outputs go.

 

Board is now ready to fit into the chassis. Special Boron Nitride filled insulators are fitted to the 4ch outputs, and thermal compound to all the other parts.

 

Channels 3&4 set up for the horns. Sensitivity has been dropped so they don’t play too loud, and distortion is very low. With horns being so efficient 1 watt will be very loud, typically 105dB.

 

 

 

Distortion residual is clean.

.

As promised, this is a standard Continuum straight out of the box, just as it arrived. Distortion is ball park for a SS reference series amp. Compare with the modified result though reduced by approx 92%

 

This shows the jagged distortion residual caused by underbiased outputs in the stock amp. The output stage biasing modifications detailed above all need to be done so this can be virtually eliminated. Picture taken at the same settings as above.