Apexjr has a decent 150W amplifier that can be customized for use as either a full-range amp or as a subwoofer power amp.  The good news is that at $50 it is a great buy, but the bad news is that it has a 220V transformer.  

The amp is a conventional 150-300W design, using the same type and number of output transistors in the ApexSr amp.  It's got a generous heatsink so with a decent transformer and ±65v rails it should be able to output around 300W at 4 ohms.  However, the power supply capacitors are a bit small for a 300W amp, (2x6800uF@80v), so call it an honest 150W.  

The amp comes in two pieces--the amp itself and a control panel that includes the low-pass filter needed to make this a complete subwoofer amp (Apexjr calls this a "pair"--a better description would be a "set").  It is possible to use the amp separately, which is nice if you need a full-range amp that can be mounted into a speaker or if you already have a subwoofer filter circuit and just need a power amp.  The instructions below show how to make the amp stand-alone and how to get around that 220v problem.

Replacing the transformer

The amp needs a 93VCT transformer to generate the ±65v rails.  As luck would have it, Apexjr also sells a very nice 250W toroid that is perfect for this amp--and it is only $25.  It's a bit big, but if you mount it vertically it will fit nicely on the heatsink.  Just connect the secondary winding in place of the main outputs from the 220 transformer that comes with amp.  The original transformer also has a low voltage winding that is used to power the ±12 circuitry on the control board.  However, we won't need that--so don't worry about those extra 3 wires.

The hard part is mounting the transformer.  I made some crude brackets using 3/4" aluminum angle stock and some plexiglass.  Not too hard to do with a table saw for cutting the plastic and angle stock and a Dremel and drill to make holes.  See below for construction details.

The toroid gets clamped between the plexiglass, and I used silicon glue on the sides and bottom of the toroid to make sure that it stayed put.

Stand-alone Operation

The amp won't work unless you connect it to the control panel assembly.  This is because there is a relay on the board that keeps the transformer outputs disconnected from the bridge rectifier inputs until -12v is applied.  The reason this is done is to eliminate turn-on and turn-off noises.  The amp itself is well behaved on power-up and power-down, but the filter circuitry puts out a lot of garbage until the voltages stabilize.  However, if you aren't going to use the control panel assembly, there is no need for the relay--just short out the contacts or else replace the relay with two pieces of wire as shown below .

With the relay out of the way, the amp will power up when you apply 110V to the power block connections.  All you need to do now is connect the input to a standard connector and provide a way to get AC power to the amp.  The connection diagram at Apexjr shows that you need to connect to pins 5 and 6 for the audio input, or else bypass this connector and connect directly to the board.  Pin 3 is actually -12v and pin 4 is the relay--but who cares--unless you add another transformer these voltages won't be there.

Comments

OK, so it's a bit of work, but it's only $75 and the amp looks great from the outside--it's a nice, simple, massive anodized heatsink.  If you need a full range amp or already have a filter/eq circuit and just need a power amp, this is a good low-cost, high-quality solution.

Update, 3/18/2004

I was curious whether the 10,000uF capacitors that Apexjr is selling for $4 would fit on the board in place of the existing 6,800uF capacitors.  Given the relatively high voltage used on this amp (±65v rails), 6,800uF should be more than adequate (remember--it's the total amount of charge stored in the capacitor that counts, and Q=C*V).  However, for driving lower impedance loads, where the total amplifier output is current-limited rather than voltage-limited, the higher capacitance would be beneficial.  

It turns out that the board is actually drilled and laid out for these larger caps, as shown above.  The 10,000uF caps have 4 terminals rather than 2, but the board will accept either capacitor.  Unfortunately, there are a couple of obstructions that you need to work around, but it turns out that these problems aren't difficult to solve.  First, there is a resistor (R25) and capacitor (C8) that are in the way when you try to install the larger cans.  No problem--you've already unsoldered the old capacitors and cleaned out the 4 terminal holes, so what's the big deal about removing two more little components?  There's plenty of room for these two parts on the back side of the board--see the picture below for how they look installed in their new location.    

The other "gotcha" is that there isn't enough clearance for the mounting screw that is near the new capacitor.  This isn't a real big problem, either:  just squirt some silicon glue into the hole and let it dry.  Actually, the case on these devices appears to be insulated from the capacitor terminals, so it shouldn't matter whether the screw penetrates the vinyl case, but it still seems a good idea to just go with a glob of silicon glue.

This mod is fairly easy to do, and with the beefy toroid the amp should be able to put out on the order of  200-300 Watts.  I've got a dummy load somewhere in my parts boxes and will update this page when I can get around to measuring the output.  

 There's only one thing that bothers me about this upgrade.  If you follow the traces on the board you will see that the two "blank" terminals are connected to either V- or the AC input to the bridge rectifier.  This shouldn't be a problem, since according to ohmmeter measurements these terminals appear to be totally insulated.  These devices from Apexjr are Nichicon type GK (M), but I can't find any information on this series.  However, they are similar in construction to the  GK (HH) devices, and the GK (HH) datasheet has this warning, which makes me a bit concerned:  

    As blank terminals are not insulated from capacitor element, they shall be mounted on independent lands. 

Well, in this circuit the blank terminals aren't mounted on independent lands.  I'm assuming that there shouldn't be a problem as long as you stick with the Nichicon GK (M) capacitors, but it would be nice to know for certain whether this is a potential problem (it certainly seems to work fine on both of my amps).  If you come across a Nichicon GK (M) datasheet, please send it my way.

Last updated Mar 18, 2004