Well I was able to run the MTF Modified SX-1280 today for most of the day at a 4 ohm load (2x HPM-100/2xHPM60) and it ran nice and cool and sounded great all day.
This unit had a problem on every board aside from the blown outputs so this was not a good unit for the do it yourselfer's like me for sure.
Anyway I couldnt be more happy it was worth the wait.
Something this cool needs its own thread so here it is.
Steve

The preliminaries:

Know that the failure of one of these meters is a pretty rare event. Since the meters are DC coupled to the driving transistors (Q1 & Q2), a failure of these devices can toast the meters. I'd replace the electrolytic caps and the two transistors just to CYA. Toasting the 'new' meters would be a bad thing.

Edit: Oh...Mark is on the case. NM.

So what's the story with that receiver? Any blown outputs?

Yes, 3 of the 4 output transistors on the right channel were shorted, but surprisingly enough the rest of the amplifier card checked out fine with original parts. Those 4.7 ohm base resistors may have saved the drivers etc...

Steve had obtained another right amp channel and I transplanted the original transistors from it into his right channel. Turns out the replacements show 30 mV bias (target: 15 mv, 10 - 25 range acceptable on second pair) without any bias adjustment on, and then the bias adjustment can be cranked without effect until it suddenly kicks in and starts going above 30.
The left channel had a severe mismatch between the idle current of the two pairs - I don't remember the numbers right now - I didn't write them down...

edit:
Strangely enough, subsequent bench testing of the outputs showed NOTHING to indicate the symptoms I saw. Diode drops were consistent across all of them, leakage on a transistor tester was non-existent, and leakage on a bench test at 30 volts Vce with base and emitter connected was also zero, on a meter range with 10 microamps resolution...

So I am using the "extra" amp as a dimensional example to install and use TO-3P On-Semi output transistors.

Remember how EchoWars did the SX-1980 transistor replacement?

This is for the SX-1280 which has it's own circuit board connecting the transistors and emitter resistors underneath.

And no, I haven't powered it up yet, nor done the electrical adjustments, that's later in the week. The transistor / board combinations are almost ready to be swapped into a real live unit.

The ORIGINAL transistors are:

2sb705/2sd745 case?? cebc 160v 10a/16a 120w 40-200hfe b:17/d:15Mhz

I'm using :
863-MJL1302AG pnp TO-3P/TO-264 bce 260V 15/30A 200W 45-150 30MHZ $2.97ea
863-MJL3281AG npn TO-3P/TO-264 bce 260V 15/30A 200W 45-150 30MHZ $2.97 ea
so $23.76 for 8 currently
These as you can see are the slightly smaller (electrically) little brothers to the transistors EchoWars used in the sx-1980,

EchoWars used in SX-1980:
863-mjl4302ag pnp TO-3P/TO-264 bce 350V 15/30A 230W 50-250hfe 35MHZ $5.25
863-mjl4281ag npn TO-3P/TO-264 bce 350V 15/30A 230W 50-250hfe 35MHZ $5.25


http://www.audiokarma.org/forums/attachment.php?attachmentid=158990&stc=1&d=1247348406

AS you can see the replacements are twice as fast as the original transistors, and the rest of the amplifier was left bone stock with the ORIGINAL transistors. Also unlike the 1980's amp, the 1280's amp has a LOT of small film type caps scattered thoughout it to stabilize things. Thus after testing I didn't find it necessary to change any component values or add anything to tame the beast.

Idle current set nicely, and consistently between the two amps, and the balance of idle current sharing between the 4 transistors per channel was more an expression of the slight variation of emitter resistances than it was of transistor mismatch. It was bang on.

In fact at 90khz (the Tek SG505's max), the response was only -0.3 dB down, with no appreciable change in distortion.

A nasty overshoot on square waves turned out to be what my generator was feeding it, and even then it faithfully reproduced it.

It illustrated to me that I need to get better, more powerful, non-inductive load resistors.

Then again, I have a spare 1280 amp to recheck the effects, plus examine any problems that may crop up when (possibly lower Ft) replacement transistors are used in the amp control and driver circuitry.

Prolly not done posting to this thread, but I'd like to see details of your board replacement.

I had suggested a while back that the board could be done away with altogether, and standing terminal strips used, such as those commonly employed in tube amps wired point-to-point.

http://www.tubesandmore.com/cemirror/inv/P-0401H03.GIF

I really don't understand, that picture shows all phases from the PC board through the final assembly, including one hold - down for the TO-3P package.

The original hookup wires were mostly soldered to the transistor pins, which I did, and the wire wrapped pins with wires on them (base connection, emitter resistor common output terminal) were transferred to the new board and soldered in. The original doubled up the power wires on one transistor's collector, routed the current through the transistor to it's other collector lead, then across the PC board to the other transistor. I took the second wire coming in from the big capacitor's hot connection and instead connected it directly to the collector of the second transistor - minimizing the resistance of the current path.

I reused the 3mm metric mounting screws, washers and lock washers. With 1/8 inch thick aluminum hold-downs for the transistors the screws had full thread engagement without bottoming out in the holes. The hold downs were about 1.2 inches long, 1/2 inch wide and 1/4 inch thick aluminum bar with appropriate hole spacing and size.

I kept the original mica insulator in place, the extra space allowed for a little maneuvering for position without shorts. Plus then they were available IF ever needed.

The pictures were from the first amp channel, I rotated one of the second channels boards 180 degrees to put the output terminals closer together so the blue wire in between them reached while tucking in to the heat sink recess provided for that wire. That just meant drilling one extra hole in the opposite side to mount the base connection wire wrap pin close enough for the original wire to reach.

The emitter resistors are about 1/2 inch square and 1/4 inch thick, and are mounted on the component (not visible ) side of the board.

This picture is of the first channel done, just before I buttoned it up for installation. The white stuff is padding to protect the anodizing of the heatsink.

Brown and yellow wires (2) are the base drives, and there is a jumper on the opposite side of the board to connect the other base.
orange and white wires (5) are power coming into the collectors(4), and going to the board to energize the drivers(2).
green and violet wires (4) are the emitters for the connections to the power limiters
blue wires (2)are the common points of the emitter resistors (1) and the output connection back to the board for the feedback and filters etc...

second picture is the other channel before cleaning and greasing for installation.

third picture is back side view of original amp assembly off of heatsink, shows resistors....

http://www.audiokarma.org/forums/attachment.php?attachmentid=165428&stc=1&d=1250747739

http://www.audiokarma.org/forums/attachment.php?attachmentid=165432&stc=1&d=1250750373

http://www.audiokarma.org/forums/attachment.php?attachmentid=165431&stc=1&d=1250749861

Your first pic didn't offer much of a view of the fabbed boards. I guess that was my complaint. Nice work. Is the new board etched?

Actually, no. It was a simple pattern on purpose.

Although later, more intricate boards will be.

It was actually quicker to score razor lines and peel the foil off than to get resist on, and etch it in Ferric chloride. Once I started the peel, if I was careful, I would get the entire strip... I've got some modified side cutters I ground so that I can get REALLY close into things, with a lot of leverage. I get a couple of hundreths of an inch up and I could grab it.

On the last of the 4, I actually was able to turn the corner in one strip on that last right angled "anti-trace".

Later on, I'm going to get some Citric Acid as an etch accelerator, and some guys have had moderate success modifying a printer so that PC boards can be fed straight through, and directly printed on. If the board is pre-heated before printing, and baked afterwards before etching, it seems to work well enough that they have used some surface mount chips.
I'm even looking at a combination of a defunct scanner assembly to move a head traverse assembly rather than moving the object of the printing.

Of course that takes time to do, and I seem to have very little of that now, versus not a huge need for pc boards.

BUT, I would like to make up a bunch of current sources. so... I'm conflicted..

edit: and, YES, I know about the laser printer to toner to pc board fusing etch resist transfer method and products.

edit2:
Back when I was of High School age, at home, without thinking, I started to etch a PC board, using an ALUMINUM PAN !! :yikes: man did that etch the PAN fast.... quite an exothermic reaction. Which SPED UP the process - now THAT was thermal runaway....
I had NO excuse.. I had had enough chemistry to know that was a no-no.... :nono:
Not much surface tension induced slowing of the reaction that time. :stupid:

I started to etch a PC board, using an ALUMINUM PAN... You made me spit my beer out! LOL!