Would like to re-use an SS power section

[ChopSauce]

Well, I couldn't find the COM goes anywhere but to the pin 3 (collector?) of the (BR) 9014 transistor - through R244 (2k7/checked) - on the upper left side of the pic, below



and I haven't (yet) traced the following part of this circuit.

First I'd like to see if it works as is & the grounding of the power chips is more mystery from this board...

(which makes much more than I expected)

... so I agree with you are better not rely on me, but the point is my ohm-meter reads 97R (for a marked 100R resistor) between the heat sink and the GND.

Also the radiator is anodized (isolated) & the chassis had a cut to clear the part of the heat sink, which is made of raw alloy.

The board was bolted to the chassis trough both the ground hole mentionned by Enzo and the isolated radiator. Only the GND was earthed, which makes sense. Of course that may sound a bit foolish to rely on anodization for isolating parts on a production amp but...

(the chassis was powder coated, also)

... here it is, anway



(almost) ready to fire up. And also - as requested - (which makes me feel guilty for having forgotten to mention the 19V CT secondary before)



Anyway, now I have to go and buy some fuses.

EDIT: typo corrected

[ChopSauce]

That thing is terrible. I'm ready to give up.

(almost) ready to fire up...

... somewhat: the input unplugged, I heard a buzz & smoked the small R227 (100R) resistor between GND and the heat sink, which now reads 90R approx.

I have some 100R/1W I could solder in lieu of this one, but what for?

Maybe bolt the anodized radiator to an eyelet soldered to GND, just to see?

My first intend was to know I was better to throw it, or keep it on my shelves, and er... guess what -

- if no one has a better idea

[Enzo]

Make sure the collectors of the power transistors are not shorted to the heat sink.

[ChopSauce]

Make sure the collectors of the power transistors are not shorted to the heat sink.

Very impressive

So does the collector of the 2SA1295.

I'm curious at how you found this, and what it means?

[Enzo]

It means someone left out the mica insulating wafers under the transistors. They insulate the transistors electrically. The collectors are the center leg, but also the metal back of the part. If you bolt it down to the heat sink without the insulating mica, then the collector is grounded. The collector is connected to the power supply on each transistor there. Since the heat sink appeared to be grounded through the 100 ohm resistor, that means shorting power supply to the heat sink means the 100 ohm resistor has to absorb all the current, so it burns up.

[nevetslab]

I saw what looked like mica insulators under each of those M-200 case power xstrs in that one photo of the module/heat spreader. I guess it was just the grease pattern, unless there really IS mica insulators there under each xstr. My appologies if that mislead you. I saw on the power transformer the Primary is 220V, and the high voltage supply is 62VAC C/T (31V-0V-31V). So, you'd have around +/- 44.6VDC for the HV Power Supply. It's actually a 38VAC C/T winding (19V-0V-19V). Though I don't see those three wires.

BTW, did you save the mounting hardware for the toroid Xfmr? One or two metal 'dishes', probably 2 rubber pads and teh long mounting bolt/washers & nut.

[ChopSauce]

I saved the metal dish, but the hardware is the easy thing, anyway. I am a metalworker fond of mechanics & have a full cabinet for just nuts & bolts (& washers...) that I collect (almost) compulsively.

I also have a heap of basket cases full of mild steel, aluminium, stainless, plus a few racks for lengthy bars, tubes, etc.

Why do you ask?

I should be able to bolt everything back almost anywhere, should I have to.

I shown you the amp I built inside a battery charger enclosure, with a custom made vertical support for tubes on one side & a turret board on the other side, etc - just to convince you don't have to worry about that.

I also have a tig welder (& know how to use it), an oxy-gaz torch...

(I do miss a sheet metal bender, yet have enough stuff to build one myself)

... that should be fine, really!

Back to the amp...

EDIT (quick inspection report): mica wafers, thermal compound paste and unmarked screw heads. Everything seems "fresh from the factory". No internal shorts for the power chips...

... so what

[ChopSauce]

 

I wish I could change the title of the original post to "Harley Benton HBW 150 power board failure", afterwards...

(anyway, here it is mentionned, for search engines)

... & I'll check for shorted diodes around the power chips but I think I'm gonna throw it if I can't fix it this week-end.

I need to clear space on the bench.

Thanks for your help, anyway!

[nevetslab]

Sorry to hear your degree of discouragement this early in the project. After all, you were beginning with an 'unknown' status that did have potential for creating what you were conjuring up for a tube front end/SS power amp back end. Most of us would have checked for bad semiconductors thruout the unit prior to bringing it up on a Variac or light-bulb-limited AC mains. If the degree of damage is slight (no burnt circuit traces, and only some failed semiconductors that caused the component failure), you're not that far from troubleshooting the unit, finding that failed, replacing them and moving on. Great opportunity for learning, if nothing else, time availability not withstanding.

[ChopSauce]

Thx a lot for these kind words of encouragement! Of course you can feel my impatience

Well, the point is I am no professional in electronics - I am just doing that for fun & having this kind of debugging done on distance is not so comfortable, as you noted.

Another thing is I could find no more short. Maybe I was too much expecting from Enzo to having found the solution to carefully check the range of my ohm-meter, but here it is: I could measure no shortings at all anymore, neither between the power chips and the heat sink, nor between any legs of the power or driver chips pins - plus the loudspeaker gently whispers each time I touch a leg of any transistor in the unit with the ohmmeter probes - which makes me guess they are all OK, maybe wrongly?

So I don't feel like I can't afford troubleshooting the rest of the unit & I am just lost in confusion.

It is too complicated for me. Maybe I'm just lacking a method?

So you suggest that I de-solder the transistors and have them checked one by one with a controller? I saw mine does something like that but I haven't used this feature yet.

[nevetslab]

Check to see if your multimeter has a Semiconductor Test Mode (Diode Junction Mode). With that function, it allows you to measure the semiconductor junctions of diodes, transistors, FET's, etc. Assume RED lead for (+), BLK for (-). On the 2SC3264 NPN, the left terminal is the base, middle terminal is collector, right terminal is emitter. With the Red probe on the base, you can check for good semi junctions using the Blk probe on the collector and emitter. You'd probably read a voltage or around 0.55V or so in both cases. Similarly on the 2SA1295 PNP Blk probe goes to the base, Red probe goes to the collector and emitter, and would see similar reading if good. If it reads 0.000V that indicates a shorted junction.

One can also test for shorts on op amps. In your case, the 8-pin IC is a Dual Op Amp. V+ terminal is Pin 8, V- terminal is Pin 4, Outputs are on Pins 1 and 7. You can get semiconductor readings between the output and both V+ and V- of each section, and will normally only read in one direction, but not the other. If you get a short indication between Pin 8 and Pin 4, then there is a short across the V+ and V-. May or may not be on the IC, though.

I hope this helps some in test methods.

The small signal transistors may have their base terminal in the middle or on the right side, facing the flat side of the part. If the part prefix uses 2SA and 2SC, the 2SC is NPN, 2SA is PNP. Test for good semiconductor junctions on all of your transistors on the module. Diodes, be it rectifier, small signal or zener diodes, the Cathode is marked with a single stripe at one end. Blk probe on the cathode, Red probe on the Anode will measure a bit higher than the power transistors...typically 0.6 to 0.7V for silicon. On driver transistors, you may encounter a low reading on Base-Emitter junctions like 0.09V instead of 0.55V. There, you have an emitter resistor in the circuit, and it would read the same if you reverse the test leads. Sometimes there will be low readings on a bias transistor. Or, a transistor could be strapped so there is a collector-base short in the foil pattern, and you only have a base-emitter junction that tests good.

When I'm checking something that won't power up properly, I go thru and test every semiconductor on the unit first, as they are the prime suspects. Finding burnt resistors will usually lead to finding shorted transistors.

BTW, what the Diode Test mode does in the multimeter is to provide a 1mA current source with a 2.0V meter range. Good enough to even turn on RED LED's, but not GRN LED's, unless the voltage range is higher than 2.0V. Old analog meters in Ohms, can also be used to test semiconductor junctions, as an ohmmeter is just a current source and a voltmeter.

Anyway, not knowing if you already know any of this or not, this is a good place to begin in troubleshooting. AND, this would be done with NO POWER Applied and NOT plugged in to AC mains !

[ChopSauce]

Thank you very much for your patience!

Yes, that makes sense to me. I'll manage to do that and report back accordingly. Expect it to be delayed, though.

Thanks again,

[nevetslab]

Testing Semiconductors with a DMM...helpful hints

Last night when I wrote the previous post, I didn't have a DMM or some amp modules to use in photographic demonstration on what you may find. I took a series of measurements this morning on two different power amp modules, both using similar Sanken MT-200 style Power Transistors as you have on your amp.

In a couple cases that follow, measuring the Base-Emitter Junction on a power transistor may show a very low reading, as shown below:



Instead of the DMM measuring a typical B-E junction, it reads really low. That's due to the driver stage emitter resistor (being 2.7 ohms), and both that resistor and the power xstr's emitter resistor of 0.1 ohm are attached to the output bus, so it shows up as a very low reading. Only if the 2.7 ohm resistor were lifted off the output bus would you then get the typical semiconductor junction reading, as we see in the following:



Next, a typical reading on a diode...in this case a rectifier diode:



Then, here's what you would find if you encounter a shorted power transistor. B-E, B-C, C-E junctions are all shorted:



So, while checking all semiconductors that are soldered into place in the circuit, some readings require a closer look to see if there's a reason for seeing something other than a typical good semiconductor junction as you would read on a free-standing transistor.

Also, on the heat sink, there is a clamp in the middle, between the two power transistors, which looks like it is holding TO-220 Driver transistors down and not sure what else. you'll need to check those parts out as well. There should be NO voltage on the heat spreader bracket to which the power and driver transistors are attached.

[ChopSauce]

Well, I am really glad I could read and learn from these detailled troubleshooting instructions you wrote.

The smoking resistor shown there is some voltage on the heat spreader bracket, though - doesn't it?

I could find no short however, around the power chips. I have no time at all for more, so far.

Might that exist such thing as a short appearing when tension or heat is applied?

So thanks again, and...

... yes, the heat sink "embeds" 3 diodes. They are passing through the heat sink - like if they had to be cooled - but there also I could find no short at all.



You did your best, anyway!

[nevetslab]

Those three diodes are the Bias Diodes for the output stage, and are clamped to the heat sink so they thermally track the temperature of the hat sink, maintaing 'nominal bias for teh output stage.

Normally, there should NOT be any voltage on the heat sink, so somewhere you have a connection to a DC voltage, and it must have been high enough to overheat your 100 ohm resistor. Always some puzzle to work out when troubleshooting. You can always use the ohmmeter to see if there is any resistance between the heat sink and the two HV power supply terminals (either at the collectors of Output Xstrs or the buss caps that feed them). Now, none of us have seen under the middle clamp that holds the two TO-220 Driver Xstrs and the three Bias diodes. Unless those Drivrs are insulated caes, they would use smaller mica insulators. Since there aren't small screws with nylon shoulder washers thru the TO-220 xstr heat sink tabs, you'll have to look to make sure there isn't any continuity between either driver's collector (metal tab & middle terminal) AND the heat sink.