power supply fets keep blowing

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
need help guys, been trying to trouble shoot this for a while now
working on a 225 hcca (digital reference) taken it to 2 repair shops also and no one knows whats up.

whenever i try to replace the fets in the power supply they blow immediately as soon as i connect power....

the transformer windings arent shorted, i know this because i had purchased another amp of the exact same model...a working one with no problems.

if anyone has worked on this model amp, they should know of the double sided tape orion used to hold the fets to the retaining rail that screws down onto the heatsink....this is where my headaches started....

like an idiot i tried to remove the securing rail off the fets again, the tape stuck just like the first time and some of the power supply fets got destroyed again...the fets are smp60n06-18 ( i cant find datasheets on these fellas either)


so now i have two 225 hcca's that are scrap, just to see what happens i replaced the damaged fets on the second amp and the same thing happened -power supply fets fried immediately after connecting power.....i tried ssp60n06 and rfp70n06 but both types blew, i then tried taking the remaining good fets(the original ones) from one amp and transferred them to the other, same thing happened...


so my question is what is causing this? do i need to disconnect the drive transistor in the output section- connect power-disconnect power- then reconnect the drive transistor?

does anyone know what needs to be done to correctly fix this, because there is obviously a series of steps i need to follow to get these amps running again.

any help greatly appreciated :xeye:
 
The first thing that you should have checked is the gate drive section. You will blow as much MOSFETs as you can afford if the gate drive circuit is malfunctioning and turning both banks at the same time or for a too long time. Also, you may be unadvertedly damaging the MOSFETs when you mount them (too much ESD), there may be some insulation problem between the tabs of the devices and the heatsink, or some lost solder ball may be unadvertedly stuck in the wrong place...
 
The first thing you need to do is to get some sort of current limiter for the power supply. I use a 2 ohm 50 watt resistor. Others have suggested using automobile head lamps. Even a low value fuse (5A) will save the FETs.

As Eva mentioned, the drive section could be at fault. There could also be something in the audio section that's drawing too much current from the supply. Either can kill the supply. If the FETs fail instantly, the problem is likely in the power supply itself.

These are some common problems with Orion power supplies that can cause the FETs to overheat/fail:

Defective driver transistors. The PNPs fail almost every time the FETs fail. The NPN transistors sometimes fail also. Both are inexpensive. Change both the NPN and the PNP transistors when replacing the FETs. They use MPSA06s and MPSA56s if they are too burned to read.

The 10 ohm resistor between the emitter of the PNP transistor and ground opens. This prevents the PNP transistor from turning the FETs off and will cause them to fail.

The zener diode connected to the gate of the FETs fails. When it does, it will prevent the FETS from being driven high. If only one is shorted (for 1/2 of the FETs), the FETs on the other side will run hot and eventually fail if the amp is driven into a load.

Sometimes, some of the gate resistors open. This will cause one or more FET to remain on which will cause it to overheat and fail. When it does, others will follow.

Check the output transistors for shorts.

Check the large capacitors on the 12 volts side of the power supply (~3300uf). Sometimes the legs burn when the supply fails. If they are glued down it may not be obvious.
 
ok, but what really trips me out is that these were perfect working amps before i got curious and pulled the securing rails off and damaged the fets because of the tape...

a month before that i had changed the capacitors in the rectification portion to Elna 1200uf from 1000uf....also the two smaller caps 1000uf to 820uf that are laying next to the horizontal caps...sound improved ALOT....

then i got itchy and wanted to see what else could be modified, pulled the rails and have been stuck in hell since then...

i dont hink that changing these caps should effect the power supply, but if im wrong someone tell me...everything worked fine and the amp even ran a bit cooler...

The first thing that you should have checked is the gate drive section. You will blow as much MOSFETs as you can afford if the gate drive circuit is malfunctioning and turning both banks at the same time or for a too long time. Also, you may be unadvertedly damaging the MOSFETs when you mount them (too much ESD), there may be some insulation problem between the tabs of the devices and the heatsink, or some lost solder ball may be unadvertedly stuck in the wrong place...

what is ESD???

:confused:

The zener diode connected to the gate of the FETs fails. When it does, it will prevent the FETS from being driven high. If only one is shorted (for 1/2 of the FETs), the FETs on the other side will run hot and eventually fail if the amp is driven into a load.
i cant see any zener diodes in the powersupply section, but there is a light grey with black rim diode that is connected to the gate traces.....doesnt look the same as a traditional black diode.

there is one part of the amp that i think is suspect, i dont know what kind of circuit it is:

the layout of the fets and transistors are like this:

Power supply>f-f-f-f-f-d-n-p-p-p-n-n-n>>>>>> rca side

f=fet
d=diode
n or p = bipolar transistor

the 'n' between the d and the p....what is this guy? there are diodes on the base and emitter, the middle collector runs straight into a mylar cap, and the trace goes into the ground plane for the speaker outputs(i think)....this trace plane also has the rectification caps on it...on the other side of the pcb in the same position is the 'p' version of this trans...that side hasnt blown any fets yet.

the reason why im so confused is because these were perfect working amps...before i took the transistor securing rails off, there were no problems with them before, when i installed the new fets i checked all my soldering spots and made sure nothing would arc.

i replaced 'all' of the transistors and fets on the first amp and the same thing happened, although i'll admit i didnt check the drive fets when i put them in...but these were perfect working amps.

why is this happening?

:bawling:
 
The zeners are gray/silver.

The N device is the current-boost transistor for the regulator. If you look closely, there is a diode on each leg. One diode is on the 'input' pin (the center pin). One diode is on the third pin (the output of the regulator). Both of those are standard diodes (1N400x types). The first pin of the transistor has a zener diode on it. There is a P counterpart on the other side of the heatsink. The NPN is the positive reg. The PNP is the negative reg.

When you initially pulled the transistors off of the clamp, you likely wedged a screwdriver between the clamp and the tab of the transistor. This can rarely be done without destroying the transistors. What you need to do next time is this...

While the board is still screwed down, pull the screws out of the clamps and insert a T20 driver in the hole in the clamp. Be sure that the driver does not extend down into the hole in the heatsink. While applying at least 50 pounds of downward pressure, pry back away from the center of the board. Start in the holes near the ends and work your way down the clamp. You will rarely break transistors when you do it this way. If you are applying enough downward force, the transistors will remain flat on the heatsink while the clamp pulls away. See atached image.
 

Attachments

  • orionampclamps01.gif
    orionampclamps01.gif
    16 KB · Views: 611
ok , gotcha those grey/silver are zeners
i got the A06...A56...zeners...10ohm R's today...but do you think i should change that 'n' transistor for current boost too?

if so, think it would be ok to use a 'ST' BD9112/911 instead of the 2n6488/91 instead because thats all i can find (crossreferenced)....as a precaution i would probably change the diodes on them also...

just to double check....the current boost transistor has a zener on the left pin...no diode that i can see on the middle pin...and a regular diode on the right pin...just want to make sure were on the same level...

the original zener diode in the power supply is a
1N47-42A-202....all i could get was...1N47-42A-806...that ok?

ok one more question :) the zener on the left pin of the current boost transistor is a 1N47-44A-208...would it be ok to use the 1N47-42A-806...since i got 20 pcs in one package when i bought them...

oh, and whats that vertical standing 2N6488 transistor for ???

im really close to doing this >>> :smash: to the amp

double thanx
 
ok , gotcha those grey/silver are zeners
i got the A06...A56...zeners...10ohm R's today...but do you think i should change that 'n' transistor for current boost too?

**** If it ain't broke, don't fix it. Unless you physically damaged it, it's likely OK. I've never seen a bad one in all the ones I've repaired.


just to double check....the current boost transistor has a zener on the left pin...no diode that i can see on the middle pin...

**** It's likely about 1/2" away from the transistor but it's connected to the middle pin.


the original zener diode in the power supply is a
1N47-42A-202....all i could get was...1N47-42A-806...that ok?

**** Unless it's shorted, leave the originals. The numbers beyond the 42a are likely a date or manufacturing code.


ok one more question the zener on the left pin of the current boost transistor is a 1N47-44A-208...would it be ok to use the 1N47-42A-806...since i got 20 pcs in one package when i bought them...

**** If it's not defective leave it. If it is defective, you need to replace it with the same value. The 42 is a 12 volt zener and the 44 is a 15 volt zener. Using the lower voltage device will result in lower regulated voltage for your preamp section.


oh, and whats that vertical standing 2N6488 transistor for ???

**** It's likely the regulator for the gate drive. They use a crazy pulse VOLTAGE modulation instead of the normal pulse 'width' modulation. If you look at the gate pulses (with an oscilloscope) when the amp is driving a load, you can see the amplitude of the pulses change with demand (higher amplitude for higher demand).
 
ok found the middle pin diode, the top was on the same trace as the transistor, the bottom of the diode went to another trace on one of the PS rails... :) interesting design...

but the current boost transistor is going to have to be changed, the back of it is discolored like its burnt, already pulled it and the tape thing happened again (before you told me the right way to do it) :(

think its ok to use the ST microdevices BD911 and 912 for these current boosters? im not going to be able to find 2N6488/911's unless i buy 1000's...

do you have any recommendations for the power supply fets?
smp60n06-18's cant be found around here, but i have alternatives
like irfz44n, buz12, rfp70no6, ssp60n06, more...

i cant find a data sheet for the smp60n06-18, all i can find is a toshiba crossreference...the smp60n06-18 is listed as 60v 60a 125 watt device with .018rds

gonna be a couple days before i actually start replacing components, have to go pick up a new dmm...not reading 0 ohms when leads are touched.

really appreciate your help
 
If you only have one meter, set the meter to ohms, pull one lead and touch the tip of the remaining lead to the contacts in the port where the other meter lead was. Do the same for both leads. If both leads read the same, the meter is likely defective but you could go one step further and try jumping between the meter lead ports with a short piece of ~16g wire. If the meter still reads a high resistance, the meter is definitely defective.

Keep in mind that there are many meters that won't go to zero when the leads are touched together. If you're measuring a device that has less than ~1 ohm, you need to touch the leads together to see what it reads. If it reads 0.2 ohms, subtract that from the reading you get across the device.
 
ok, got the second amp working.... jeez im so dumb... for a current limiting power supply, i used my battery charger....completely forgot i could use it for this purpose...
variable from 6-60 volts with 8 amperage adjustment settings...100 amp max

but i installed a 4 amp fuse in line with it just incase...amp came on at 7.6 volts (could hear the fets saying ooooh ahhhh)....

when turned on at 12.4 volts, , the ammeter on the charger jumps to around 15 amps then settles back down somewhere below 10 amps...this sound right???

i havent tested the actual output yet though, no rca cables, have to wait till tomorrow...

anything i should test for before i actually put it on a battery?
------------------------------------
for the other amp (another hcca 225) i took the transformer out a long time ago, to test for a short in the wires...i unravelled the wires and have to rewind it again....if i can still find where i left the original core...

if i use a different core, are core materials really that different from one another?

and

the primary windings on the core were 4 strands bonded parallel together like this >>> llll

and the secondaries are 2 strands bonded together in parallel

if i twist them together instead,think that will be ok?

almost there man, almost there....

:D
 
The idle current should be ~1.5 amps. anything more than ~3 amps indicates that there's a problem.

The battery charger likely has no filter capacitors. It may cause the caps in the amp to run hot if you try to do more than run the amp at idle on it.

Core material is important. The material that is used for inductors won't work well for transformers. If you can't find the core, try the following site:
http://www.cwsbytemark.com/prices/toroidal.php

You'll want to use either 'F' or '77' material.

Twisting them is OK but try to lay the windings out precisely as they are in the original.
 
ok checked the amp again with a more precise ammeter and it idles around 1.8 amps.....the ammeter on the charger wasnt very detailed for anything lower than 10 amps...

gotcha on the core materials:)

tested the amp today , one weird thing is happening...output is very low for the first 10 minutes or so...then increases all of the sudden....then plays normally...does this both in stereo 4+4 ohm and 2 ohm mono.

while the output (not input) is low, the clipping indicators light up if input is driven hard.... i havent had a chance to probe all the components on the pcb yet, but would you have any idea what this could be? this only happens when the amp is cold, and started up in the morning.....id say the output is 2-5 % of what full output is, at a given ohm load.

and im curious, do you know what op-amps they use for the preamp/feedback section?

lastly, what the heck are those white opti chips in the power supply for?

thanx, id buy you a beer if i could. :)
 
The clipping indicators don't seem to have much to do with the actual speaker level output.

It sounds like the muting cap isn't charging. Measure the voltage on pin 3 of the muting transistors. It should be ~-10v. If it's near 0 volts, the amp is in mute.

The following diagram shows you the opto-isolators. I didn't double-check the schematic so let me know if you find an error.


http://www.bcae1.com/images/swfs/orionpartial02.swf
 
ummm, which ones are the muting transistors and muting cap? :confused:

are they the ones in the schematic?

in the schematic; the MPS A06 that connects to the TL494, has the thermostat inline going to the TL 494, any chance the thermostat is going bad?

when i measure the muting transistors (once i figure out which ones they are :) ) do i need to keep the black probe of the dmm at the negative power wire ? and probe the legs of the transistor with the red probe? dmm switched to dcv?
 
There are two 2n5639 transistors near the NE5532. Those are the audio muting transistors.

The mute delay cap is near the third 2n5639 which is near the positive regulator. When the amp is initially powered up, the cap charges towards the negative rail voltage. If the voltage isn't going negative, the cap could be defective.

The thermostat on that amp is simply a switch. There is little chance that it's defective. You should read ~0 ohms across it when it's cold.

Touch the black probe to the shield ground.
 
shield ground? as in the rca cable ground? sorry, still a little ignorant as to some terminology...

ok, i spotted all of those transistors lastnight...but whats that NE5532 (not at home now)

that cap by the positive regulator is a relatively small one if i remember correctly, i think it was around 10uf...

should finish this amp up tonight if everything goes right.


sure am glad its not the op-amps, like i originally thought....do you know what brand they are? ive never seen this kind before.

triple thanx
 
arghh.....bad day.....changed the muting cap thats near the positive regulator there was another cap in the same vicinity (same value) changed both....still the same...

measured the 2n5639, both gave me 0.0 volts

any other ideas?

btw how do you correctly check caps?and A06, A56, 2n5639 transistors while still on the board, or do i need to pull them?

and those A06, A56, 2n5639 transistors, if the part number is facing up, the pins go from left to right 1,2,3, right?...well checked all of the pins and they all gave me 0.0 volts, with the black probe on the shield ground.

:bawling:
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.