Matching output transistors on an audio amp?

[budwich]

Re: Matching output transistors on an audio amp?

continuing to do checks on transistors that have been remove. There are ones that appear to check "well" with the transistor checker in terms of the output data shown BUT when measured with resistance to the collector (case) have differences... some resistance as opposed to "infinite" which I take as some form of "leakage". i would think that means they are not good.

However, the thing that appears to get me now is that some transistors have the plastic insulator sleeve in the screw hole areas of transistor while two out of the 5 that I have removed, have none.... seems rather weird. Looking closer at the heat sink mount area (beyond all the thermal grease), the transistors that don't sleeves have a "full insulating gasket" (ie. that full shape of the base of the transistor). Why the difference? should they all have "gaskets"?

[petehall347]

Re: Matching output transistors on an audio amp?

maybe the screw insulators are broken ? they are top hat shape

[R_J]

Re: Matching output transistors on an audio amp?

There may not be an insulating sleeve if the bolt is the used for the connection of the collector of the transistor to the pc board.
Take some pictures so you know how to reassemble them.

All the pictures I have seen (of a Crown DC300A) do not use the white plastic bolt insulators, If you insulate the bolt head, there is no connection for the collector to the foil. and all transistors have insulators between them and the heatsink and the bolts should all have star washers to make a good connection

[petehall347]

Re: Matching output transistors on an audio amp?

whoops 2 posts broken ...
the screw insulators are top hat shaped or ought to be so they might be broken .
are the heat sinks connected to ground ?

[budwich]

Re: Matching output transistors on an audio amp?

MY eyes are not that good... :-( closer inspection shows that I need better glasses.
Two of the 5 transistors that I pulled supposedly had no screw "insert".... BUT what it was is that the plastic "insulator gasket" (of the shape of the base of transistor) stayed stuck on the heat sink... oops :-) Looking at the transistors with the "insert"... after clenaing the silcon better, has the gasket still stuck to their base. So all transistors have this "full size gasket" insulator.

sorry for the confusion... I need to do better inspections.

[larrymoencurly]

Re: Matching output transistors on an audio amp?

The Harbor Freight multimeters they often give out free with the right coupon and any purchase have a transistor hFE tester, but I don't know how well it works for matching transistors.

Here's Walter Jung's review of a power amplifier and his efforts to match its input transistors: https://cdn.badcaps-static.com/pdfs/...45f4ffe112.pdf

[budwich]

Re: Matching output transistors on an audio amp?

thanks... it appears to indicate some "improvement" for matching on the input side but also indicates that no matching is required on the output (for the amp in question). I do believe crown matching (selects) its input stage.

[budwich]

Re: Matching output transistors on an audio amp?

still more questions about output transistors. I got some mj15024 to replace crown original 2n5631.

My question is around c-e resistance testing. On existing transistors (most) the resistance, is infinite. On ones that appear to have "issue", the resistance is very low and some case higher (ie. in the megaohms). Testing the replacement transistors, the same way, indicates most with megaohm c-e resistance. I kind of expect them all to be infinite but is this expectation wrong. Does the c-e junction need to be infinite or is some form of high resistance OK? Checking the transistors, with a capitance / transistor checker appears to show things as "good".

[eccerr0r]

Re: Matching output transistors on an audio amp?

C-E resistance rating when B is unconnected is basically its leakage. Silicon transistors should not leak and the resistance should be very high. Depending on your multimeter, most of them should show up as infinite. Germanium transistors leak more.

There are some transistors with built in freewheel diodes, these will show up a low resistance short if you connect the meter one way, but should be open circuit the other direction.

Likely it will still work even if the resistance is very high though I'd suspect some QC issues there. All of my low frequency silicon transistors show up as infinite. My germanium transistors leak and do not show up as infinite. I don't have any uncommitted silicon RF transistors to test but I suspect they may leak some.

[budwich]

Re: Matching output transistors on an audio amp?

thanks... makes me feel a bit better about them as "off shore sources" don't always give the most confidence. I can see what you are saying about an unconnected B having a possible "influence". The ones that I tested "in circuit" had only the C and E isolated as I would have had to demount them totally to isolate the B. The failed ones that I measured as "partially isolated in-circuit" had obvious issues with c-e near zero. However, a couple on the "distorted channel" had some "few mega ohms" resistance compared to others in the channel (same type) that all measured open / infinite resistance... that's why I was somewhat concerned about the replacements having "similar" characteristics.... albeit higher megaohm resistance but certainly not open / infinite.

[eccerr0r]

Re: Matching output transistors on an audio amp?

Incidentally, my germanium power transistors leak tremendously, though they work, my DMM gives resistance ratings in the kiloohms scale never mind megaohms scale. When connected in a circuit as a common emitter, the transistor does shut off and the load does not get powered with no base drive.

[budwich]

Re: Matching output transistors on an audio amp?

OK... so I might have another problem on the distorted channel that might not be related to the two transistors that were "different characteristics" than the other 8 on that side... :-(. That will be further down the list... I first need to replace the "known bad ones" to get the thing to power up and go from there.

Thanks for the "data point" / experience.

[eccerr0r]

Re: Matching output transistors on an audio amp?

Hopefully the circuit is well designed and compensates for transistor mismatches. Especially when paralleling, mismatches can cause premature failure, connecting a bunch of n transistors of x in parallel does not give you n*x, nowhere near that much...

[budwich]

Re: Matching output transistors on an audio amp?

My fingers are crossed (along with my eyes). I have replaced what output transistors that I measured c-e resistances and compared against others... basically a total of 5 at this point. I am trying to limit damage from my "fingers" as soldering and related skills are not "tech grade" so I did not want to replace all 20 output transistors ...:-) I used a "bulb limiting power wiring unit" to limit the current going into the amp. After lots of checks and compares against the two channels circuit (dry tests), I powered it up... bulb lit bright for a second or so and then dimmed to "darkesh" and the amp switch light stayed lit so things are at least some what "electrically happy". I need to do some more "low current measurement checks" along the various circuit areas to see if I missed or blew any other areas during earlier "fingers". Hopefully it will check out and the mj15024 will be "play happily" with the remaining original output transistors that I did not replace.

[eccerr0r]

Re: Matching output transistors on an audio amp?

Kind of funny, I have a MOSFET that I didn't have enough gate drive due to a crap circuit I was using. It got so hot that the plastic shoulder insulator melted and plastic goo everywhere.

Luckily the heat sink was already electrically isolated, and the MOSFET still sort of works...