Acoustic 370 output power vs 470

[johnk_10]

Like crossover distortion? And if so, did you try adjusting bias (by pot or resistor adjustment) ?

the 370 doesn't have a bias pot but I always install one in them anyway. drawing 45 watts at idle it was oscillating and very dirty sounding with the MJ15003G's. at 45 watts with the stock output trannys, it's perfect.

[J M Fahey]

Ouch!!!
Pity they didn't work.
Probably because modern parts are much higher bandwidth and may oscillate or at least be unstable if plain replaced in old stuff.
Please scope the amp out while driving speakers (it might be stable driving a resistor and go crazy with an inductive load) and check for instability, which is horrible sounding and veryu visible.

You might also have an open Zobel network, a relatively common problem, which would be masked by old transistors.

FWIW old 2N3055H and 2N3773 are slower (800kHz) vs modern transistors (3 MHz) ; that said old Acoustics can and do work well with modern transistors , but you may have to add some compensation caps.

Please post some scope screens, literally "what you see is what you get" , horrible sound is easily visible and waveform hints at what kind of problem it may be.

Also old Acoustics did NOT use current balancing emitter resistors at all, they just put all power transistors in parallel, literally (even the emitters) which is a terrible design practice.

They got away with it because they used tightly matched transistors, as you see a recipe for disaster when random unmatched transistors are used.

But let's see your scope pictures, you *need* them for proper troubleshooting.
And the problem might be something as simple as a dead Zobel network.

[johnk_10]

Ouch!!!
Pity they didn't work.
Probably because modern parts are much higher bandwidth and may oscillate or at least be unstable if plain replaced in old stuff.
Please scope the amp out while driving speakers (it might be stable driving a resistor and go crazy with an inductive load) and check for instability, which is horrible sounding and veryu visible.

You might also have an open Zobel network, a relatively common problem, which would be masked by old transistors.

FWIW old 2N3055H and 2N3773 are slower (800kHz) vs modern transistors (3 MHz) ; that said old Acoustics can and do work well with modern transistors , but you may have to add some compensation caps.

Please post some scope screens, literally "what you see is what you get" , horrible sound is easily visible and waveform hints at what kind of problem it may be.

Also old Acoustics did NOT use current balancing emitter resistors at all, they just put all power transistors in parallel, literally (even the emitters) which is a terrible design practice.

They got away with it because they used tightly matched transistors, as you see a recipe for disaster when random unmatched transistors are used.

But let's see your scope pictures, you *need* them for proper troubleshooting.
And the problem might be something as simple as a dead Zobel network.

since the amp works perfect with the proper original transistors I find no need to go any further with trying to get the MJ15003G's to work in it.

[nsubulysses]

Also old Acoustics did NOT use current balancing emitter resistors at all, they just put all power transistors in parallel, literally (even the emitters) which is a terrible design practice.

They got away with it because they used tightly matched transistors, as you see a recipe for disaster when random unmatched transistors are used.

But let's see your scope pictures, you *need* them for proper troubleshooting.
And the problem might be something as simple as a dead Zobel network.

This is an aside but how would you go about choosing values of emitter resistors. I often see different values of emitter resistors and they are usually low resistance in output sections (under 50 ohms). Would it be worthwhile to add emitter resistors to 370s for better reliablity if you are unable to match transistors?

I've wondered before if it's ok to replace 22 ohm emitter resistors with 33 or 47 or another similar value if I change ALL the emitter resistors. How critical is the emitter resistance choice in these output sections?

[Jazz P Bass]

Ballast resistor value is calculated.

Re1 = (β+1)(Vc−Vbeo(1−γΔT1)/
Rb1+Re1(β+1)

I am not an EE, so you may want to read this.
Source:Transistors in parallel - Electrical Engineering Stack Exchange to post #3)

[Enzo]

I think you will find those ballast resistors are 0.22 or 0.47 ohm, rather than 22 or 47 ohm.

[g1]

i'm going to try 8 new 2N3055HOM's in it when they get here.

Please let us know if they work properly. The HOM is some NOS part? Is it different from a current manufacture 3055?
It would be good to have a fix for these amps using off the shelf currently available parts.

[nsubulysses]

RCA - 2N3055HOM - Transistor, NPN silicon. P/N: 2N3055.

I have used these in 370s and Sunn Concert Basses with good results. I believe this is what the OP is referring to.

And yes Enzo you are right I re-checked schematics and it's .33 emitter resistor, not 33 ohm.

[g1]

Yes, they are surplus old parts "Selected version of the old standard". What I'm wondering is whether regular new generic 3055's (or any other modern part) will work without modification.

[J M Fahey]

Those are old technology types (notice the '85 date code) , when they're gone, that's it.

Hometaxials have not been manufactured for many years, my last batch was 300 from UR/USHA (India) which had acquired obsolete dies and technology from USA, go figure , they lasted until 2008,

I could never ever get H type again, only E (epitaxial) and everybody thought me crazy whem I asked for the old ones.

Same happened with 2N3773 which dried out around 2010 or so.

Modern 2N3055 are very different from the old ones, much faster and with smaller SOAR (safe area), so it pays to learn to use modern ones.
I can get as many as I want of these, in closed boxes straight from the official ST distributor, but they are E type, so useless to repair my old amps.
Can be used in new designs, if you consider the weaker parameters.



These "Toshiba" are "almost good" fakes:

can stand a lot of voltage (good) but dissipate less , the chip is good but internal heatsinking is poor, so in extreme cases , **if you have space in the heatsink/thick backpanel** you can safely use, say, 6 of these instead of original 4 and so on or in old 50W Acoustic which used 2 or 100W which used 4 , just don't even approach 100W per pair (which the H ones could handle if well heatsinked).
Same in old Kustom amps.

[gbono]

I believe the issue with the 2N3055H vs E is that the early planar epitaxial devices would exhibit worse BVCEO "latch up" or secondary breakdown and fail under high output levels (and temperatures). If you look at a BJT in BVCEO latch up you will notice that the IV plot shows avalanche,"negative" resistance and secondary breakdown regions.This is for common emitter operation. reference On Semi application note AN1628/D. The SB would cause the part to have a lower SOA than expected.

The Isb for a "modern" MJ15003 is 5A (VCE=50V) hard to believe a 2N3055H has/had a higher Isb rating.

[johnk_10]

just wanted to add that I just got in some RCA 2N3055H (homotaxial) from Electronic Surplus.com for $1.95 each ( RCA - 2N3055HOM - Transistor, NPN silicon. P/N: 2N3055. ) and they work perfectly in my 370. they have almost 1800 of them left in stock so I can recommend them.
I also got some NOS RCA 2N3773's from ebay and they work perfect as well.

[Sinclairx]

Acoustic 370 Power section

just wanted to add that I just got in some RCA 2N3055H (homotaxial) from Electronic Surplus.com for $1.95 each ( RCA - 2N3055HOM - Transistor, NPN silicon. P/N: 2N3055. ) and they work perfectly in my 370. they have almost 1800 of them left in stock so I can recommend them.
I also got some NOS RCA 2N3773's from ebay and they work perfect as well.

Hi

[Sinclairx]

Does anyone have an explanation of the power section theory of operation? I have clipping on one section only of the Wave output on my 370.

[gbono]

At $1.95 these are cheaper than the current 2N3055 or MJ15015 from On Semi...................