OnSemi MJL3281A/1302A direct replacement for SK 2SC3856/2SA1492 ?

[JohnSz]

Well, after a year or two, of parts just sitting around, finally getting back to some amp mods. During this time frame, I have misplaced my notes While I can still handle a soldering iron, am extremely rusty on circuit design and modifications, and do not trust myself. I would appreciate help from this forum, while I continue to search for my old notes. Thanks.

The mod in question, is to an Adcom 545II, of which I have two of these amps. I had received 25 pairs of MJL3281A/1302A's from Onsemi, back when I was first contemplating this mod. Are they a dirrect replacement for the SK 2SC3856/2SA1492's? or do they require a couple of minor resistor changes?

The Adcom 545II uses 3 pairs in parrellel, per channel. The OnSemi devices are physically larger, but will fit on the heatsinks, without redrilling. Here is a snipet of the amps schematic. Please note that this is two pages that are pasted together, ie the verticle line. See attached file.


OnSemi MJL3281A/1302A Specs

I suspect, that my original thoughts on this mod, was to increase the power output, especially for 4ohm performance. Can someone confirm this? BTW, the power supplies are being upgraded, as well. Same rail voltage, just higher current capability.

[IMG]
http://johnsaudio.homestead.com/file...5IIAmpSecB.jpg[/IMG]

[N-Channel]

I am familiar with these Output units. Their linearity from 100mA to 7A makes them excellent candidates for this mod.

I am considering this mod for a NAD 2200PE 100W/Ch amp. It uses 6 2SC3281As and 6 2SA1302As per channel. I also have the New On MJL4302As and MJL4281As (350Vce) that I am contemplating using instead.

I think that upgrading your 545II with either of these transistor pairs would significantly improve its operation and sound. Just my humble opinion............

Regards,

Steve

[JohnSz]

N, it was too big to attach, so put on my website. See bottom link, in first post. Sorry.

Adcom 545II Schematic

[N-Channel]

Yep, I got it . I edited my first resaponse before you posted your reply.

[anatech]

Hi John,
I am curious. Why are you going to change the outputs? I'd leave them if they aren't blown. There will be no difference in sound and you do not have to change resistors or biasing to use them. You would have to reset the bias to the same spec.

If you want to improve the sound, you'd get further by matching the outputs, drivers and diff. pair. There are some caps you could change, but the difference in sound would be small.

It's not a bad amp as it sits, better than many for sure. I used to do warranty service on these and many other high end brands. These would put up with all kinds of sillyness before they blew up.

If you have the entire manual, I'd be interested in a PDF. Sold my shop and manuals 7~8 years ago.

-Chris

[JohnSz]

Chris,

Why are you going to change the outputs?

I am assuming that my original analysis was that the OnSemi devices would provide more current, but quick looking at specs, hasn't convinced me that is correct.

There will be no difference in sound and you do not have to change resistors or biasing to use them. You would have to reset the bias to the same spec.

OK, just to be absolutly clear, if I did change them, I just have to re-adjust the bias current (pot), with no other component change, correct?

If you have the entire manual, I'd be interested in a PDF. Sold my shop and manuals 7~8 years ago.

I have the spiral bound service manual, which I purchased off of ebay, a couple of years ago. I believe they still can be purchased, there. FWIW, the 545II manual was more detailed than the 555II manual. email me for other possibilities.

[AndrewT]

Hi Johnsz,
a couple of observations on the part schematic.
Q611 limits VAS Ipk to exactly 2times Iq. There may be situations when slightly more current could usefully be drawn without risking reliability of the VAS.
I suggest you keep the total resistance of the R638+R641 the same but change the ratio by making R641 about 2 times R638. i.e. R638=22r and R641=43r or 44r4(47r//820r) if you can find one!!

Secondly the Vbe multiplier has the adjustment on the collector side. If the pot wiper goes open circuit the bias in the output stage will go sky high.
The pot should be a high reliability type that can NEVER go open circuit. I recommend that you try to find a way to move the pot location to the lower arm of the Vbe multiplier (needs some new component values as well as PCB mods) or more simply, change Q617 to PNP type and again resistor values need to be changed to keep the same range of voltage adjustment.

[anatech]

Hi John,
The On Semi devices have a higher SOA area, but you haven't blown the amp up so it doesn't matter. The newer devices also get rid of heat better, also not a concern at this point.

You can use the new devices in the same circuit without doing anything but readjusting the bias current.

Anthony's point about the bias pot is valid. I haven't seen a failure yet on these due to this, but it will start happening at some point. That's a long standing complaint I have on many amp designs. The pot is unlikely to go completely open (that would be good, the fuses would blow quickly). Unfortunetly, the bias current will just increase to a high level and trip the thermal - or blow the outputs. I haven't seen this situation yet in these amps.

As far as increasing the peak current from the VAS, don't know. It would only matter when driving difficult loads. You could try it and see. I think the bias pot issue deserves more consideration.

-Chris

[anatech]

Hi John,
Your email is turned off in your profile. Your web address comes back invalid too.

-Chris

[JohnSz]

Chris,

Email is turned on and corrected error in website address. Thanks for pointing it out.