Transformers and transistors, can I build an amp?

[ric-paul]

Hi,

Please shoot me down if this is not the place or not the way to ask this!

I have stuff that's doing nothing. I'd like to build an amp or something else useful with it.

Can you guys think of a useage?

Toroid: 250VA 48V~ transformers, 4 of them, so 1kW (matching ~1980 (!) capacitors (2 sets of 36,000uF 75V))

Toroid: 750VA primary 200-220-240, secondary 0-52-60-85V transformer (thinking of using this for the BTL TDA7294 amp, giving me 2x150W@8 out of +/-25V (85-52V~ winding) or 2x170W@16ohm out of +/-39V (0-52V winding), virtual earth)

Transistors, among others:
Ten BDX67B (NPN 150V/16A Darlingtons).
Only one BDX66B (complementary PNP version of the BDX67).

One pair of Motorola 2N5883/2N5885 (NPN and PNP, 60V, 25A, 200W)

Twelve Moto/RCA 2N6678 (NPN 300V, 15A)

Eight MTH40N10's, N-channel mosfet 40A, 100V

Any of these pairs (5883/5885 combo?) suited to turbocharge a TDA7294?

I don't have the ability to make or draw PCB's, nor the skills. My last endeavor (ORCAD/ULTIBOARD, think 1995?) resulted in a blown TDA7294 and a fried voice coil...

Thanks for any ideas even if you're just gonna talk me out of it...

 

[Tekko]

The BDX66/67 darlington transistors sound like those used in old "hifi" amplifiers, and should work fine for a basic blameless running from 2x35VDC.

This circuit is simple to build and is still commonly used in many consumer level amplifiers and surround receivers: http://i.imgur.com/AjAWI.png

The single pair of 5883/5885 would only offer up to 25-50W into 4 ohms due to the low Vce, which would work decently for a 6.5-8" subwoofer for a small computer stereo and similar.

The twelve 2N6678 could make a nice version of quasi's nbip300, or a half power stereo version running off half the supply voltage.

The mosfets you'll have to check if they even spec a soa for linear operation, if not theres prolly not much point in even trying. They could work nicely for a high power class D subwoofer amp though, provided the gate charge and turn on/off speed is reasonable.

 

[djk]

"The twelve 2N6678 could make a nice version of quasi's nbip300,"

Those only have an SOA of 25W at 100V, they are a switching transistor, not suited for an audio amplifier.

 

[ric-paul]

Thanks Tekko! Not that much I can use out of the transistors I have. Too little sets of complementary ones...

And djk, how about eight MJ802's (100V, 30A, 200W) instead of the 2N6678's? They come out of a single sided amp (with 100V outputtransformer), as do the 48V toroids...

Well, if I can get my amp fixed, find my contact-spray and finish the TDA7294 boards, I will already have way more amps (10 or so) than I know what to do with 🙂

Thanks guys.

 

[Tekko]

Its not single ended but rather quasicomplementary.

the MJ802's with four per channel would make a good 50-100W into 4 ohms quasi amp, you can use quasi's nbip300 schematic as a base.

For that, i'd use a psu with 2x35-40VDC.

However if you feel you already have more amps than you know what to do with, you may consider selling the transistors.

 

[ric-paul]

Its not single ended but rather quasicomplementary.

Since posting I searched the design and read up on it. I should have done that before assuming and writing it was a single ended design...🙄

As for selling, I have most of them in the swap section. But most are used and very few complementary pairs, so I doubt there is that much demand for them.

Trying to clear out, as I had a close call last year in the hospital, and don't wanna leave the wife with all this 'crap' 😛 (Things are well now, so I have some years left to get rid)

 

[Tekko]

Myself i am surrounded by amp projects, a p channel mosfet amp(needs a chassi), a complementary lateral mosfet amp(my primary every day stereo rig), a 2.1 channel amp with full complementary output(a work in progress), a N channel mosfet amp(still prototyping and testing), a high power full complementary subwoofer amp(also prototyping and testing), and on top of that more complementary transistors still new in box and/or semiassambled output stages.

Building audio amplifiers is a bit of a passion for me.

 

[AndrewT]

Do you have any dual secondary or centre tapped secondary transformers to create dual polarity DC supplies for the modern style of Output DC coupled amplifiers?

If not, then look at using a single polarity supply with a low impedance virtual POWER GROUND. Quad shows one method.

 

[ric-paul]

If not, then look at using a single polarity supply with a low impedance virtual POWER GROUND. Quad shows one method.

I saw the Quad virtual ground in another thread/answer. What I found strange is that there are quite high resistors used.

My idea was to use some 250ohm 5W resistors between ground and the + and - supply, but the Quad design uses 2k2 resistors before the transistors 😱, so even if the transistors are fully 'open' or conducting, it's still only 2.2k between Vcc and GND...

My gut told me you had to be able to equalize reasonably high currents, but provided the Quad is not a BTL design (and since I seem to remember you could bridge it, I guess they are not!!!), this goes against my instincts. But science is often more precise than my instincts, so I'll figure Quad thought this through...!

My plans:
750VA transformer with (after rectification) 50VDC output (~10A ac loadable as not all windings are used...) will be used on a BTL 2x150Wrms @8ohm TDA7294 amp. Virtual ground for +/-25V, even an option to up this to +/-36V for 2x170Wrms@16ohm using different windings.

250VA 48V~ should be OK for a stereo TDA7294 amp (2x80W@8ohm)
I have 3 of those boards and 4 of those transformers. but it indeed needs virtual ground...

And there's two 300VA toroids with 2x13.5/2x27Vac, 2x15/30AC or 2x16.5/33VAC selectable.

Those could work for a stereo TDA7294 amp or for a turbocharged TDA7294 with outputtransistors...

Well, pondering about the options also keeps me busy...I'm not always that fast in implementing things (mounting my solarpanels should have a higher priority for one...)

 

[djk]

"Toroid: 250VA 48V~ transformers, 4 of them, so 1kW (matching ~1980 (!) capacitors (2 sets of 36,000uF 75V))"

I would use these for two 48-0-48 supply, ±67V DC. A dual-mono amplifier with these would be 320W/4Ω, maybe a bit more.

New semiconductors are so much better and more affordable than the older ones that I would not try and re-use any of the old semiconductors, although the eight MJ802 would do well in a 50A 13.8V supply (the 200W SOA starts to slope above 30V).

"Toroid: 750VA primary 200-220-240, secondary 0-52-60-85V transformer (thinking of using this for the BTL TDA7294 amp, giving me 2x150W@8 out of +/-25V (85-52V~ winding) "

A QSC style output stage would work with this transformer with 85V AC giving 120V DC. This would make a single 350W/4Ω, 600W/2Ω amplifier (good for a subwoofer).

"(matching ~1980 (!) capacitors (2 sets of 36,000uF 75V)"

Due to the age I would probably not use these. I have cases of this kind of stuff I would give away to anyone willing to collect in person.

 

[ric-paul]

I would use these for two 48-0-48 supply, ±67V DC. A dual-mono amplifier with these would be 320W/4Ω, maybe a bit more.

A QSC style output stage would work with this transformer with 85V AC giving 120V DC. This would make a single 350W/4Ω, 600W/2Ω amplifier (good for a subwoofer).

If I ever finish my plans, I will need something to drive two Aliante subs in my home (350Wrms@4ohm), so sounds good 😀

(but my 160Wrms JVC amp would be plenty too since I don't exactly live in the middle of the woods...)

although the eight MJ802 would do well in a 50A 13.8V supply (the 200W SOA starts to slope above 30V).

Cool, I guess some good old 14V~, 50A transformers I have access to wouldn't hurt for this design either. 😀 And three 90A diodes on a heatsink from a Winchester powersupply from the 80's...

"(matching ~1980 (!) capacitors (2 sets of 36,000uF 75V)"

Due to the age I would probably not use these. I have cases of this kind of stuff I would give away to anyone willing to collect in person.

I know, I've had them for ~20 years 😀 I've offered them for free many times. Can't bring myself to throw them out. It's a pity they may be too old really, 78,000, 36,000 and 5x37,000uF, that's almost +/- 150,000uF@75V😀

 

[AndrewT]

reform them and measure the leakage. Then, if you have the resources, measure the esr.

 

[wg_ski]

If they've been sitting around without bias in temperature controlled environment - and don't have that many running hours on them they may be useable. After re-forming, of course. If they've been alternately baked and frozen every year for 20 years they probably won't form in any reasonable time and you'll have to pitch them.

 

[wg_ski]

"The twelve 2N6678 could make a nice version of quasi's nbip300,"

Those only have an SOA of 25W at 100V, they are a switching transistor, not suited for an audio amplifier.

He could make an original Flame Linear 400 clone. Complete with flames. It would take an 8 ohm load, in light service - but heaven help you if you start partying.....

 

[Tekko]

Or a Class A amp running from a single 24-35V rail.

 

[djk]

"He could make an original Flame Linear 400 clone. "

The MJ410 used as the driver in the PL400 has 70W SOA at 70V, 50W at 100V. The outputs were more robust (I don't have the curves for them).

Switching transistors work for switching, not for linear use.

 

[wg_ski]

I thought the XPL909 *was* a switching transistor - not much better than a 6678. The Fairchilds were better, but still couldn't hold a candle to a 15024.

 

[djk]

So you think the drivers were more robust than the outputs?

I have successfuly use the 2N3773 in the PL400, it has 150W SOA at 100V, higher than the MJ15024.

 

[wg_ski]

Too bad they didn't come with 3773's. And too bad ric-paul doesn't have any laying around.

 

[djk]

Too bad they didn't come with 2N6259s either, 250W SOA at 100V.

Altec used them in the 9440.