Darlingtons used as Audio output devices....?

[peteb2]

OK back in the 80s i was right into building high power amplifiers, the bigger the better and that meant WATTS of power... mainly for the Disco 'Industry'... DJs would find some understanding person who'd build them a plinth with two turntables, a basic mixer, a huge amp internal and maybe a coloured light show plus some dodgey big speaker bins set up! 

There were plenty of designs around and i must have munched through 100s of 2N3055s... the go to transistor of the day...

I'd an interest in HiFi because CDs had recently arrived but were horrendously expensive and no one had a player.

Then one of the hobby magazines published a design using MOSFETS and the game changed. The cost of making WATTs of Audio Power came down and for a while i was munching through quite a few of the new 'weird' things... ahhh and blowing up a few i might add! (they self oscillated and even made a mess of the upper VHF radio band if you didn't watch it)...!

I read articles in HiFi magazines on how the MOSFET gave a better tone, similar to tubes (valves). I honestly did not notice that aspect but i recall seeing the hundreds of AudioPhools spouting their preference for Bipolar transistors only.... which i thought strange because even then the holy grail were tubes....

Then my career changed, bye-bye audio stuff hello TV Broadcast cameras.... (another story).

Cut to today and someone brings me a knackered Tuner-Hometheater Amp made by Sony that had been to so many parties. It was being used as straight amp or sometimes the FM radio was simply left running. They used it on two speakers only, joining them to the FRONT marked terminals. Worked for years said the owner.

This is a 7.1 surround sound thing meaning there's 7 main speaker terminal sets and another driver for a SubWoofer and it all weighs a ton. Obviously dropped a few times too going by the dents. I class it as genuine dumpster stuff but the owner 'really needs it".

It randomly goes "SPLATT!" shuts down, puts up the words in in very dim display ("RECOVERING").

So i tore it down and discovered where what must be 90% of the worlds aluminum reserves have ended up. The thing's heatsink is MASSIVE. And bolted to it a pile of fat transistors and i go hunting on Google and find the pdf for these transistors.

I made the assumption the output drivers would be huge Bipolar Specials or MOSFETs like when i found that sort of trend in the mid 1980s but NOPE.... they are DARLINGTONS!!!

Talk about live under a rock. Darlingtons in my day were always power-switching of loads like driving ignition packs on 4 cylinder motorcycle engines or H-Bridge arrangements on 3 phase motors... I never saw an audio circuit using Darlingtons because they would have lots of extra noise and AudioPhools would start on 10,000 rpm head spins over the shocking sound quality....

So here i am looking at a row of big 2SB1647/2SD2560 TO-3P 150V 15A devices.... 7 individual channels big push-pull driver output stages... Man what i could have made for the DJ guys back in the John Travolta/BeeJees 'Saturday Night Fever' days if this stuff had existed... 

[Cyberdragon]

Wow, you have been under a rock. Darlingtons have been used as output drivers since the 70s. I have a 70s set that has "DARLINGTON POWER PACK" in huge letters across the output chips.

[duak]

I recently transplanted some darlingtons into a 22 KW, 3 phase motor drive for a CNC lathe.  They're Mitsubishi parts from about 1984 and rated for 400 V, 150 A and are switched at a couple of KHz.  I don't how linear they are or what their ft is but they'd probably be fine for an industrial strength sub amp.

Cheers,

[T3sl4co1l]

Heh, old industrial drives used transistors from time to time (and perhaps SCRs more often, say during the 60s to 80s?), usually being a triple Darlington device.  The last transistor is huge, high voltage and low hFE, and doesn't turn off until the first two turn off completely -- turn-off time was measured in multiple microseconds!

Tim

[Circlotron]

You need to post some pics of this doomsday amplifier!

[Zero999]

The truth is, as long as an amplifier is well designed and it isn't driven into clipping, it will sound the same, as any other, irrespective of whether the output devices are MOSFETs, BJTs or thermionic valves. No one will be able to tell the difference, when subjected to a double blind test. If anyone tells you otherwise, they're mistaken.

MOSFETs will be better, than BJTs for higher output voltages, because they can drop a higher voltage, at a higher current, without going into secondary breakdown. The safe operating area is greater. BJTs are better at lower voltages.

Guitarists like the old thermionic valves/tubes because like the sound they make when they're driven into clipping. Newer solid state designs, with BJTs and MOSFET output devices, clip harshly, causing a nasty buzzing noise. It's easy to hear the difference between a solid state and vacuum amplifier, when they're over driven, but that's not what you should do, if you desire high quality, unadulterated sound.

[David Hess]

Bipolar designs all rely on multiple stages to get enough current gain.  Darlingtons and the three transistor equivalent are the most common but there are other configurations which are sometimes used.  I would only expect integrated Darlingtons to be used in cost sensitive designs because they lack the high frequency performance of a discrete Darlington.  But even with that disadvantage, a proper design can still achieve low distortion.

[peteb2]

Wow, you have been under a rock. Darlingtons have been used as output drivers since the 70s. I have a 70s set that has "DARLINGTON POWER PACK" in huge letters across the output chips.

Yes agreed... but i never ever had anything in my experience that was specifically HIGH POWER AUDIO using big Darlington transistors... (back in the late 70s / early 80s). At the time i recall is was almost obsessed, reading as many articles as i could from whatever electronics magazines i could get my hands on. It seemed every month there was another design and in the hobby mags there was another project. Not once did i see a rant on say a push-pull using Darlington devices though.  I am certain there would have an obvious recognition of using Darlingtons and i am certain there'd have been associated ravings so all i can imagine is that your 1970s "DARLINGTON POWER PACK" is by a high-end specialist company.... Kind of interesting indeed!

[Red Squirrel]

First thing that came to mind. 

[Cyberdragon]

Not really, just a Panasonic SA-400. It uses STK0040 hybrid modules (Sanyo?).



They made a range of different hybrid modules. Mine is only 45W per channel, but they made up to an STK0105, whatever that power ouput is.

[IanMacdonald]

Use of darlingtons by maimstream manufacturers was quite rare. The usual choice for pro audio gear was the 2N3773 in a quasi complementary arrangement. This is a slow transistor and doesn't yield yield very open or clear treble, but at a disco who cares? The BDY56-58 series yield noticeably better sound quality, and have less of a problem with SOA.

As for MOSFETS, they seem like a good idea in theory but have two major issues, one that the SOA is usually much moire limited -as in can handle a jawdropping 600V or 50A, but only 2.5A and 30V when applied at the same time.  Secondly, that they are a lot less linear then BJTs. Whereas the gain of a BJT drops off moderately with increasing current, the gain of a MOSFET increases dramatically at higher currents.  That means in linear circuits they are likely to deliver worse THD levels unless high standing currents are used. (and they usually are, making for an inefficient amp at low volume levels)

[Zero999]

As for MOSFETS, they seem like a good idea in theory but have two major issues, one that the SOA is usually much moire limited -as in can handle a jawdropping 600V or 50A, but only 2.5A and 30V when applied at the same time.

I think you've got that backwards. MOSFETs generally have a wider safe operating area, than BJTs and are capable of handling far higher currents, at the full voltage specification. Check out a few data sheets for high voltage MOSFETs and BJTs and you'll see what I mean.

[peteb2]

Under half an hour left of 2017 as i type this. Managed to poke my iPhone into a crevice of this Aged Sony home theatre thing and took a shot... this is just one channel. Apologies it's such a kooky angle of view.... Looks like some kind of big possible hybrid module doing the pre-driving and then these two Darlingtons the rest.... there's 14 of the things on a huge heat sink....

Happy New Year everyone!

[David Hess]

As for MOSFETS, they seem like a good idea in theory but have two major issues, one that the SOA is usually much moire limited -as in can handle a jawdropping 600V or 50A, but only 2.5A and 30V when applied at the same time.

I think you've got that backwards. MOSFETs generally have a wider safe operating area, than BJTs and are capable of handling far higher currents, at the full voltage specification. Check out a few data sheets for high voltage MOSFETs and BJTs and you'll see what I mean.

For obsolete lateral MOSFETs this is almost the case but not for vertical MOSFETs.  At low currents, the temperature coefficient of the gate threshold voltage reverses so current sharing is not enforced even on a single die.  At low drain voltages this is irrelevant because power dissipation and heating are low but at high drain voltages, it results in thermal instability and current hogging.  Most MOSFET datasheets do not bother to show this in their safe operating curves because they are intended for saturated operation where this does not apply.

https://www.fairchildsemi.com/application-notes/AN/AN-4161.pdf
http://www.irf.com/technical-info/appnotes/an-1155.pdf
http://www.ixys.com/Documents/Articles/Article_Linear_Power_MOSFETs.pdf

[Zero999]

As for MOSFETS, they seem like a good idea in theory but have two major issues, one that the SOA is usually much moire limited -as in can handle a jawdropping 600V or 50A, but only 2.5A and 30V when applied at the same time.

I think you've got that backwards. MOSFETs generally have a wider safe operating area, than BJTs and are capable of handling far higher currents, at the full voltage specification. Check out a few data sheets for high voltage MOSFETs and BJTs and you'll see what I mean.

For obsolete lateral MOSFETs this is almost the case but not for vertical MOSFETs.  At low currents, the temperature coefficient of the gate threshold voltage reverses so current sharing is not enforced even on a single die.  At low drain voltages this is irrelevant because power dissipation and heating are low but at high drain voltages, it results in thermal instability and current hogging.  Most MOSFET datasheets do not bother to show this in their safe operating curves because they are intended for saturated operation where this does not apply.

https://www.fairchildsemi.com/application-notes/AN/AN-4161.pdf
http://www.irf.com/technical-info/appnotes/an-1155.pdf
http://www.ixys.com/Documents/Articles/Article_Linear_Power_MOSFETs.pdf

I didn't say that the safe operating area is not a limitation with MOSFETs, just that it's generally better than BJTs, especially at higher voltages. Check out the data sheet for a typical HV MOSFET such as the IRF740B. Can you find a BJT with a safe operating area of 400V at over 300mA?
https://www.mouser.com/ds/2/308/FairchildSemiconductor_1614842276095-1191888.pdf

[Yansi]

Some mosfet technologies are a bit limited regarding DC operation but, Infineon still makes fully DC  operable high voltage mosfets.

For example IPW90R1K0C3. I have used those in my high voltage power supply.

Some manufacturers (like BK Precision) think it is okay to use non DC rated mosfets in linear mode. Just look up one of the last Shahriars video, where he repaired the blown HV PSU. There was a STW15NK90Z used.

[Circlotron]

Some mosfet technologies are a bit limited regarding DC operation but, Infineon still makes fully DC  operable high voltage mosfets.

For example IPW90R1K0C3. I have used those in my high voltage power supply.

Just looking at that datasheet, the output characteristics graph shows 5 amps and 4 volts which means RDSon is 1.25 ohms. The SOA graph though shows 5 amps and 13 volts on the RDS limited part of the graph, so this would indicate a RDSon of 2.6 ohms. Why the discrepancy? I noticed the same thing with a SPW47N60C3 datasheet I was looking at.

[Yansi]

You need to take into account Rds change with temperature. If I am not mistaken, the output charachteristic is measured using short pulses, so the whole is measured at the amb. temperature, where the SOA is specified as Tc=25°C, not Tj=25°C. 

The SOA is the limit, where Tj = Max (150°C) and Tc = 25°C - given by the thermal resistance of the package.  You then do the derating of the current depending on the real case temperature.

If you look at fig 7  the Rds vs temp graph, it shows up to 2.7ohm at 150°C .

//EDIT: Note that there are two output characteristic graphs! One for Tj 25°C and second for 150°C. Which one have you compared the SOA to?

[madires]

I still got some old amp modules based on a pair of a BDV64 and a BDV65 doing 100W pure sine wave without any problem. IIRC, they were used for active speakers and satellite/subwoofer combos.

[Zero999]

I still got some old amp modules based on a pair of a BDV64 and a BDV65 doing 100W pure sine wave without any problem. IIRC, they were used for active speakers and satellite/subwoofer combos.

100W should be no problem for most transistors. Assuming an 8 Ohm load, it's only peak output voltage of 40V, which, with a Darlington output stage, would require a supply voltage of at least 43V, with 48V being optimum. When the voltage is at the peak, the current will be highest, but the voltage across the transistors would be only 10V.

Secondary breakdown is more of an issue at much higher powers, several hundred Watts, hence why you typically won't find an amplifier with a BJT output stage, rated to above 500W RMS (real power, not fake power).

[Yansi]

100W into 8ohm amplifier requires definitely more, than 43V. You need to also account for the voltage drop accross emitter resistors, and do not forget to account for the limits of the VAS of the amplifier.  48V, or maybe even 50V may be reasonable values of the unloaded voltage on the filter bank. But definitely not 43. Mind also the voltage ripple on the filter bank when loaded!

[Yansi]

Secondary breakdown is more of an issue at much higher powers, several hundred Watts, hence why you typically won't find an amplifier with a BJT output stage, rated to above 500W RMS (real power, not fake power).

Why not?  500W in class AB is still quite usuall, certainly not ideal.

With BJTs, you can make even 2kW/4ohm if you want. We haven't specified any limits on the output topology, did we?  Generalizing that you can not make more than 500W using BJT is nonsense.  It is only true for power stages using only paralleled BJT devices.  But BJTs can also be series-paralleled to achieve much much bigger output powers.

In fact, most of the PAs used in last 10 years befor the class D took the PA market, almost all of those high powered amps used BJTs using series-parallel combinations and multiple supply rails to obtain higher powers and efficiencies.

[Zero999]

Secondary breakdown is more of an issue at much higher powers, several hundred Watts, hence why you typically won't find an amplifier with a BJT output stage, rated to above 500W RMS (real power, not fake power).

Why not?  500W in class AB is still quite usuall, certainly not ideal.

With BJTs, you can make even 2kW/4ohm if you want. We haven't specified any limits on the output topology, did we?  Generalizing that you can not make more than 500W using BJT is nonsense.  It is only true for power stages using only paralleled BJT devices.  But BJTs can also be series-paralleled to achieve much much bigger output powers.

In fact, most of the PAs used in last 10 years befor the class D took the PA market, almost all of those high powered amps used BJTs using series-parallel combinations and multiple supply rails to obtain higher powers and efficiencies.

Sorry, I should have been more specific, I was talking about for an class AB amp, driving a 8 Ohm load and I didn't say it's not possible to design a high powered BJT amplifier. It just starts to make more sense to use MOSFETs, over a few hundred Watts.

To get 500W RMS into an 8 Ohm load, you need a peak voltage of 90V, which means a 100V supply. If you look at the safe operating area graphs of most BJTs, you'll find they can't take high currents at voltages of over 50V for long. Of course you can use a cascode, but then it's easier to use MOSFETs.

For higher power amplifiers, don't use class AB. Use class G or H, which has multiple power supply voltages, that way the excessive voltage drop across the output devices can be avoided.

[Kleinstein]

With high power it need high supply voltage. At voltage significantly above 50 V the choice of BJTs can be limited due to second breakdown, the reduces the usable power per transistor. There are a few transistors good for 150 or 200 V operation, but there start to get expensive. However many MOSFETs have a similar limitation to second breakdown - just a different name. So one also has to look for special rate MOSFETs.


Anyway, who need so much power for a single speaker ?

At very high power it can be more effective to switch over to class D or at least a class H (kind of similar to cascode with a second transformer tap) amplifier. This reduces the demand on cooling - thus better SNR since one might get away without a noisy fan.

[Dr Wobble]

HH made some decent amps, the TPA 25, 100 etc that used Darlington's in the OP stage, they also of coarse made huge heavy power amps using mosfets, the "V" range, up to 900w I think..

I recently built a 70w amplifier that uses Darlington's ( Mospec BDV66/67B) out of an old Elektor mag. It's a 70w into 8 ohm and sounds quite nifty. Mind you anything you build yorself sounds good, once you've stopped it oscillating that is.The nice thing about Darlington's is that they are quite over looked by those that ascribe magic powers to certain mosfets, Vfets and what have you and tend to be cheaper as a result.

A.