Here is a Challenge

[latala]

Here is a Challange

Just suppose the pnp transistor had not been perfected/produced . Could any of you design a Dc coupled amplifier using NPN devices only !!!
Target power is 30 watts rms or more in 8 ohms at less than 0.1 % distortion midband
I have produced a no of small class A designs with succes
You can use any technique multiple supplies etc any thing goes but no PNP devices.
It seems to me that most of the designs I have seen on this site are all generic and feel that the challange will stretch our minds a little further
Regards Trev

 

[kASD]

Re: Here is a Challange

Just suppose the pnp transistor had not been perfected/produced . Could any of you design a Dc coupled amplifier using NPN devices only !!!
Target power is 30 watts rms or more in 8 ohms at less than 0.1 % distortion midband
I have produced a no of small class A designs with succes
You can use any technique multiple supplies etc any thing goes but no PNP devices.
It seems to me that most of the designs I have seen on this site are all generic and feel that the challange will stretch our minds a little further
Regards Trev

I love to break challenges and also the people who set the challenges

I can build with both NPN or N-channel FET as well!!!!!!!

All 100% Single polarity.

 

[latala]

Go0 on then Trojan

Lets see a concept circuit I dont mind if its fet or bipolar it could even contain vacum tubes any thing but PNP devices .
Please do not thing I am being smart or agressive its just I have not myself been able myself to come up with a dc coupled circuit that would have low and stable offset!
regards Trev

 

[lumanauw]

Isn't all tube amps using output transformers are all using Nch devices (as there is no P tubes)?

 

[latala]

Tubes don,t have to use output transformers!
What about the futterman type circuits they don.t use transformers
I desined a hybrid that was dc coupled that used 4 transistors and 4 tubes that was dc coupled and has an offset of 50 mv at the spk terminals.
Three of the transistors were pnp though so no prize for me with that 1
regards Trev

 

[teemuk]

The interstage transformer coupled designs (pre early 70’s) commonly used only PNP transistors because no decent high power NPN ones we available. The transformer performed the required phase inverting that sort of turned the PNP device into NPN (think Sziklai pair). You can easily take one of these designs and convert it to use NPN transistors instead, but I suppose you’ll have plenty of problems finding the required iron nowadays.

 

[Elvee]

Hi

The output part is relatively easy: you just have to use something based on the "virtual complement" concept, introduced by Bengt Olsson (Electronics World, Dec 1994). The idea had been used in the tubes days, but without specific denomination (quite understandably).
The phase splitter is equally easy.
The real problems begin with the voltage shifting: if you want the DC coupling, this means no capacitors, and without PNPs, the only solution is "rigid" level shifters, using zeners or similar devices. In tube op-amps, it was done with gas-filled tubes, but it is not very satisfactory from the PSRR point of view.
Any idea of improvement on this point?

 

[SpittinLLama]

The chip amps from National, think LM3886, are all NPN and the DC offset is manageable with no input capacitor or feedback cap. The popular Gainclones people built can be done without caps for more pure sound. I forget what the DC offset is but it is not typically a problem. The chip amp forum can be searched for such.

-SL

 

[Conrad Hoffman]

LM3886 all NPN?? Where did you get that idea? The datasheet shows lots of PNP devices in the "equivalent schematic".

 

[latala]

The output part is relatively easy: you just have to use something based on the "virtual complement" concept, introduced by Bengt Olsson (Electronics World, Dec 1994). The idea had been used in the tubes days, but without specific denomination (quite understandably).
I vaguely remember reading this article but did not that only refer to the output fets I seam to remember that it used pnp transistor however I am not sure
I agree the main problem is level shifting
Also I stated earlier that I built a hybrid but I used 1 more npn I forgot the bias
We could level shift using opto device ?
Regards Trev

 

[megajocke]

Is a DC servo (all NPN of course) allowed or does that count as AC coupling?

 

[Radioman62]

Much talk but no schematic yet

 

[SpittinLLama]

sorry, was thinking output stage only. My bad.

-SL

 

[latala]

I see no problem with a dc servo ! Anything other than complimentary devices in this case I have said NPN only
The amplifier can have a input capacitor and a feedback to gnd capacitor
the spk is to be dc coupled to the output stage
I have as yet no schematic of my own the idea is to get away from what are esentlialy only variants on standard schematics!
level translation could be via diodes ,zeners/opto devices
capacitor or even galvanic
Just no pnp devices
regards Trev

 

[Tim__x]

Whoop, this is the result of ten minutes. The output stage could be drastically improved and scaled for a typical load, but that's the easy part, this is just an example topology.

Around 40db open loop DC gain climbing to 80 some db, currently uncompensated and marginally stable with a 260mhz gbw.

 

[Steve Eddy]

I see no problem with a dc servo ! Anything other than complimentary devices in this case I have said NPN only
The amplifier can have a input capacitor and a feedback to gnd capacitor
the spk is to be dc coupled to the output stage
I have as yet no schematic of my own the idea is to get away from what are esentlialy only variants on standard schematics!
level translation could be via diodes ,zeners/opto devices
capacitor or even galvanic
Just no pnp devices

By NPN only, are you being literal in that you only want NPNs as in bipolar NPNs, or does that also include N-type FET devices? If so, here's a concept for you.

An externally hosted image should be here but it was not working when we last tested it.

se

 

[latala]

Tim X I can not open your attachment all I get is a wiring list ?
please give us a diagram !!!
The other one by steve eddy is interesting ! and very unexpected
lots of iron though! still its food for thought
I shall submitt my concept sketch tonight
regards Trev

 

[DRC]

This OP stage is mostly NPN !!

The inner diff pair / current source can be readily replace with NPN but this leaves a level shifting problem - maybe a current sourced NPN E-F and a resistor to shift the level ?

 

[Tim__x]

It was an LT Spice simulation file, earlier I wasn't on a computer with image editing software installed.

Obviously I ignored the power goal, and only focused on making a workable topology with only NPN devices, I chose to go DC coupled throughout including the input. The two caps in the second stage serve only to enhance AC gain, there is still considerable DC gain and complete control of the output stage down to DC.

 

[Tim__x]

I should note: it's VERY unpolished, it is just a work of ten minutes and ONLY an example of topology, not a fleshed out buildable schematic. DON'T build it as is..