Death of Gain Clone

Jump to Latest
#141
Hi,
Thanks for the pointer on Donberg which has all the needed transistors (if a little pricey). If I decide to go ahead this will be the source I will look to.

I like the look of this design, my dad always had Quads and I remember they sounded nice and balanced. I think he still has a 404. This design looks like a simpler implementation, without an old op-amp, which should be good for sound quality.

Unfortunately it will be up against my valve amp, which will be a tough act to follow. If I do go ahead it will be out of interest rather than a great desire to better my already excellent amp.

Shoog
 
#142
lineup said:
If using 2SK389, dual, it does not fit a PCB made for single JFETs
unless this option is at card.
Click to expand...
I have another PCB - exactly for 2SK389 😀

IRF9610 is more linear than 9620;9630
9610 is a good choice - linear and cheap.
The best alternative are 2SJ76;77;78;79, but they are more expensive...

IRF620 can be replaced with IRF520;530;540;620;630;640...any TO-220 nMOSFET with Uds>80V and Id>2A
 
#143
lineup said:
For IRF9610/IRF620 I think there are several options.
Maybe even IRF9510/IRF510 which are 100V
I dont think it makes too much difference using
IRF510,520,530,610,620,630 and the IRF9xxx of the same.
Click to expand...

The N-ch part can be any of a wide variety of MOSFETs as it's operating as a curent source. It's capacitances do not present much of a problem in this circuit.
The P-ch part benefits a LOT from lower Cgs and especially Cgd, and high transconductance. You really want either IRF9610 or 9510 in this place - 510 has slightly higher transconductance and capacitance, which to a point compensates each other, so it will probably not be much of a difference. 2SJ76 looks interesting at first, with slightly lower Cgs and signifficantly lower Cgd, until you see that it's transconductance is 1/10th of the IRF9510 or 9610. This means the current mirror has to swing 10x the voltage on the combined Cgs and gain * Cgd, which effectively makes these capacitances look much higher. On the other hand, 2SJ76 is a lateral MOSFET and is very linear, so if the OL bandwidth does not suffer too much, it may be a good, if expensive, part to use.
 
#147
I’ve breadboarded one channel with slightly different components as I had them at hand.
BC560C has become BC415, BD911/12 is replaced by MJE15030/31 and the 2R2 drain resistors are 2R7 for the moment.
The 120pF is 100pF and the parallel capacitors 3300µF/4.7µF are one single 2200µF. The 10ohm/100nF combination at the output is omitted.
The 4.5µH coil is made from a two times 18 windings of 1.6mm copper wire and an inner diameter of 6mm. Not sure I got the correct value though.
At first I forgot the CE 100nF capacitor on the BC550C which made the amp oscillate.
But it plays and idles at 180mA. Offset is only 30mV.
I’d like to know what makes the coil so important.
I understand it’s, together with the 47ohm, 120pF and 2//1k some kind of filter.
A picture? Sure but it's ugly as hell.

/Hugo 🙂
 

Attachments

  • mvc-446f.jpg
    mvc-446f.jpg
    83 KB · Views: 1,764
#148
These values are very critical, and should be trimmed for optimum settings .. It forms a bridge circuit so that the current dilivered by the currentdumpers is compensated for by increasing the feedback...


here is a schematic which sows the principle .. the 4 resistors (r2, r3, r4, r5) form the bridge.. the ratio of the values is very importand..

http://www.diyaudio.com/forums/attachment.php?s=&postid=793335&stamp=1134915935

now you can exchange 1 resistor with a coil, and another with a capacitor.. and the ratio stays the same (correct?) .. the circuit will still work, and will have less power wasted due top resistors in series with the loiad...
 
#150
I couldn't resist doing a google on the word 'widowmaker' jsut to see what comes up. What I found on the first few pages were:

A film about a russion submarine incident
Another film on something else
A pop/rock (?) group
A sci fi novel
A mountain climbing championship
Some kind of role playing clan
Something Star Trek related (I think), maybe also role playing
A bar/restaurant in the US
A brand of billiard cue cases


Seems mostly harmless. 🙂



EDIT: After posting this I found that all the post this was intended to comment on were moved to Texas, so don't worry if you don't understand why I posted it.
 
#152
Netlist said:
.... BD911/12 is replaced by MJE15030/31 ....
/Hugo 🙂
Click to expand...

MJ15030/31 is often used to drive heavier power transistors.
but are also power transistors with high qualities.

Using like 2-4 parallelled would make one good output
with high performance.
Case is TO220 standard which make easy to find isolator bricks for heatsink.
They are available and does not cost too much.

To follow Netlist example might not turn out too bad.
🙂

Some data:
MJE15030 (NPN) MJE15031 (PNP)
Complementary Power Transistors in TO220 Case

Max 150 Volt CE
Power= Max 50 Watt at infinite heatsink
8 Ampere continous, 16 A peak
fT = 30MHz at 500mA/10VCE

Max junction temp 150 C
Rth = 2.5 C/Watt
 
#155
Netlist
I can not say I undestand how this current dump with inductor works.
I only have a vague idea.
Inductors does not not behave as normal resistors 😀

Nordic
I agree should be built with commonly found semiconductors.
There is no real successful amplifier project using very exotic components.

LM3875 chip amp
Nelson Pass amplifiers
Leach amp
ESP Elliots amps

They all are to 90% based on what people can easily find in real world.
It is also no disavantage to make amplifier work with 2 optional transistors.
And if needed, tell if any resistor needs to have another value for alternative choice.

Some PCB comes with alternative fittings for different transistors.
All does not have same PIN configuration,
and all power transistors doesnt have same distance between pins.
 
#156
Lineup,

I hear what you are saying, but really, what are the exotic components here? Widowmaker makes the point that they are all cheap; there is no audiophile 'grade' stuff in there, and the inductor is hand wound, very similar to the output inductor a conventional amp must use too for stability at HF.

While I concede that 90% of 'popular' DIY amps are conventional, there's room for this design too, especially as it's cheap to make and likely to perform very well.

Cheers,

Hugh
 
#158
AKSA said:
Lineup,

I hear what you are saying, but really, what are the exotic components here? Widowmaker makes the point that they are all cheap; there is no audiophile 'grade' stuff in there, and the inductor is hand wound, very similar to the output inductor a conventional amp must use too for stability at HF.

While I concede that 90% of 'popular' DIY amps are conventional, there's room for this design too, especially as it's cheap to make and likely to perform very well.

Cheers,

Hugh
Click to expand...

Yes, indeed...
Prices in BG - 2SK301R=0.26 euro; IRF9610=0.75 euro...BD911/912=0.45 euro (each)...
 
#159
AKSA said:
Lineup,

I hear what you are saying, but really, what are the exotic components here? Widowmaker makes the point that they are all cheap; there is no audiophile 'grade' stuff in there, and the inductor is hand wound, very similar to the output inductor a conventional amp must use too for stability at HF.

While I concede that 90% of 'popular' DIY amps are conventional, there's room for this design too, especially as it's cheap to make and likely to perform very well.

Cheers,

Hugh
Click to expand...

There are no exotic components in this amplifier.
And that is one good thing about it.
It has potential to be of interest to many.


And generally speaking again, of course one can not and should not restrict an interesting circuit
just to keep design possible for everyone to build.

Exotic and advanced designs can help us to increase knowledge.
Even if most of us will not actually be able to build them.
 
You must log in or register to reply here.
Top Bottom