but its hard to argue with the convinence and performance of the ad797 or lt1028
Even harder to argue with a cheap amorphous Lundahl
Which model and how much?Even harder to argue with a cheap amorphous Lundahl
Tom
EE
There is a variation of the Leach head amp at http://www.st-and.demon.co.uk/Audio/700/700page5.html . For a head amp you need transistors with very low base spreading resistance. This usually means power transistors or special (high priced) transistors designed for low base resistance.
Regards,
Ray
Regards,
Ray
tmblack said:
Which model and how much?
Tom
EE
i have a 9206. don't remember the price but it was lunch money ($100?). Yes, there are better transformers but not under $300-$400.
Fantastic result!I measure the noise with my sound card and a ultra-low-noise pre amp based on 4 parallel 2SK369 jfets. It has a noise floor of less than 0.35nV/rtHz
Can you tell us what software is used to make the measurement or maybe there is a link to find that out?
Tom
EE
not on topic exactly ...
ray:
that was an interesting link to surf around in. thanks for sharing it.
mlloyd1
ray:
that was an interesting link to surf around in. thanks for sharing it.
mlloyd1
rayfutrell said:There is a variation of the Leach head amp at http://www.st-and.demon.co.uk/Audio/700/700page5.html .
...
Regards,
Ray
rayfutrell said:There is a variation of the Leach head amp at http://www.st-and.demon.co.uk/Audio/700/700page5.html . For a head amp you need transistors with very low base spreading resistance. This usually means power transistors or special (high priced) transistors designed for low base resistance.
Regards,
Ray
This circuit is so pathetic soundwise it's not even funny. I built a variation at a time when i was too young to understand that huge electrolytic caps are never too kind to microvolt level signals.
A single common source fet is so much better. But it's always nice to have choice
I write my own stuuf in LabVIEW, I you really want to, I could post a .zip file of the executable. But you also need the runtime library of about 30Mb.
The trick I use: use a whole second of measurement samples for the FFT, so your resolution is 1 Hz. It's much easier to calculate the nV/rtHz.
Dazyweb spectrum analyzer is also nice, but it has a few bugs and FFT samples are fixed to 16000.
The trick I use: use a whole second of measurement samples for the FFT, so your resolution is 1 Hz. It's much easier to calculate the nV/rtHz.
Dazyweb spectrum analyzer is also nice, but it has a few bugs and FFT samples are fixed to 16000.
Indeed matlab can do the same thing and if you're a bit experients with matlab : go for it!
In matlab if you take 44100 sampes (1 sec.) and take the square root of the output FFT (only of the 2nd -to - 22050th number) I think you get V/rtHz.
letme know how you're doing.. I would be interested in a matlab solution!
In matlab if you take 44100 sampes (1 sec.) and take the square root of the output FFT (only of the 2nd -to - 22050th number) I think you get V/rtHz.
letme know how you're doing.. I would be interested in a matlab solution!
What's the point of having 2 complementary transistors operating in parallel in these amplifiers?
ummm where to start, too hard for me to put into words, edit - nearly pressed send - here is one reason: we have to float the base(s) because we do not want offset on the cartridge, the symmetrical circuit on turn on would probably put 'less thump' through the cartridge than a single ended one. Maybe the designer just likes symmetry. Possibly ways around it, perhaps a jFET version, smaller caps...
a couple comments:
It is a ingeneous, simple, elegent solution to the problem.
I like the implementation of Leach's circuit better than the '700' (thanks for the interesting link) since it avoids the cartridge signal going through a large elec capacitor.
Unfortunately there are still electrolytics - grounding the bases, and coupling the output.
Electrolytics, possibly ok to ground things but signal though them? I know there some good sounding units out there with electrolytics, but it is not my path.
Bill
Electrolytics, a necessary evil at times!
The Leach circuit uses two transistors to reduce 2nd order distortion.
It can be improved by dropping the supply voltage from 9V to 3V or less. After all the output level is just a few millivolts.
If you simulate it, you will see that the gain is strongly dependant on supply.
This has the side effect that AA batteries can be used and give a lot better lifetime than PP3s
2N4401,4403 are the best cheap transistors as they have reasonable gain and very low base spreading resistance, as they are intended for switching applications. This means excellent performance at 100R source impedance.
It can be improved by dropping the supply voltage from 9V to 3V or less. After all the output level is just a few millivolts.
If you simulate it, you will see that the gain is strongly dependant on supply.
This has the side effect that AA batteries can be used and give a lot better lifetime than PP3s
2N4401,4403 are the best cheap transistors as they have reasonable gain and very low base spreading resistance, as they are intended for switching applications. This means excellent performance at 100R source impedance.
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
- Home
- Source & Line
- Analogue Source
- Leach MC Headamp