Nick G modified Lentek Head Amp

[Light Dependant Resistor]

Oh dear..... It's not class C, it's class A. The "cartridge signal" couldn't turn anything on or bias anything....

Common base would work equally well yes, and many variants of the Lentek circuit and similar common base ones are extant. Common base would make it virtually impossible to change cartridge loading though, and the gain would vary considerably with cartridge source resistance. I've not simulated it but a glance at the circuitry says to me that the standard Lentek input loading is roughly 90 Ohms.

Dual rail supplies are what I was hinting at to Marco earlier but getting the full benefits are rather more awkward than at first glance in this topology...
Whilst there are all sorts of obvious compromises, I can see exactly why, under the circumstances of it having to be small, powered by just a PP3 for a good long while and relatively cheap, they choose the blend of compromises that they did. Poor noise performance and needing electrolytics or tantalums in sensitive areas are the most salient down sides of the compromises.

For best results this circuit would also need the transistors to be not only selected for low noise (partial get round by choosing known very low noise devices and taking pot luck), but also carefully matched for gain and Vbe. It can give an output offset voltage of several millivolts, which could upset some following MM stages if DC coupled. Matching would help hugely with this, as well as overall SQ, but the devices should be thermally bonded for good thermal tracking.

Cascoding would be irrelevant to this topology and just be noisier and twice as complicated etc.

Good luck with building a mosfet head amp BTW

The transistors as provided in your schematic are not in their active region, as there is no DC bias to the base.
Either you have the schematic drawn wrongly, or need a few minutes refreshing your knowledge of transistor biasing
if still asserting Class A. Here is a nice tutorial https://www.youtube.com/watch?v=c6cmkm3UPUI

Where a transistor lacks bias to the base, as your schematic shows tied approx 150-200 ohms from ground,
( that is a strange arrangement for each channels stage sharing a resistor prior to ground )
- the transistor circuit is Class C but would be requiring a preceding amplifier stage and would be hopeless
in terms of audio fidelity... we both agree I think. Perhaps you should look at how the transistors
are achieving bias, as there is a bit of mystery at the moment.

The other exception and would be quite a good circuit, is the common base form allowing voltage gain
Here the emitter of the transistor is the input and the collector the output. Best though with a dual rail supply
http://www.allaboutcircuits.com/text...ase-amplifier/

Cheers / Chris

[337alant]

Probably the biggest improvement one could make to a standard Lentek would be to replace the emitter capacitors with film types such as polypropylene. In Nicks mods he increases them from the standard 10uF ( I haven't been able to read the lettering on these caps so I don't know 100% that these are originally 10uF but I would put a lot of money on it that I'm right!) with 47uF ones , as I would also, and the issue here is a purely physical one... together, the 4 caps would be bigger than the whole head amp!! Re-boxing it in a much larger case would be required.

Edit: I may have lost a lot of money on that bet as well!! In close up photos of the original innards all the tantalum caps look identical (it's a pity the photos at the beginning of the thread are gone!) and as most are marked as 10uF in Alan's photos (which have some errors BTW) I have assumed all the identical looking originals are 10uF....
HOWEVER.... If this is correct then the original version was -3dB @ 50Hz!! The 47uF ones are suitable...
As with most cap types, tants get bigger as they go up in capacitance and/or voltage rating. It is possible that the originals were indeed 47uF but at probably 6V rating whilst the identical looking ones at 10uF are rated for 10 or 16 Volts... It would be interesting if anyone can confirm. Neither is even slightly critical and the more capacitance the better in both cases. All other values I am certain about as clear enough views of the parts were available.

Jez
Only the 2 inner tantalum were 10uf
the 4 outer ones are 47uf as identified in my photo
In post 120 I listed the caps
http://theartofsound.net/forum/showt...152#post790152

Alan

[RothwellAudio]

Well done Jez for working out the circuit from those two pictures - not easy at all.
However, I don't get it. I think I'll be staring at that diagram a while before I work out how it works

[Arkless Electronics]

Jez
Only the 2 inner tantalum were 10uf
the 4 outer ones are 47uf as identified in my photo
In post 120 I listed the caps
http://theartofsound.net/forum/showt...152#post790152

Alan

Thanks Alan. Circuit sims said they had to be really but it's nice to have it confirmed Some close up photo's I saw which unfortunately didn't show the values had all the tants looking identical and I at first assumed the increase to 47uF was due to Nick having made the decision to change them.
I;ve now added a note to the post with the circuit diagram correcting this.

[Arkless Electronics]

The transistors as provided in your schematic are not in their active region, as there is no DC bias to the base.
Either you have the schematic drawn wrongly, or need a few minutes refreshing your knowledge of transistor biasing
if still asserting Class A. Here is a nice tutorial https://www.youtube.com/watch?v=c6cmkm3UPUI

Where a transistor lacks bias to the base, as your schematic shows tied approx 150-200 ohms from ground,
( that is a strange arrangement for each channels stage sharing a resistor prior to ground )
- the transistor circuit is Class C but would be requiring a preceding amplifier stage and would be hopeless
in terms of audio fidelity... we both agree I think. Perhaps you should look at how the transistors
are achieving bias, as there is a bit of mystery at the moment.

The other exception and would be quite a good circuit, is the common base form allowing voltage gain
Here the emitter of the transistor is the input and the collector the output. Best though with a dual rail supply
http://www.allaboutcircuits.com/text...ase-amplifier/

Cheers / Chris

"Better to remain silent and be thought a fool than to speak out and remove all doubt" Lincoln/Twain

[Paul Hynes]

I have to side with Jez on this.

The Lentek circuit is simple and elegant in its concept. Input DC bias is from one base to the other and the base terminals are referenced to ground by the 100R to ground in parallel with the cartridge impedance. The output swing is limited to 2x base/emitter voltages less the collector/emitter saturation voltages of the transistors, but this is more than sufficient for the signal levels in operation. The collector/emitter current is provided via the 15K resistors, which also balance the battery voltage around ground to give a split rail supply to the transistors. The output voltage is developed across the 4K7 collector load resistor.

The resistors used in Alan’s photos are thick film and metal oxide types which were available from RS components back when the Lentek was manufactured.

Paul

[Juha]

Found on the Internet. This is supposed to show one channel for the Lentek.

[Arkless Electronics]

As this simple circuit seems to be causing consternation and confusion I'll explain how it works.

It is a pair of common emitter amplifiers working in push pull. The collector of each device provides a very high impedance active load for the complementary device and renders the whole a virtual transconductance amplifier (it would be more nearly a true transconductance stage if common base but close enough). The 4k7 resistor to "ground" in effect sets the gain by deriving an output voltage proportional to the current according to Mr Ohm.. The 100R resistor to "ground" from the middle of the 47uF caps on the emitters also helps to set gain by providing some AC only degenerative feedback. Without the 4k7 at the collectors there would be no overall change in current from which to derive this current-controlled-voltage-applied feedback and so the 4k7 and 100R work together to set gain.

Getting down to mega basics, a transistor (npn) is forward biased when we reach the Vbe of approx 0.65V + a wee bit to give some base current. Now Vbe is "Voltage between base and emitter", all that is required is for the base to be 0.65V positive with respect to the emitter .... you can ground the base and then take the emitter 0.65V NEGATIVE to achieve the same effect

Power is applied from the 9V battery via the two 15K resistors to the emitters and it is these resistors that set the bias and in effect act as a voltage divider with equal upper and lower resistor values, this assists in forming a floating "ground" (hence the quotation marks), in association with AC decoupling from the two 10uF tants. In effect it is working from +4.5V and -4.5V supplies and the whole thing is floating, the AC midpoint for each channel being derived from the centre point of the 47uF tants via the 100R resistor.

The 15K resistors provide a huge amount of DC degenerative feedback to each transistor and hence bugger all gain at DC (the 47uF tants decouple this at AC), which is good as we don't want to amplify the offset voltage, and the complementary npn pnp pair means the offsets cancel at both input and output, to the accuracy of the transistor matching anyway!

Simple but in a satisfyingly elegant way!

And incidentally very similar to the input stage of "The Paradise" MC phono stage.... except that doesn't need to "float".

[Arkless Electronics]

Well I'll go to the bottom of our stairs...
Another explanation and a circuit diagram (which is correct... a few values are different probably due to being a different "Mk") have appeared whilst I typed away... stopping for coffees and tabs here and there, over an hour or more... Took me about 3 hours to reverse engineer it and do that schematic 'n all...

[Arkless Electronics]

I have to side with Jez on this.

The Lentek circuit is simple and elegant in its concept. Input DC bias is from one base to the other and the base terminals are referenced to ground by the 100R to ground in parallel with the cartridge impedance. The output swing is limited to 2x base/emitter voltages less the collector/emitter saturation voltages of the transistors, but this is more than sufficient for the signal levels in operation. The collector/emitter current is provided via the 15K resistors, which also balance the battery voltage around ground to give a split rail supply to the transistors. The output voltage is developed across the 4K7 collector load resistor.

The resistors used in Alan’s photos are thick film and metal oxide types which were available from RS components back when the Lentek was manufactured.

Paul

Spot on Paul and yes indeed max output is +/- one Vbe peak to peak so roughly 1V peak to peak and therefore 350mV RMS. Agreed.

I seem to recall those 2% resistors being sold as "thick film metal glaze" at the time....