Lightspeed Attenuator a new passive preamp

[georgehifi]

Heinz, you did not look too hard at the Perkinelmer site, you will find one as low as 1.5ohm at it's max led current of 40ma at 10ma it's still 5ohm which plenty, mine are 50ohm (different brand) which is still great, it will not go to absoulte zero volume, but that's no problem to me as it is just a whisper.

Cheers George

[georgehifi]

Quote:George,
please can You tell how equal in which range Your stereo-attenuator works?

Regards
Heinz!


Heinz, my attenuator after calibration is at worst +- 1db from min to max level at best I can achieve +- 1/4db from min to max level if my ldr matching is good, I use a 25 turn 500ohm pot in series with the led resistor on the strongest channel to calibrate this, it could almost be used as a balance control.

Cheers George

[gfinlayson]

George,

Do you use anythig fancy for the 100k pots on the LEDs, or just standard carbon track items?


Thanks,

Graeme

[powerbecker]

Originally posted by georgehifi
Heinz, you did not look too hard at the Perkinelmer site, you will find one as low as 1.5ohm at it's max led current of 40ma at 10ma it's still 5ohm which plenty, mine are 50ohm (different brand) which is still great, it will not go to absoulte zero volume, but that's no problem to me as it is just a whisper.

Cheers George

George,
indeed, my "knowledge" stems from long years back experiments
with some "no name" LDR`s!
Regards
Heinz!

[powerbecker]

snip

Heinz, my attenuator after calibration is at worst +- 1db from min to max level at best I can achieve +- 1/4db from min to max level if my ldr matching is good, I use a 25 turn 500ohm pot in series with the led resistor on the strongest channel to calibrate this, it could almost be used as a balance control.

Thank You George,
these results are very good for this simple arrangement!
But as You surely know it`s not really "hihi".
(I take the quality of a switched attenuator as a ruler.)
If I like to design such one with LDR`s I would try to insert
this passive arrangement into a servo circuit which compares in/output and steers the led`s.
A little bit simpler to do this can be the using of a high frequency
pilottone with a notch in the output.
Of course this is much more input so the LDR using lost its simplicity!
And of course one had to sell it for "$20kUS"
Regards
Heinz!

[st2_998]

Originally posted by powerbecker

Of course this is much more input so the LDR using lost its simplicity!
And of course one had to sell it for "$20kUS"

Probably I missed the point but I thought that LDR use had its advantage in avoiding contact and moving parts, not in being "simple".

High-frequency pilottone filtered out without spilling into audio band, i.e. by harmonics combination? Doesn't seem hi-end as well

[georgehifi]

Well I suppose now I have to give my long winded explanation of why I believe it has beaten everything I've put up against it, here goes.

I while back a freind of mine who worked for the CSRIO (massive research company in Australia) borrowed a giga hz storage oscilloscope worth 10's of K$ and we started to do some experiments.
On all the best pots mentioned in my first post, what we found was that during fast pulsed millisecond transisents at cd levels, audio band squarewaves, then stored and reviewed frozen on the scope, the leading edge was getting corrupted sometimes ringing, this leading edge has a very fast rise time > 100's volts per microsecond, and what we found was that the wiper on all of the pots were getting bounced, if you like off their tracks during these fast transients, (it was as though the contact area of the pot had become a diode and it was trying to rectify the ac signal) if we pushed on the wiper to give it harder contact to the resistive track the corruption stopped. Naturaly this would wear out the track in no time with this pressure. This also happened to switched resistor pots and also input switches, though to a lesser degree.
Now maybe you all can see why myself and later on Mark Porzilli (of Melos fame) were so obsessed with getting rid of small area contact devices (pots&swithes) from the singal path as they give some form of transient distorion. Hence the use for us to the led/ldr arrangment(zero contact area)

And Graeme you asked

George,
Do you use anythig fancy for the 100k pots on the LEDs, or just standard carbon track items?

No nothing fancy mate, the good old $2 carbon pots will do fine, but a good super smooth regulated 5v supply for the leds is a must.

Cheers George

[powerbecker]

snip
Probably I missed the point but I thought that LDR use had its advantage in avoiding contact and moving parts, not in being "simple".

....and how do You avoid the contacts of the connectors?

With "simple" I mean of course the difference between the use of
LDR`s alone compared with a additional servo arrangement.
Regards
Heinz

[powerbecker]

snip
High-frequency pilottone filtered out without spilling into audio band, i.e. by harmonics combination? Doesn't seem hi-end as well

In principle I agree!
On the other hand it depends from circuit design and the frequency....about 500kHz.
Regards
Heinz

[powerbecker]

snip George
On all the best pots mentioned in my first post, what we found was that during fast pulsed millisecond transisents at cd levels, audio band squarewaves, then stored and reviewed frozen on the scope, the leading edge was getting corrupted sometimes ringing, this leading edge has a very fast rise time > 100's volts per microsecond

How this??...after a 20kHz filter! Possible done through a internal
overload?

snip
This also happened to switched resistor pots and also input switches, though to a lesser degree.

I have never notice this effect, though I built (non audio) amps to 10MHz.
What about the contacts in the scope attenuator?

snip
if we pushed on the wiper to give it harder contact to the resistive track the corruption stopped

In this case one need stronger/cleaner contacts!

snip
Hence the use for us to the led/ldr arrangment(zero contact area)

Why not, but at the expense of less stability and more non linearity (especially with high input voltages!)

For a "hifi" attenuator I`ll prefer therefore precision resistors connected with hermetically closed relays.

Regards
Heinz