Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Analog Line Level Preamplifiers , Passive Pre-amps, Crossovers, etc.

A precision LED/LDR-based Attenuator
A precision LED/LDR-based Attenuator
Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 1st September 2011, 08:06 AM   #151
AndrewT is offline AndrewT  Scotland
R.I.P.
 
Join Date: Jul 2004
Location: Scottish Borders
If you require a constant input impedance/resistance and a constant output impedance/resistance, then a two part, shunt & series resistances, cannot provide this.

A T attenuator is one way to get there. This T requires three accurately chosen resistances for each attenuation level.
There are probably other ways to maintain this constant input/output impedance, but I suspect all require more than two resistances.
__________________
regards Andrew T.
  Reply With Quote
Old 12th September 2011, 12:23 AM   #152
wapo54001 is offline wapo54001  United States
diyAudio Member
 
Join Date: Nov 2008
A precision LED/LDR-based Attenuator
Someone posted the ladder attenuator schematic with resistor values as shown in attachment one. It has a fixed 10K input impedance and, it appears, an output impedance that is suitable to drive a 10K load.

I found the formulas used to compute the resistor values for this ladder, and did a spreadsheet. Attachment two is the resulting graph for a constant 7K input impedance and constant output impedance.

Quesstion is: Is this approach the way to go in designing an LDR volume control, or should I simply do a conventional pot where the total resistance of the top and bottom halves of the pot always equal some desired value, say, 10K?

Any thoughts?
Attached Images
File Type: jpg 10K constant Z ladder.jpg (130.6 KB, 733 views)
File Type: jpg 7K constant Z.jpg (187.5 KB, 716 views)
  Reply With Quote
Old 12th September 2011, 10:35 AM   #153
AndrewT is offline AndrewT  Scotland
R.I.P.
 
Join Date: Jul 2004
Location: Scottish Borders
I don't understand what you have done.

Look at the relay schematic.
Discard all to the left of -64dB relay.
You are left with 10k in series, 10k as load and 6r31 as shunt resistor.
This is a T resistor network with 10k input impedance and 10k output impedance.

Now do the same for any single relay. It too presents 10k for both input impedance and output impedance.

I cannot correlate the T resistors with what you are showing on your graph.
__________________
regards Andrew T.

Last edited by AndrewT; 12th September 2011 at 10:51 AM.
  Reply With Quote
Old 12th September 2011, 02:43 PM   #154
wapo54001 is offline wapo54001  United States
diyAudio Member
 
Join Date: Nov 2008
A precision LED/LDR-based Attenuator
Quote:
Originally Posted by AndrewT View Post
I don't understand what you have done.

Look at the relay schematic.
Discard all to the left of -64dB relay.
You are left with 10k in series, 10k as load and 6r31 as shunt resistor.
This is a T resistor network with 10k input impedance and 10k output impedance.

Now do the same for any single relay. It too presents 10k for both input impedance and output impedance.

I cannot correlate the T resistors with what you are showing on your graph.
The trailing 10K (load) resistor in the T you describe is the 10K input impedance value of the load amplifier. Thus, the T is completed by the load amplifier.

I should have changed the spreadsheet to match the 10K of the original ladder design before posting it. Look at this version, with Z adjusted to 10K.

My attenuator does not go to -63dB because it is limited by the 40 ohm minimum value of the LDR. Using this design, I can only get into the 40~50 db range without paralleling LDR devices.

Look at the 1dB rung of the ladder schematic, and the 8dB rung. The resistor values are exactly the same as my spreadsheet (rows 9 & 23) and graph. The formulas I used are the same as this (apparently) well-respected attenuator design, and the curves I show are the curves created by that design.

At this point with hardware rebuilt with some improvements included, I can now focus on the software -- which includes deciding upon which resistor values and attenuation curve to apply to both the series and shunt resistors (which are the same electrical design as the ladder attenuator shown earlier).

The main advantage of using a PIC to control the LDRs is that the PIC can control each LDR as separate devices with different resistance curves between series and shunt values. Thus, the curves shown in the attached spreadsheet are (theoretically) quite possible using LDRs.

My question is, is it worth the extra effort to design around these curves as opposed to simply emulating a mechanical potentiometer where the two resistive values always total a fixed value?
Attached Images
File Type: jpg ScreenShot001.jpg (227.8 KB, 668 views)
  Reply With Quote
Old 12th September 2011, 06:03 PM   #155
AndrewT is offline AndrewT  Scotland
R.I.P.
 
Join Date: Jul 2004
Location: Scottish Borders
If the 10k is the Rin of the next stage then that receiver stage does not see the 10k as it's Rs.
The receiver sees the remainder of the T without the third reisistor. The attenuator is not constant output impedance.
__________________
regards Andrew T.
  Reply With Quote
Old 12th September 2011, 06:13 PM   #156
wapo54001 is offline wapo54001  United States
diyAudio Member
 
Join Date: Nov 2008
A precision LED/LDR-based Attenuator
Quote:
Originally Posted by AndrewT View Post
If the 10k is the Rin of the next stage then that receiver stage does not see the 10k as it's Rs.
The receiver sees the remainder of the T without the third reisistor. The attenuator is not constant output impedance.
Well, you know, you are just reminding me (as though I needed reminding!) that my expertise is in the digital and DC realm, not audio. I can 'shape' the resistance curves of two LDRs to any appropriate values using a PIC, but I don't know which curves would be most useful. I have tried several times in this thread to ask for help in identifying what those curves should look like, but the extent of the responses seem either to address questions I didn't ask, or just tell me I don't know what I'm doing. I know I don't know what I'm doing -- that is why I'm trying to get someone knowledgeable to help me by telling me what I should be doing in the way of designing these curves!
  Reply With Quote
Old 12th September 2011, 06:18 PM   #157
AndrewT is offline AndrewT  Scotland
R.I.P.
 
Join Date: Jul 2004
Location: Scottish Borders
Low source impedance pays dividends for almost every type of receiver.
A power amp likes to see a low source resistance.
If that power amp is AC coupled it does not care if there is a small variation in the source impedance. 0r0 is OK. 10r0 is OK. 50r0 is OK, 200r0 is OK. It does not need to be constant.

The transmitter likes to see a load it can drive. Anything over 10k is usually high enough to not cause problems for most transmitters, provided there is not too much capacitance inparallel with that resistive load.

It comes back to what was said way back, constant input and output impedance offer very little when driving a power amp.
__________________
regards Andrew T.
  Reply With Quote
Old 5th November 2011, 03:07 AM   #158
wapo54001 is offline wapo54001  United States
diyAudio Member
 
Join Date: Nov 2008
A precision LED/LDR-based Attenuator
Well, trying to develop a precision LDR potentiometer with voltage control of what is a current-operated device has been a red herring and a waste of time. It would have been a simpler circuit if it had been possible, but, while control was possible, precision was not. In previous attempts using voltage regulation to control the LDRs, LDR resistance was not very stable over time. With current control, however, I find that LDR resistance is very stable and can therefore be programmed with accuracy.

Iím shifting gears. I have found a voltage-to-current converter chip and a precision op-amp which, when paired, appears to resolve all the issues and does it in a very elegant way. Each chip will operate two channels which means one of each chip per audio channel will control the series and shunt resistors nicely.

The attachment is the new board configuration.

The converter chip will deliver up to 15ma of current without a boost transistor and is internally limited at that level. The other end of the current spectrum is set by a single resistor, and if that resistor is a multi-turn trimmer, very fine control of the minimum current is practical. It appears that about 250 ohms will deliver the current required to drive an NSL32 to about 10K ohms of resistance. A 500 ohm multi-turn seems ideal for the job, and Ĺ watt pots are available so power considerations are met. No other support circuitry is required, just the one resistor per control channel. Full rail input to the converter will drive the converter to full current output, or 15ma out. Slightly above Gnd, say, 10mv, will be ideal to drive the converter to minimum current as set by the variable resistor. I have successfully set the LDR ohmage level to values of 5K, 10K, 20K, and 25K by varying the value of the trimmer. Very elegant solution.

The input of the converter is rather low impedance and therefore must be driven by a very capable circuit. My original mosfet & resistor voltage divider wonít do the job. Iíve found a precision op-amp which is unity-gain stable and will drive to within about 2mv of the rails given the input impedance of the converter. That means that of the 1023 counts available with the PICís ADC pins, approximately 1018 counts will be available to control the level of the op-amp which will control the converter. This is the best case scenario that Iíve been able to come up with so far.

Another problem that Iíve run into in trying to control the LDR is that the LDR response is very slow compared to the response of the PIC, and the lag has caused problems in software control. In this new scenario, the PIC will drive an op-amp and converter with near instantaneous response. The LDRís lagging response will have no impact on the feedback to the PIC, so this problem will go away.

The very high input impedance of the op-amp will be a vanishingly small load on the capacitors controlled by the PIC, so very small and infrequent control adjustments will be adequate to maintain precise control voltage levels to control the op-amp inputs. If the circuit is successful, I will probably replace the .47uF metal-film caps with 1.5uF tantalums or even 15uF tantalums with matching changes in the resistor limiting PIC drive to the capacitor. This should result in fast, accurate adjustment by the PIC followed by very long stable periods when the capacitor does not require adjustment due to the high impedance of the op-amps.

In addition to the changes to the circuit resulting from the added ICs, Iíve added a facility that will allow precision testing of LDRs using the boardís circuitry simply by moving one jumper. This board therefore can be used to evaluate and select LDRs which could then be marked and sorted and selectively installed to the board Ė a one stop solution. The LDR test facility is always available even after all LDR positions are filled and the board is fully functional as a volume control. If fast, repeatable testing of multiple devices is desired, it would be easy to set up a rotary switch with precision resistors to program current output at multiple levels in any series desired. If this approach is followed, the mA meter could be omitted and replaced by a jumper at the J3 pins 1~2.

Iím switching to two LDRs per position Ė LEDs wired in series, resistors in parallel. I find that when two LDRs are wired that way, 15 milliamps results in a resistance of about 25 ohms. That means the LEDs will be loafing at 75% of full rated brightness, and the resistance will be very usable. Due to the low maximum current used, once calibrated this configuration should last for a very long time if not forever without degradation of the LED.

The power rails -- +5V and Gnd Ė are wired in a star-configuration with different parts of the circuit connected directly back to the regulatorís output capacitor. Various sections of the circuit should not interact and voltages should remain very stable. The regulator circuit contains a multi-turn pot for setting the voltage very accurately. In theory 5.000V is desired, and in practice, this circuit varies by .001V only.
Attached Images
File Type: jpg P1030646.jpg (571.1 KB, 616 views)
  Reply With Quote
Old 8th March 2012, 10:04 PM   #159
cab is offline cab
diyAudio Member
 
Join Date: Aug 2009
any updates?
  Reply With Quote
Old 8th March 2012, 10:30 PM   #160
wapo54001 is offline wapo54001  United States
diyAudio Member
 
Join Date: Nov 2008
A precision LED/LDR-based Attenuator
I have the chip successfully managing one LDR with fast, accurate control up to 10K ohms, looks like it might go to at least 15K, maybe 25K if anyone wants a pot with that kind of resistance range.

Have just populated the board with the other three LDRs with a view to testing out the control of four devices at once, which is the standard required for stereo. Working on the pseudo-multitasking aspects of the software now.

BTW, the board which I have shown as the latest turned out to have a fatal flaw, am talking about, and currently working with, an older version of the board. If the multi-tasking works out well, I'll build another prototype of the board.

Last edited by wapo54001; 8th March 2012 at 10:37 PM.
  Reply With Quote

Reply


A precision LED/LDR-based AttenuatorHide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
LDR photoresistor attenuator preamp, in general romana Analog Line Level 54 14th December 2013 02:32 AM
TUBEpre and Lightspeed LDR attenuator tinitus Analog Line Level 5 31st January 2010 05:08 PM
LDR attenuator questions Carlp Analog Line Level 15 21st January 2010 01:01 AM
LED to LDR amplifier electro samurai Analogue Source 8 21st March 2004 05:07 PM
Spectrol, high precision, 10k attenuator. Peter Daniel Swap Meet 3 4th February 2004 05:15 PM


New To Site? Need Help?

All times are GMT. The time now is 04:37 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2018 DragonByte Technologies Ltd.
Copyright ©1999-2018 diyAudio
Wiki