Mismatched LDR's
Harry,
Try the R2E for series and the R2A as the shunt elements. This will increase ovrall attenuation when the pot is closed and give more control over at high attenuation.
Even closely matched quads do not present a flat impedance or attenuation rate. The resistance versus current level function is not linear. In all mine the impedance is lowest in the middle. But it is only maybe +/- 15% at worst.
You are very lucky that all three of each match. My experiance with Allied is that three of a grade will be match up close enough to make a pair.
I think Georgehifi startedf using unsorted devices. That would be a pain, I could see it requiring 20 to get a pair. ut if you had a 100, that might get 30 pairs and a few quads.
Good luck with your build!
George
Harry3 said:
Harry,
Try the R2E for series and the R2A as the shunt elements. This will increase ovrall attenuation when the pot is closed and give more control over at high attenuation.
Even closely matched quads do not present a flat impedance or attenuation rate. The resistance versus current level function is not linear. In all mine the impedance is lowest in the middle. But it is only maybe +/- 15% at worst.
You are very lucky that all three of each match. My experiance with Allied is that three of a grade will be match up close enough to make a pair.
I think Georgehifi startedf using unsorted devices. That would be a pain, I could see it requiring 20 to get a pair. ut if you had a 100, that might get 30 pairs and a few quads.
Good luck with your build!
George
Not that bad
Six pairs will not require anywhere that many. If you use sorted to begin with, I think that 25 should be enough.
Unless you are using them in a commercial project, it is much cheaper to use the sorted. The impedance seems to be all over the place with these. Unsorted would be a LOT of work.
You mentioned six pair. If you are going six channel, at least the front would be better with LDR series and shunt. For the surrounds, a LDR shunt and fixed series should be okay, the level and frequency extremes are reduced already.
Are you bulding a six channel with the GB Symasym?
George
Six pairs will not require anywhere that many. If you use sorted to begin with, I think that 25 should be enough.
Unless you are using them in a commercial project, it is much cheaper to use the sorted. The impedance seems to be all over the place with these. Unsorted would be a LOT of work.
You mentioned six pair. If you are going six channel, at least the front would be better with LDR series and shunt. For the surrounds, a LDR shunt and fixed series should be okay, the level and frequency extremes are reduced already.
Are you bulding a six channel with the GB Symasym?
George
matched in what sense?
Just to ensure everyone is on the same page regarding matching these LDRs, I'm using 3 resistance point 10K, 100K, 0.5M with a regulated 5v off a 12V SLA battery.
It's really PITA to get a decent match. I can hardly get 1 pair off to match the 3 resistance points properly (after measuring 20 pcs - I've got 30 more to go). BUT, what's you criteria of matching? Within how many ohms? Do you match all 3 points, or just the critical lower 2 points will do ?
Just to ensure everyone is on the same page regarding matching these LDRs, I'm using 3 resistance point 10K, 100K, 0.5M with a regulated 5v off a 12V SLA battery.
It's really PITA to get a decent match. I can hardly get 1 pair off to match the 3 resistance points properly (after measuring 20 pcs - I've got 30 more to go). BUT, what's you criteria of matching? Within how many ohms? Do you match all 3 points, or just the critical lower 2 points will do ?
Matching
Will,
Without know the details of your setup, the resistances you are using are too high. If you plan on a 100K pot the highest resistance it can see is 100K. Same on low end.
Try 200 ohms, 10K, and 50K. I looked for slope and intercept matches. Since volume is a log function the match is very forgiving. A 10% match is fine, but tighter matching is easy to accomplish.
Try measuring the volme pot you are now using. Most are very sloppy in matching as rotated. I have seen some that are off by 30% in places, yet seem to maintain channel balance in practice.
George
Will,
Without know the details of your setup, the resistances you are using are too high. If you plan on a 100K pot the highest resistance it can see is 100K. Same on low end.
Try 200 ohms, 10K, and 50K. I looked for slope and intercept matches. Since volume is a log function the match is very forgiving. A 10% match is fine, but tighter matching is easy to accomplish.
Try measuring the volme pot you are now using. Most are very sloppy in matching as rotated. I have seen some that are off by 30% in places, yet seem to maintain channel balance in practice.
George
George,
My setup rig is pretty straight forward. 12V SLA battery feeding into 7805 with 470uF at the input and output, and the 5v feeds the metal film resistors of 10K, 100K and 0.5M. Each of the resistors are connected to a crocodile clip.
I don't quite get your meaning of "I looked for slope and intercept matches".
My setup rig is pretty straight forward. 12V SLA battery feeding into 7805 with 470uF at the input and output, and the 5v feeds the metal film resistors of 10K, 100K and 0.5M. Each of the resistors are connected to a crocodile clip.
I don't quite get your meaning of "I looked for slope and intercept matches".
Current vs resistance
Will,
I measured the voltage drop across the resistors to get the current with the three resistors used to match.. Then plotted it out with current on the x axis and resistance on y.
The differnce between them is not always linear. Some plot out flatter, other are lower resistance at high current and higher resistance at the low current level. This what I meant by slope and intercept.
Plus none plotted with a linear response.
George
Will,
I measured the voltage drop across the resistors to get the current with the three resistors used to match.. Then plotted it out with current on the x axis and resistance on y.
The differnce between them is not always linear. Some plot out flatter, other are lower resistance at high current and higher resistance at the low current level. This what I meant by slope and intercept.
Plus none plotted with a linear response.
George
Does anyone know the output impedance and voltage of the LC Audio ZapFilter Mk2? It seems to be a very popular modification for CD/DVD/SACD players. I have mailed the company and some modders but have not received any replies...Here is the link
http://www.lcaudio.com/index.php?page=6
Regards, Peter
http://www.lcaudio.com/index.php?page=6
Regards, Peter
abq-pete said:
I asked this of Lars from LC and he said it's about 100ohms. let me say on my system, that in balanced input current mode, single ended output, it is the tightest most dynamic output stage I have heard, crystalline highs and gut thundering bass very open mids, this is an awesome piece of hardware, but it can sound soulless at times maybe that's the truth, who knows.
Cheers George
Multichannel LDR Preamp
I am in the process of designing a multichannel self calibrating LDR preamp.
I am driving the LDRs (should not need to be matched) with 16 bit current DACs from a microprocessor. I have also included mercury relays for the input switching although would be happy to consider changing these to optical switches if anyone can suggest a suitable method.
I originally started this design to fit on the back of my Aleph P1.7 clone PCBs but as a result of this thread thought that I could make it either an active preamp with LDR volume control or passive or even a combined passive/active operation thanks to the microprocessor control. It can support both balanced and single ended signals.
The multichannel element extends to six channels at the moment but could consider more channels if required. In the current design the micro will interogate the DACs and decides how many channels are connected on power up and arrange the sofware accordingly.
In addition the design has the usual Infra Red control, LCD display support and a RS232 interface for futureproofing.
Ryssen, if you are prepared to wait a while I might be able to help you.........
Thanks George for the enlightenment (that could be a good name for the final PCB!) and encouragement
I am in the process of designing a multichannel self calibrating LDR preamp.
I am driving the LDRs (should not need to be matched) with 16 bit current DACs from a microprocessor. I have also included mercury relays for the input switching although would be happy to consider changing these to optical switches if anyone can suggest a suitable method.
I originally started this design to fit on the back of my Aleph P1.7 clone PCBs but as a result of this thread thought that I could make it either an active preamp with LDR volume control or passive or even a combined passive/active operation thanks to the microprocessor control. It can support both balanced and single ended signals.
The multichannel element extends to six channels at the moment but could consider more channels if required. In the current design the micro will interogate the DACs and decides how many channels are connected on power up and arrange the sofware accordingly.
In addition the design has the usual Infra Red control, LCD display support and a RS232 interface for futureproofing.
Ryssen, if you are prepared to wait a while I might be able to help you.........
Thanks George for the enlightenment (that could be a good name for the final PCB!) and encouragement
georgehifi said:
Georgehifi,
Thanks for taking the time to respond.
Ouch! All the way up until the end, I was thinking that you were going to say that it is the greatest thing since sliced bread. Can you elaborate on the souless part? I assume this was dependent upon the recording and not all material.
I just purchased a Sony DVP-NS9100ES player and am planning to have it modded with the ZapFilter Mk2. The DAC in the machine is a BB DSD1796 that outputs in differential current mode. It seems to be a similar configuration to that which you heard? The impedance at 100ohms seems to be in the ballpark for the Lightspeed. What about the output level? Is it sufficient (too high or too low)?
There are some othe tube-based output stages but I think their output impedance will not match well with the Lightspeed.
Regards, Peter
Peter I didn't mean to give the wrong impression with the soulless part, it that some cold recordings don't get the artificial warmth that a opamp based I/V converter tends to give.
As you may be aware the I/V stage on 99.0% of dacs is handled by an opamp this is not the perfect way of converting a dac that is current output to voltage, the best way is a transimpedance amplifier, and the closest thing is the Zap in current mode as the balanced differential output from a current output dac sees not the bass of transistors in the Zap (when it's in current mode) but it sees the emitters, and so sees what a current output dac should see which is a SHORT! This is the perfect loading for a current output dac, and keeps them linear.
The gain in my setup was 20% higher using the Zap and the Lightspeed Attenuator loves it.
Cheers George
As you may be aware the I/V stage on 99.0% of dacs is handled by an opamp this is not the perfect way of converting a dac that is current output to voltage, the best way is a transimpedance amplifier, and the closest thing is the Zap in current mode as the balanced differential output from a current output dac sees not the bass of transistors in the Zap (when it's in current mode) but it sees the emitters, and so sees what a current output dac should see which is a SHORT! This is the perfect loading for a current output dac, and keeps them linear.
The gain in my setup was 20% higher using the Zap and the Lightspeed Attenuator loves it.
Cheers George
georgehifi said:
Georgehifi,
Ahhh! I understand.
Perhaps you can help me with some math here. With the assumption that I will be getting a Lightspeed as the "preamp", what would be the minimum amplifier input impedance requirements if the CD player was tube-based and outputting about 3V at 600 ohms?
It's kinda funny because I am selecting components determined by best compatibility with the Lightspeed that I have yet to hear!
Regards, Peter
I have a friend that has a MF Limited Edition Electra Signature CD Player Tube (6922) output stage, at around the 600ohm.
Using the Lightspeed Attenuator he says that he can just hear the difference from 100k to 150k on the power amp but not 150k to 200k, so he uses 150k on the input of his tube SET's.
But when he uses his Belcanto cd player which is 50ohms he said over 50k cannot be heard
Cheers George
Using the Lightspeed Attenuator he says that he can just hear the difference from 100k to 150k on the power amp but not 150k to 200k, so he uses 150k on the input of his tube SET's.
But when he uses his Belcanto cd player which is 50ohms he said over 50k cannot be heard
Cheers George