Lightspeed Attenuator a new passive preamp

[MagnumOpus]

Sounds good. Thanks for the reply.

[bpcairns]

Perhaps someone here could help me with a bit of advice. I'm using a beresford Caiman DAC with all of Stan Beresford's Mods. This includes the passive mod that involves removing the op-amp board and using a DC blocking capacitor. This was fed into a Linn LK280 power amp via a 10K Alps pot and about 20uF of capacitance was enough to get a low frequency cut off around 20Hz. I fancied remote control so bought one of those far eastern remote controlled pots from ebay. It was terrible so I reasoned that perhaps I could use it to control a lightspeed (still really thinking remote control at this point. Thought you were all talking through your ***** about LDRs). This was suggested by a friend who I think was winding me up. I bought Uriah's kit for a lighter note, a lightspeed all but the balance method. This thing sounds unbelievably good. I had no idea that a decent pot could smother the sound like that. A revelation from which there is no going back. Many thanks to George for telling us how it is done and to Uriah for sorting and supplying the LDRs. Anyone who thinks reports of how good this is are exaggerated has to hear one. It sounds like there is nothing between the source and the power amp. Every aspect of the sound is better. Clearer, more detailed, (much) sweeter treble, Deeper cleaner Bass, vocalists sound like they are sitting in the room. At least in my setup they do . Easily the biggest difference I've made to my system in years and the cheapest by far, ever. My problem is this however:- An LK280 has a witheringly low input impedance of 5K. Yes that is five kiloOhms. The LDrs I got from Uriah are around 12K giving a pot of around 13-14k ish. I now seem to need around 100uF of blocking capacitor in the Beresford DAC to get my bass back. The sound is utterly fantastic in every way except for a little thickness caused by all that capacitance in the signal path. Today I replaced the resistors on the inputs of by beloved LK280 (took quite a bit of courage to mess with it I don't mind admitting). The sound fell apart. No treble. Had to put the old resistors back. I have a circuit diagram for this amp but Linn don't support it and won't answer questions on it. I read on another forum somewhere that the LK280 and its relative the Klout are a rather odd design. Anyway I can't increase the input impedance using the method I've seen George suggest many times on this very thread (I have read it all). So it seems to me I have three options. 1. Live with it. Could experiment with different caps or even put the op-amp board back in. The sound is really great in most ways. Better than it has ever been. I don't hear any of the symptoms it has been suggested I should. Only problem I can hear is the caps. 2. Build a buffer. 3. Try a lightspeed with a lower resistance, mine is twice what george says his commercial lightspeed measures at. My questions then are 1. Would a lightspeed with lower resistance help? I think it would because of my comparison with a 10K pot but perhaps there is something at play here I've missed? 2. Would someone be willing to look at the circuit of the LK280 and possibly suggest another way of modding it? 3. Even if I changed the LDRs to lower resistance ones would it still be worth trying a buffer? Even if no one can help, my system still sounds much better than it ever has.

[Blues]

Seek and ye shall find...

[georgehifi]

Originally Posted by bpcairns

My questions then are
1. Would a lightspeed with lower resistance help? I think it would because of my comparison with a 10K pot but perhaps there is something at play here I've missed?
2. Would someone be willing to look at the circuit of the LK280 and possibly suggest another way of modding it?
3. Even if I changed the LDRs to lower resistance ones would it still be worth trying a buffer? Even if no one can help, my system still sounds much better than it ever has.

Hi bpcairns:
1: No, the difference between your 10k and my production Lightspeed at 7k is minimal for your problem.
2: Post it up lets take a look at it, it may have the input loading resistor incorporated into the feedback network or something weird, for it not to work.
3: In your case yes and I would do the Nelson Pass buffer he did for the Lightspeed, back a few thousand posts.
(dam these low impedance amps, they make you use unnecessary extra transparency killing buffer circuitry)

You also may have another problem, it also sounds to me you have no idea what the output impedance of the dac is with the mods that have been done, this needs to be found out as well, as it has to drive the added combination of the Lightspeeds input impedance together with the Lin's input impedance which in your case is going to be very low, maybe even down to a combined 2k or so that it will see.

Cheers George

[bpcairns]

Here is the diagram.

I think you're right about the 2K. That's suggested by the corner frequency calculator I was using, based on how much capacitance I need. DAC seems to cope pretty well, sounds good at all volumes. I'll try to measure it properly though and let you know.

I'll start putting together what I need for the buffer.

Thanks again.

Brian

[georgehifi]

I can see where you've come unstuck a little.
With the 10k out of your Lighter Note plus the 1k series input resistor on the Lin this is giving around 11k series resistance, this will give a HF fiter when combined with the 4n7 (4700pf) HF cap. That has a cut off frequency of -3db at around 3khz!!!! No wonder the sound "fell apart" You'll need to change the 4n7 also for something much smaller around 500pf.
I know you said you like the Lin, but, I would think about another amp with higher (>47k) input impedance, as it is going to give this trouble with any passive and most tube preamps as well.
Cheers George

[bpcairns]

Hi George

I wondered about that cap but didn't know what to replace it with. Do you think it is doing anything other than DC blocking? Do I need it at all? If I replaced it with a link would the amp still work . There will always be a DC blocker on the output of the DAC so should never need one here? I only have one source and no interest in any others at present. An oportunity for one less cap in the signal path?

When I made the changes I started with a 120K in place of the 4k7. This reduced the gain as well as the HF cut-off so I put a25.5K (closest I could get to 25K in a decent type) and the gain returned but obviously not the HF. So it would be 25.5k where the 1K is at present. Does this change the value of the capacitor I should use, assuming I NEED to have one at all? Also my lightspeed is nearer 14K. the 10K was my old Alps pot. Does this make any difference?

I'd like to persevere with the Linn. I've had it apart a couple of times now and am getting quite confident with the changes we're discussing here. I'm also not sure what I'd replace it with. While there are many fine amps out there I suspect anything of this caliber would be out of my budget range now. I'll look at a replacement if I blow it up with these changes. Worst case, I either live with the loss of transparency caused by caps or the loss of transparency caused by a buffer. But I might stumble upon a mod that sorts it.

Thanks again.

Brian

[georgehifi]

Originally Posted by bpcairns

Hi George

I wondered about that cap but didn't know what to replace it with. Do you think it is doing anything other than DC blocking? Do I need it at all? If I replaced it with a link would the amp still work . There will always be a DC blocker on the output of the DAC so should never need one here? I only have one source and no interest in any others at present. An oportunity for one less cap in the signal path?

When I made the changes I started with a 120K in place of the 4k7. This reduced the gain as well as the HF cut-off so I put a25.5K (closest I could get to 25K in a decent type) and the gain returned but obviously not the HF. So it would be 25.5k where the 1K is at present. Does this change the value of the capacitor I should use, assuming I NEED to have one at all? Also my lightspeed is nearer 14K. the 10K was my old Alps pot. Does this make any difference?

I'd like to persevere with the Linn. I've had it apart a couple of times now and am getting quite confident with the changes we're discussing here. I'm also not sure what I'd replace it with. While there are many fine amps out there I suspect anything of this caliber would be out of my budget range now. I'll look at a replacement if I blow it up with these changes. Worst case, I either live with the loss of transparency caused by caps or the loss of transparency caused by a buffer. But I might stumble upon a mod that sorts it.

Thanks again.

Brian

It's not a dc blocking cap. but a HF filter cap that now goes together with the Lighter Notes 14k "you say" and the 1k (15k total). 15k with the 4700pf to ground gives a HF limit of -3db @ 2250hz. The cap needs to be much smaller with your 15k series resistance to get say what Lin had in the first place with just the 1k series resistance of -3b @ 34khz, to get this with a 15k series resistance the cap will have to be lowered to 300pf and the 4.7k resistor still made say 51k, but your dac still has a question mark over it's output impedance, all is still not worked out.
Cheers George

[AndrewT]

Bp,
R2, R4 & R5 all have to be set in conjunction with each other.
You cannot just change any one of them, you must change all three in exactly the same ratio. and that also involves a change to C2 to maintain the designers choice of ratios.

But, if you do this, the amplifier may behave differently. Be prepared to put it back to standard if the changes don't work.

[georgehifi]

When you have a an input resistor higher than the sources output impedance, the input resistance becomes the output resistance of the source.
As I've said in previous post, with an amp that is bipolar input like my own is, which had low input impedance (2k) I changed it to 100k but it gets over ridden by the output impedance of my Lightspeed which is 7k so in effect the input bipolar transistor of my amp sees 7k in my case even though I have a 100k there, which is fine, I just don't switch it on with no input attached, as you will get dc some offset at the speaker outputs if the "dc servo control if fitted" is out of adjustment range.
If the amp is fet input, all is usually quite safe for raising the input impedance to 100k just like tube poweramps.
Anyway this is becoming all too messy, and I don't want to be the reason you blow your amps like I said get a buffer after the Lighter Note, or change amps.

Cheers George