LC AUDIO MILLENNIUM XP - HELP if possible from Lars Clause!

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Hi!
I just received (bought) a complete kit of Millennium Xp amplifier from Lc-Audio. But... I can't use it...

I discovered, as you can easily see from the joint schematic, that the input impedance is 2K !! Yes, 2 KOhm!!

Now, I have a simple valve preamplifier, SRPP with a coupling capacitor of 0,47 uF..... I have big problems:
first because the gain of this stage has hardly decreased from this low resistence,
secondly because It cuts all the slow frequency!
I calculated, I need at least a coupling capacitor of 20uF (!)
Too big, to expensive a good one, and too poor will be the sound with this big capacitor.


Now, my questions:

1. Is it possible that the big (really) Lars Clausen, has made a circuit of a PWA with only 2 KOhm input Z? No Power Amplifier has so low input impedance...

2. According to who is so kind to help me, is it possible to increase this input resistance up to, for example, 20-50 KOhm? What will the consequences be?

3. You know what is the gain of the VAS? Because LC audio didn't decleare it on their "cook book"...

4. Any other suggestion?

Thank you very very much in advance.

P.S. I already send TWO e-mail to LC-Audio, of course, but after 10 days I didn't receive ANY answer. It's odd, but It is.
 

Attachments

  • mill sch.pdf
    44.5 KB · Views: 602
Hi,
you should have expected that order of input impedance.

It is a balanced input stage typically driven by output stages that can rise to the demands of 600ohm load impedance. Most are in the range 10r to 50r.

As dark says, raise the input resistors but accept the extra noise.

Regarding the DC blocking cap in your preamp;-
try using a 100uF low esr Electro // 4u7F polypropylene, all times two for your balanced output. Or eight 4u7F polypropylenes giving a high pass of 4.2Hz (F-3), 8Hz(F-1).

How has your pre-amp organised the balanced source to match your new loading?

Can your pre-amp current ability/output impedance drive very long cables?
 
I can't see you'll run into any significan noise problems by changing to something like 47 Kohm resistors. resistor noise cant be an issue at linelevels

I suspect you are running the tube amplifier singleended so the gnd and ind- are shorted ??

The ultimative problem could be that your tube preamp might also see the 47 Kohm as a hard load. Unless of couse you are using anodefollower output or a supertriode like 6H30.

Best of luck

Michael
 
andythefirst said:
Hi!
I discovered, as you can easily see from the joint schematic, that the input impedance is 2K !! Yes, 2 KOhm!!

__SNIP some questions__

P.S. I already send TWO e-mail to LC-Audio, of course, but after 10 days I didn't receive ANY answer. It's odd, but It is.


Hello World!!

- Lower Impedance, the more control is needed from the preamp, the better possibiliti'es for optimal control and sound quality... - And who in the world would use such antiquated things as tubes for other stuff than putting in museums??? ;)

There is no problem in changing the input Z resistor to any wanted value. Please keep it under 100K though, and PLEASE do not let the input wander off uncontrolled ( Keep something shielded lownoise attached to the input, no "floating" with high / very high impedances.

The VAS is answered already, no point in repeating that...

Regarding Emails to LC Audio..: - You are probably using the wrong Email Address - where / who did you write??... Try writing to support@lcaudio.dk , and i will answer your qquestion. Only in weekends does it appear that mail grow older than 48 hours ( And maybe if the questions is _very_ weird - but then they are replied by "i'll investigate that - please wait" )...

Sincerely, Lars / Viborg - Tech Support for LC Audio at support@lcaudio.dk
 
Andy,

The reason for low IP reference resistor is mainly OP voltage offset.

For a power amp you have to assume it will be driven by a source
with a cap coupled OP with a relatively high value ground resistor
after the cap.

As such, the IP bias current from the Millenium's BJT stage will cause
an offset voltage. Even though the IP stage is complimentary and
the bias currents theoretically cancel there will be some mismatch
in components and thermal drift which can cause some IP offset. If
the base to ground IP resistor is large, say 50k, the offset can be
considerable.

If the Milleniums servo opamp has say 14V of OP swing, then it can
effectively provide 15V/(301k+499k)= about 18 micro amps of
correction max. This servo feeds the voltage stage which has a
30k load so that 18 micro amps will result in about 0.5V of correction

So if the amps gain is 60 then the tolerable IP offset is less than
10mV for the servo to be able to correct the OP.

The IP to gnd resistor has -nothing- to do with the noise of the
amp.

Cheers,

Terry
 
"The IP to gnd resistor has -nothing- to do with the noise of the
amp. "

Then I must be deaf, and RCA must have made up the data in their SSD204 amplifier handbook showing a 16dB improvement in S/N ratio when going from an 18K input impedance to 2K.

Tube guys must be daft too for buying expensive grid chokes instead of just using a resistor.
 
djk said:
"The IP to gnd resistor has -nothing- to do with the noise of the
amp. "

Then I must be deaf, and RCA must have made up the data in their SSD204 amplifier handbook showing a 16dB improvement in S/N ratio when going from an 18K input impedance to 2K.

Tube guys must be daft too for buying expensive grid chokes instead of just using a resistor.

djk,

Relax, and do some more noise theory study.

Any decent preamp should have an OP Z of <1k, usually it is in
the range of a few hundred ohms.

As such the pre's OP Z dominates and the noise produced by
the amps IP to gnd R is shunted by the pre because they are
effectively in parallel.

Maybe you are thinking of a case where the amps noise is
measured without a pre amp connected and without the IP's
shorted. This is in fact a meaningless noise specification because
it does not relate to a real world condition.

WRT your comments on grid chokes, this is a separate and
unrelated issue. The grid choke is in -series- with the IP and
-does- contribute noise separate from whatever pre is driving it.

Also, we are not discussing any sonic effect that such series
resistors may have. We are discussing the noise contribution of
the resistor from base to ground.

Cheers,

Terry
 
"WRT your comments on grid chokes, this is a separate and
unrelated issue. The grid choke is in -series- with the IP and
-does- contribute noise separate from whatever pre is driving it."

I'm not sure if we're talking about the same things here. The choke I am talking about replaces the typical 470K resistor from the input to ground with a choke that has high AC impedance and low DC resistance.

The idea WRT using an input transformer with the LC amp is similar. The DC off-set is kept within the range of the DC servo by using a 2K bias resistor, this not being a problem to drive for SS preamps. By using a transformer the input impedance seen by the tube preamp may be raised and the DC resistance of the input transformer might be in the range of 100 ohms. Depending on the output impedance of the tube preamp this may give a signifigent reduction in noise, as well as eliminating any ground loop issues that may or may not exist.

Rowland is one company that comes to mind that uses high quality input transformers, even though they use FET inputs with very high input impedance.

My personal experience in this area shows a worthwhile improvement in sound.

YMMV
(Your mileage may vary.)
 
Hi Terry D,
OP Z of <1k, usually it is in the range of a few hundred ohms.

As such the pre's OP Z dominates and the noise produced by
the amps IP to gnd R is shunted by the pre because they are
effectively in parallel................
the noise contribution of the resistor from base to ground.
I follow your explanation - thanks, I have almost changed my stance.
But first a question.
Is your parallel resistor source impedance affected by DC blocking caps?

Presumably a high source impedance will be noisier, but at what level does it become a concern (0dbu sensitivity power amp)?

See my next post on transformers.
What happens here with secondary resistance and reflected source impedance all in parallel with the 2k0?
 
Hi Djk,
you said
Use a 10K:10K input transformer and change the resistors to 10K.
Did you mean impedance matching transformer? Maybe 10k to 1k (3 times turns ratio).
Then you can keep the 2k0 resistors or even reduce them to 1k0 if the circuit performs better.

But, keeping the 2k0 and 3:1 transformer gives an effective input Z=18k. This would suit the proposed Pre-amp much better and could go even higher (5:1) for a valve (tube) source.
 
To Nillerviborg:

1. Thank you for you reply!
2. I didn't want to be polemical with LC Audio. Not at all! Instead I consider the Firm very serius and competent!
3. It is than sure I wrote to the wrong e-mail (jp@lcaudio.dk).
4. I will keep your one for the future. Thank You!

To all, thank you for you kind and valuably replies! I will keep them, and use, and I'll let you know...

Whishes!
 
AndrewT said:
Hi Terry D,

I follow your explanation - thanks, I have almost changed my stance.
But first a question.
Is your parallel resistor source impedance affected by DC blocking caps?


You have to break the equation down into two conditions, DC and
AC.

First DC:

This is where offset comes into play. If the pre amp has a low OP Z
but also an OP cap then you are just looking at the reistances
-after- the cap to ground.

So if the pre has a 100k R to ground after the cap then -this- goes
in parallel with the amps IP Resistors for DC drift and offset calcs.

next AC:

In this case we consider noise as noise is an AC signal. So the
pre amps actual OP Z (before the OP cap) is in parallel with the
amps IP resistors.

Yes, at lower frequencies this impedance is less because of the
reactance of the caps but the OP caps should be of such a size
to drive reasonably low amp source impedances and as such can be
considered a short at AC.



Presumably a high source impedance will be noisier, but at what
level does it become a concern (0dbu sensitivity power amp)?


It depends whether you are referring to the pre amp as the source
impedance or the amps IP R's. Proper noise analysis is fairly
complex. Generally the pre amps self noise will be dominant. The
Milleniums IP referred noise will be quite low, probably around
10nV/rt Hz or less. Most pre amps will have way more OP noise than
this.




See my next post on transformers.
What happens here with secondary resistance and reflected source
impedance all in parallel with the 2k0?


OK WRT transformers, you have to do as above, look at DC
conditions and AC conditions separately.

At DC the transformer has very low resistance so it will minimise
amp offset. Especially in this case because the Millenium is a
true differential IP (balanced). So the transformer will work very
well here.

At AC the transformer reflects the pre amps impedance to the amp
and the amps IP impedance to the pre amp. The reflected
impedance will be turns ratio squared. So if the IP transformer is 2:1
and the amps IP Z is 4k (2x2k - balanced) then the pre amp will see
4 x 2^2 = 16k.

Coversly the amp will see the preamps OP Z / 4. However the real
issue for offset will be the DC winding resistance of the
transformers secondary which should be from 10's to 100's of ohms.

WRT transformers and noise it is a lot more complex than just to say
they reduce noise. In fact transformers often increase noise,
especially in low level IP stages compared to say a low noise Jfet
alternative. But that is beyond the scope of this thread.

Cheers

Terry
 
andythefirst said:
To Nillerviborg:

1. Thank you for you reply!
2. I didn't want to be polemical with LC Audio. Not at all! Instead I consider the Firm very serius and competent!
3. It is than sure I wrote to the wrong e-mail (jp@lcaudio.dk).
4. I will keep your one for the future. Thank You!

To all, thank you for you kind and valuably replies! I will keep them, and use, and I'll let you know...

Whishes!


- Yo Andy (1st.) ;) - No worries, mate, if it seemed so, i did not mean to be rude to you!! - I know exactly haw it can feel to be asking into oblivion. Your question might suddenly by it self appear from the mail server, without any logic explanations - Weird, Ay?? - Then i might answer it again...

The LC Audio company should be serious and competent, but unfortunately the company is under serious changes, both in production, goods and strategy. The change is absolutely for the better, but i seriously fear that it might get a little worse before all is solved. At present all after sale service is handled by my company in another town, and over the next year, most of the manual production is probably moved to my shop also, as the present location is way too big and way too expensive to maintain. it started use, when all was made locally, now almost all is made out-of-denmark, only the secret stuff is done localy...
Hopefulle the result vwill be a much stronger company with many more goods of high quality, and a very much better service and support!!!

- If you need to know more about any tech stuff, then don't hesitate to mail me!!

Lars / Viborg
 
LC Audio The End3.1 input impedance for passive application

Hi everyone!
My first post in this excellent forum btw.

My question is similar to the original thread so i take my chances and post it here.
I have a LC Audio The End 3.1, balanced, which wont work properly with a DACT 10Kohm passive attenuator. I had 2 of NLE12 buffers with them, but statics killed one of them.
Now i only use it passive, but it doesnt work right, althogh i somehow like the sound of it. My guess is that the problem is i cant get enough gain.
What values on the input resitors should i replace the original with? THe original are 301E. my "research" says that they are 301 ohm (?) with 1% tolerance. I have ordered 4pcs 5K and 10K Vishay's, but i actually have no idea what i'm doing here :xeye:
Help would be very appreciated :)

Many thanks in advance!
 
Hi Olsen,

I'm a "millenium XP" owner and I don't have an end 3.1 schematic. Are those 301R the serial input base resistors ? If so, don't replace them. It will be of no help for gain setting. You should rather look at two resistors wich are in // with the inputs. Or simply measure resistance between each input and ground with a dmm. If close to 2K, those will have to be increased.

Best regards,
Francis
 
Hi,
If the Dact is a conventional stepped attenuator, then the output impedance varies between 0ohm and [10k+Rs]/4~=2k5.
The minimum load for that range of source impedance is 5times i.e Zin>=12k5
I would try for 20times i.e Zin=50K.
But, only change the 2k0 resistors, not the 301e.

BTW,
does 301e =300r or 301r? i.e. does the 1 represent adding one extra zero to the two preceding digits? not that it makes a jot's worth of difference to the operation of the circuit.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.