X-Can v3 Channel imbalance

[bingcrosby]

Hey guys,

I was listening to my X-can v3 the other day and noticed that at very low volumes the left channel outputs audio but the right channel doesn't. I'm not sure, but this slight imbalance seems to remain throughout the audio spectrum. I actually have an MF A 3.5 integrated amp in the repairer at the moment for this.

Its not a problem with my HD 650s, this doesn't happen on other devices. And I don't think it is a source issue either, it happens on a laptop and tape outputs of my integrated. Any suggestions for repair/realignment?

On another note, I was wondering if you guys could recommend some mods for my A 3.5 CD and integrated amp that a qualified technician can do, like upgrade capacitors, upgrade op-amps in output stages, upgrade transformers, better regulated PSUs, new burr brown IC's, valve output stage for the CD ...

Is a circuit diagram and instructions going to be released for Little Pinkie V3i also?

Congratulations on a wonderful forum and website!

[diho]

Potentiometer is probably the cause.

[Deleted]

bingcrosby
Do you have access to a Digital Multi Meter (D.M.M.) ?
If you do, the potentiometer is normally divided into 2 sections, each with 3 output tags.Taking one section, measure between the 2 outer tags and write down the readings. Turn the potentiometer to minimum volume, and read between the centre tag to each of the other 2 tags.Write down the 2 readings. One reading should be the same as the very first reading you took, and the other only a few ohms. With the DMM connected to the 2 tags that just gave the lowest reading, move the volume setting to half way, and write down the readings.
Repeat the procedure on the other section, in exactly the same order,and post your results.

SandyK

[bingcrosby]

I don't have access to a digital multimeter at the moment, maybe I could convince the technician in the physics lab to lend me one.
I read on the forum that this is a standard problem and is only "solved" by changing the gain resistor to a higher resistance and maybe getting a better volume knob. With changing the resistor, how would I know where I am on the log scale output of the amp, as not to enter clipping region?

[Deleted]

bingcrosby
Changing those resistor values will not correct channel unbalance. It only reduces overall gain so that maximum volume setting isn't so abrupt. I presume that you are referring to the V2 with those suggestions ?
SandyK

[bingcrosby]

Sandyk

I am referring to this discussion and in particular PinkFloyd's first reply.

rockgrotto.proboards.com/index.cgi?board=q&action=display&thread=4204

Is there any standard way of determining safe volume knob ranges in which clipping doesn't occur? It can be hard with to tell a bit with speakers becuase as you drive them a bit harder they increase their distortion levels too.

Thanks for your help btw.

[Deleted]

bingcrosby

The X-Can V3 circuitry may not be the same as in the V2 that the gain reduction mods were designed for. The V3 may indeed be similar, but I would need to see a schematic to be certain.
I posted the original gain reduction mods based on the V2 schematic.

SandyK

[PinkFloyd]

ALL of the X-Can amps (be it V1 / V2 / V3 or V8) have the same channel imbalance at very very low Volco levels... always the left channel..... MF put this down to "crosstalk" and it's not usually a problem unless you listen to your music at zero volume levels.....

The gain tweak helps slightly with the V2.... until a schematic becomes available for the V3 the best thing to do is live with it and accept it's part and parcel of the animal..... surely you don't "listen" at such low levels?

[Deleted]

Mike
It sounds more like poor earthing techniques to me.
If the earth end of a volume control is properly connected to the 0 volts rail, and the potentiometer has SFA resistance at the minimum setting, as a good volume control should have, there is no way that there is any signal to amplify. I agree that the problem is made far worse by the unusually high level of gain that Antony Michaelson has used. If Michaelson's designs are so bloody good, why do so many DIYers modify them, and in particular, why do so many tweaks that you have done end up being used by MF ?
Alex

[PinkFloyd]

Michaelson hasn't designed since the early 80's.... all of the MF gear is designed in China

[Deleted]

i157.photobucket.com/albums/t69/pinkfloyd2007_bucket/DSCF1143-2.jpg[/IMG]

It would seem that listening to this ,has not only given the reviewer a "hard on" , but the amplifier as well , if the photo is anything to go by !

[PinkFloyd]

Oct 14, 2009 21:49:54 GMT @sandyk said:

i157.photobucket.com/albums/t69/pinkfloyd2007_bucket/DSCF1143-2.jpg

[/IMG]

It would seem that listening to this ,has not only given the reviewer a "hard on" , but the amplifier as well , if the photo is anything to go by ! [/quote]

[bingcrosby]

Maybe this is a part of all MF products as my A 3.5 integrated has this problem, as did a A5 integrated I listened to, and a previously ultra high end valve pre-amp did (I can't find the name atm, as the MF site in under reconstruction but it was early 2000's). So all their design's come out of China? Wow. I think its Cyrus or Naim for me next time - but they cost sooo much more.

Surely this isn't a property of the circuit design? And although I listen on the X-cans v3 at about -54 dB to -46 dB, I feel like with my integrated that perhaps the stereo image is slightly off centre, especially with vocals. And if I can here this then I don't think it can be entirely contributed to crosstalk.

If the repairer can't fix my integrated, do you guys know of a good pre-amp (valve or solid state, kit or off-the-shelf) I can use to drive the integrated with (it has pre-in's, which could be my saving grace)?

Cheers.

[traf]

Hi Bingcrosby

Oct 14, 2009 6:44:18 GMT bingcrosby said:

On another note, I was wondering if you guys could recommend some mods for my A 3.5 CD and integrated amp that a qualified technician can do, like upgrade capacitors, upgrade op-amps in output stages, upgrade transformers, better regulated PSUs, new burr brown IC's, valve output stage for the CD ...

Amp tweaks:

I've been playing with a couple of MF X80s and an X150 recently. These both share the power- and pre-amp circuits with your A3.5, but don't have such a beefy power supply and the X80 only has 2 output devices per channel, rather than 4 in the A3.5 and X150.

Keeping the first X80 unmodded as a reference, the second was incrementally modded with:

1. Schottky diode rectifier bridge for the preamp and control supply
2. 14 pin socket for opamp rolling in the pre-amp stage (I'd previously experimented with 14-pin opamps in my VCan so know which ones to try..). Poly bypass caps fitted between the supply pins and ground.
3. Most of the power supply and pre-amp stage electrolytic capacitors replaced with Panasonic FM&FC that I had lying around.
4. The big power amp supply caps were replaced with Panasonic TSHA

The other X80 and the X150 ended up with only the Schottky diode rectifier and opamp socket upgrades.

Regarding opamp rolling, I finally settled on the OPA4227 with the amp feeding my Castle Severn 2s. The LME49740 I tried gave a little more detail and extension at the top end, but appeared to suffer from a subtle dip in the upper bass region to my ears. Using two LME49720HA wired to fit the 14pin arrangement improved this but the 4227 still provided a more natural overall balance into my speakers (the answer may of course be different when I finish mending my Quad esl63s). I am toying with using 2 Audio-GD Earths wired into a 14-pin socket as I've found this 'opamp' to provide a very balanced yet detailed sound in the DAC output buffer stages of my VDac and KW250S.

The 2 modded X80s sound very similar if not identical suggesting that the capacitor changes do not affect the overall sound to the same extent as the other mods.

The fitting of the TSHA had previously made a significant improvement to bass power and extension with an XPreV3 I once owned but in this application their effect was not noticeable.

The X150 sounds bigger and more powerful which would be down to the bigger transformer and possibly the additional output devices. All are a definite improvement on the stock versions though - lots more detail & extension, more natural and less fatiguing sound (these two traits don't always go together). Fortunately the 'big' sound is still there, which is what I really like about this family of MF amps (my KW250S and the A5 have basically the same power amp circuit as the A3.5 and X150, just MUCH bigger power supplies and nicer boxes).

In theory you could of course fit bigger (possibly outboard) transformers to your amp but this gets expensive and might require bigger heatsinks if you regularly push it hard!

CD tweaks:

Again I think the biggest gains result from a) improving the low voltage power supplies by fitting Schottky diodes in place of the low voltage bridge rectifiers and b) opamp rolling in the DAC output stage. Although replacing capacitors here and there can yield improvements, its a bit hit and miss and you have more control over the sound by changing opamps.

The clocks used by MF are pretty good so I'd leave these well alone.

I'm not familiar with the DAC output stage of the A3.5 CD but fitting IC sockets will again allow you to play tunes with the opamps used for I/V, filtering and output buffer. I use LME49722MA on browndog adaptors for I/V+filter, Audio GD Earth or LME49720 for output buffer, but it is of course easy enough to try others to get the sound you want out of your system in your room.

You could of course just add a seperate DAC. The VDAC responds very well to (some pretty substantial) tweaks that enable it to be turned into effectively an XDacV8 without the tube output stage and nice box.

Valve buffer

As a trial, you could knock up a headphone jack-phono plug adaptor and use your XCanV3 as a pretend X10V3 valve buffer (adjusting the volume control for zero gain). The circuits are pretty similar and it would give you an idea of any improvement to be had. You could also try feeding the headphone socket output into the HT input of your amp, bypassing the preamp stage in the A3.5 completely. Mr Michaelson is always going on about his headamps being so good, they would make brilliant preamps, so it may be worth a try Indeed I think the XCanV8p has preamp outputs.

hth, cheers & good luck!

simon

[bingcrosby]

Thanks Simon

I am interested that you are planning to drive Quad ESLs with the MF amps, as I read in a hifi test for my A3.5 that the amps fuses were blown at 2 ohm speaker impedance measurements (on a resistive bar I might add). The repairer also said sometimes they change the caps and transformers with a standard service because they don't want them coming back with blown ESR 85 degrees components, when it wouldn't have cost that much to change to Panasonic ESR 105 degrees.

Would a more highly regulated power supply exist that would allow low impedance drives - maybe something inbetween an A3.5 and A5 that would allow for replacement in the same unit? If not could I fit more heatsinks in the place of the transformer and put it in an outside case, or even make the thing completely dual mono? And if there is more current, will this shorten the life of the PCB? I couldn't imagine so because I think these are standard from A3.5 to A5.
Finally, would I need better transistors for the amp?

Thanks heaps Simon,
Peter.

PS I like the big sound too of the MFs. If only it could be more Cyrus/Naim like in other areas.

[traf]

Hi Peter

One of my X80s was happily driving a single ESL63 speaker to surprising levels the other day without any problem - I don't think the 63s impedance goes down quite as far as 2Ohms anyway - the original 57 might. I just need to get around to rebuilding the 2nd 63....

As regards transformers the +-15V side is ok. Its only driving the control electronics and the quad opamp. If you were to tackle the power amp transformer side you could either keep the existing tranny and just use the low voltage windings, or put a dedicated +-15V / 30VA or bigger tranny in its place

The power transformer side is something different though. A good route to better regulation is to use a bigger transformer (which also allows for more power). It would be quite feasible to use two big (300-500VA) transformers, knock up a clone of the rectifier/smoothing stage of the KW250S/A5 (rectifier, fuses and capacitors) and, providing the heatsinks were up to the job, arrive at the same power amp stage as the monster amps (ie 250Wpc). The output devices & circuit are fine as they are (except for the output inductors which might need to be rewound with thicker wire and as mentioned before the heatsink may need to be upgraded to cope with the extra heat if you use the extra power available).

Now whether all this would in fact make financial sense is another matter Off the top of my head:

2x300VA transformers (I think my KW uses 75V windings - the X150 and X80 are only 35V) = £70
New/bigger bridge rectifiers £15
Additional/replacement heatsinking = £20
New power supply capacitors - 8x80V, 4700uF = £40
New box = £25+
Other bits = £10

So that's less than £200ish to bring an A3.5int up to A5ish spec (albeit without the phono stage) BUT the difference between 2nd hand A3.5 and A5 amps is probably about £200-£300 so it may not be cost effective to go down this route. The A3.5int is quite a rare beast so may well carry a premium on the 'bay, especially with the state of £/Euro exchange rate.

hth

simon

[bingcrosby]

Simon,

What approximate value do the A3.5's reach on ebay?

Do you know what value the stock transformer is? Would it be possible just to go halfway and update to say one 350 VA and clone the rectifier/smoothing stage of the A5 in the same box? Can you also get a say 90 % efficient as opposed to 80 % efficient transformer with the same output (like for a computer PSU)?
And do you know why that the fuses may blow for low impedance drives?

Also, have you had any channel imbalance experience with the MF amps? I was thinking about maybe using a passive attenuator instead if the tech. can't fix it.

Kind of trying to get the amp right - I'm thinking of getting some audiovector S3 signature's in the next few months maybe, hopefully ... (they are 91 dBm 8 ohms nominal so shouldn't be a problem)

Thanks Again.

[traf]

Hi Peter

"What approximate value do the A3.5's reach on ebay?"

- There is an X150 on the bay at the moment which is up to £445 and I would expect an A3.5 to go higher than this (provided you're happy to send it to the continent where the euro makes it makes it an attractive buy). In contrast a 'UK only' sale of an A5 might reach £700+

"Do you know what value the stock transformer is? Would it be possible just to go halfway and update to say one 350 VA and clone the rectifier/smoothing stage of the A5 in the same box? Can you also get a say 90 % efficient as opposed to 80 % efficient transformer with the same output (like for a computer PSU)?"

- No and MF are unlikely to tell you, although you can get an idea of the total VA for all the windings from the max power consumption figure on the back of the unit. Unfortunately, the physical size of a toroidal transformer isn't neccesarily a good guide to its VA rating either as core size can vary a lot.

"And do you know why that the fuses may blow for low impedance drives?"

- Purely a function of a higher current into low load. It may also happen if the wrong fuses are used of course.

"Also, have you had any channel imbalance experience with the MF amps? I was thinking about maybe using a passive attenuator instead if the tech. can't fix it."

- Thankfully not

Kind of trying to get the amp right - I'm thinking of getting some audiovector S3 signature's in the next few months maybe, hopefully ... (they are 91 dBm 8 ohms nominal so shouldn't be a problem)

Thanks Again

- best of luck

Simon

[bingcrosby]

Hey guys,

I noticed some tweaking warnings from MF about increasing jitter by changing power supplies to DAC's, and creating instability and oscillations in amplifiers and damaging speakers on these pages,

www.musicalfidelity.com/technical-and-sonic-overview.asp
www.musicalfidelity.com/tuning-information.asp

I'm assuming this is not the case and that MF are just trying to scare people into using their uber-expensive service. Or do their warnings carry some truth?

Cheers.

[robertkd]

Now aren't musical fidelity full of their own shyt, not only do they sting the punter when they buy there overpriced merchandise, they say to the punter well we can make it even better at a price,...

Reality likely some back yarders will ferkem but to imply that only they could get it right is a joke as there are a lot of telented people out there. Krap they would have done it 'perfect' the first time,... it's not what you market, it's how you market it.

MF are seriously full of their own shyt,...

gets of soap box and pushes it back into the corner ;D

[PinkFloyd]

Bloody unbelievable, they have copied most of my tweaks to the letter (which I have been doing for years) and have the cheek to warn about "third party mods" ;D

[bingcrosby]

PinkFloyd,

If its any consolation, I heard of a guy who bought a big kW amp (550 integrated I think), had it tweaked by MF before he got it, and instantly one of the channels blew when he plugged it in. Apparently the local tech. told him they had shorted it by mistake, but luckily none of the components were damaged.

"Understand the full technicial ramifications of any modifications" - when I learnt how to make a series and parallel circuit when I was 5 I learnt not to short it, although sparking 1.5V batteries are fun. Not to 4 or 5 figure speakers though.

And even if jitter was increased, amp. instability etc., (which I don't believe) if it sounds better with tweaks doesn't it have better "Musical Fidelity"?

Peter.

[ignatius]

Edited this bec. my problem was easily solved.
It was just the phono jack so it was trivial

I now have to train my ears to appreciate the differences a h/phone amp makes to the sound quality. That may take a few sessions, and some A-B testing.

Thanks, Ignatius

If any kind person wants to email me the MF XC V3 schematic it would be appreciated. Inevitably there will be a more serious fault. One day.