Oxford Crystal Reference restoration

[peter]

Interesting you say that Peter

Now i'm just wondering if it's worth modifying the Krell to keep those digital circuits running & at a set temperature? I have been thinking about this for a while

If you can't tell any difference hot or cold then it's probably not worth the hassle. You need to conduct some listening tests. I perceive it as bad imaging and fuzzy soundstage.

I am only just beginning to get into this but looking at the data sheets it seems that on the 94, the SAA7210 does interpolate up to 1 uncorrectable error, and the SAA7220 up to 8 sequential errors (based on EFAB (pin 4) which can be nailed). Grounding that pin should (I imagine) cause an audible "tick". If not a LED, scope or counter can be connected and the errors observed during warm up (or other tweaks), all with the intent of reducing uncorrectable errors (unreliable samples).

On other Phillips ICs I have seen, the error information is a data stream similar to to serial data, with a number of bits positions in the error word indicating more detailed information, but on the SAA7210 it simply shows that the current sample is "reliable" or not.

It could be that if one has bad tempco parts on say your Demod PLL filter it is less effective when cold at maintaining lock or similar. Also that the various adjustments were made when the transport was at operating temperature, and the transport is sub-optimal at other times (or the supplies are not stable etc).

I think the CIRC does not intruduce any artefacts but the interpolation (error concealment) does, and this is the only way the sound can be effected (as from there I go via coax to an offboard DAC, permenantly powered). After all, the concealment/filtering is there to hide itself. This just makes it more difficult to detect by ear.

Perhaps unreliable samples are much more common than we think, that it takes time for the brain to form the soundstage perception, and that process is spoiled by "bad information".

peter

[peter]

Its the TDA 1541 DACs they improve when fully warmed up.

Well, I wasn't using the onboard DAC at first, but what you say sounds promising, and I'm looking to improve that later with some tweaks that Dave mentioned, The clock mods look most promising for improving the HF and low level details.

The NOS hacks I've seen are also interesting, but mild oversampling seems less evil than aliasing artefacts that I know are very bad when they occur in visual images, especially when we are trying to improve the high frequencies.

This is all very interesting to me, there are just so many things that need work...

peter.

[Spectral Morn]

Well, I wasn't using the onboard DAC at first, but what you say sounds promising, and I'm looking to improve that later with some tweaks that Dave mentioned, The clock mods look most promising for improving the HF and low level details.

The NOS hacks I've seen are also interesting, but mild oversampling seems less evil than aliasing artefacts that I know are very bad when they occur in visual images, especially when we are trying to improve the high frequencies.

This is all very interesting to me, there are just so many things that need work...

peter.

One area of weakness is the thin bottom plate. If you have access to quarter inch aluminium plate then replace it with an exact copy just much thicker. I did this well over twenty years ago and it improved the sound a lot. I had access to a workshop back then.



Regards D S D L

[peter]

Something else intriguing:

SAA7220 p7 Filtering

"to simplify the design of the digital-to-analogue convertor a d.c. offset of +5% is added to the accumulator".

What? How? Where? When?

What?

"Audio samples are available for DAO after interpolation, attenuation and muting, but before filtering"

Oh ok. that means that the DAO is not oversampled. Is that always the case? Thinking about it, it seems like it must be as the transport does not know what kind of anti-alias filter is used in whatever D/A it is connected to.

[peter]

If you have access to quarter inch aluminium plate then replace it with an exact copy just much thicker. I did this well over twenty years ago and it improved the sound a lot. I had access to a workshop back then.

This is for rigidity not weight, correct?

Or can I just rest/fix it on something really really heavy. Does it have to be bolted on, and is the material important.

I ask because I ditched the horrid plastic feet and replaced them with thick black sorbothane isolation disks/feet that were all the rage in 1988. These isolation feet are now well glued onto the case and the wooden equipment table shelf, and I fear would not survive a seperation.

The isolation is very good though

[Spectral Morn]

This is for rigidity not weight, correct?

Or can I just rest/fix it on something really really heavy. Does it have to be bolted on, and is the material important.

I ask because I ditched the horrid plastic feet and replaced them with thick black sorbothane isolation disks/feet that were all the rage in 1988. These isolation feet are now well glued onto the case and the wooden equipment table shelf, and I fear would not survive a seperation.

The isolation is very good though

Hi Peter

The thin plate is a source of vibration. To my mind at the time it seemed daft to build the innards of the player into a nice heavy die cast box only to compromise that with a thin plate, which would rattle.

The one I made is bolted on using the original screws, its also brushed and painted with a varnish.

I used the original feet but filled them with Miliput which is an epoxy mix that dries solid and like clay. This was better but cones are more suitable imho.

However it sounded best with cones (point pointing down)...one directly under the transport, in the middle, one under the mains transformer and a cone to balance. Three is better than four and even with the new bottom plate you could still hear an improvement in sound quality as you move them around. Once you know where is best fix them with double sided tape. The effect with the thin plate base was more pronounced.


Regards D S D L

[DSJR]

The TDA1541 based machines I tried at the time (Meridian, Marantz/Philips, Quad 66, Arcam) were very sensitive to rf muck coming in on the mains and I'm told the machines themselves could also spray loads of muck out via the screens of the signal leads, the AVI S2000 DAC being a notable exception. use of cable ferrites and a high current mains filter really works, together with getting the 2 pin mains plug round the right way. I'm stunned how good my CD player still is and the transformer coupling to the output helps too.

[peter]

the SAA7210 does interpolate up to 1 uncorrectable error, and the SAA7220 up to 8 sequential errors (based on EFAB (pin 4) which can be nailed). Grounding that pin should (I imagine) cause an audible "tick". If not a LED, scope or counter can be connected and the errors observed during warm up (or other tweaks), all with the intent of reducing uncorrectable errors (unreliable samples).

OK folks, some results

On a warm player, playing an average disk, I get no movement of the EFAB pin except when starting/stopping switching tracks, FF/REW etc. I can get errors by interfering with the disk rotation physically, but no reasonable amount of shock causes the slightest problem.

With a deep scratch applied to the disk using a screwdriver (to a CD-R copy I hasten to add), I typically get two error samples per rotation. If I ground the pin (so the error indication does not get through), there is no clear audible artefact (the sample will just be a repeat of the last one), so no easy way to hear the errors.

Other signals indicate that the transport is basically giving two fingers to any attempt to disturb it , the PLL filter and motor speed waveforms are rock solid and show the transport is in control. I can apply various patterns of scratches to the disks and I can see these as dropouts on the scope but the CIRC is doing its job perfectly.

Judging from reported variations in transports (and own recent experience) I was expecting the error concealment signal to be chattering like a geiger counter, but this is not the case.

I will repeat the tests tomorrow when the machine is cold.

On the analog side I can see how the various mods being discussed would improve the D/A section. While the D/A chip has a good spec, I doubt it acheives its potential in the environment it is placed, on that single sided PCB which is a joke in todays multi-layer, ground planed, EMC aware world.

It beats me they did not put the entire section in a metal can.

peter.

[peter]

The thin plate is a source of vibration.

Normally I would be suspicious of this kind of thing except I came across my own box issue with the Primare (which has stalled further work on turntables and the like).

Martin has pointed out that the worst kind of noise is HUM, and I have to admit that this was a problem (its audible between tracks) so I tried to reduce the level of hum by improving the filtering in the PSU, which has about 800mV of ripple on the +/- 30V supplies. Changing to a pi filter arrangement before the regulator reduced the ripple voltage to 100mV but actually increased the hum because the ripple current had gone up as a result of the extra capacitors. In fact everything I did increased the audible hum. After regulation there is no measurable hum, so it's probably induced anyway. In the end I decided I would build a proper hefty external supply later and would live with the hum for now as it didn't seem as bad as I first thought.

As I slid the steel cover back on I was still listening to the preamp and noticed that the more the cover was in place, the louder the hum was getting. It also changed the sound of the hum. Download from here (hum.wav 3Mb)

Primare MC input,Hum and Noise, with cover

Now, mega-gain has been applied here. plus the preamp is working at maximum sensitivity so this isn't quite as bad as it sounds.

It's as if the hum leaks out of the box without the lid but is trapped inside with the lid on.

I thought I could filter it out with a simple 50 and/or 100 Hz notch filter but if you try this you will find the hum is still there. This is probably because it's ripple, which is a sawtooth waveform, not hum, and stretches all the way up the frequency band.

peter

[Spectral Morn]

There was a noticable improvement in sound quality.


Regards D S D L