+ Reply to Thread
Results 1 to 10 of 66

Thread: How hi (-fi) can you go?

Threaded View

Previous Post Previous Post   Next Post Next Post
  1. #1
    Join Date: Oct 2012

    Location: The Black Country

    Posts: 6,089
    I'm Alan.

    Default How hi (-fi) can you go?

    An EE’s progress over the last few months for system improvement.
    A lot has happened over the past months with sometimes startling success.

    Firstly I want to champion the case for synergy. It’s not something that you can order or compute but it can have such an impact on ultimate results, particularly if you haven’t got it. For example a power amp not coping with an awkward part of a specific speaker load, it doesn’t need to be over a wide frequency range to upset things. Or a mismatch of impedances between pre and power, or with a passive pre.

    Certain design approaches can mitigate possible problems, in general low output impedances and high input impedances go a long way to help prevent problems.

    The other important factor IMO is having a chain of components that are of equal performance, such that improvements throughout the chain can be heard and appreciated. Having low distortion wide bandwidth amplification helps here.

    I love valves for small signal amplification. Choosing the best can be tricky, see the developments made further on. Good circuit design is crucial for best results.

    As a techie I do have test equipment (obviously) but only use it to confirm functionality, check bandwidth and measure gain figures. The ultimate tool for performance evaluation is a good pair of ears and here I have a secret weapon. A younger and more acute pair of ears by the name of Bigman80, who has been extremely helpful, apart from pushing me down the path of lightening the wallet to achieve the improvements.

    A quick run down of the kit in use and some background details.
    The speakers. The ubiquitous Quad ESL57, not perfect by any means but what they do well is beyond reproach. Exceedingly low colouration and no enclosure resonances. As a lot of folks say, mid-range to die for, but the extremes can be bolstered as shown later.

    The power amp. My own design dual mono valve hybrid using a valve for amplification and MOSFETS for output drivers. Good low output impedance, a bit of overkill with 90W driving Quads so easy going on the output levels. Not much to look at but the case was a donor.




    The sub woofer. A copy of the REL Stadium, a kit enclosure from Wilmslow Audio many moons ago fitted with a 10 inch car subwoofer with a good response down to 25Hz.
    Downward firing, the legs are the originals from the ESL57’s.




    The sub amplifier. Rather than build it into the sub enclosure I chose to have it in a separate enclosure, the original one wasn’t pretty and the twin toroidal transformers had an annoying mechanical hum. So recently I rebuilt it into an AIR case and replaced the mains transformer with a single quiet one. MOSFET output putting 50W into 4 ohm.




    The preamp. My own design KIN valve hybrid with ‘super linear’ triode amplification stages. The unit uses a mu circuit that tricks the valves into thinking they are in valve utopia. The result is better linearity, more gain and vanishingly low distortions. MM and MC capability via a JFET cascode input, improvements later.




    The turntable. A humble Goldring Lenco GL 75 in stock form apart from new V blocks and a delrin ball in the main bearing. Oh and additional weight on the counterweight necessary for the fitting of a DL-103, it’s a heavy beast.




    The cartridge. One of the recent additions, a fabulous ZYX R-100. It does everything that the 103 does with fantastic drive but so much more. More detail, separation, frequency extension and image depth, fabulous. A bonus was that I didn’t have to remove the extra counterweight, with the combined weight right up against the bearing assembly the R-100 balanced perfectly. Synergy or what?





    So the improvements over the last few months:

    The purchase of the Z-100. I won’t be looking to better this, all the detail and tremendous imaging are first class. Helped of course by the balance of performance with the rest of the kit, see further on.
    This really is a stella performer, with the final improvements the music detail and focus are some of the best I have ever heard.
    ___________________

    Re-building the sub amplifier. I managed to lower the bleed through of the internal variable clock and the background noise as well as removing the transformer hum.
    ___________________

    Adding tweeters to the ESL57’s. One of the limitations is the treble dispersion and adding the additional tweeters has so overcome this. The crossover components are mounted behind the tweeters and the variable L pads at the bottom rear of each panel within easy reach for adjustment.






    ___________________

    Re-visiting the design of the KIN preamp. After attending the bake off at Huddersfield it was realised that the performance was lacking and could be improved.
    One comment made was a lack of downward dynamic range (DDR), a term used by Nick Gorham. It’s difficult to detect if you are not sure what to listen for or it’s not something that is pointed out. An analogy is contrast in a picture, you might get a lot of brightness in the picture but without portraying the darker areas correctly.

    After consulting the design notes for the mu circuit I realised I had made too much of a compromise in limiting device dissipation, both valve and FET. Increasing the current levels brought back the DDR. I also lowered the anode voltages to better match published characteristics. More image depth and better dynamics ensued, this really was a worthwhile step.


    Another initial compromise, as it turned out, was the choice of valves to use, originally chosen to enable JFET cascode operation for MC, with a very simple circuit. (KISS principle)
    Anyone who has had any Firebottle kit will know I avoid the most common valve types because obtaining the best examples can be fearsomely expensive.

    As it turned out the PCC189 originally used is bettered in the linearity stakes by the PCC88, so with the ‘super linear’ circuit the improvements are doubly good, as has been heard.




    So now we have PCC88’s fitted, getting top class NOS examples is easier. I have a Telefunken and a Tesla fitted to the KIN and 2 Tesla’s in the latest Vivant build.
    It’s not that I wasn’t a believer in tube rolling, as everything matters, it’s just that I have been trying for the best sound without reliance on the best performing valves.
    So when Bigman80 insisted I go down this route I was very pleased with the improvements made. Again more image depth and width and more detail, a palpable sense of placement accuracy and focus.
    ________________________

    One more big improvement to come, but some details of the stand alone phono amp design in the latest Vivant build first.
    The circuitry is essentially identical to that in the KIN after the improvements, but with very minor component value differences in some areas. Now fitted with Tesla PCC88’s it is indistinguishable from the performance of the KIN.








    Exceedingly good performance but ….. another small compromise was initially made to keep the MM to MC switching simple (KISS principle again). This was potentially restricting the ultimate performance achievable with a JFET input, as the operating characteristics couldn’t be altered. So to explore the alternative of using an SUT directly into a valve MM input (just the flick of a switch) I purchased a Denon AU-320 SUT.




    Discussion and testing has been undertaken on the loading of MC cartridges. As resistive loading is altered then different cartridges have different changes to their performance. Some seem to lose focus above say 300 ohm yet others are happy with loadings up to 1000 ohms. What I don’t quite understand is how a reflected impedance (through the SUT) of 470 ohms can can have such a dramatic improvement on the whole sound stage of the Z-100 in particular. I know it is now a reactive source (the cartridge coils) into a reactive load (the SUT) but the improvement is literally phenomenal.

    The design mods to the KIN gave good improvement, the change to PCC88’s gave good improvement, the change to quality NOS gave good improvement, but the change to quality SUT into MM has given an improvement equivalent to all the previous combined.

    I’m staggered, over the moon and an extremely happy bunny all at the same time.

    It’s brilliant.
    Last edited by Firebottle; 06-11-2017 at 08:07.

+ Reply to Thread

Posting Permissions

  • You may not post new threads
  • You may not post replies
  • You may not post attachments
  • You may not edit your posts
  •