The Google machine is your friend here. All I could do is say what's already out there and readily available. Suffice to say, a 741 is an old op amp with a 1MHz gain bandwidth product. The 301 came before that. The 301 was used all over the place in things like the vintage ARP 2600 synthesizer. Ever play one of those? Ever tried keeping one in tune for more than 10 minutes?
Unboxed a world designs kt88 integrated Amp kit just now, to build for a friend. Everything looks good. Will report back on how it sounds.
Nad C515 minor update, Just changed the four power supply capacitors on the op amps from 100uf to 470uf. No great change I could be certain of. But my goodness I am so enjoying the moded 515. It is wide open, diction is all clear and I'm hearing new sounds in everything. I took some internal pics of the C538, the current up to date model, but for some reason PostImages has blocked me so I can't show you. It is as I feared though, it's all gone surface mount micro circuitry. The power supply is switched mode, the main board is about 4cm square with surface mount micro components that can only be soldered by ants wielding micro soldering irons, can barely see the resistors. The capacitors are some brand I've never heard of certainly no Elna, no polymers at all. There's no service guide available so no circuit diagrams.
Many folks love the sound from the small AR-4x bookshelf speaker, 1965-1974. Just a 2-way. Many buy them and refurbish them. Problem with them over the years is failed tweeters. Original AR tweeters aren’t made; they have to be repaired or working tweeters salvaged or bought as a take-out from damaged cabinets. It’s a 2.5 inch paper cone tweeter. A common replacement many use is the inexpensive phenolic ring tweeter (PRT). About $20 each. They are a drop-in replacement, fit-wise. But they are a bit harsh compared to original AR tweeters. So I decided to buy a pair of these PRTs, to see for myself why they don’t sound very good. The surround at the cone edge is too stiff, not allowing the cone to move. That sets up some nasty artifacts, as the cone center moves, but not the cone edge. So I decided to fix that. As-found AR-4x speaker: The original AR tweeter: The Phenolic Ring Tweeter (stock): PRT surround removed, wool felt added under the cone to attenuate the back wave: A new surround and seal. Made from high quality silk, treated with butyl rubber. Very flexible, but damped. It will allow the cone to move as it needs to, keep it centered, and damp any ringing or “bouncing” from the voice coil movement. How does it sound? Smooth and clear. The original PRT has a harsh, fatiguing sound that wears on the listener after a few minutes.
That's dedication, apart from the nice wood I would be tempted to lob the rest in the bin. I wonder if some armour all or similar treatment would loosen up the woofer.
It’s all good. The walnut veneer cabinets will cleanup nicely. The drivers will clean up nicely as well. A microfiber towel and a little isopropyl alcohol will clean off the dirt. Doped cloth surrounds, so no refoam job necessary. Replace the caps and level control pot. Sand and repaint the tweeter face. These still stack up well against modern bookshelf speakers once repaired.
Just auditioning the tweeter… Mounted tweeter disabled, midrange disconnected, wires run external up to tweeter from the crossover: Mounted in foam on top of the AR-2ax cabinet:
Just a YouTube clip. Sound switched to mono. AR-2ax tweet and mid disabled. My project tweeter on top of the cabinet, set in foam, using the internal high pass mid-tweet crossover. A little Emmylou Harris using the Shure V15-III cartridge with original Shure nude elliptical stylus:
BTW, for those that like to see measurements, below is what the latest version 2 phono stage shown above looks like in REW, top is the frequency response from 20 to 20kHz, both channels, and below is the distortion and noise. The scale on the frequency response is 0.1dB per division, so response is approximately +/- 0.1dB from 20-20k, and left/right channels are within 0.1dB at low frequencies, rising to almost 0.2dB at high frequencies. The distortion is almost all second order, as expected in a single-ended design like this with no loop feedback. It's pretty low at this level, but will rise at higher levels. I've had a little problem getting completely rid of the 60Hz noise, but it is pretty low here. The measurement is performed with a RIAA calibration file. The phono stage gain is set to about 42dB ...
cheapo dust cover Finish closing it up when I got time. Maybe never. Better than my old shirt and it's good for when it's playing.
Just taking things to the next level! I was planning to just make a new top board for my old Xerxes as a straight copy of the original, but in beech block board. Then looking at Rega’s ideas of reducing mass while increasing stiffness I came up with this. However wood is not ultra stiff laminate bonded to polyolefin foam! With this in mind I didn’t think that copying a P8 or 10 plinth with it’s large void cut outs would work very well as wood tends to flex the thinner the section. Instead I looked at mass reduction using variable hole cuts to also intercept resonance while maintaining stiffness by virtue of thickness and retaining the strength of a honeycomb...of sorts! A lot of head scratching went into this arrangement, but there is a fair bit of sanding and shaping to do yet. The new board retains all the same geometry of mountings as the original, which enables me to drop it straight on to the suspension towers of my existing lower chassis. If the results are good enough I will proceed with an all new base and motor mounting to match the new aesthetics. Hopefully I can maintain or improve on the current motor isolation by cribbing a few ideas off later Roksan’s. This was a late night lash up so early testing has yet to reveal all the strengths and weaknesses when I turn up the volume, but initial impressions were good. There is some fine adjustment and tweaking to do yet and a nice rich wood stain to finish if this design is a keeper!
It will look pretty cool, even if the performance isn't on a par with the new Rega Naia! You need to do a vibration modal analysis to determine where best to make the holes
Thanks Davey. Some years ago I did a crude vibration test of the standard top board by making my own jig and covering the board with salt granules. By attaching a motor close to the rear isolator support point you could see the granules move away from the most vibrant areas. However for this new board I just ‘winged it’. The holes vary in size and are positioned within the cut out shapes of the latest Rega skeletal models in an attempt to tap into some of those elements. Basically I cheated! I could also experiment by plugging some holes with denser or damped materials to see how this might influence or ‘tune’ the sound. The danger of wood is that it can sweeten the sound at the risk of homogenising everything, but I prefer what beechwood does in preference to MDF! The circular arm mounting cut out allows trial use of other materials so that will be my next objective. Another feature is that the arms of the star shaped plinth will have the edges flat sanded to give the plinth a trapezoidal section alternating up and down. Everything can make a difference, including the jig saw and belt sander which dramatically speed up the creative process. I also bought a complete 3 metre beech worktop so that I could make a series of these boards to experiment with. However, listening to it again this morning the backgrounds do sound nice and quiet so I may run with this one.
Ok, time to review some results of the new design of top board to my Xerxes after extended listening of a wider range of LPs. First the good points; clarity has noticeably increased and retrieval of low level detail is very apparent. Resolution of acoustic guitar and other stringed instruments is spectacular. I am also hearing a layer of information on some records that was previously blurred into obscurity. So the theory of a lighter board with resonant dispersion holes appears to be working. Negative points? The mid to upper bass has become slightly recessed while very low level frequencies are there like a sub woofer has been introduced. This gives an illusion of a complete frequency range on some material, but there is the ‘suck out’ of the kick drum, which is possibly the influence of a solid wood board albeit a composite of alternating grain blocks. Is this the trade off in some low to mid frequencies in return for clarity that solid wood is creating? Solutions? The Corus Black stylus I am using here can be prone to ‘thin presentation’ in given circumstances. My 2m Blue for example does a nice line in mid LF dynamics in the Tabriz arm on the standard top board. It would be nice to drop in a warmer bass heavy cartridge with a rolled off HF, but the turntable / arm combination has to work with a wider range of transducers. I don’t want to fix a broken leg with a sticking plaster to coin a phrase The current results were on my DIY three tier chassis with TMS Mk3 blobs on top of 9kg rated domed isolators. I still have a standard two tier chassis with the standard M5 screw insert mounted Xerxes Mk 3 blobs and to try, but my suspicion is that some sort of composite or mass loading to the new top board is required to neutralise the interference in the bass response. Another solution might be to remake the top board altogether with less mass reduction and honeycomb effect and just a few strategically positioned holes? There are some things I can still do to this ‘star shape’ such as finishing the trapezoid edging and changing the arm mount disc to another material such as ebony or teak to tune out the excess of beech. Our local discount store also has some solid Acacia wood cheese boards of near enough the right thickness!
What a difference a day makes? In my quest to fix the slightly ‘unbalanced’ sound the above beechwood top board I was reading more papers on sound wave propagation and absorbency tests on different woods. Alas after much searching there was nothing about the influence of shaped wood on resonance to support my experiments. I had already arranged for the direction of the wood grain and bonding of the wood block in my new board to run front to back of the turntable to reduce motor breakthrough from the rear suspension/isolator mount towards the arm base. Then I found a research paper that found acoustic energy (or vibration) travelled ‘across grain’ better than longitudinally in hard woods. Doh! Whatever, I have used cheaper mass produced kitchen worktop board which would not come from the finest or most seasoned stock, but the arm board insert comes from my stock of 30 year seasoned wood with a noticeably narrower growth ring. Given the cheaper wood (around £18 or $22) for this 400 x 450 MM slab, I decided that I was being too precious about my creation and had been trying to blend everything into a nicely curved shape by sanding mainly for aesthetic reasons. And so yesterday I stripped the turntable down again and set about altering the new top board again with my drill press and belt sander. If my idea of creating an angled edge section was to have any chance of working then I needed to go more extreme. Below is part of the current enhanced if rough finished result adjacent to the front right isolator mount (under the faint pencil marks). 0 At the rear of the board shown below, I bored an extra hole between the rear isolator (under the pencil circle right of the platter) and arm base and planed more of a bevel. There are more alternating bevelled edges out of sight in this view. The grey cap at the bottom of the picture is an old film container lid, which previously formed a silicon damping dish for the original Xerxes top plate (now surplus to requirements). Vertical antenna is from my FM tuner on the shelf below. Last night after vacuuming away all the dust I re-assembled the turntable. The standard Xerxes has felt discs stuck inside the isolator sockets under the top board. Aftermarket felt pads are on the thin side so I doubled up the layers. The main bearing is held into the top board by a aluminium collar, which is threaded for three thin bolts that reach up through the board to pull the collar down onto a flange on the bearing housing. Perceived wisdom is not to over tighten the bolts on the standard Xerxes board as this is thought to over damp the sound. I usually pinch up the bolts firmly, but loose enough so that I can grip the bearing housing from below and just turn the housing against the friction of the collar. I think that the untreated beechwood is slightly softer than the veneered MDF Xerxes board and so I figured the bolts needed a pinch more torque for the new top this time round. Today we had a good listening session including; ‘United Kingdom’s Ultramarine’, ‘Baby by Yello, ‘United States of Ambient’ double LP, and ‘Dire Straights Love Over Gold’ original ‘82 pressing. I am pleased to report quite a transformation from the first set up. The sound is now open, clean and refined with more immediate bass slam, within the limitations of the old XPS1 motor controller. The new top board is also a significant step up from my previous Xerxes set up even with the beech wood insert arm mount. All I can suggest is that the one extra hole cut plus more pronounced angled edge facets have somehow cancelled out an interference wave within the board structure!? More careful assembly with attention to details also helps. Use of resonance intercepting holes in plinths appears to be catching on, but bevel edges to reflect resonances are possibly new. If multi faceted forms can work for radar, maybe they can work in audio to repel resonance? If so maybe you saw it here first? No unrealistic claims here over Rega’s excellent decks, but maybe I don’t need braces if I have ‘trapezoids’? The possibilities could enable me to make a new base that can incorporate as much isolation or resistance to noise within two tiers and that performs as well as my three tier chassis. Now that would be cool.
Part four of War and peace! After a couple of days listening to the ‘organic sound of wood’ of the DIY top board for my Xerxes, I began to notice more that the frequency balance was not entirely natural with a low down hump and upper mid overemphasis. This worked wonders with the right material. Natural acoustic works or dark sounding recordings like Vangelis’s ‘Themes’ LP just lit up and hypnotised, but other LPs were becoming irksome for a slight lack of neutrality. That’s the game here in plinth material selection and proportion. It seems that you either have to go fully inert or isolate materials that might colour the sound to excess. My spring suspended decks seem to carry off this trick better than my Roksan did if only because I had grown a perceived aversion to MDF. This material is widely used for it’s consistency, ease of machining and a neutrality.....of sorts. I decided that I needed to add a thin layer of something to even up or ‘temper’ the excesses, however alluring, of my beech slab! I had space for a 1/4 inch (6mm) layer of something, but rejected plyboard for it’s flakey nature. Much as I hated returning to MDF, I reasoned that it would only be a 6mm layer under the 28mm of beech. I also had to cut or plane off the area above the motor mount to obtain clearance, and I wanted something that could be turned around quickly. Below is the freshly bonded layer of fibreboard. At the centre is the recessed hole for the main bearing bolts. This was just one glitch I had to overcome....... While installing the main bearing on the board I accidentally tipped the bearing too far and lost the tiny ball bearing! Before installing the sub platter you should always check. The prescribed methods is inserting a cotton bud on a stick into the bearing housing, which will clearly show the dented impression of the ball....if it’s still there! I would have to order a replacement at whatever the cost today. The previous one I bought several years ago was £25 for a 2.5 MM ball sampled to a claimed 1 micron! As luck would have it I still had a MK 1 main bearing and shaft on the shelf so I decided to give that a go. There is no visual difference between the bearings, but once lubed the older bearing shaft takes less than 2 mins to sink down onto the ball bearing as opposed to 10 to 15 minutes minimum of the later bearing! Once seated the old bearing seats ‘hard’, but the later Mk 2 bearing almost ‘soft lands’ onto the ball, such are the tighter tolerances. That being said, I had used the latest Roksan oil, which is probably thinner than of old. Maybe a drop of Harley Formula Plus would help? By coincidence I had been distracted from hifi matters that afternoon by a Yutube video of the 2024 ‘King of the Baggers’ racing on TV. The sight and sound of 600lb Harley and Indian touring bikes hitting 185mph round the Daytona banking is ‘epic’ to my mind! Below shows the extra 6.3mm layer of MDF under the natural wood after sanding and contouring to a smooth finish. The cut away for the motor mount is to the right. So what did this sound like. Well steangevas it might seem to some, yes it did neutralise the sound to a significant degree. Despite being at the base and not connected to any critical component the MDF layer does nothing, but alter the resonance characteristics of the natural wood above. The suspension towers, main bearing and arm mount are all still bearing on natural beech via the hole cut outs that mirror the board above. Is it better? Well the very deep base is not quite so deep, but is less coloured and the high frequencies do not ‘sing’ as much even though they are arguably also less coloured. I can hear the influence of the MDF, but only time will tell if this is a good compromise. Given time I would like to try a Half inch ply board bonded to a thinner possibly 3/4 inch section of beech!? Plenty more wood in the shed!
Philips cd723 cd player further adventures. Started modding these back on page 17. Have just applied a few more mods addressing the op amp feedback and output circuits. Someone on here suggested the 2 watt vishay resistors were the best sounding, thanks to them, so below bottom right the big resistors on the output. I know resistors can't make a difference well that's what I thought too but they do. Obviously it will never get near 2w or even 0.1w but it's about the sound. The upgraded opamp with it's vishay 220p disc capacitor and vishay resistor in the feedback circuits. I went with the stock l7805cv regulator but read up best practice, it should have 330nf on the input between pins 1 and 2 so that's fitted under board, and ~10uf on the output so below is the 10uf which fitted parallel through the same holes as the existing 100nf. Nichicon kz all round. Heatsink added to the regulator but it doesn't need it. What else, there is a resistor in the supply line before the regulator which I shorted out for max power, the whole board is fused anyway, and 100 ohms before the DAC chip and 220ohms before the opamps, these replaced with 2.2ohm fusible resistors. Sometimes I wonder how the components got any power at all! Added 1nf on the DAC chip and 100nf on the opamp under board. Result, great clarity with very good separation of sounds. Well worth replacing stock feedback components.
Nice work. You could consider replacing the 7805 regulator with a pin-compatible regulator using the more modern LT3045. These newer regulators have much lower noise. You can get a 7805 pin-compatible module using the LT3045 here: Best Low Dropout, Ultra Low Noise Voltage Regulators
OpAmp feedback circuits. On the left the enemy, 390pf of cloudy congestion (ugh why). On the right a lot better. Suggest change these before OpAmp rolling.
A few pages back I outlined my plans for some experimental double tonearm pods. The panels arrived from Acoustand today and it all seems to work. Very heavy and solid, so should be stable.
I finally started my DIY 2 way horn speakers. I bought the PE 15" knock down kit a few years ago and just finally got around to starting the pre-fit. This will have the Radian 475BP CD's and MCM 55-2963 15" woofers.
