The Analogue Addicts Phono Preamplifier 2006 Edition

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Konnichiwa,

First a little history.

The original "Analogue Addicts Phono Preamp" may very well be the most popular Phono Stage design on the net, it is certainly talked about and referenced a massive lot. I have seen this cropping up in all sorts of unlikely places.

Originally the design was posted to the UK based "Analogue Addicts" E-Mail list in either 1996 or early 1997, as response to moans that there where no simple phono stages for DIY and nothing affordable and semi-decent to buy.

In the last 10 Years Vinyl has taken a resurgence few would have predicted. In this day and age you cannot throw a brick or swing a cat at a HiFi show without hitting a few Phonostages, from the sub $ 100 NAD PP1 to the sub $ 30,000 Boulder 2008.

Still with all these riches many DIY'ers talk about my little "semi budget" gem from the mid 90's. A republication of a later re-write in article form is found on my (long not updated) website:

The Analog Addicts Phono Preamp

I have of course moved on from this design, nice as it is. However, due to some current commercial work I decided I needed to "roll back" my system into something that is a lot more like commercial mid-priced High End (< $ 20,000 single source system).

I have build a speaker for that Job that pulls together what I liked best in technical terms about the SLS S8R Monitor, the Dynaudio BM15 Studio Monitor and the Sonus Faber Extrema.

I also wanted an "average" Phonostage, now I might be doing the AA Phono a disservice by calling it average, but in many ways it a simple, workman like design, with decent sonics.

Digging around in my fundus of stuff saved for future projects, samples, leftover parts from modification work and so on I had all I needed.

I'll list details about the practical implementation in another e-mail (it is likely quite interesting), short and sweet, while working up a list of the parts I did not readily have (mainly resistors) to pick them up from the local Maplin (sorta like Radioshack UK) I got to think if I had learned anything since the original AA Phono that could be easily applied.

In addition I wanted a simple AC powered solution and I wanted to pick up on some comments that have rolled in since and various discussions and debates since.

So (drumroll), may I introduce to the genteele reader

"THE ANALOGUE ADDICTS PHONO STAGE 2006 EDITION"

Here the signal circuit schematic:

An externally hosted image should be here but it was not working when we last tested it.


To go with this the powersupply:

An externally hosted image should be here but it was not working when we last tested it.


I'll discuss the differences and my implementation more detail later.

For the case I used once salvaged from an ancient, but very well build Hayes modem like this:

An externally hosted image should be here but it was not working when we last tested it.


The power supply is another IT Services throwout and once supplied a small HP Jinkjet Printer with 30V/400mA DC, I downgraded this unit to an AC only.

Finally, I have quite a number of Preamp Boards from the Shanling SP-80 Monoblocks which I tend to pull out like rotten teeth during mods. As these have space for four regulators, various rectifiers, capacitors and two Op-Amp's (with the PCB Patters alread offering an OPA627 option) and enough space and holes for the pre-pre section in the place that houses the PGA2311 volume control....

An externally hosted image should be here but it was not working when we last tested it.


An externally hosted image should be here but it was not working when we last tested it.


I'd show some photos of the implementation, but it's not overly neat.

I encountered one key problem, as originally the case was not screened (alu parts not connected to ground) the MC setting especially tented to pick up noise. I solved this by using screened wiring for the inputs and thereabouts and I also found a way to connect the Alu case sleeve to the screen. There is still some slight sensitivity to external hum fields, but it's now okay enough.

I fitted captive leadouts for the line out and jacks for the inputs. Parts quality is noted on the schematics.

More another day.

Sayonara
 
AA Phono Pre V2K6 - Moving Magnet Signal Circuit description

Konnichiwa,

Okay, time to go and explain a little what exactly the differences are between the old and new AA Phono.

As said earlier, I did consider many other options but came back to the active EQ Op-Amp MM Stage with an added Headamp as it works best for what we are trying to achieve (Maybe I get time in this thread to go into more detail why).

Any passive EQ solution introduce noise and headroom penalties and while in Valve circuits we have loads of headroom and often circuit resistances dominating the noise, with Op-Amp Phonostages we typically have a +/-12V supply, which limits output swing severely, typhically to less than 20V P-P which is around +20dbu. Using a passive EQ wedged between two Op-Amp's produces a humungous headroom penalty, even just having the 75uS HF eq between two stages drops some headroom.

Apart from the noise and headroom issues another key factor in favour of the single stage active RIAA when using Op-Amp's is that the open loop gain falls with a first order funcion above a fairly low (usually below 100Hz) frequency.

This means also the available feedback to correct amplifier non-linearity drops by 20db every time we go up a decade in frequency. So if we have a "flat" amplifier with 40db gain (as we might use in a passive EQ Phono Stage) we have for an Op-Amp with 120db open loop gain and a 10Hz corner frequency (a fairly common construct) at 100Hz a feedback factor of 60db, at 1KHz 40db and at 10KHz 20db. Worse, at 100KHz we are right out of negative feedback. So, between 100Hz and 10KHz the feedback factors changes by some 40db!!!

This means distortion will rise strongly towards higher audio frequency and if any supersonic noise is present (groove noise, needle resonance in MC's, RFI/EMI) things will really go haywire, also the distortion profile will vary with frequency and so on.

Comparing that with a 40db Gain @ 1KHz RIAA Equalised stage, here at 100Hz we will have a Gain of 52db and a feedback factor of 48db, at 1KHz we have a gain of 40db and a feedback factor of 40db, at 10KHz we have a gain of 26db and a feedback factor of 34db with 100KHz in theory still retaining this 34db of feedback factor (in reality we will in fact have a further breakpoint at 50KHz which reduces our 100KHz feedback fatcor to around 25db)...

This means in our active RIAA equalised Op-Amp Phonostage we have a variation of the feedback fatcor between 100Hz and 10KHz of only 14db instead of the 40db for the flat gainstage.

So, back to the best OpAmp for Phonostages (Burr Brown OPA637 or Analog Devices AD8610). We use these in a Phonostage with an active EQ, we minimise the capacitance applied to the output as load to limit slewing issues (1), we remove any coupling capacitor we can (2) and we have the original AA Phono.

(1) By minimising capacitance we maximise the impedance of the RIAA Network and we need to limit this to one that does not add noise over the self noise of the Op-Amp - the OPA637 needs to see less than around 1K impedance for the whole feedback network to avoid degrading it's noise performance materially, our impedance is around halve that, so the NFB RIAA Natwork impedance is optimised between noise and extra load on the output.

(2) The NE5532/34 single op-amp phono stages usually need three coupling capacitors, one on the input to prevent input bias currents flowing in the cartridge, one in the NFB to ground leg to minimise offset and finally still one on the output. Using a precision FET Input Op-Amp allows two of these three to be discarded.

I only made a few tiny, but significant changes to this circuit.

I actually included the output coupling DC blocking capacitor in the negative feedback loop. This means the sonic impact of this capaictor is reduced more or less by the feedback factor, in our case using the OPA637 by around 50db or more below 10KHz!

To do this I inverted the sequence of the RIAA EQ sections and connected the 4.7nF Capacitor not directly to the Op-Amp output but instead to the point after the coupling capacitor. The DC loop is still as before, but for AC (music signal) we now correct the capacitors (mis)behaviour.

The next trick which relies on the first (3) is to use a modest value (0.22uF was what I had lying about) coupling capacitor and to bridge it with a series combo of resistor and large value capacitor (100uF Black Gate in series with 1k in my case) which makes sure the large nasty capacitor is more and more removed from the circuit at higher frequencies.

The circuit I use is optimised for loads of 10KOhm or greater, if you must drive lower impedance loads you need to scale the values and of course, neither capacitor can have too much capacitance, if the general quality can be retained.

(3) Where the output coupling network not included in the feedback loop BTW this little trick could not be used like that, it would result in a shelved down response below the "takeover" point between the two capacitors (~ 700Hz in my case), the degree of shelving down depending on the load impedance.

My next stop was to seek to remove the need for biased, back to back electrolytic capacitors in the output coupling circuit (4). I instead provide around +0.5V Bias on the OPA637 Output by adding a 1.5MOhm resistor from the positive supply line to the Offset nulling input pin 1. This use is specific to using the OPA637 or OPA627.

(4) These go back to the seminal Jung/March "Picking Capacitors Article) - http://home.comcast.net/~wjungieee/wsb/Picking_Capacitors_1.pdf http://home.comcast.net/~wjungieee/wsb/Picking_Capacitors_2.pdf

The final setp was to push the Op-Amp output into operating as single ended class A emitter follower, rather than as class B push-pull follower. By simply connecting a 6K8 resistor from the output to the POSITIVE Supply (5), this forces the other halve of the output stage into cutoff and draws around 1.7mA through the emitter follower. The good thing is that if happen to need to drive a low load that needs more current than this happily provides the push-pull output stage will kick in and drive the heavy load.

(5) For most Op-Amp's the resistor or CCS needs to connected to the negative rail, in case of the OPA637 it is the positive rail however as the simple emitter follower part of the output stage uses a PNP Darlington pair).

Often seen is the use of a constant current source (CCS), as this reduces the load on the Op-Amp, but the only handy single item CCS's I had around caused the OPA637 to latch up on the positive rail, so I used the resistor and if we are realistic the difference between CCSA and 6K8 resistor, given all the other relative impedance and the drive ability of the Op-Amp is minimal.


Well, this covers the MM Signal circuit with the changes over the original AA Preamp and many other similar solutions published. I decided to give the headamp a seperate page, as it is very simple, but the reasoning behind it is very complex....

Except to mention that I changed the input termination on the Op-Amp from 680K to 100K simply because my case is only partially shielded and even using fully double layer shielded wiring on the inputs managed to pick up more hum than I was happy with. Improving the screening and reducing the termination resistance killed enough of the hum to make not worry any more.

With 100K terminating the input of the Op-Amp another 100K resistor terminates the RCA jack, the 220pF Capacitor is needed for higher inductance MC cartridges to make sure the LC resonance with the inductance boosts the treble back to flat. The exact value is really a job for trial and error.

I keep nowadays some "Y" RCA adapters handy, plug these between input and arm cable and plug test loads fitted to RCA plugs into the Y-Part, once you have found what works best simply solder this into the preamp, remember as you build the preamp you are NOT voiding your warranty (in fact, you never had any in the first place).

Sayonara
 
AA Phono Pre V2K6 - Moving Coil Preamp Stage Signal Circuit description

Konnichiwa,

Okay, we have gotten our actual Phono Stage circuit, good for MM, low noise and all that, we cover the power supply later. Given that we can deal with cartridges that put out at least around 1.5mV @ 5cm/S modulation, which covers the highest output MC cartridges and most/all moving magnet pickups.

Here a quick excursion into the way LP's are made. We have according to the international standards a 0db level of 5cm/S, HOWEVER, it also specifies a MAXIMUM level of 25cm/S, in other words the LP's equivalent to CD's 0dbfs is actually 25cm/S modulation. And I know from experience usually much of this reserve (called the same as what Bill Clinton called the Oval Office - namely Headroom) is used during cutting.

So, this leaves us with the problem that the phrugalphile Analogue Addict cannot use this hidden treasure of a Pickup called Denon DL-103. I have said this many times, used and set up correctly (maybe a topic for another day) the DL-103 is a cartridge that gives most sub $ 1,000/£ 500 "high End" moving coil cartridges a hard time to justify their existence.

No, it is not the worlds best cartridge or the only one, but it is much better than it's detractors make it out. If you told me I could from tomorrow only use the DL-103 as pickup but gave me a nice Car (say a BMW M3 in the HK ralley conversion) in exchange for the promise I'd not hesitate one minute.

We could, perhaps, simply increase the gain of our MM Phono Stage by connecting a 68R resistor across the 560R resistor to ground on the inverting input. In fact, this works (I tried) when using the OPA637 and non too bad, commercially the Grado PH1 Phonostage does exactly that!

However stealing another 20db from the feedback does make the sound worse and noise is higher than I like. Further, with >60db DC gain as in the AA Phono you can already see quite significant levels of output offset, increasing the DC tenfold again can turn a 0.5V offset into a 5V one, seriously affecting overload performance, Op-Amp's with less DC precision than the OPA637 may even simply latch to one supply rail. The only way around is a capacitor in series with the feedback resistor and this would have to be a VERY large value.

So, we would like around 20db gain or stepup ratio. A number of more or less expensive stepup transformers exist, but even with fairly prosaic cartridges and the AA Phonostage those I had a chance to try where non too good.

The best I ever found where ribbon microphone stepup units salvaged from a "Grampian" console Amp of the 1950's/60's. Some other "equipment pulls" including Neve did more or less well, afforable MC stepup stransformers from current production definitly did not too well for my ears.

I eventually had a custom model designed which is now generally available as "TX-103" from Stevens & Billington. These beat the grampian pulls handily. While a pair of TX103's with an MM only AA Phono would be definitly tasty, it does rather severely skew the sense of perspective and misses the objective.

So, in order to keep the cost down an active "headamp" was decided upon. After reviewing the various new IC's available for the job, the wide range of circuit designs using discrete components and so on I re-settled on a slightly modified variant of the l'Audiophile single J-Fet Headamp used in the original.

Many Headamp circuits exist, including variants of the common base Marshall Leach Design, the John Linsley Hood circuit, Douglas Selfs hybrid input stage and many others (an overview of some here: http://sound.westhost.com/project25.htm and here http://users.ece.gatech.edu/mleach/headamp/ ).

One thing all these circuits share is the fact that very small signals are coupled through very large value coupling electrolytic capacitors (usually several) even for supposedly DC coupled circuits, not something I like from either a philosophical or sonic view.

So, back to what the japanese call so charmingly "Single (gem)stone pre-pre". It is still sublimely simple and good. The only drawback of sorts is that inverts polarity. If you care about that kind of stuff simply invert the cartridge connection at the cartridge, that will corect polarity.

I used NOS 2SK147BL which I had around as leftover matched pairs from earlier projects. I tend to buy around 10 devices every time I need a pair and have not much stock remaining. J-Fets, even if used with a lot of degeneration, need to be matched.

You can use a number of different types, currently easiest to get is the 2SK170 and it's Interfet version LSK170. You can also use 2SK147, 2SK369 or the Dual versions 2SK146 & 2SK389, which mean you can save yourself the matching process but which in turn cost enough more to make buying a lot and matching them worth your while.

Schuro in germany sells the 2SK170BL for 58 Euro Cent each:

http://schuro.de/preisl-jap-trans.htm

Cricklewood Electronics in Cricklewood London ships readily internationally but is more expensive at £ 1 per J-Fet, still I tend to shop directly there and combine it with a curry at the Pink Rupee next door, an excellent place for authentic Nepalese food.

http://www.cricklewoodelectronics.com/Cricklewood/customer/search.php?xid=51e8b90cbf2831b746cda61ca916e0e8&substring=2SK170

In this circuit matching just for 0V Idss suffices, so make a simple rig with a 9V battery connected with a series resistor of 100 Ohm (1%) on the positive (+) pole. Connect up a small transistor socket with the 1R resistor connected to the drain.

Connect drain and source connectiosn together on the socket and connect with a piece of wire to the 9V battery negative pole. Measure the voltage across the 100R resistor with a digital multimeter, you get 0.1V/mA. Simply find the closest matched pair, keep the rest for next time.

Theoretically the 2SK147 has less noise than a 100 Ohm resistor. In our circuit the 68 Ohm Gate Stopper and the 68 Ohm source degeneration resistor add enough noise to make the effective noise figure around 3db worse, still very little noise in absolute terms, looks like around 0.2uV RMS in the 20Hz-20KHz bandpass.

If we have a Denon DL-103R (I do) the unweighted self noise of the headamplifier is around 62db below "0db" and 76db below "full scale". That is about the same as the $2,500+ Dynavector PPA200 Stepup Amp.

As the following RIAA EQ attenuates anything above 50Hz progressively, this noise is also attenuated, leaving only the very low frequency, sub 100Hz components to content with.

If we use "A-weighting" (which accounts for the ears reduced sensitivity at low frequencies) the significant noise is attenuated another 14 - 20db at 100Hz which means we would have an input referred, A weighted S/N ratio closer to 80db and 94db to full scale. In fact, the MM Preamp will likely remain a significant contributor in noise terms.

Looking a little ahead on the powersupply, this circuit has actually virtually NO supply noise rejection, all supply noise appears on the output. My current off-mains supply however contributes so little noise that it is significantly below the circuits self noise by at least 6db or more at 20Hz with progressively less noise as the frequency rises.

Anyway, this complete this little, simple but very nice sounding headamplifier. In fact, if I compare the MM input of the Phonostage to the MC input using my DL-103 I am hard pressed to observe ANY differences in sound quality.

Sayonara
 
Re: AA Phono Pre V2K6 - Moving Coil Preamp Stage Signal Circuit description

Konnichiwa,

We are reaching the home stretch. On to the power supply. I could have build something pretty elaborate, really extreme regulators, used batteries again, but I did not. Really, anything too elaborate is against the spirt of the excercise.

I decided to see what could be done with the opposite approach, namely with trying to get good sound out of the generic PSU circuitry.

As mentioned before, I started with a transformer and case from an HP inkjet supply, giving around 40V rectified DC. With schottky diodes, around 45mA total current consumtion and a pair of series 6800uF capacitors we get 120mV Peak-Peak ripple.

As the the supply I had did not feature a centertap, but very good primar/secondary separation and low leakage it needed some work to get a symmetrical supply. The way I have arranged things it means I have different currents in the positive and negative supply as well (1) and I needes to equalise the currents in the positive and negative supply.

The 1k resistor added between the audio ground and the negative supply draws the neccesary current to balance the currents. The 1k resistor is what worked for me, but it will need adjusting/selecting for an equal voltage across the two capacitors, you need a fairly powerfull resistor (2W or so).

(1) The 7915 regulator has 1.5mA quiescent current, the 7815 has 10mA, also we have around 8mA extra current from the MC Headamp and 10mA more for the 7812 regulator, so the negative supply draws around 25mA less current than the positive supply.

Now, using 7815/7915 regulators in audio is frowned upon and for good reasons. (2) However, not all is bad, if all we want to do is to kill a good few 70db or so of ripple and get the DC voltage approximatly to the right voltage to an RC filter chain they do just fine.

(2) Check the Audio Amateur Article by Walt Jung on Audio Supply regulators for the noise levels of the 7815 & 7915 http://waltjung.org/PDFs/Regs_for_High_Perf_Audio_2_A.pdf especially the 7915 is a noisy bugger....

For the Op-Amp supply I have used a RCRC filter chain that drops around 3V, the 10R/150R values are actually sub-ideal, a pair of 75R resistors are a better choice, giving around 10db more noise suppression above 100Hz.

The 120mV peak-peak overall ripple gives in effect around 20mV RMS ripple before each regulator. The 7815 & 7915 each knock 100Hz ripple down by around 70db typhically, so after the regulator the ripple we get will be knocked all flat into the regulators noisefloor.

At 100Hz the 7815 has around 200uV selfnoise, while the 7915 has nearly 4mV selfnoise!!!! The filter as shown will drop the noise at 100Hz by 43db, so on the supply pins of the OPA637 we will have 1.4uV on the positive supply pin and 28uV on the negative supply.

As the frequency rises the RC filters become much more efficient (82db noise suppression at 1KHz) so the high noisefloor of the 7915 and the rising noisefloor of the 7815 are not really a problem, the negative supply line will have 0.4uV noise and the positive line 0.05uV at 1KHz further falling at higher freqencies to theoretically around 1.5nV at 10KHz.

In terms of noise this arrangement is superior to the best active regulators Walt measured in his AA article by quite a bit, especially above a few 100Hz, all at the cost of a few pennies worth of resistors and capacitors.

On the negative supply the OPA637 has 120db supply line noise suppression, given the around 60db gain used at 100Hz we loose around 60db of this, so our 28uV become 28nV on the output of our OPA637, which means this supply line noise will disappear in the devices noisefloor.

At 1KHz we have 110db supply line noise supplession and 40db gain, so we have actually 70db supply line noise suppression on the negative supply which turn our 0.4uV into fractions of a nV, so again we are fine on noise.

On the positive line the OPA637 has much less supply line rejection (105db/100Hz, 85db/1KHz, 45db net adjusted for op-amp gain), but as our regulator on the positive line is much less noisy we still have similarly low noise feedthrough as the negative line. All in all even the worst noise we have in the supplies is way below the OPA637's self noise.

So, this leaves the MC pre-pre. As said, this has got no supply line suppression to speak of, hence we have a more complex filter after the regulator. The three RC cells, with the final one separated by channel kill our 200uV at 20Hz by 55db and at 100Hz by 95db, giving us respectively 300nV noise at 20Hz and 3.5nV at 100Hz with even less noise at 1KHz.

Referenced to a 500uV 0db level we have over 60db S/N ratio at 20Hz, given the rising noise of the 2SK147/170 et all below 100Hz our supply line noise again simply drops below the active devices self noise, so we are okay.

Again, this illustrates the power of a few pennies worth of passive components compared to extensive active circuits when it comes to noise suppression, again a cheap noisy regulator and a handfull of capacitors and resistors beat "super", "hyper", "mega" and other regulators.

Well, that completes the comments on the power supplies, again like many other features in the AA Phono they are a good illustration of the "less is more" zen principle in electronic design.

Of course this phonostage is not the final word on the subject, nor is it even that extreme, however it is easy to add a lot more circutry and things and to achieve a much worse perfomance. So, if you need a cheap and easy to build phono stage that performs pretty well give this one a try.

Sayonara
 
Analogue Addict's phono Pre-amp

Hi Kuei Yang Wang

Bravo!! Nice work here. Also, the DL-103 vindicated yet again!

If I may be permitted, do you know anything about one of the phono pre's I use - a dc Nagatron HA-9000 dating from around 1978 and which I have been using for years? The only info I have is that it was reviewed in the American journal Stereo Review by Julian Hirsch (Hirsch-Houck Labs) in December of 1978. It is a rectangular 'black box' with 4 rca phono's at the back + earth tag and is powered by 4 C-cells/ch. Aso at the back, is a tiny analog battery condition meter and at the front, it has a switch which changes colour at switch on. I play my Garrards every day for a few hours and the Nagatron's batteries last many months before recharge is needed. No info on the circuit but a nominal gain of 40dB (voltage gain of 100X) and a frequency response of 10 - 200,000Hz +-1dB is quoted. Total harmonic distortion of 0.01% at 1,000Hz (at an unspecified level). Input impedance is 20 ohm. In 1978, the HA-9000 Nagatron cost US$275, which was not inexpensive for that time.

Other snippets from the review quoted " It is clear that the HA-9000 wil never limit the dynamic range of the music system in which it is used".

Some years ago, I took a tt, the Nagatron and other stuff away from here to demo to some friends the difference between analogue and digital sources. One of these guys was ham-handed enough to bump the Nagatron's earth tag, which became loose.

I then discovered there was NO WAY to open it to make a repair but after several attempts over another few years, I somehow managed to do a proper job, but the components and circuit remain a mystery as it has a folded and glued inner mu-metal case and an outer copper case.

Are there perhaps any other user/s of this equipment?

bulgin
 
Error Warning

Konnichiwa,

MY BAD.....

I realised, I drew the Powersupply rectifiers the wrong way around;

An externally hosted image should be here but it was not working when we last tested it.


So, please turn them back to front like this:

An externally hosted image should be here but it was not working when we last tested it.


Sayonara
 
Hi Kuei,

I built one of the earlier versions for a friend and it is a sweet and pure sounding little phono stage- better than his Burmester 838...Thanks for the design.

One criticism is that it is a little lacking in verve and does not "do" large scale dynamics as well as I have heard.

I tried a few little mods myself including biasing the opamp with a j-fet configured as a ccs. This was good even though I tied the ccs to the wrong rail -blush...

This phono stage is direct coupled, offset is fairly low so no problem in his particular system. Is there any reason we couldn't apply some offset nulling here (noise maybe)?

I am interested in your subjective take on the improvements with the new version. I wonder whether the dynamics issue I mentioned is a function of the battery power supply in the case of the one I built. Sometimes batteries are criticised for lacking a little "drive". Penny for your thoughts...

Rob.
 
Konichiwa,

Robert F said:
This phono stage is direct coupled, offset is fairly low so no problem in his particular system. Is there any reason we couldn't apply some offset nulling here (noise maybe)?

Yes, noise may become an issue, the other reason I never went for that is that you would really need a servo and I don't "like" Servos. Wrapping the capacitor into the feedback loop should do equally well.

Robert F said:
I am interested in your subjective take on the improvements with the new version.

There are nearly 10 years between this one and my earlier version, which has long passed on to a friend, so - no idea, honestly. Sounds very nice though.

Ciao T
 
Hi kuei,

Choosing 78xx/79xx regulators instead of the 317/337 has been done for simplicity and cost efectiveness or there are technical reasons for this ?

The 100 u BG at the output could be a cerafine ?

Did you build the whole circuit in one box ?

Do you feel that the 1 u decoupling Wimas are enough for the 637 in this circuit ? Could the "blue" MKP Siemens bricks used instead ?

What about your preference of Micas in the RIAA circuit ? is it still valid ?

Thanks for the excellent write up and for sharing (once more),
Marinos
 
Konnichiwa,

Marinos said:
Choosing 78xx/79xx regulators instead of the 317/337 has been done for simplicity and cost efectiveness or there are technical reasons for this ?

Yes, there was a serious technical reason for this. I wanted to use one of the Preamp PCB's from the Shanling Amplifiers (I have a box full of these around from modding the Amp's - the best mod is to just pull these rotten unneccesary things out ;) ) and they feature a pair of 78/7915 regs and another pair of 78/7905 regs with various decoupling caps etc.

Just leaving the 7X15's in place, soldering some cap's below and adding the RC filtering (which would still have been needed for the 317/337 anyway) reusing the already present decoupling capacitors made the job of sorting this out very quick and easy.

There is no reason not to use better regulators if you have them, but equally there is no real reason to do so. Use whatever is in the project box and takes your fancy.

Marinos said:
The 100 u BG at the output could be a cerafine ?

Or silmic or anything you fancy. With 1K in series it's sonic signature will be minimised. My reason for the BG was that it is very small.

Marinos said:
Did you build the whole circuit in one box ?

On the shanling PCB shown earlier on. The two seperate Op-Amp positions alread provisioned for OPA627, so that was the Op-Amp's taken care off, where there used to be the PGA2311 Volume control IC I had two rows of 8 positions "veroboard" after cutting/peeling some traces which is ideal to wire up the headamp.

Marinos said:
Do you feel that the 1 u decoupling Wimas are enough for the 637 in this circuit ?

These are the ones local to the Op-Amp, remember there is also a shared 470u per rail within an inch or so of PCB trace.

Marinos said:
Could the "blue" MKP Siemens bricks used instead ?

I selected the Wima's based on size/resonance frequency. Larger size capacitors are a bad idea IMHO and need then again further bypassing. I wanted to keep it simple. I'd suggest any physically small 1uF Film Cap.

Marinos said:
What about your preference of Micas in the RIAA circuit ? is it still valid ?

Yes, in fact I might try mix some Mica's back in to add a little more bite to the Circuit. For now I literally used "what I had around".

My ideal preference for RIAA Circuit capacitors has long been non-magnetic NOS Silver/Mica for the HF portion (75uS/3.18uS) with a Tinfoil & Polystyrene for the LF portion (3180uS/318uS).

All Polystyrene sounds a little warmer and laid back, all Mica can be a little etched and "High End HiFi" sounding, it depends on the circuit. Valve circuits generally seem to prefer all Mica RIAA EQ's (exceptions apply).

Sayonara
 
Konnichiwa,

fred76 said:
What R load do you prefer for the DL-103R? Thanks.

For now I selected 1K based on recommendations by people I trust (the needed 1nF or so are nearly supplied by the J-Fet), but I'll try some more.

Easiest thing to do is to get a pair of "Y" adapters and to solder some selection of components into some cheap RCA Plugs and to listen.

Sayonara
 
For now I selected 1K based on recommendations by people I trust (the needed 1nF or so are nearly supplied by the J-Fet), but I'll try some more.

Easiest thing to do is to get a pair of "Y" adapters and to solder some selection of components into some cheap RCA Plugs and to listen.

Thanks. I'd try 1k for the active MC section in my ss phono. There are binding posts in the MC section for R loading. However in my particular phonostage, the MC output signal would also pass through the MM stage’s 100pF input cap, unlike the AA's MC stage which bypasses this since its MM cap load seem to be directly terminated at the rca input jacks not on the pcb...

A bit OT: I use Y-adapters for a Valve phono for some time now (R rolling). What Rload/s do you recommend for the 103R when used with step-up trannies?? For now I use 12k-12.5k to get an Rload of ~100 Ohms using 1:10 stepups for an MM phono input of 51.1k//100pF.

I selected the Wima's based on size/resonance frequency. Larger size capacitors are a bad idea IMHO and need then again further bypassing. I wanted to keep it simple. I'd suggest any physically small 1uF Film Cap.

Does this also hold true for coupling caps in phono sections not just decouplers? Thanks.



regards,
fred
 
thanks...

Thorsten, I enjoy a lot your hieratic, concise and controversial posts.

I have played a lot with the ¨vintage¨ El Cheapo, surprising with it even the most obscene and rustic vinyl detractors.
From time to time I was wondering about a suitable ac power supply to go with it, and now here it comes, with the added benefit of a newly developed version of the original preamp...

Thank you very much for your effort.

Cheers

Federico
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.