Correction for the original layout, after the opamps were switched to the underside of the board. When I mirrored the diagram I forgot to flip the regulators back to the top side.
The opamps are on the underside so that they can be mechanically fixed to the case, and the whole board can held down with just 2-3 screws.
The opamps are on the underside so that they can be mechanically fixed to the case, and the whole board can held down with just 2-3 screws.
Attachments
Operational Phonoclone
It works.
After I had carefully replaced the two disembowled Vregs and equally carefully re-wired all the others correctly I discovered that all but one of the remaining ones had blown as well, just less dramatically. So back to the parts store for another handful of new regulators (fortunately they are only 30 and 60 cents each for the LM7812 and 7912 respectively) and, later, another agonizing session with the soldering iron ...
The end result aint pretty, but it works.
At this stage I can say it delivers what it was designed to do, with no weirdness. Noise is acceptible, the standard hush. Gain is high, in line with the calculated 75dB. Frequency response appears accurate.
The "how does it sound" part will have to wait until the caps have broken in. The Black Gates in particular need at least 50h to settle. But ... we are encouraged. (Its pretty cool.)
-R
It works.
After I had carefully replaced the two disembowled Vregs and equally carefully re-wired all the others correctly I discovered that all but one of the remaining ones had blown as well, just less dramatically. So back to the parts store for another handful of new regulators (fortunately they are only 30 and 60 cents each for the LM7812 and 7912 respectively) and, later, another agonizing session with the soldering iron ...
The end result aint pretty, but it works.
At this stage I can say it delivers what it was designed to do, with no weirdness. Noise is acceptible, the standard hush. Gain is high, in line with the calculated 75dB. Frequency response appears accurate.
The "how does it sound" part will have to wait until the caps have broken in. The Black Gates in particular need at least 50h to settle. But ... we are encouraged. (Its pretty cool.)
-R
In the process of trying to copy the PhonoCube, you have designed something different from it. That, I think, is a better outcome than redoing what has already been done.
As it continues to break in it keeps getting better and better.
On the homepage I decided to change some of the values. Increased R3 to 2k2 from 680R to make it a more reasonable load for the first stage to drive. R4 is 2k2 to balance, and the RIAA loop is recalculated to keep the gain of the second stage at 30dB.
I'm still using the original values of course.
Messing about with the simulation, I came across a set of values that are not too hard to accomplish - 750k, 110k, 1nF and 2.9nF - which happen to match the RIAA response exactly. I mean, within 0.01dB which is exact to all intents and purposes, aiming for within 0.1dB is already well into overkill territory.
2.9nF can be had by paralleling three slightly under-spec 1nF capacitors. This has the additional advatange that only one value is needed for the RIAA loop, rather than two.
-R
PS. also put up some pictures of the finished unit
On the homepage I decided to change some of the values. Increased R3 to 2k2 from 680R to make it a more reasonable load for the first stage to drive. R4 is 2k2 to balance, and the RIAA loop is recalculated to keep the gain of the second stage at 30dB.
I'm still using the original values of course.
Messing about with the simulation, I came across a set of values that are not too hard to accomplish - 750k, 110k, 1nF and 2.9nF - which happen to match the RIAA response exactly. I mean, within 0.01dB which is exact to all intents and purposes, aiming for within 0.1dB is already well into overkill territory.
2.9nF can be had by paralleling three slightly under-spec 1nF capacitors. This has the additional advatange that only one value is needed for the RIAA loop, rather than two.
-R
PS. also put up some pictures of the finished unit
Hi,
is there any possibility of making a "Group-buy" for the PCB's of the Phonoclone?
Greetings Barossi
is there any possibility of making a "Group-buy" for the PCB's of the Phonoclone?
Greetings Barossi
Hi rjm
yes I can since I have one which I would like to use and that is an Micro Seiki Lc80w mc which has a 82uv out and dc resistance of about 7.5 ohms , manual says 2.2 .
A 1 volt out would also be sufficientI think to drive my A400/Smg a combo.
regards
keith birss
yes I can since I have one which I would like to use and that is an Micro Seiki Lc80w mc which has a 82uv out and dc resistance of about 7.5 ohms , manual says 2.2 .
A 1 volt out would also be sufficientI think to drive my A400/Smg a combo.
regards
keith birss
Wanted, a few brave men.
This has been on my mind for some time. Its just not possible for me to properly beta-test this by myself, there are two many variables for one person to get around. The results are good enough (it is impressive) that I'd like some other people to play with it a little.
I wonder if there are enough people interested. I'd need about 10 I think for it to work.
I'm willing to put the effort into getting the PCBs made and shipping them to the beta-testers. I can also get the Black Gate caps and voltage regulators relatively cheaply. I can sell them to the beta testers at cost + shipping. The cost of the "kit" would depend on the numbers, but roughly $30-50 I expect.
Beta testers are responsible for supplying the rest of the passives and the op-amps, since I can't get those any more cheaply or easily than you can, as well as the power supply and the rest of the hardware.
Interested persons can email me at rjm003.geo@yahoo.com. I'll leave the subscription period open for 2-3 weeks. We can discuss what components people what to use, and I'll modify the layout to accept the top few choices.
I'd like someone to suggest an online place to get PCBs made up. I know a couple, but haven't tried any personally yet.
-R
This has been on my mind for some time. Its just not possible for me to properly beta-test this by myself, there are two many variables for one person to get around. The results are good enough (it is impressive) that I'd like some other people to play with it a little.
I wonder if there are enough people interested. I'd need about 10 I think for it to work.
I'm willing to put the effort into getting the PCBs made and shipping them to the beta-testers. I can also get the Black Gate caps and voltage regulators relatively cheaply. I can sell them to the beta testers at cost + shipping. The cost of the "kit" would depend on the numbers, but roughly $30-50 I expect.
Beta testers are responsible for supplying the rest of the passives and the op-amps, since I can't get those any more cheaply or easily than you can, as well as the power supply and the rest of the hardware.
Interested persons can email me at rjm003.geo@yahoo.com. I'll leave the subscription period open for 2-3 weeks. We can discuss what components people what to use, and I'll modify the layout to accept the top few choices.
I'd like someone to suggest an online place to get PCBs made up. I know a couple, but haven't tried any personally yet.
-R
I guess I have to tackle that one sooner or later.
The two are closely related of course. The Phonoclone circuit is the VSPS, inverted, plus and inverted input stage. If you built the VSPS dual mono, with OP27's, and separate regulation on each op-amp, and use the same set of passive components, the only significant difference is the inverting vs. non-inverting configuration of the gain stage. They would in all likelyhood sound extremely similar.
However, the placement of the two could not be more different. Where the goal in the case of the VSPS was to have pleasant, enjoyable sound, the Phonoclone is an all-out stab at high-end soundstage and detail resolution. Keep in mind the VSPS is for MM cartridges, while the Phonoclone can tackle the lowest of lowest output moving coils.
With that, here's the test setup: my VSPS (NE5532, 40dB, lowZ version) was modified to have an input impedance of 100 ohms so as to connect to the Denon DL103 directly without any step up transformers. The Phonoclone was with the original values, R2 set give a total gain of 60dB with the DL103. The difference in gain was made up by turning up the volume on the amplifier. I was using the 6BM8 Elekit amp rather than my NI gainclone, a choice which favours the Phonoclone in terms of tonal matching. Further, with the volume turned up to accomodate the VSPS, the impedance seen by the grid of the input triode was enough to indroduce audible buzzing noise.
Test LP was Mozart K339 "Vesperae solennes de confessore" (Colin Davis / London Symphone Orchestra and Chorus). Besides everybodies favourite Laudate Dominum, the other movements are a stress test providing every shade of instrumental timbre all mixed together. It's "difficult" music to reproduce enjoyably by a hi-fi system, in that regard it's a choice biased in favour of the Phonoclone. With less complex pieces I expect the differences to be less remarkable.
Once again, I remind you that what's being compared here is NE5532 vs. OP27, dual vs single, one regulation unit vs. four, inverting vs. non-inverting, and, trivially, a different set of RIAA caps. I used the same power supply, simply plugging it into whichever unit was under test.
First, the VSPS is balanced lower and warmer than the Phonoclone. Despite the Allen Wright response built into the VSPS, the Phonoclone is subjectively faster, lighter, and more open.
Second, it becomes clear that the VSPS has comparitively no soundstage. All the instruments are grouped together in a "ball of sound" in the middle between the speakers. The Phonoclone produces a soundstage that is not only extremely wide, but provides some illusion of depth, too, probably owing to the significant retrievel of low level ambient detail.
Ah yes, detail. This isn't just about being able to pick out more instruments out of an orchestra, but rather about each one sounding so much more realistic. Vocals are especially lifelike, projected with considerable force out into the room. In comparison textures rendered through the VSPS are a little fatter and softer than they should be, and the midbass is particularly slurred together.
Although the Elekit amp has poor low frequency response, the additional slam and control of the Phonoclone in the lower regions is evident. (With the Gainclone substituted bass performance becomes truly stunning.) No matter how complex the music got, the Phonoclone was unconcerned. To sum up, it is a powerful and precise machine.
However, the VSPS remains a very enjoyable phono stage. Its more forgiving and less fatiguing to listen to. The reason I'm using my tube amp with the Phonoclone is I find myself needing to inject some extra warmth to compensate for something I can't quite put my finger on .. its very subtle, but someting I might describe as a slight metallic etch to the treble.
From past experience, here's my take on how to read all this:
For the most part, the poor performance of the VSPS can be blamed squarely on the dual op-amp and common regulation. Basically, not enough isolation between channels. I'm not convinced the NE5532 per se is the problem, though the non-inverting topology might be a factor.
If I was to bet, I'd blame my remaining concern about the Phonoclone on the cheap metallized polypro RIAA caps I used. Cheap metallized caps have just that "taste". Or, less likely, I might blame the OP27s. They are typically rather cool sounding. The NE5532, rather like the LM3875, is atypically "warm" for an op-amp, which is why I like it. Third, least likely possibility is the DL103s output impedance is interacting unfavourably with the Phonoclone.
So, the bottom line is I feel that the Phonoclone has awesome potential, and already in a different league than the dual-channel VSPS, yet missing that last little push.
-R
Hi rjm,
I appreciate very much the effort you made, and I find your findings both interesting and inspiring. I already have the parts for a phonoclone lying around since a few weeks, only lacking time...
I hope to find time in about 2 or 3 weeks to actually build my own version, but with p2p wiring, because there are some things I want to try. I will fire it with batteries, because I have them around.
I'm quite curiuos about the develpment of your clone, and looking forward to post my impressions later,
Rüdiger
I appreciate very much the effort you made, and I find your findings both interesting and inspiring. I already have the parts for a phonoclone lying around since a few weeks, only lacking time...
I hope to find time in about 2 or 3 weeks to actually build my own version, but with p2p wiring, because there are some things I want to try. I will fire it with batteries, because I have them around.
I'm quite curiuos about the develpment of your clone, and looking forward to post my impressions later,
Rüdiger
A move to low output cartridges
Greetings to All:
A source used by BrianGT for chip amp circuits is Advanced Circuits. The following quote concerned proto boards for the LM4780 chip. I would assume the phonoclone board would be in the same range:
"Originally posted by BrianGT
If an order is generated, here are some prices that I worked up, based on a quote from Advanced Circuits. The price is including tooling, shipping from vendor, and paypal charges.
25 boards - $27.75 each
50 boards - $17.50 each
100 boards - $12.25 each"
I also solicit suggestions for an entry-level cartidge to be used with the phonoclone on my Technics SL-1700 turntable. I would also entertain tweaks to that arm, though that could go WAY OT and is probably better if discussed elsewhere.
Greetings to All:
A source used by BrianGT for chip amp circuits is Advanced Circuits. The following quote concerned proto boards for the LM4780 chip. I would assume the phonoclone board would be in the same range:
"Originally posted by BrianGT
If an order is generated, here are some prices that I worked up, based on a quote from Advanced Circuits. The price is including tooling, shipping from vendor, and paypal charges.
25 boards - $27.75 each
50 boards - $17.50 each
100 boards - $12.25 each"
I also solicit suggestions for an entry-level cartidge to be used with the phonoclone on my Technics SL-1700 turntable. I would also entertain tweaks to that arm, though that could go WAY OT and is probably better if discussed elsewhere.
http://www.olimex.com/pcb/index.html ...very cheap for single sided boards...and it looks like it will fit on single sided?
$26 for a SINGLE SIDE EURO FORMAT PCB 160 mm x 100 mm (6.3" x 3.9") board..maybe around 4 phonoclones could fit? Maybe even 6..brings it to around 5 dollars per person. Repeat order is even cheaper....for obvious reasons
I have done business with them before...fast service..good comms
The proto boards are obviously not as nice as the double side BrianGT boards...with the through hole metalization...etc....but are good quality
$26 for a SINGLE SIDE EURO FORMAT PCB 160 mm x 100 mm (6.3" x 3.9") board..maybe around 4 phonoclones could fit? Maybe even 6..brings it to around 5 dollars per person. Repeat order is even cheaper....for obvious reasons
I have done business with them before...fast service..good comms
The proto boards are obviously not as nice as the double side BrianGT boards...with the through hole metalization...etc....but are good quality
Once channel would be about 2.25" to 3.25" on a side. Last version was spacious at 2.7" square. Should be possible to fit 4 on 160x100mm. Several people want to use larger caps and resistors, limiting the amount of possible shrinkage.
Just wondering: how hard it would it be to cleanly cut the finished boards into the individual channels?
Just wondering: how hard it would it be to cleanly cut the finished boards into the individual channels?
If they are not scored you mean? Quite hard...fibreglass is tough to cut. In the past I used a dremel..not sure if that is the best tool though..
cutting circuit boards
IF I were to attempt cutting boards without scoring, I would opt for a thin kerf, high tooth-per-inch blade. My first choice would be a scroll saw with variable speed. Second choice would be a band saw with a metal cutting blade for the tooth count. The fiber glass is the problem. The finer cut will leave less need for touch-up sanding to remove burrs.
Hope this helps.
IF I were to attempt cutting boards without scoring, I would opt for a thin kerf, high tooth-per-inch blade. My first choice would be a scroll saw with variable speed. Second choice would be a band saw with a metal cutting blade for the tooth count. The fiber glass is the problem. The finer cut will leave less need for touch-up sanding to remove burrs.
Hope this helps.
How about leaving the regs of the board? That we might be able to squeeze a hell of lot more onto a prototype. And leave the regs open to some experimentation? Surely the 7815 or whatever can be bettered...or was I lazy again (don't wanna go throught the whole thread again.)...and are they part and parcel of the package for some reason
- Status
- Not open for further replies.