Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Building PPL's 2001 Space Oddity/Apheared 42
New Posts  All Forums:Forum Nav:

Building PPL's 2001 Space Oddity/Apheared 42

post #1 of 30
Thread Starter 
Well... I'm taking the plunge. I've decided to build PPL's 2001 Space Oddity (as Apheared says it should be called...). Of course, having gotten my electronics education from someone who calls himself Shifty Bugman... (Author, The Basement Bugger's Bible)

I plan to try the whole design out on breadboard, then get an ExpressPCB made. If you want to see an example of horrible first-time PCB design work, I've got the preliminary design on my server, hal9000.drsue.net/files/spaceoddity.pcb

I based the placement of resistors and caps on that of the original Cmoy design, since I know little about designing amps other than that which is on page 33 of the aforementioned book, and on the datasheets of the components used in the Oddity.

I do have a few questions, though. Is the 301 ohm +input resistor Apheared mentioned in the original article eqivalent to R2 in the Cmoy design? What the heck is the "parallel of the feedback"? Where do I put the stabilizer cap? And which is the output resistor Apheared was trying to get perfect?
post #2 of 30
Could you post some links to the articles that you are talking about please? This sounds interesting but people around here don't seem to like to link to things I guess.

I mean I know you're talking about a book but I want to see this amp.
post #3 of 30
Thread Starter 
post #4 of 30
I've read through what I could find, and the most relevant post in that thread seems to be the last one: if you can't figure the circuit out from the description, you probably don't have any business building it. It seems to me that several salient points about the circuit are implied, and others are matters of taste to be decided by an experienced builder.

A good example of the the latter is that Apheared never says where the buffer goes, exactly. One post in that thread by someone else says it goes inside the opamp's feedback loop, but Apheared doesn't say that, nor does he confirm or deny it later in the thread. Reading between the lines, it sounds more like he had it outside the loop, and from what I can see in the photos I believe I'm right. Apheared is being very humble in this thread, which tells me he thinks it's a very simple circuit. Since putting the buffer outside the feedback loop is simpler, I infer that's how he had it set up.

Another example is the matter of input capacitors: Apheared never mentions them, but in the protoboard pics you can see a pair of box caps that are in the right place to be opamp input capacitors. Yet, he also specs a 308 ohm input resistor. That resistor and the input cap (if it's present) form a high-pass filter, rolling off low frequencies. If you model this in a program like MicroCap, you find that you'd need a 47uF cap or larger to keep the low-frequency rolloff from eating into the bass audio range. (A 0.1 uF cap as spec'd by Chu Moy for his pocket amp rolls off at about 1 kHz with this resistor! That's why the input resistor is so large in the CMoy design.) Yet the box caps I see in the photo are pretty small, and they're probably film caps, which means they probably aren't 47 uF jobbies. I'd say they're more like 1 uF. So, did he stick with the 308 ohm input resistor, or did he manage to find a very small large-value film cap, or am I just misreading the photos? I can't tell you the answer to any of those questions.

More ambiguity: he says the opamp has been biased into Class-A operation, but doesn't say how. Apparently one is expected to understand the causal mechanism from the mere mention of the effect.

All is not dim to me, however:

Is the 301 ohm +input resistor Apheared mentioned in the original article eqivalent to R2 in the Cmoy design?

What the heck is the "parallel of the feedback"?
When using bipolar input opamps, you have to balance the input offset currents. See this article for how this is done, in the context of using the LM6172 opamp. The same principle applies here. The short version is that running the 332 and 4.7K values through the parallel resistor equation equals 308. See that article for why this relationship is helpful.

Since the AD823 he's using is a FET-input opamp, that's why Apheared describes this balancing as overkill. It really isn't needed, but it's clear he was going a little overboard on the accuracy angle when building this amp. (Evidence: See the oscilloscope stuff at the end of his post.)

Where do I put the stabilizer cap?
Exactly where Apheared said to: ahead of the TLE2426. I expect he's using the rail splitter as normal, with the stabilizer cap between the power input rails. As Apheared said, this may be overkill. But then, I've been putting 470 uF caps in that position in my BUF634-based power supplies, so I'm just a pot calling the kettle black here.

And which is the output resistor Apheared was trying to get perfect?
The one between each opamp's output and each buffer's input, of course. See ppl's first post in that thread and Apheared's response.

There: I've answered all your questions and raised some you never thought to ask, and do not know how to answer myself. I'd say you are in for an education if you keep at this project.
post #5 of 30
Thread Starter 
Thanks, Tangent!

Regarding the buffers, I'm inclined to agree with you because in every buffered-opamp schematic I've seen, it's been wired that way.
On the the board design, I'll leave room for some big 47uF jobbies, unless Apheared reads this thread and clarifies it a bit.

And yeah, the point of attempting this amp was so that I'd have to learn something about amp design

edit: I've updated the PCB layout, to reflect Tangent's information.
post #6 of 30
Thread Starter 
Well, I've pretty much finished the layout. I'm 90% sure I got the wiring right...
I've designed it so I could board-mount a Noble (maybe Clarostat, depending on lead spacing) or Panasonic (by bending the leads a bit) pot, since when I was building the Hansen board, wiring the pots took the longest of any component. Although I've checked the hole spacing of the Noble pot against the one in my Hansen board, I don't have a Panasonic pot to use, so I'd be much obliged if I could get some feedback on the issue

I've also left tons of space for some really big input caps and power caps, but included holes for input caps with a lead spacing down to .5". I also included a SSOP->DIP adapter board in some spare board space, since I've always wanted one so I can do another project. The tentative part values are as follows:
R2=308 ohms
R3=332 ohms
R4=4.7k ohms
R6=66 | 61.9 ohms
R1=LED/power dependent
C1=47uF film cap (any better suggestions?)
330uF electrolytic power caps (as marked)

Input voltage=???

FYI, the PCB file is here, if you're too lazy to copy/paste the URL above.
post #7 of 30
For the pot on the Hansen board, there are two rows of three holes each with 100-mil centers, but the Panasonic pots have one row of six pins with a narrower pitch: about 85-90 mils. That's eyeballing with a vernier caliper, because I can't understand the datasheet's measurements. (It says "2-5.0" for the pin spacing, whatever the hell that means.)

Anyway, take it from me that trying to get the Panasonic pot to fit into holes with 100-mil spacing is futile. Lead bending isn't enough to compensate for the difference in pitches. If you don't design for this part specifically, you'll end up running hookup wires from the board to the pot.

As for the input cap, consider leaving it out if your sources all lack DC components on their outputs. That avoids the issue of input cap selection altogether.
post #8 of 30
Thread Starter 
I meant to bend the leads of the Panasonic pot by 90 degrees, so the pot would stick out the front of the amp, not the top . I guess I worded it wrong...

Incidentally, is there a way to tell ExpressPCB to snap to a grid of 85mils? It only takes input in fractions of an inch. (in decimal form)
post #9 of 30
"Mils" are thousandths of an inch. Occasionally you will hear that term, too: "85 thousandths". And on datasheets, it's usually just written out in decimal form: 0.085".

As for bending the leads on the Panasonic pot, you might not have to do that, if you're making your own PCB. There are two variants on the basic pot design, a vertical version and a horizontal version, made for vertical and horizontal PCBs with the shaft emerging from the case horizontal. If this isn't clear, examine the datasheet drawings.
post #10 of 30
Duplicate post, please ignore.
post #11 of 30
I was hoping ppl or Apheared would jump in...

You can place the output buffers in several different configurations. I built the amp with three different buffer configurations and finally settled on the multiloop design (local gain around the AD823 with the buffers inside the feedback loop). You can look in HeadWize here for more info.

I built mine almost identical to the Jung Topology with an overall gain of 5.

I believe PPL has recommended this design in other threads.

As to the small capacitors in Apheared's design, I believe those are for power supply decoupling. I have found that with the multi-loop configuration, and the AD823 combined with the EL2001 buffers, DC offset is extremely small. No need for signal coupling capacitors unless your source has offset.

I also built my amp with a BUF634 virtual ground driver. Because, uh, I like to drive the right and left channels with-out-of-phase sine waves

Good Luck!
post #12 of 30
not sure which topology they were referring to, but i've put together (on breadboard only) the multiloop topology with the 823/2001 (2 in parr. for each channel) and the 823/2008 in the same layout... both sound very nice... i prefer the 2008 for a little extra low end reach (albeit at the expense of battery life). although, some have said i need at least four 2001s per channel to drive grados, so that may be what i'm hearing... i'm extremely pleased with the results so far with the 823/2008 using a 634 to drive the ground(the PS is waiting for a few more parts, but i'm not sure if i'll see any improvement, honestly)... this is the only setup i've tried so far that makes the bass sound tight on crystal method songs without recessed vocals. it draws about 35 ma, so i need to reevaluate my 3x9v supply
post #13 of 30
Hi. Aparheared used the 2001 within the overall 823 Loop as conventional opamp/Buffer Combo's as depicted in the BUF-634 data sheet. Ah El-2008's a Buffer after My Bass hart with the AD-8610 This Buffer really Comes alive The 2008 can sound Thin in the Bass with a 825 and warm in the Midbass with the 823. The OPA-627's Prominate mid's are too Pearcing with all Elantec Buffers. the AD8610 is way Better than the OPA-627 in all regards with any Buffer Evean the BUF-634 and the BUF-634's Recessed Mids compliment The 627's Pronounced Mid's. The EL-2001 is a Great sonic Match for The AD-823 but the Light and Airy and up front upper mids of This OPA compliment the Nutral 2001's sonics in this Range for a splendid sonic Presentation. An AD-8610/EL-2008 combo is nice and soft in the Mids yet still have Plenty of Microdynamics. The 8610 appears to have about an octave Deeper Bass extention than eather a 823,825 or OPA-627 with the 2008. All my inferences were using a multiloop Topology With all my Trappings in Powersupply and Current sourcing i takked about over the Years on this and the Headwize Fourms. See My Portable Headphone Amp Design at http://homepage.mac.com/WebObjects/F...eSharing1.html For a Portable amp you will just love to Listen to alot if built using AD-8610's and a EL-2008 Folks might just have to Pry the Cans off your head. You must use all the Circuitry and add my cascode Current source. I have a hard copy of the Actual PC Foil Pattern and Component Locations if someone would like to scan them and Put on the Net or post the Gif File on this site ill send you the hard Copies. For info on My Casscode Current source Go to this link http://www.head-fi.org/forums/showth...p+into+Class+A Hear you will also get to hear oposing vewpoints from my critics, But for my ears this works verry well
post #14 of 30
Hey ppl, how would you rate the 8610 with either the 2001 or 2002? I've got four of each of those buffers but no 2008s.
post #15 of 30
Well I have been listening to a stack of (6) El-2002's per Ch. with the Ad-8610 and cascoded Current source and the Amps Icq is more than a 2008 but the sound is verry articulate going to a stack of 2001's sofened the top end slightly and may be Better for poor source and recordings. but with q Quality source and Good Recordings like Gladiator Wow or Pink Floyds Pulse DVD, Fantastic. Digital cannons on the telarc 1812 Explosive Still lacks the ultimate slam in the Lower octaves compaired to an el-2008 but alot better Mids and highs. A Great sonic match to the EL-2001 is the AD-825 as is the 2002 but the 2001 tames the over prominate sibelence with the 825 bass is more extended than usind the 825 with the 2008, Figure that. I am now trying Mixing 2001's with 2002's with the 8610 I havent formed an opioin on this yet but sounds unique and positive so far. like neather the 2001 or 2002 alone. try this if you have Both 2001' & 2002's
New Posts  All Forums:Forum Nav:
  Return Home
Head-Fi.org › Forums › Misc.-Category Forums › DIY (Do-It-Yourself) Discussions › Building PPL's 2001 Space Oddity/Apheared 42