[antomas]

Thanks very much for your kind answer and sorry for overlooking the other post!
Massimo

 

[jarthel]

Please be aware that this chassis is heavy. Shipping cost is a bitch but you can't deny the fact it's looks so good.

Please also be aware that drilling this is a bitch too. It was the hardest part of the project!!! My hands ached from drilling.

I also bought my TKD pots off them which I would say are very nice pots.

Jayel

 

[jerikl]

BoyElroy--

If you don't mind me asking, how much do the Gilmore and Szekere's amps cost to make? Thinking about building one of the two and am wondering about cost compared to the CHA47 or CMoy amps.

 

[BoyElroy]

Hi Thomas--

You know, its really strange...I must have some sort of brain aphasia because I get all confused when I try to solder circuits on a breadboard/prototype board. Maybe its that 3D vs. 2D thing, but I have a very low success rate when doing point to point wiring projects and almost a 100% success rate when I print a PCB. I'm really bad at math and music and I would guess that I'd be horrible at 3D chess as well--which would lead me to think that its some sort of left brain/right brain/visualization thing with me...

I hope I didn't give the impression that this is in any way a substitute for a professionally produced pcb such as the Hansen board(!) I put up the "47" design I used as an example of a really cheap and quick way for novices to begin making circuits that work. The point was to use basic materials and household items to do a fast, dependable and easily modified build. I've run off about six or seven amps from this design and given them away to friends...I guess I like the fact that I can reproduce these boards whenever I want and slap them together for people who would otherwise never get a chance to listen to higher fidelity sound. Plus, its cheap!


Hi Antomas--

I'm glad you're thinking of making the Gilmore amp because its a very high quality circuit. If I could come up with two words to describe the sound, it would be "balanced" and "smooth". The bass is firm and deep, the highs are clear and never harsh and the mids are full and on the warm side. It doesn't call attention to itself and lets you listen to hours and hours of music without fatigue. A very cool thing about my Gilmore amp (non-bridged) is how the soundstage expands to points seemingly outside your head. The imaging is both precise and very, very wide. Voices and instruments separate in a way that other amps that I've made don't quite match.

-Ironing Transparencies:

I think the point with an iron is to transfer as much heat as possible onto the transparency so my answer would be, it depends on how hot your iron is. I have a pretty low-power home iron so I usually don't use anything in between the plastic sheet and the iron surface. I think that if you have a more powerful iron, you could use a sheet of copy paper on top of the transparency.

My iron has a teflon coating as well so when some plastic (such as tape that I use to hold down the transparency) does melt on to the iron surface, I can just wipe it off with some acetone.

I'm using a 100K Alps pot with the Gilmore amp which I bought at Radio Shack for $3 US. It sounds pretty good, but I'd like to move up to an attentuated ladder type soon...

My suggestion would be to just experiment with different transparecies, irons (cheap ones?) and layouts. Remember, the most expensive single item will be the copper board and if you screw up, you can just wipe away the bad toner ink with acetone and do it over again until you get a print that works. You'll be surprised how well it works (I hope!).

best of luck!

 

[BoyElroy]

Hi Cap't Bubba--

Oddly enough, this thread started as just a quick rundown on the different sound qualities of the various amps that I've the pleasure of building here and at Headwize...but...I realize now that its not an easy thing at all! Just having dealt with an amp as difficult to nail down as the original PRR-Szekeres, I'm finding that its a very time consuming process to fairly compare more than two amps at a time. That is to say, what may seem straightforward for two amps, seems to get exponentially more complicated with three or four. Just trying to remember the general characteristics, not to mention the specific ones, of several amps is a pretty daunting task! I certainly have newfound respect for all those professional reviewers out there.

What I've decided to do is just spend a few days each with the four/five different amps I have just checking out different tracks, then sitting down and comparing the sound of perhaps 3 or 4 common tracks on each one. I'm tending towards using some of the following tracks:

1) Stereophile Test CD 2--Acoustic Drum solo
2) Stereophile Test CD 2--"Mapping the Soundstage" imaging test
3) Empire of the Sun OST-- "Suo Gan"--Excellent high freq. track
4) Empire of the Sun OST --"The British Grenadiers"-amazingly low kettle drums that have bottomed out every speaker I've auditioned except the Talon Khorus (!)
5) Bach Cello Concerto #2 -- Robert Cohen
6) Ella and Louis-"Moonlight in Vermont"-- Excellent vocal recording.
7) Melon Remixes-"Stay" (U2 Underdog Mix)--great bass, many left-right pans, echoes, background vocals.
8) Groove Armada -- "Inside my Mind"--Nicely mixed electronic track

If anyone has any suggestions or reviews they'd like to post, please feel free---I'm afraid that I won't have anything up until next week or so...

my apologies for the delay--

 

[antomas]

ByElroy, Jarthel,

Thanks very much for your answers! You make me convinced that "it can be built", and "it can be beautiful, too" (I'll work on it next summer). I also have a crazy idea to add a phone/preamp switch to take the power off from two of the output transistors, but I don't understand if the additional complexity is worth the effort. For the power supply, I think I'll try to use a simpler power supply with just a LM317/LM337 couple.

Thanks once again for all.
Massimo

 

[BoyElroy]

Well, here's the first installment of my highly non-scientific amplifier survey---

I'll limit my first reviews to the basic SZEKERES and the GILMORE(non-bridged) for now and add the 47 later.

A Disclaimer: I'd like to emphasize to anyone reading this that all the following comments refer solely to the specific amps that I built over the past couple of months. As is true of any DIY project, component choices, build quality and layout will differ from amp to amp and all review comments should be read with this in mind. Just because my homemade Szekeres sounds like "abc" doesn't mean anyone else's will sound the same way. I will try to compensate for this by detailing the specifics of each amp and leaving it to the reader to extrapolate from that.

My Szekeres--

This is the basic Szekeres built around the 510 mosfet. It has a 1uF Hovland Musicap as the input coupling capacitor and a 1000uF Panasonic FC electrolytic output cap. My R4 is a 12 ohm Dale wirewound resistor with aluminum cladding and the mosfets are mounted on two large (5" by 6") heat sinks. The power supply is a highly regulated 1.5 amp Teac laboratory unit ("PSII") with multiple adjustable outputs. All listening was done with the power supply set to 15 Vdc. The volume pot was a 100K ohm Alps pot purchased at Radio Shack.


My Gilmore

This amp was made with mostly high end parts; Vishay and Dale resistors, Canare internal wiring and generic gold plated rca plugs. The volume pot was also a 100k ohm Alps unit from Radio Shack. The power supply was built according to Kevin Gilmore's "Ultra Regulated Power Supply Design" and set to output +/- 16.40 Vdc.

All listening was done on my Grado 325's with the bowl ear pads reversed.

The Sound:

To begin with, the Szekeres is a really fine sounding amp. I went back to my basic Szekeres (rebuilt) after flirting with about 5 different variations and was taken aback with how nice and smooth the amp played a wide variety of music. It was far superior to the headphone jack on my 5 channel Sony amp and easy to listen to for hours on end. It's strength is definitely in the midrange and adds a powerful bass foundation to the music as well.

After listening to the "Bass Decade 200Hz-20Hz" track on the Stereophile Test CD2, I would go so far as to say that the Szekeres goes even lower than the Gilmore. The difference isn't great, but the Szekeres does play noticeably louder at the final 30-20Hz decade than the Gilmore.

However, that being said, the Gimore's bass is much faster and quicker than the Szekeres. On certain tracks with a strong bass line or drums, each instrument comes through clearly on the Gilmore, with the transient attacks taking on a sharp, crackling, almost electric character. The Szekeres, on the other hand, presents a much more veiled image, with drums and bass guitar sounding more like a generic "thrum, thrum, thrum" than a bass string being snapped against the surface of the bass. This sense of a veil being pulled across the soundstage extends across the frequency range as well. Where the Gilmore gets across even faint sonic cues such as a glass tinkle here, a performer's breathing there, the Szekeres sounds just a bit muddy and congealed.

Indeed, one of the most striking differences lies in how well the two amplifiers re-create the space surrounding the perfomers. With the Gilmore, voices and instruments emanate from a palpable soundstage. Singers and instruments come forth from definite, clearly identifiable places. You can hear the echoes of chairs moving, people coughing and drums fading slowing into deep silence--the resolution is quite remarkable. You can close your eyes and explore the texture and layers of the song in a rather unsettling way. Its almost like taking a stroll through a forest of voices, sounds and instruments--the imaging is that vivid. On one track, "Moonlight in Vermont" sung by Ella Fitz. and Louis Armstrong, you can close your eyes and drift over to where Ella's singing, then walk over to the piano and finally over to Louis as he joins the duet. The instruments come out from such a strongly defined and wide soundstage that you just grin yourself silly and pull out more and more of your favorite CD's. It is something that I've never quite experienced before, even with my old ARC tube amp and Audio Physic Sparks, a combination known for excellent imaging if not much else.

The Szekeres, in comparison, tends to collapse the image, back to front and side to side. The sense of space surrounding the performers diminishes and you are left with the sonic image of voices and instruments coming more directly from the headphone drivers. The result is a noticeably flatter, more forward presentation. (I wonder if this comes from its rolled off high frequency response. I would guess that many of the sonic cues that are necessary to paint an aural image are in the high freq. range and the Szekeres suffers a bit from this lack of information.)

That is not to say that the Szekeres is not fun to listen to. It has a warm, upfront presentation that lends itself well to a most pop and vocal recordings.

On classical tracks and more complex musical tracks, though, the Gilmore is much more involving and incisive. It has outstanding resolution; its clear highs, agile bass and pin point imaging makes it fun and engaging to listen to orchestral passages, jazz quartets, vocals and indeed, any track other than a congested, badly recorded top 40 pop song. Even on the vast majority of pop songs, the Gilmore introduces a sense of space and rythmn 'n pace that the Szekeres can't quite match. You can clearly hear the individual drums, guitars, triangles, voices, etc. interact and play off each other on the Gilmore while the Szekeres tends to combine them into a more compact and combined whole. It is very difficult to get that sense of air and ambient space on the Szekeres--even on well recorded pieces. The detail just isn't there.

On more bass heavy tracks, such as the Shirley Bassey remix of "Goldfinger" by Propellerheads, the soundstage on the Gilmore opens up with instruments extending far outside the headphones. And even through this expanded soundstage, the bass and vocals remain clearly discrete and focused. The Gilmore has that common trait shared by many fine musical amps where you can turn up the volume to insanely high volumes without causing painful distortion. On the Stereophile Test CD 2 recording of "Ave Maria", the lead violin keeps climbing higher and higher until you think,"sheesh, when's this guy gonna stop?", and then it keeps climbing some more. The funny thing is, no matter how high it goes, it doesn't cause the sort of high pitched, etched in glass sort of pain that many badly placed/badly designed/badly purchased loudspeakers can cause. Through the Grado 325's, the highs were razor sharp and incredibly extended but never harsh or congested.

The Szekeres, on the other hand, seemed to go rather suddenly from soft, warmish highs, to a kind of painful take-the-headphones-off highs. And it wasn't that the volume was turned up too much, it was more that the amp had a "peakiness" that cut in at certain levels and made it somewhat disagreeable at higher volumes on certain tracks.


In conclusion, I think that I'd have to say that the Gilmore is certainly the more detailed, involving and arguably more enjoyable amplifier of the two. If one were to judge these two amps irrespective of price, then the Gilmore would be the clear preference. It has a level of transparency, quick, agile bass, clear highs and wide soundstage that places it at a different level of performance than the Szekeres.

The Szekeres, on the other hand, is an outstanding amplifier for what it is--a simple, easy to build and extremely affordable amp for driving low impedance headphones. A beginning DIY'er can put one together in a day and be enjoying music on a level that far surpasses most stock headphone jacks. It has excellent low bass response, very good midrange and smooth, slightly rolled off highs. It is pleasant to listen to, with a warm and relaxed presentation that lets you listen to music for hours at a time with minimal headphone fatigue.

The Gilmore may be a bit more time consuming to make, both in terms of amp construction and power supply, but the improvements over the Szekeres are not small. The extra time and effort are easily worth it and in the end, the DIY'er will end up with an amp that he or she probably won't need to upgrade for a long time, if ever.

Thanks for your patience!

Addendum:

I added 50K dual mono stepped attenuators to my Gilmore amp today and the already terrific sound got even better. The top-notch soundstaging, imaging and channel separation of the amp jumped up to yet another level. If anyone is considering moving up to a stepped attenuator from cheaper plastic film volume pots (not the high-end ones like the Alps Black Beauties) , my advice would be to go ahead and do it--the difference is not small!

 

[Possum]

Great sound comparisons! I'm now interested in trying the Gilmore amp since I also have a pair of SR-325's.

Do you happen to have any Sennheiser headphones to use for listening tests?

 

[BoyElroy]

Hi Possum--

I have the closed ear Sennheiser HD25-1 which is a 75 ohm 'phone that sounds pretty different from the larger Senns (from what I've read). I've been trying to talk a lawyer friend of mine into buying some HD600's so that I can borrow them, but no luck so far.

The gain on the Gilmore seems pretty high, though. On my 325's, I don't think I ever went past 10:00 O'Clock on my 100k log volume pot (where 7:00 o'clock is maximum attenuation). It seems to me that the Gilmore should be able to drive a pair of HD600's, although I don't know anything re: their sound quality. I'm actually pretty curious myself about how'd they sound with the Senns.


Hi Jerikyl--

Re: the cost of the Gilmore or Szekeres, its kind of hard to come up with a definitive number because so much of what I used (chassis, power supply parts, connectors, etc.) were scavenged from other projects and/or bought used. If I had to guess, I would supply the following ballpark figures:

Szekeres:

Heat sinks -- $10-$15
Vol. pot -- $3.00
15 Vdc PSU -- $30-$50
510 Mosfets -- $2.50 ea.
Resistors -- $5.00
Chassis -- anywhere from free to 100's of $
1 uF caps -- anywhere from $1 each to $20 ea.
470uF electro. caps -- Panasonic FC's at $3.00 ea.
Board -- $5
Other connectors -- $10-$15


For the Gilmore, I think the prices are posted already earlier in this thread. In general, I'd say that the Gilmore amp section would cost about $30-$40 and the power supply another $100 or so (with the servo and REF02 chips). It would be much, much cheaper to build the PSU without the opa548 and REF02 sections, but then you'd miss out on all the regulated goodness of KG's dual tracking design...

Hope that helps and good luck!

 

[BoyElroy]

5/21/02 *Note*

**I made a pretty embarrassing error in my original notes re: the use of the term "CHA 47". The "H" in "CHA" refers to C.E. Hansen, who has been producing professional boards for the Headwize/fi community based on the "47" circuit. The "C" refers to Cmoy and the "A" refers to Apheared who originally designed the amp. (Thanks Thomas

). Since I have never heard or built the Hansen-based version, a better description of the Apheared designed amp described below would be "A47" or just "47". My apologies to all concerned! **


Here are some comments on the "A47"/"47"

Once again, I'd like to make the disclaimer that all the amps that I am writing about are DIY projects and as such, differ from one another in important areas such as component choices and layout design. No two projects will ever be exactly alike and my comments refer only to the specific amps that I have built over the past couple of months.


The board I'm using was etched from a design integrating Apheared's amp section and Cmoy's power supply design. It is a basic, no-frills amp without a crossfeed or buffer.

I recently rebuilt my "47" amp using 1uF Multicaps (ppfx) for the input coupling caps and 1000uF Panasonic FC electrolytic caps for the power supply. Power came from batteries and a 19 Vdc laptop computer power supply. The opamps used were OPA2134's, one per channel.

All listening was done on Grado 325's.


The Sound:

To begin with, the "47" is a booty-kicking portable amp. It sounds great with the Grados or my Ety4's and they've run for 2 weeks off my 8 pack of AA cells. The 47 is clearly superior to my 5 channel Sony amp's headphone jack. I just made one for my upstairs neighbor and he said today that it sounds so much better than his stereo that he was going to buy a super long cord that'll let him walk around the apartment with his 'phones on (the plan worked!!!).

The 47 has excellent bass, good midrange and nice highs that only get congested at very high volume levels. Using my trusty analog Radio Shack level meter and the "Bass Decade" track on Stereophile's Test CD2, the 47 was just a shade down in bass output compared to the Szekeres on the 20 Hz test tone. Subjectively, also, the Szekeres seemed to go a bit lower than the 47, although this was not immediately noticeable on the majority of songs I listened to. The Szekeres and the 47 had very similar bass performance and again, the weightier sound of the Szekeres came through only on a select number of tracks.

On the treble test tone track, the 47 and the Szekeres put out almost identical levels up to 20KHz and this, too, was borne out in subjective listening sessions.


Perhaps the greatest difference between the "47" and the Szekeres is the soundstage. By soundstage, I mean a couple of things; the left to right and back to front imaging of the amp, the perceived physical width of the sound field and the ambient information of the space a particular song was recorded in.

In the case of the "47", I was struck by how much narrower the soundstage was compared to the Szekeres. That is to say, if I may use a optics analogy, listening to the 47 was like looking at something through a 100 mm lens while the Szekeres felt more like a wide field 35 mm. The soundfield of the 47 felt collapsed somehow compared to the Szekeres. Images did not extend as far left and right, depth was much more restricted and the presentation was definitely very forward and focused to the center. This was pretty much true of any track; the 47 tended to push everything towards the middle with the listener being placed in the front row.

On the acoustic drum recording on the Stereophile Test CD2, through the Szekeres, you can hear the drummer moving from left to right, up and down on his extended solo. Your eyes will actually dart around, following the impact of the drumsticks on the various drums and cymbals. There is a sense of space, too, surrounding the drummer and the soundstage extends about 180 degrees from ear to ear. On the 47, the impact of the drumsticks is slightly softer, but more than that, the drums are compressed into the middle so that the drumstrikes all appear closer to each other and mostly near the center. There is little or no ambient information so it is difficult to get a sense of space around the drummer. You hear mostly the sound of the drum, front and center, with the soundstage extending perhaps 60 degrees or so from center. The sound field never quite makes it from ear to ear, with all sounds appearing to come from in front of you, to varying degrees.

The Szekeres, on the other hand, does manage to convey a feeling of ambient space and a soundstage within which voices and instruments appear. To continue with the live theater analogy, with the Szekeres, you are sitting several rows back from the stage as opposed to the 47 where you are right there at the front of the stage. Just as the 47 tends to flatten the soundfield and narrow it, the Szekeres tends to add depth and extend it left and right.

Additionally, the Szekeres is able to portray separate instruments and voices within the soundfield more clearly. The 47 tends to combine them into a slightly muddier composite picture.
Listening to "The British Grenadiers" (Empire of the Sun OST) on the Szekeres, you can hear the ringing of the bells way on the right and snare drum way on the left. The horns open up in center and the kettle drum begins banging away to the left of center. The Szekeres manages to keep all these instruments more or less discrete and separate. The 47, as good as it is, can't quite match the Szekeres in his area; it is difficult to locate the source of the bells and snare drum and the kettle drum sounds more like a dull "boom-boom" that almost overwhelms the other instruments.

Again, I was especially struck by the differences in portraying a sense of space. On the closing bars of "Suo Gan" from the soundtrack for "Empire of the Sun", there is a dramatic moment when the choir builds to an ear-splitting climax then falls silent. The chapel continues to reverberate for a few beats afterwards. The Szekeres is able to convey at least a part of this decay from loud to sudden quiet. The 47, on the other hand, never quite gets across the sense that the singers are in a room to begin with so that when the voices stop, the end of the song feels as if the recording engineer simply turned the record levels down to 0. That feeling of being within a space and hearing music coming from that space certainly comes through much more on the Szekeres.

I suppose mostly due to the 47's tendency to narrow the soundfield, I don't think it would be such a good amp for jazz or orchestral recordings. It is not an easy thing to follow individual instruments on the 47 and its even more difficult to follow them as they interact with other instruments. For example, on the "Ave Maria" track from the Stereophile Test CD2, the lead violin as heard through the Szekeres is clearly but one of several instruments and there is a complex counterpoint being played out between them. The violin is center right and other instruments occupy extreme left and extreme right. On the 47, its hard to follow this interaction because the amp moves everything together towards the middle. The recording sounds more like a violin solo and once again, that sense of air and depth is shortened considerably.

So what does sound good on the 47? I would say that solo vocals and bass heavy dance tracks/pop songs would sound fine on the amp. Its high end response, while not super smooth or extended, still beats 99.9% of pcdp's out there and its certainly more portable than a Szekeres. Moreover, the gain on the 47 is much higher than a Szekeres and it can arguably drive a wider range of 'phones than the Szekeres. I use my 47 every day and I'd probably go nuts without it. It sounds great for what it was designed for--portability and high quality for minimal cost. I'll be making several more of these for friends and family before the home production line closes.

 

[tangent]

That lack of "space" you're hearing in the A47 is due solely to the op-amp. Upgrade to an AD843 or an OPA627 or something else in that class and suddenly your Suo Gan track will move from an indefinite enclosed space to being in a chapel that is precisely 84 feet wide by 211 feet from front to back and 18 feet high.

This isn't to say that the A47 is the be-all and end-all of op-amp headphone amps. Just that you always have to consider the op-amp in these designs, because they're responsible for a huge portion of the overall sound. You can do several other upgrades (faster P.S., biasing op-amps into class A, adding output buffers, etc.), but a simple op-amp upgrade is probably the single most important.

You also say that there may be some bass attenuation with the A47. That could either be due to too-small input caps, or it could also be a power supply speed issue.

 

[thomas]

Boy elroy: excellent reviews, i've had the same findings about the apheared 47 and szekeres, so your comments about the gilmore amp have me intested... i'll probably try it this summer (wired point to point

I actually do much better when i have the parts on hand to arange around for the best layouts, and if i had to do stuff away on a computer they probably wouldn't work too well )

IMO, the both amps have a pretty warm sound, but the CMOY varients tend to sound grainy and lack real resolution, while the Szekeres is smooth and has much better resolution all the way from the bass up to the treble (hence the better soundstage)... The sound is IMO a perfect match for Senn HD600...

Tangent- i haven't tried the AD843, but one of my earliest amps was a apheared 47 with dual 627 biased to class A (though that was before there was an apheared

) and it doesn't come close to a good szekeres or apheared 42 amp... The output stage of the 627 doesnt seem to do a very good job of driving complex loads like headphones, and distorts a lot... Adding a BUF634 or a discrete output stage (szekeres works well) to the OPA627 does wonders for the sound...

 

[tangent]

I haven't tried the 627 in class A. Bet your last dollar I will with the META42, though.

How much voltage did you give the 627? I find that unless you get it up over 12V, you can get nasty distortion pretty easily. There is a lot of detail, which may actually be grain, I'm not sure. Either way, it's gone with the AD843 -- it's a very smooth chip, but the important resolution is still there.

 

[ppl]

Tangent: are you refering to the dual 823 or the Sing 843 thay are different Ic's.

 

[thomas]

+/- 9V

going up to +/- 15 improves it a bit more, but impractical for a portable amp...

Its definately a very detailed, accurate chip, and can have a nice sound if used properly, but its output stage really can't drive headphones properly IMO, even with 2 opamps in parallel... A buffer in the feedback loop improves sound a lot, allowing the 627 to use its input and gain stages which are excellent...

But even with a buffer, many newer chips, esp from AD, offer similar performance (though very different sound) for much less...