Piano World Home Page Forums Our Most Popular Forums Digital Pianos - Electronic Pianos - Synths & Keyboards The DIY Digital Grand stand

Ever since I was a kid I wanted a grand piano at home. There were so many reasons not to that I ended up just going for a keyboard and calling it a day. Until now. In the spring I advertised in the local buy and sell for a "broken" baby grand piano. The only real requirements were cheap and relatively good looking. I got a hit and picked up a free 1922 Adam Schaff 5 foot baby grand that had been sitting for years after being in a fire. My goal was to create a very realistic looking and sounding cabinet for my newly acquired Yamaha CP5.

Now that the project is nearly complete I decided to share a bit of my experience and progress with you all. A couple of months back Dewster posted about his progress on a stand with speakers and I really enjoyed watching that. So here goes....

Sadly I didn't take an official "before" picture but I glued together a few shots taken from the previous owner.


I looked around on the net and found a pic of a similar Adam Schaff piano - hopefully the final product will look similar.


After removing the lid, legs, and harp I couldn't wait to figure out where I would put the speakers...


It took a good portion of the summer to clean up all the pieces. I attempted to use paint stripper but that ended up so goopy that I decided to just scrape all the gunk off. In hind sight I probably should have been a bit more gentle but I didn't realize that most of the finish was on top of walnut veneer. Live and learn I guess. I ending up doing all the smaller pieces first, painting and varathaning as I went along. Finally I got to the body just the other week.


One issue that was difficult to manage was the width of the CP5. Because of the mod wheel the sucker is really long. It ended up taking 8 hours to gouge out a 6 inch by 16 inch 1.25 deep trough in the sides of the piano to make it fit. OUCH!


Now that the main unit was stripped I could get to the work of creating new upper and lower sound boards to hold the speakers. Here I am deciding which ribs to keep and where the speakers would fit best.


Over the last two days I go the sound board stained and the main body painted. Next I will varathane the body to make it more impervious to scratches and the like.

Great project! Looks great.

Can't wait for the final pictures! Nice project.

Great project! Photos when you're finished ....

Awesome project! I am helping a friend and his dad fit a second hand Yamaha P80 in to an old birdcage upright as we speak (except we plan on integrating the digital piano's components in to the upright, losing the DP body entirely). These projects are fun!

Wow, this already looks terrific - great work!

James
x

Absolutely amazing!

(Now how again are you going to get it in the VW?)

That's so cool CMac! Let us know how it works out. Make sure the amp(s) you use to drive the drivers have enough head-room in them so that you get nice clean power to the drivers and peak power. In other words, don't go cheap on the amps!

Nice project CMac!

What speakers will you be using and what was your criterion for choosing them? I assume they'll be operating in an open baffle configuration (no box to contain the backwave)?

Fascinating....looking forward to updates....

Hi Dewster, I didn't spend a lot of time calculating box sizes. My main concern was choosing high quality speakers of similar output sensitivity and having enough of them to recreate the "wall of sound" effect from the soundboard radiating area.

The Box:
I chose to use an infinite baffle design (closed box) with two large 12 inch drivers shooting downward without a crossover network limiting their frequency response. The main idea here was that the CP5 is a sampled piano and I was not convinced that there would not be sensitivity issues with the bottom end. I do not know enough about a real piano and how loud the base strings are and the physics of the wave behavior (constructive and destructive interference's at varying frequencies). If I chose an infinite baffle design I could control that and have a degree of confidence for the overall frequency response. (I owned Magnepan speakers in the past an remember the bottom end sensitivities when placing them within three feet of a hard surface.)

The Drivers:
We have an audio supplier down the street from where I work and decided to choose from their catalog. (I like to support local business:) The tweeters are Goldwood GT-510S which have a frequency range of 2k-20k with a fairly flat response. They are also fairly efficient at 90db 1w/1m. The woofers are Goldwood GW-6PC8 which have a frequency range of 46hz-3k. Their efficiency is a bit less at about 88db but should be sufficient. Crossovers are Phillips ADF 3000/8 which cross both the top and bottom over at 3k (6db per octave on the low pass and 12db per octave on the high pass).

Soundboard and speaker placement:
I had to remove the original soundboard in the piano as it was in very poor condition. I went for a 1/4 piece of one good side maple plywood. Time will tell if that choice was too thin or not. I decided I wanted to keep the weight down as much as possible and possibly introduce a small level of soundboard resonance. Usually a bad idea in speaker design but I have a hunch that it may make for a more realistic piano sound overall. Perhaps the speakers will behave somewhat like a tactile transducer. The piano ribs are still attached so there should be enough support for the weight of the speakers. Since I am driving the speakers with an A/V amplifier I have channels A (right and left), B (right and left), front and back at my disposal. The 12 inch woofs are alone on channel B firing down, and the other 4 pairs of drivers are on the soundboard shooting up in a configuration that should allow for a bit of the stereo imaging that was there from the samples and covering a broad enough area to hopefully provide for that "wall of sound".

Amplification:
I purchased a used Yamaha RX-N600 A/V amp along with the speakers that housed the 12 inch woofers for $80. It will provide separate 95 watt amplification to the various channels which should be more than ample to drive everything. It has a nice feature which can simulate surround sound given a stereo input pair so that should provide an "in between" signal for the front and back. I don't perceive that stereo imaging is going to be much of an issue but hopefully it will be an approximation of what is happening from a real piano sound board.

Possible additions:
I may wish to place attenuated high frequency drivers on the music rests to recoup some of the stereo separation generated from the synth sounds of the keyboard. I appreciate the warm atmosphere that stereo provides and I am concerned that I may loose some of this with the loss of the point sources of two main high frequency drivers. I will play with this a bit after I have everything hooked up to see if it would be beneficial or not.

I'm concerned about the 1/4" soundboard. Seems awfully thin. Also, I don't think it will simulate the effects of true piano soundboard. There's nothing to properly couple sound from the speaker to the board. In a piano, the bridges drive sound directly into the board. Kawai ties their transducer directly to the board. Your speaker cones won't do that.

I agree about adding tweeters somewhere "above board", up front. I think you'll need them.

While I agree that there is no "direct" connection I think the sound coming out of the rear of the speakers into the baffle cavity will "couple" them to some extent. I am not looking for the resonant amplification factor that one gets in a true grand piano - just for a little extra cabinet color. How it sounds will be another story I am sure. The thinness does not concern me so much as I only removed two or three ribs to make the speakers fit and the speakers themselves have solid cast ribbing.

I've found that matching sensitivities isn't that much of an issue really. You want a tweeter that is at least a little more sensitive than the mid woofer so that you can level match with a series resistor or resistive divider at the crossover. And if the tweeter is a lot more sensitive it will be able to effectively handle more power than the mid woofer in a system due to this attenuator. Tweeters are often the first thing to blow when driven hard.

The sensitivity of the 12's isn't that important either since they will be driven by a separate amplifier. I'd spend a bit of time using a free tool like WinISD to at least confirm your choice of box sizes for the woofers.


Sounds good, I suppose the boxes and 12" drivers will be hidden beneath the soundboard then?


Will the 6 1/2" drivers have boxes to contain the rear wave, or will they run open baffle? If open baffle, I think I'd pick more of a PA type driver with a stiffer cloth surround, rather than rubber. 3 kHz seems like a reasonable crossover point, but I'd have to experiment with the actual drivers to know if that and the slopes were somewhere near optimal. It's a lot more work to design your own, but I'd recommend you stay away from off the shelf crossovers unless you've seen a project on the web that uses them and turned out well (and measured well). There's just too much interaction going on around the crossover point, what with the natural roll-off of the drivers, the rising impedance of the woofer, etc. complicating things to simply use by-the-book formulas.

If you do run the mids open baffle you might want to install a series cap to keep accidental DC and low bass from injuring them.


Will you be gluing the soundboard to the ribs? If not, what will you do to suppress unwanted resonances and rattles between the ribs and soundboard?

Right now I have a very simple design of all the speakers sharing the same box - which happens to be the full size of the soundboard by the thickness of the main structural ribs of the piano which are around 6 inches thick.

There were a number of factors for me here. I wanted to keep the weight down, so sharing the main box seemed like a reasonable choice. I like the longevity of the rubber surrounds over the foam ones for sure and I like the aesthetic appearance of the rubber over the cloth, and seeing how this is the main noticeable piece of the piano when people realize that it is not acoustic I stuck with the rubber.

This brings about another design decision that I have made. Seeing as how I was loosing most of the stereo imaging, and I was firing the drivers up in the air and into the floor I questioned being too particular off the hop. I have a good stereo in the same room as the piano so I can just hook that up if I want a more traditional sound. Also, I have decided that I will be keeping the volume levels below that of a traditional grand piano too. This doesn't negate the desire for a good sound coming from the instrument of course so I did choose reasonable drivers. The 6 1/2 inch woofers are surprisingly stiff and given their efficiency I cannot foresee driving them very hard. The choice to use off the shelf crossovers basically fit this model so I went with that. I don't foresee the need to port at this point either as the amp has a number of options for tailoring the sound to fill any gaps if they happen to arise. My main reasoning here is that all grand pianos sound distinctly different. With high fidelity speakers the desire is to reproduce the sound as accurately as possible. In this configuration what is accurate becomes more a matter of personal preference. Once it all comes together I will tweak until it suits.

This one is new to me. Are you suggesting that the larger drivers may be pushing too much air against the smaller woofers and possibly injure them? If so, will the new caps change the sound at all?

Ah yes. Been spending a bit of time with that question. I think I may end up using silicone to attach the soundboard to the ribs. Silicone is a good sonic insulator but I wonder if it may steal the desired cabinet resonances that I was hoping to get. My concern with glue is that if it ever lets loose (like the original piano did.) the buzzing may drive me crazy.

I hope you'll be using inner partitions to isolate the 12" woofers from the 6 1/2" drivers. Even in traditional 3-ways the drivers are all acoustically isolated in the rear. The tweeter nearly always has it's own sealed rear resonant cavity, the mid either has a closed back or its own small box, and the woofer takes the lion's share of the remainder. If this wasn't the case the tweeter would be instantly blown out by the woofer backwave, and the mid would really take a beating too.

You should also internally acoustically isolate the left channel speakers from the right, otherwise the stereo imaging will suffer.


I always pick rubber over foam surrounds for the reasons you state. But I think I would still pick cloth roll over either for this particular application - the compliance is lower so you don't have to worry so much about long term gravity induced "cone droop" (which would make you unpopular with the girls). Dayton and Eminence make some nice models.

What are you thinking of doing to protect the speakers? Vertical mounting will make them collect dust like the dickens, and they'll be much more susceptible to accidental damage too.


One of the hardest thing to tweak in terms of frequency response after the fact is mismatched drivers, which is why people spend so much time trying to get the crossover as optimal as possible. And I'm not trying to tell you not to experiment and have fun, but it's been my experience that Murphy's law works overtime when it comes to acoustics. Chances are the resonances and other things introduced by unorthodox speaker construction won't sound so nice when reproducing piano sounds. So I'd aim for a flat frequency response and take it from there. I think your choice of 1/4" ply is way too thin for the largest drivers - too much of the backwave will escape, making the bass sound thin, and I would fully expect some wild, uncontrollable resonances.


Like I said above, the mids will hate and perhaps be destroyed by the woofer backwaves, so they should be acoustically isolated from them. I would use a cap in series with the mids only if they are running open baffle, otherwise the spring effect of air in the sealed box behind them will be sufficient to protect the drivers from being damaged by DC or low bass.

I think you are correct if this were a conventional enclosure. But if the builder places all these drivers on an "open baffle" then there is no pressurized cabinet. You said to place each driver in it's own space. I think a better more generalized
advice is "Don't place then in the same cabinet."

If the builder is using a grand piano sound board as an large open baffle then those problems don't apply. as there is no cabinet

I think Yamaha's Advent Grand is like this

I am not so sure the AvantGrand is like that.
I found this article that shows low frequency drivers firing into the floor.
http://www.popsci.com/gear-amp-gadgets/article/2009-02/yamahas-digital-grand-piano

Interesting, thanks!

Though from the article: At home, Cameron plays on a Yamaha Model C -- a six-foot piano selling for about $40,000. So then I asked the crucial, rubber-hits-the-road question: If the $19,000 AvantGrand had been available when you were on the market, would you have bought it? "Definitely!" he said, with a force that made me think he meant it.

Ugh! Of course Cameron plays a high end Yamaha at home (not e.g. a Chickering). Why do all of these product puff pieces have such a transparently awful cookie cutter feeling to them? I need a shower...

[Edit] I clicked on the link to Cameron in the article: Mr. Cameron is a Yamaha Artist since 2008 so it all makes sense. I need to stop reading these industry ads masquerading as unbiased info articles.

Strange view of the woofers underneath, the nearest one in the picture looks different than the other two. Can't tell if it's a passive radiator or just a trick of the light:



Anyone with this AG care to comment on that?

All previous speakers I have made have been two driver designs with the tweeters completely sealed hence they were in essence one speaker per box. I understand that my current project is an unorthodox design with 6 woofers sharing one box. I'm not really worried about loss of stereo imaging given that a real piano has an infinite number of sources. I will mostly be playing the instrument at levels far below that of a real grand piano so this shouldn't be an issue but the possibility of "destroying" the drivers does concern me.

This is a complicated issue from what I have read. When multiple drivers (fed by the same source) are placed within the same cabinet they require twice the volume of a single speaker. It is a valid design for speakers to share a single box. In my case I have 6 woofers each being fed by a different signal where two of the drivers are larger than the others. This is unlike a single boxed 3-way system where just a single midrange must take the backwash from a single larger woofer. I could see there being a much higher probability for midrange destruction in that case. In my system every driver will be sharing the backwash load with 5 other drivers within the cabinet - all being fed similar frequencies (the smaller speakers are not midrange drivers - they are just smaller woofers). One of the 5 will be the other 12 inch woofer and possibly both of them. I will have complicated waves pushing back on every driver most of the time for sure. This situation is not ideal for "accurate" reproduction of the original signal. Then again most boxes have some sort of back wave that they must deal with. So the question is - how much is too much?

Apart from removing all the ribs and constructing a labyrinth of boxes or partitions are there any other options that would work for me given my design constraints? How about rock wool insulation? It is of higher density than the typical poly or fiberglass often used and one layer between the 12's and the 6.5's would be a type of acoustic insulation. From the literature that I read enclosure insulation effectively reduces the effect of the back waves within the enclosure by dampening standing waves and the like. Perhaps a port would also reduce the stress on the smaller drivers. Ported designs are generally more complicated though and I leave that as a less desirable option.