Have been listening to my panels instead of experimenting, and sometimes trying to follow the theoretical discussions here, but cannot say I'm able to follow much of it lacking both a lot of knowledge and focus.
Been thinking about the exciters and amps needed. It seems like I can push more than what 4xDAEX25FHE-4 on one plate. Using 8 exciters doesn't seem like a great idea, so it might be worth for me to use 4xDAEX30HESF-4 instead, giving 160w per plate instead of the 96w I have now. I'm anyway leaning towards investing in equipment for vacuum bagging and with carbon/nomex plates they should be able to handle that while having similar efficency.
I will probably want 4 speaker stacks with 2*18" subwoofers, and then 4 plates on each stack. I will probably use active subwoofers, and it would be a lot more practical run run line level to each stack, so probably make one amplifier box per stack for the panels.
Each box will then have a total rated power for the exciters of 640w. So something with around 1000w total power would be ideal I think.
There are a few options I have considered. I'm running it with a TPA3255 amplifier now that sounds good to my ears (Ayima A08). It would be a very cheap option, even if of course you cannot use the cheapest possible board to run att full power for long time. But running one plate per channel might be pushing it though if I have 160wpc, and having one amp board per plate it is no longer that cost effective.
A common choice for pro audio manufacturers seems to be ICE Power and Pascal modules. They are actually quite cheap, at least when you compare the total cost of many of the speakers using them.
For example it seems like RCF speakers that are around €2000 use Pascal boards, which is same board IMG Stageline use in their €365 STA-1000D.
ICE Power have a 1000w board for €350 including PSU: https://www.soundimports.eu/en/icepower-1000asp.html
Now to the question...have anyone done any testing of made any observations regarding how much choice of amp affects the sound?
When for example reading discussions on Hypex NCore vs ICE Power, there any many comments that they are like night and day when compared, ICE is only good for bass and don't have good mid and treble and so on...
Is it true that DML speakers are much easier to drive for an amp, and is it perhaps unlikely to be any noticable advantages of say using an NCore amp instead of the ICE Power?
Been thinking about the exciters and amps needed. It seems like I can push more than what 4xDAEX25FHE-4 on one plate. Using 8 exciters doesn't seem like a great idea, so it might be worth for me to use 4xDAEX30HESF-4 instead, giving 160w per plate instead of the 96w I have now. I'm anyway leaning towards investing in equipment for vacuum bagging and with carbon/nomex plates they should be able to handle that while having similar efficency.
I will probably want 4 speaker stacks with 2*18" subwoofers, and then 4 plates on each stack. I will probably use active subwoofers, and it would be a lot more practical run run line level to each stack, so probably make one amplifier box per stack for the panels.
Each box will then have a total rated power for the exciters of 640w. So something with around 1000w total power would be ideal I think.
There are a few options I have considered. I'm running it with a TPA3255 amplifier now that sounds good to my ears (Ayima A08). It would be a very cheap option, even if of course you cannot use the cheapest possible board to run att full power for long time. But running one plate per channel might be pushing it though if I have 160wpc, and having one amp board per plate it is no longer that cost effective.
A common choice for pro audio manufacturers seems to be ICE Power and Pascal modules. They are actually quite cheap, at least when you compare the total cost of many of the speakers using them.
For example it seems like RCF speakers that are around €2000 use Pascal boards, which is same board IMG Stageline use in their €365 STA-1000D.
ICE Power have a 1000w board for €350 including PSU: https://www.soundimports.eu/en/icepower-1000asp.html
Now to the question...have anyone done any testing of made any observations regarding how much choice of amp affects the sound?
When for example reading discussions on Hypex NCore vs ICE Power, there any many comments that they are like night and day when compared, ICE is only good for bass and don't have good mid and treble and so on...
Is it true that DML speakers are much easier to drive for an amp, and is it perhaps unlikely to be any noticable advantages of say using an NCore amp instead of the ICE Power?
Hello Leob
Happy you have reach a performance with which you have pleasure to listen .
I understand this thread is difficult to follow. This part of the difficulty of DML which has a low background.
About your question, 2 things :
- you are looking for a very unusual power (far from my needs). This thread might be not the best one.
- I have been building amps for years. The one I had a very long time was an Hiraga 20W. We are now in an other age where the ICs have reach a very high quality. I have now cheap TPA3116 playing. For them I was surprised by the choices made by the supplier. Changing some capacitors and settings transform a poor quality amp in a good one. So to come back to your question, if you have the possibility to a more "internal" view, it will help.
Christian
Hello,
I have updated the "history file". As before the content reflects my sensibility to this thread... so probably far to be objective.
Nevertheless, it should help finding information.
Any feedback is welcome.
If we could increase the number panel examples, the detail of each... it will be great.
Christian
I have updated the "history file". As before the content reflects my sensibility to this thread... so probably far to be objective.
Nevertheless, it should help finding information.
Any feedback is welcome.
If we could increase the number panel examples, the detail of each... it will be great.
Christian
Thanks Christian,
Ignore my ramblings about the plans for the whole setup and power requirements, the question is really if anyone has experience with how much choice of amp affects the sound quality when power exciters. Should I think about it differently than with a regular speaker?
I don't really want to modify boards, but get them ready assembled with high quality components. I guess that can be possible for cheap china boards like with the 3a audio TPA3255, but still not sure I would trust to have it max powered non stop for days. The ICE Power board is at least tried and tested in PA applications.
This seems like a very good deal for an NCore solution though I must say: https://www.soundimports.eu/en/hypex-fa502.html
2x500w but cheaper than buying 2x400w NCore boards, and includes PSU, DSP and cooling so both saves a lot of work and money when building a box if it should be NCore based. Might be worth it compared to the ICE Power even if I might not be able to hear the difference on exciters. Not only for the convenience and peace of mind know that it should be great sounding amp. But it is 93% efficient compared to 83% for the ICE Power board, so only that could make it worth the extra cost.
Leob.
You seem very fixated on how many watts a set of exciters can handle ?
Surely you should be concentrating on how many db a panel can output without problems.
If one of two exciters(depending on panel materials) drives the panel into saturation point how is adding more exciters going to help ?
A panel that is used too low in frequency is going to have problems quicker than one that XOs higher up.
The panel material is very important, and hard to predict.
You need to find a material that meets your needs, then figure out what it needs to drive it to the db levels it can safely stand.
That's how I see it anyway ?
Steve.
You seem very fixated on how many watts a set of exciters can handle ?
Surely you should be concentrating on how many db a panel can output without problems.
If one of two exciters(depending on panel materials) drives the panel into saturation point how is adding more exciters going to help ?
A panel that is used too low in frequency is going to have problems quicker than one that XOs higher up.
The panel material is very important, and hard to predict.
You need to find a material that meets your needs, then figure out what it needs to drive it to the db levels it can safely stand.
That's how I see it anyway ?
Steve.
Agreed, and that is what I'm doing. It is not so easy at high SPL though...will have to do testing outdoors eventually. They are far to loud to listen to and test with a mic indoors. Definately way above 120dB...I would guess closer to 130dB, and if I listen from another room sound seems clear also at extreme levels. But my impression is that 4xDAEX25FHE-4 will not saturate my EPS panels., and they can handle even more.
Actually I have not seen anything about saturation points for panels. Tectonic have 400w on one panel and hear they mention running 1000w on them without problem. And that might not be the limit of the panel, but other factors probably decided how much power that they run per plate.
EPS panels does seem to meet my needs apart from the response above 10k, which is not that important IMO, but will play around with composites and see if I can get a better FR.
Again, I understand if power handling is not something most users experiement with or have any interest in, but very curious about you experiences with amplifier performance when it comes to DML, also at normal power levels.
Hello Leob,
I see..
Even if it is in my do list, I haven't taken the time to measure the impedance of a panel. It is not at the top of my list for 2 reasons :
- I am not sure there are information to extract from it that will help in the design. For standard loudspeaker this is not true if you want to design passive crossover. The impedance might inform about some resonance frequencies
- I don't worry about the load of a panel to an amp... this should answer to your question. Basically the impedance of a panel is the DC resistance of the exciter, in HF it is dominated by the inductance and at some resonance frequencies, the impedance increases on a small band around those frequencies. So in my opinion it is a quite easy impedance to drive, much easier than some badly design passive crossover. As the main change in the impedance are increase above the DC resistance, it is possible to add parallel elements, resistance, zobel network to limit the effects. The drawback then in your case is the wasted power and so the heating in those elements.
About the amp, I am in the opinion that a DSP is a good thing with DML. It is in the steps ahead for me. I am convinced that if we can avoid the problem that occurs with a delay (like a very badly damped mode), all the remaining humps may be corrected. I am not experienced on such EQ but I would prefer to have the possibility of a FIR. There are other architecture to do that than in the amp... with a PC for instance.
If you have ideas about it or link to amp with lower power, I will appreciate to discuss that.
Christian
Thanks, I have understood that impedance is less of an issue, but really don't have sufficient experience to know how much that actually matters with class-D amplifiers. I have understood that for example Hypex is touted as having a design that gives an even response also with an uneven load, but how much do I care about that with DML? Perhaps older designs used on cheaper boards actually will sound just as good?
I will definitely add DSP anyway...it makes sense when you striving to push every possible dB out of a plate.
For building it in to an amplifier box I would have to use a separate DSP board like a ADAU1701 if not using an all-in-one solution like the Hypex Fusion amp plate. MiniDSP HD has FIR filters and can be bought as a board or standalone unit or is in some amp plates as well.
Not sure the cost of the MiniDSP is worth it though in my case since I would need one per stack. If I was running passive or for home use it would be interesting since I would need only one.
I currently have a special DSP card on my computer that is for music production that I also do room eq with, but on my laptop I use Equalizer APO which comes with some filters, but you can load any VST plugin, so plenty of different FIR filters available.
Both MiniDSP and Equalizer APO support loading in filters calculated by REW as starting point for room/system equalization.
Leob.
I have used various amplifiers with my exciters with no problems.
Class A AB B and digital plus valve amps.
The only other exciter I have used was the hdn8 which was supposedly a 50watt unit.
I was hoping to be able to drive my ply panels to louder volume.
I think that most of that 50watts was used up trying to drive the very rigid coil suspension, which was a hard plastic , and moulded in with the outer case, not my favourite exciter.
Although it did quite well screwed into my stud wall beam.
Why aren't you looking to use pro audio amplifiers for your PA ?, some have dsp and supposedly sound pretty good ?
I would always recommend listening to an amplifier before buying , if you like the sound what's the problem.
I have heard supposedly good sounding amplifiers that I thought sounded bad, and supposedly bad sounding amplifiers that I thought sounded good(digital).
My only worry about using EPS outside is that they are quite delicate compared to other materials.
Stronger winds can damage them as can dropping them on the ground ,the exciters will just dip off and leaves a big hole
The coils must not be aloud to get too hot , as this will melt EPS.
These are problems I have had 😕
Steve.
I have used various amplifiers with my exciters with no problems.
Class A AB B and digital plus valve amps.
The only other exciter I have used was the hdn8 which was supposedly a 50watt unit.
I was hoping to be able to drive my ply panels to louder volume.
I think that most of that 50watts was used up trying to drive the very rigid coil suspension, which was a hard plastic , and moulded in with the outer case, not my favourite exciter.
Although it did quite well screwed into my stud wall beam.
Why aren't you looking to use pro audio amplifiers for your PA ?, some have dsp and supposedly sound pretty good ?
I would always recommend listening to an amplifier before buying , if you like the sound what's the problem.
I have heard supposedly good sounding amplifiers that I thought sounded bad, and supposedly bad sounding amplifiers that I thought sounded good(digital).
My only worry about using EPS outside is that they are quite delicate compared to other materials.
Stronger winds can damage them as can dropping them on the ground ,the exciters will just dip off and leaves a big hole
The coils must not be aloud to get too hot , as this will melt EPS.
These are problems I have had 😕
Steve.
When you say you heard good and bad sounding amps, is that for your DML panels? Do you have examples of amps that sounded bad?
Using regular rack mounted amplifiers would give me a lot more choice...however it would be much more of a mess in practice since I will be using active subwoofers. It would make things a lot easier to work with if I can just run line level signals everywhere.
If I go with chip amplifiers it is not as easy to try them out and returning them if unhappy, so would have to take the risk with a sample, but I'm very much leaning towards the NCore fusion plate amplifiers now for convenience anyway. It is hard to find a cheaper rack mounted quality amp with DSP than the NCore fusion plates it seems. It is even cheaper than getting just the boards, without PSU and DSP, which is a bit strange.
I'm not so worried about the 25mm plates breaking in the wind, and if I use a LPF the DAEX25FHE-4 hardly seems to get over ambient temp no matter how hard I push it. But would have to be very careful when transporting them and I agree EPS is far from ideal for anything that is not a fixed installation. Plan is to get started with vacuum bagging and make nomex/carbon plates. Cost for materials per 575x400mm plate will be around €70, so not that much in the grand scheme of things.
Peob.
did you look at the soundright site ?
http://www.soundright.org.uk
They show some simple plots and some advice.
The amplifiers I didn't like are out of production now as well as the ones I did like.
I now mainly use a 10watt max digital amp to drive my 10watt exciters.
This drives the panels far louder than my TLS could ever go with an 80watt amp ,above 100hz that is.
my 20watt class A valve amp is just sitting there now days, I'm always turning my amp on and off changing things around, not good for a valve amp.
I can't fault the digital amp sound though , so I am happy with the situation at the moment, I just miss that glow😁
Steve.
did you look at the soundright site ?
http://www.soundright.org.uk
They show some simple plots and some advice.
The amplifiers I didn't like are out of production now as well as the ones I did like.
I now mainly use a 10watt max digital amp to drive my 10watt exciters.
This drives the panels far louder than my TLS could ever go with an 80watt amp ,above 100hz that is.
my 20watt class A valve amp is just sitting there now days, I'm always turning my amp on and off changing things around, not good for a valve amp.
I can't fault the digital amp sound though , so I am happy with the situation at the moment, I just miss that glow😁
Steve.
Taming/Damping/Selectiveive Stiffening of Lightweight 3 ply Plywood Panels (esp Poplar)
Thought I'd share some experience I've gained through using 3mm 3ply poplar
Poplar is very light, and in the 3 ply 3mm version, is quite bendy, esp in the transverse direction. If not appropriately stiffened/supported , it can get out of control under excitation.
I normally aim to match the XY stiffness ratios with the XY panel dimensional ratios (as best as I can). This requires treatment of the ply.
Originally, I adopted the use of spray-on acrylic vehicle lacquer which does a good job but is quite expensive. One or two light coats of lacquer will produce a significant increase in transverse stiffness without adding much damping/extra mass. It also of course increases the longitudinal (with grain) stiffness but the effect is more marked in the transverse direction.
Lately, I've switched to shellac which is trickier to apply, but does an equal if not superior job in stiffening. I find it also to be surprisingly tough - it's harder to sand than acrylic lacquer. It can be applied in very thin layers by reducing the amount of lac in the spirit solvent. It dries super fast so only short delays in overcoating.
Another technique I'd like to share is one I adopt for the panel rear surface. Having been associated with design offices for all of my working life, I have a handy supply of various materials, and one in particular has proven to be a winner for applying to panels.
It's the ubiquitous YellowTrace. This for those who are unfamiliar, is a very popular tracing paper still used in design offices throughout the world. It typically weighs only 27gsm, and has a very high wet strength.
To use this, I cut the paper to cover the area (whole or partial) I want to stiffen - chamfer corners if desired - If to an edge, keep about 10-15mm away. I then thoroughly wet BOTH sides of the paper and hang in a safe place until the surface is ready.
Clamp the edges of the panel down to a flat surface.
I brush an approx 50/50 dilute solution of water/latex (Lancolatex (an SBR additive - but dilute PVA would suffice ) to the surface then carefully lay the paper on the wet surface - You can lightly fold (without creasing) the paper in half and lay it in the correct position - If not quite positioned correctly it can be lifted and re-laid quite easily. Using the brush you applied the latex/PVA with, work the paper to remove bubbles and wrinkles - if an edge or corner needs adjustment, use the brush bristles to gently lift and tease as required. Continue to work the paper down - it can take reasonable force but don't overdo it. It smooths out quite well, but if you get a few tight wrinkles it won't be readily visible. It is important to to remove all air-bubbles.
Allow to dry thoroughly.
When wetted, the paper expands, and as it dries, it shrinks and tensions the panel in the same manner as doped silk surfaces used to do in early aircraft. If needed/desired, another layer may be added but I haven't found it necessary. I usually either leave it as is or overcoat the dried paper with a light coat of shellac.
Applying this technique to the rear surface creates some asymmetry in the fore/aft stiffnesses - adding extra damping to the anti-nodes - I can't measure the effect of this but intuitively it must be of assistance in lessening nulls.
Obviously, all treatments will affect the tone/balance of a panel, so just be aware, tap test as you go and don't overdo it. I have found that a slightly deeper tap tone will result, but one that doesn't appear to affect the HF. It provides a much more controllable panel and in the case of poplar, the result is still light weight compared to stiffer base materials.
Cheers
Eucy
Thought I'd share some experience I've gained through using 3mm 3ply poplar
Poplar is very light, and in the 3 ply 3mm version, is quite bendy, esp in the transverse direction. If not appropriately stiffened/supported , it can get out of control under excitation.
I normally aim to match the XY stiffness ratios with the XY panel dimensional ratios (as best as I can). This requires treatment of the ply.
Originally, I adopted the use of spray-on acrylic vehicle lacquer which does a good job but is quite expensive. One or two light coats of lacquer will produce a significant increase in transverse stiffness without adding much damping/extra mass. It also of course increases the longitudinal (with grain) stiffness but the effect is more marked in the transverse direction.
Lately, I've switched to shellac which is trickier to apply, but does an equal if not superior job in stiffening. I find it also to be surprisingly tough - it's harder to sand than acrylic lacquer. It can be applied in very thin layers by reducing the amount of lac in the spirit solvent. It dries super fast so only short delays in overcoating.
Another technique I'd like to share is one I adopt for the panel rear surface. Having been associated with design offices for all of my working life, I have a handy supply of various materials, and one in particular has proven to be a winner for applying to panels.
It's the ubiquitous YellowTrace. This for those who are unfamiliar, is a very popular tracing paper still used in design offices throughout the world. It typically weighs only 27gsm, and has a very high wet strength.
To use this, I cut the paper to cover the area (whole or partial) I want to stiffen - chamfer corners if desired - If to an edge, keep about 10-15mm away. I then thoroughly wet BOTH sides of the paper and hang in a safe place until the surface is ready.
Clamp the edges of the panel down to a flat surface.
I brush an approx 50/50 dilute solution of water/latex (Lancolatex (an SBR additive - but dilute PVA would suffice ) to the surface then carefully lay the paper on the wet surface - You can lightly fold (without creasing) the paper in half and lay it in the correct position - If not quite positioned correctly it can be lifted and re-laid quite easily. Using the brush you applied the latex/PVA with, work the paper to remove bubbles and wrinkles - if an edge or corner needs adjustment, use the brush bristles to gently lift and tease as required. Continue to work the paper down - it can take reasonable force but don't overdo it. It smooths out quite well, but if you get a few tight wrinkles it won't be readily visible. It is important to to remove all air-bubbles.
Allow to dry thoroughly.
When wetted, the paper expands, and as it dries, it shrinks and tensions the panel in the same manner as doped silk surfaces used to do in early aircraft. If needed/desired, another layer may be added but I haven't found it necessary. I usually either leave it as is or overcoat the dried paper with a light coat of shellac.
Applying this technique to the rear surface creates some asymmetry in the fore/aft stiffnesses - adding extra damping to the anti-nodes - I can't measure the effect of this but intuitively it must be of assistance in lessening nulls.
Obviously, all treatments will affect the tone/balance of a panel, so just be aware, tap test as you go and don't overdo it. I have found that a slightly deeper tap tone will result, but one that doesn't appear to affect the HF. It provides a much more controllable panel and in the case of poplar, the result is still light weight compared to stiffer base materials.
Cheers
Eucy
Hello eucy
Very interesting. We can see behind this post some long experiences. Let me cut it in chunks.
In other words, does that mean if the panel as a 3/2 ratio (in dimension, pure example), your design rule you try to adjust the stiffness ratio the same? Is then the highest stiffness in the long side direction?
what is the origin of this design rule?
You are here giving inputs to an area of the panel construction which has not as far as remember deeply discussed. The use of PVA was very often mentioned but not solutions for a nice finishing. I think there is a recent post from Leob also about Shellac.
Do you have a way to qualify the stiffness increase? I have in my do list to propose something but it is still an intention...
You send me far back in time... one important part of my education was industrial drawing (drawing mechanical parts on a drawing board) at a time of paper, ink, ruler, eraser, much before 3D CAD. So as I didn't know what it is, I had a look to "YellowTrace". We had "tracing paper" (FR: papier calque). I don't know if it is the same. Internet tells me 27gsm is 27g/m² which is very low, tracing paper is more something like 180g/m² in the uses I had. I have never drawn on a so light paper. Is there a link with what is also called "Japanese paper" or if I remember when I was interested by air-model "silk paper" (pure translation from the french words)?
Some additional questions...
- which kind of suspension (if any) do you use?
- when applying the acrylic, the shellac or even more the paper, the plywood is subjected of warping. What is your technique to avoid that? Here my solution was in the suspension all around the panel. It was able to counteract to the warping created by the varnish... but I am wondering if a full perimeter suspension is the best solution (I know some are in the opinion that the best suspension is no suspension... not my case for now, the suspension has several function: mode damping, keep the panel flat, link to the frame as panel protector and probably reduce the acoustic short circuit at thee dges)
Eucy, if you want to share some pictures, they will be welcome of course.
Christian
Thank you NaRenaud
There are new things for me :
- the glue type : Titebond liquid polyurethane glue (information here, example of source)
- the whole process with a 5g adding mass by the glue and the Nomex
- the idea to sand the Nomex at the edges. I hope the Nomex dust is not a problem...
- how the vacuum bag is powerful to keep the parts together and even to press non parallel sheets
- no bracing on this guitar model, meaning that the bracing is not mandatory for the guitar sound. I have read that in other places when speaking about the doubletop; The main function of the bracing is to resist to the string strength. For pure wood soundboard like in the guitar or the piano i think it gives also more close stiffness in both direction.
Does somebody can tell more about this glue? Advantage or drawback? The Titebond website said it is an alternative to the epoxy bi-composant... for a newbie like me using almost PVA, neopren or superglue it sounds like something easier to use than epoxy.
The process with the plastic sheet is also interesting by its simplicity. Any advice on the choice of this plastic?
One idea somewhere in my head is to play with some thickness change over the panel area to create a dissymetry to avoid the pure cancellation of the non (odd,odd) modes.
Where I have doubt seeing this video is the resulting stiffness which in my opinion seems a bit high for our application in DML.
I found this paper MEASURING A FREQUENCY RESPONSE OF A GUITAR showing a first guitar mode around 100Hz where a DML target is 70/80Hz or even 50Hz (rule of thumb of a 2.5 ratio between the first mode and the usable frequency)
There is also this video from the same author showing in the first minutes (just watched that for now) the full process starting at wood preparation (somebody asked some posts ago how to glue edge to edge the wood)
Hi Christian
Polyurethane glue is very good in its intended purpose (boatbuilding and the like) BUT - it's a moisture curing glue and it MUST be adequately (heavily) clamped because it expands and bubbles when curing - If not clamped, it pieces will push apart and you'll be left with a bubbly mess. The Canna guy mentions this I think in one of the comments to the video.
Not for us I think
Cheers
Eucy
Imagine a node point on a rectangular panel - now break the panel down into two equal width strips, one longitudinal, one transverse passing through the node (forming a cross). Now displace the node. If the stiffness of each strip is the same, the bending stress on the short transverse strip is larger than on the longitudinal strip by a cubic factor. The short span resistance to the node movement will therefore dominate the panel - this will also transfer load to the side supports - or if not supported, will cause 'rippling' damping along the edges which will in turn cause the node movement will be somewhat 'washed' out.
Hence (I believe) it is important for the stiffnesses to be in as closely in balance with the proportions of the panel as possible. It could be tested by cutting 2 strips (one longitudinal and one transverse (cross grain) ) and checking static deflections for each under the same load. However, I just do it by feel.
For example, a panel of 3units long by 1 unit wide would need the transverse stiffness to be 1/27th of the longitudinal stiffness to be in balance. As the thickness is the same in both directions, E lat would need to be 1/27th of E long. Yellow poplar has an Et/El ratio of .043 - so Et is 1/23rd of El.
WR Cedar is .055, or 1/18th, Spruce is similar to poplar etc etc.
Plywood makes it tougher to get it close to the ideal but it's why I favour either solid timber or 3 ply for rectangular panels, as 5 ply has a much more even spread of stiffnesses - So 5 ply is better (imo) for panels approaching square.
As above , by feel
Google "YellowTrace paper"- lots of links - You should be able to get it in France
I'm nowhere near the end of that trail yet - My first major build has minimal support - my current project is quite different and has resilient support along both long edges and lesser along top and bottom.
I also think that Spedge's canvas supported plywood panel has promise - If tensioned enough and of the right width, it may perhaps be best approach of all - Certainly worth some experimentation - unless life gets in the way
I have been able to correct warping quite well so far. Usually it occurs transversely in 3 ply and when it does, I force a gentle reverse curve in it under weights and clamps (about 2x the warp bend), give it some passes with a heat gun and leave it overnight. This works well with Shellac also as the heat tends to soften it slightly and it resets to correct the curve. In the longitudinal direction it is usually very slight and is handled by edge supports
Hope this helps
Cheers
Eucy
Last edited: