Question on cutter head driver pre-loading

[pentlandsound]

I'm in the process of designing / building a new cutter head, and have been getting some ideas from the excellent set of postings by KNOP and the 'how to' guides posted by grooveguy and others. I'm following the traditional route of two small tweeter speakers mounted at right angles to each other, the whole at 45° to the disc surface. Grooveguy mentions the use of tiny ball-and-socket joints in the pushrod assembly to allow for some deviation from true axial movement of the pushrod, but also notes that this is not entirely a Good Thing as some loading force is required to keep the ball and socket tightly coupled - effectively putting DC bias on the driver.

My understanding - and please correct me if I've got this wrong, for my understanding is still rather shaky - is that the driver/pushrod/torque tube assembly must be put together and fixed in position (using glue or setscrews or whatever) so that the driver maintains its neutral, 'rest' position. That is, the driver coil must not end up pushed inwards while fixing, as would happen, say, if the pushrod were a little too long to fit.

All well and good so far. My question is this: once the assembled head is mounted in position, and any weight is put on the disc to be cut (even before any signal is sent to the head), would this not knock the drivers out of their neutral position straight away? Should the head be assembled so that the drivers are slightly 'pulled out' (not sure how this might be accomplished accurately), so that, on applying cutting/embossing weight, they return to the neutral position?

Or does this even matter?

Thanks!

[dmills]

Technically yes, but speakers have a XMax value, and as long as the displacement is well within that it should not be a big issue. Generally you will add some even harmonic distortion due the the non linearity of BL(x) in real drivers, and possibly the non linearity of the suspension.

If you are running a DC current based coil temperature sensor you could probably set the polarity on that to oppose the displacement due to the cutting weight, but the current is generally small so the force to be had will probably be insufficient.

[pentlandsound]

Thanks for quick reply dmills. I'm trying to reduce the moving mass as far as possible by using very small speakers such as the Visaton BF 32. However, with these only a very slight physical force is needed to move the cone and coil inward. I can't find a figure for Xmax for this unit but I suspect that, although other posters have reported good results from using it for cutting at say 10 - 15 grams' force, the extra weight required for embossing would cause it to be displaced beyond Xmax.

I have in the past had fairly good results embossing with the somewhat larger CUI/Digikey drivers CMS0401KL-3X but have not attempted cutting with them yet.

[markrob]

Hi,

If you have your moving mass way down, then you might consider adding external stiffness to your system to move the system resonance higher and keep the drivers from deflecting so much under static load. If you are planning to run the head open loop, that will make the head less affected by the cutting load at the expense of sensitivity at low frequencies. If you know the driver BL, and max power capability, you can see how much stiffness you can add to allow the head to still defect maybe +/-100 to 150 um at full peak drive current. +/-150um is the Neumann limit, IIRC, to give you an idea on the peaks you might need to attain. That will determine if you can fully modulate the groove at low frequencies. High frequency sensitivity will remain the same beyond resonance since you will be mass controlled in that region and that should still just about the same. Hope that makes sense.

Mark

[dmills]

Note that peak power capability is way up on the drivers RMS rating, how far above depends on how much you dislike the smell of melted voice coil!

The fact that the target movement is only +-150um suggests that there may be something to be said for mounting the stylus into the torque tube behind the driver connection point?
Never seen any design do that, but given you are working with a fixed Xmax it is a thought.

[pentlandsound]

Thanks for your advice Mark. Total moving mass is about 3.9g - this includes torque tube, setscrew, V-spring, pushrods (in reality cut-down perfume funnels of aluminium) and in the absence of figures from the datasheet I've included 0.5g per driver in the total. 3.9g is a little more massive than I'd aimed for - the funnels contribute almost a gram each; I might in the end substitute carbon fibre rods which have a much lower mass of 0.09g. The torque tube is 3D-printed in aluminium, and anchored to the rear plate of the head by 1mm steel wire; to increase stiffness I was pondering adding extra springs or guides to the pushrods, but this would also increase mass and the trade-off might not show any net benefit.

There are no figures on the Visaton datasheet for BL, Cms/Kms, Mms or Xmax. I suspect that the only practicable way to measure the excursion at full power may be with a dial indicator on the completed head.

Dmills - yes I learned fairly early on in this game that the fuse is indeed the recordist's best friend!

[markrob]

Hi,

Note that the total torque tube mass is not seen by the drivers. Since it pivots about the leaf spring/drag rod its load more about is mass moment of inertia rather than is total mass. Still 3.9g might be on the high side. It really depends on the driver BL and power handling. Have you measured the system resonance? If you obtain that along with the stiffness, you can estimate the total moving mass. If you have a dial test indicator, you can measure the max amount of deflection by applying DC current to the drivers to cause either vertical or lateral motion (drivers in phase or anti-phase). That number and the driver BL will allow you to calculate the stiffness.

Mark

[2bitcomputer]

Does anyone have an idea or good guess as to what a safe and proper Xmax might be for a DIY stereo head cutting at 12 grams?
If I am understanding what was said so far, the 150um limit is not about the maximum excursion of the cutter but the inherent stiffness of the heads build?
So what is the maximum excursion distance that needs to be allowed for, and how would the Xmax affect that?

Thanks...

[markrob]

Hi,

The 150um figure is the max mechanical excursion limits for the Neumann head. My point was that if you can get within reason to that, you should be in good shape. You might be able to have limits quite a bit less on a non-pro DIY head and do quite well.

150um is a very small distance as compared to most any off the shelf small driver Xmax spec (if its even provided). What dmills was pointing out is that if you deflect the driver from its natural centered position due to a static mechanical load, it could result in increased distortion. You want to stay well away from the driver Xmax to keep the distortion in a reasonable range. The only way to do that is to add mechanical stiffness to the system to counteract the deflection. I suggested that adding stiffness to combat that would affect the required drive sensitivity, but might not be a bad thing as long as the drivers can push against that stiffness (and the cutting load) to fully modulate the groove. Furthermore, stiffening the head, makes the it less affected by the cutting load. On a feedback head, the stiffness is increased when the loop is closed. Open loop heads don't get that benefit.

From my experience, you would need more 12 grams of down force (IIRC, more in the 20-30 gram range) to cut a lacquer let alone emboss. Is the figure you are quoting the down force of the head or the moving mass?

Mark