V1.0 of Bryan's 3D printed head with feedback

Anything goes! Inventors! Artists! Cutting edge solutions to old problems. But also non-commercial usage of record cutting. Cost- effective, cost-ineffective, nutso, brilliant, terribly fabulous and sometimes fabulously terrible ideas.

Moderators: piaptk, tragwag, Steve E., Aussie0zborn

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

V1.0 of Bryan's 3D printed head with feedback

Post: # 31550Unread post EpicenterBryan
Thu Sep 25, 2014 12:43 am

First off, I need to thank everyone for posting all their ideas - and for sharing their failures, triumphs and works in progress. I have seen so many post from so many people that if I start to give credit I'm sure I will miss someone...

So let me say this.... What I'm about to show you is this sneak preview of what I'm messing around with. I did not invent any of it. I just applied things in a different way from things I have seen on this forum. This is a combination of everything many people are working on in the forum, but with a little twist. I ran across a driver that is a different than what is being used in the current concepts. And because it is so different, we may not have to wind our own coils to have the ability to use feedback in a DIY design.

There are guys working on winding their own coils for their own driver that have this feedback function. I just don't have the eyesight or patience to do what they are doing... So I went with this non standard path...
IMG_2916.JPG
This driver I ran across is full range mini-speaker with a pretty flat curve from 200hz to 20,000hz. It is 5W, 20W peak. I mentioned it in another post. The unique thing about this driver is that it is hollow in the center which allowed me to run a push rod through it's core into an identical driver which could be used for feedback.

My goal was to make all the parts on my 3D printer. A few things needed a bit more interaction from the builder.
IMG_2908.JPG
Enough said for now. These photos were from my final fit checks for photographic purposes. It will be a few weeks until the final hardware arrives, and I have time to run it through the paces...

Bryan
You do not have the required permissions to view the files attached to this post.

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31551Unread post EpicenterBryan
Thu Sep 25, 2014 12:53 am

This photo should help. In the background is the front and the back of the driver with no mods. In the foreground is a push rod through the center of the driver into a second unit being used as a feedback transducer as discussed.
IMG_2920.JPG
You do not have the required permissions to view the files attached to this post.

User avatar
opcode66
Posts: 2700
Joined: Sun Nov 08, 2009 10:56 pm
Contact:

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31552Unread post opcode66
Thu Sep 25, 2014 12:56 am

Cool. I'm printing parts as we speak for mine. Nice work.
Cutting, Inventing & Innovating
Groove Graphics, VMS Halfnuts, MIDI Automation, Professional Stereo Feedback Cutterheads, and Pesto 1-D Cutterhead Clones
Cutterhead Repair: Recoiling, Cleaning, Cloning of Screws, Dampers & More
http://mantra.audio

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31554Unread post EpicenterBryan
Thu Sep 25, 2014 1:13 am

opcode66 wrote:Cool. I'm printing parts as we speak for mine. Nice work.
I LOVE your version and your approach. Man, I just can't fathom winding my own drivers and feed back coils. You are a younger man than I and I can't do "wire wrap" let alone flat wire coil winding on bobbins! Thank you so much for your all the updates on your project! I follow each update and breath you share!

Bryan

User avatar
opcode66
Posts: 2700
Joined: Sun Nov 08, 2009 10:56 pm
Contact:

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31555Unread post opcode66
Thu Sep 25, 2014 1:46 am

Watch for a new video tomorrow or Friday! Should be cutting.
Cutting, Inventing & Innovating
Groove Graphics, VMS Halfnuts, MIDI Automation, Professional Stereo Feedback Cutterheads, and Pesto 1-D Cutterhead Clones
Cutterhead Repair: Recoiling, Cleaning, Cloning of Screws, Dampers & More
http://mantra.audio

User avatar
emidisc
Posts: 220
Joined: Mon Jun 07, 2010 7:57 pm
Location: lancashire

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31559Unread post emidisc
Thu Sep 25, 2014 4:39 am

Bryan that looks great love the idea of off the shelf drivers,
Do you mind me asking for details of the drivers?
I am midway through a DIY head build but have found it difficult to locate drivers of the correct rating and size & they look just perfect .....I am not planning to include feedback at this stage
Regards
Emidisc

User avatar
opcode66
Posts: 2700
Joined: Sun Nov 08, 2009 10:56 pm
Contact:

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31560Unread post opcode66
Thu Sep 25, 2014 5:00 am

I will have feedback on mine soon. All I need to do is add another coil. The magnets I'm using are strong enough for both drive as well as to induce feedback. I already have the pins and sockets for feedback.

I went with rolling my own coils because I wanted to learn the art. And, I'm able to make extremely strong transducers this way in exactly the shape and size I want. I don't have to sacrifice anything in my design. I can also incorporate Neodymium magnets.
Cutting, Inventing & Innovating
Groove Graphics, VMS Halfnuts, MIDI Automation, Professional Stereo Feedback Cutterheads, and Pesto 1-D Cutterhead Clones
Cutterhead Repair: Recoiling, Cleaning, Cloning of Screws, Dampers & More
http://mantra.audio

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31565Unread post EpicenterBryan
Thu Sep 25, 2014 1:16 pm

emidisc wrote:Bryan that looks great love the idea of off the shelf drivers,
Do you mind me asking for details of the drivers?
The drivers I bought are these: AURA COUGAR NSW1-205-8A 1" EXTENDED RANGE SPEAKER DRIVER.
I got them from parts-express.com, but you can find them on Amazon as well. The first few cost about $15 each, but I bought some spares at around $10 on sale. They have radial neodymium magnets so the center is hollow, and normally there is a screen over the hole which is easy to remove. They say the open design helps cool the drivers via air flow. The cone is Titanium, and surprisingly tough to poke a hole in.

The full spec sheet for the driver is at this link on the Aura wedsite at:
http://www.ari-web.com/aurasound/pdf/NSW1-205-8A.pdf
61HViMgsg0L.jpg
You do not have the required permissions to view the files attached to this post.

User avatar
markrob
Posts: 1623
Joined: Wed Nov 28, 2007 1:14 am
Location: Philadelphia Area

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31566Unread post markrob
Thu Sep 25, 2014 1:40 pm

Hi,

Nice job on the head construction. You and Todd are taking thinks to the next level.

The driver looks nice for the application. Low moving mass of around .5g. Should work well. Note that the acoustical frequency response plot is meaningless as you are mechanically coupling to the stylus. You will be raising the system resonance into the 1Khz range once you couple to the stylus due to increased stiffness. Also note that adding feedback is a real challenge. The trick is to avoid crosstalk from the drive coil (especially at higher frequencies). Once you wind the feedback coil, you have created a loose coupled transformer with the primary driven by currents in the amp range and reading motional EMF in the millivolt range. Its easy to think that you are detecting motion when you are really seeing crosstalk. To test, block the driver from moving and do a frequency sweep over 20-20Khz at normal drive levels and look at the feedback output. Then unblock the driver and see if you can detect the motion above the crosstalk in the audio range.

Mark

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31567Unread post EpicenterBryan
Thu Sep 25, 2014 3:08 pm

markrob wrote: The trick is to avoid crosstalk from the drive coil (especially at higher frequencies). Once you wind the feedback coil, you have created a loose coupled transformer with the primary driven by currents in the amp range and reading motional EMF in the millivolt range.
Hi Mark.
Here is a question for you. They say these drivers are magnetically shielded. And after checking them I find a large magnetic field inside, but none on the outside so my gut says I won't see fields from the coil on the outside but I really don't know. In these photos you see I have steel rod going through the drivers. Do you think I have a better chance of minimizing the effect you mention if I use nylon push rods? My threaded nylon rod didn't arrive until today and the plan was to use it but it's a bit more pliable than I thought it would be - shorter is better with this stuff. I was also thinking about making the spacers between the drivers shorter to make it more compact (and also eliminate the excess sticking out the ends). Do you think I should keep the spacing the way it is now to minimize possible coupling since the field drops as some function related to the square of the distance? (I don't remember the details)
markrob wrote: To test, block the driver from moving and do a frequency sweep over 20-20Khz at normal drive levels and look at the feedback output. Then unblock the driver and see if you can detect the motion above the crosstalk in the audio range.
Great idea! I'll check that first. That should tell the story. Thanks for the inputs, and I'll let you know. Should have time this weekend.

User avatar
opcode66
Posts: 2700
Joined: Sun Nov 08, 2009 10:56 pm
Contact:

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31568Unread post opcode66
Thu Sep 25, 2014 3:57 pm

Why not print rods?
Cutting, Inventing & Innovating
Groove Graphics, VMS Halfnuts, MIDI Automation, Professional Stereo Feedback Cutterheads, and Pesto 1-D Cutterhead Clones
Cutterhead Repair: Recoiling, Cleaning, Cloning of Screws, Dampers & More
http://mantra.audio

User avatar
markrob
Posts: 1623
Joined: Wed Nov 28, 2007 1:14 am
Location: Philadelphia Area

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31569Unread post markrob
Thu Sep 25, 2014 4:21 pm

Hi,

I didn't catch that you had two drivers stacked. That's a neat idea. It will be interesting to see if it works. You might try flipping the feedback sense driver and see if it has better shielding in one direction or the other when you do you blocked test. You can also try adding copper or mu-metal shielding between the two drivers to shield the magnetic field. Having distance between the drivers is a good thing on the coupling end, but bad from a sensing standpoint. The spring mass of the sensor coupling rod could mechanically decouple from the main driver at higher frequencies In the ideal setup, you want to sense the motion of the stylus, so closer you get to it the better, but that invites more crosstalk. Some other downsides to this approach is that you have now doubled the moving mass given two drivers (not to mention the mass of your connecting rod). This will tend to kill the drive level by a factor of 6 db (F=MA). That means 4 times the power required for the same cut level as with one driver. Also the motor constant of the speaker as a velocity sensor is not optimum. You would like more turns of really small gauge wire to give you higher output per cm/sec of motion, but that also increases the transformer effect. Keeping moving mass low is really important. Also, it looks like you are going to attach the rod to a hole in each speaker cone with adhesive. By not connecting directly to the voice coil, you might have some issues with the speaker diagram flexing and causing loss of coupling. Everywhere there is a connection, you have a spring mass system. Once you get on the other side of resonance you loose coupling. It takes some doing to make sure these secondary resonances are out of the audio range. I would start with an open loop design first and get a feel for how the system performs before you attempt to go closed loop. Make sure your head follows the classic open loop response of a moving coil head first. That is, velocity response rising at +6db/oct below resonance and falling -6db/oct above resonance.

Mark

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31570Unread post EpicenterBryan
Thu Sep 25, 2014 6:22 pm

markrob wrote: You might try flipping the feedback sense driver and see if it has better shielding in one direction or the other when you do you blocked test.
Mark, your whole comment is full of great inputs. Checking both directions for the feedback shield is a wonderful idea too!

I see your point about attaching to the coil rather than the cone. In this version I made the threaded flat disc such that it is all the way to the edge of the surround and only contacts there. Then I gooped the cone, inserted the disc and rod through the driver, put an alignment tool in the back of the driver and put a nut on the back side to put pressure on the disc / surround and keep it centered. Then I turned the driver disc face down so the epoxy on the cone would run down and bond to the disc at the extreme edge.

After doing the checks you have mentioned, the final issue will be to put locktite on the threads, perhaps a dot of epoxy at the unions. That brings up the issue of steel or nylon for the push rod. It's one continuous rod. Had I thought about it I could have made the disc thicker so I could use 2 segments - they wouldn't both need to be the same material. But now that the discs are glued on the drivers, which way would you go? Nylon or steel (assuming no issues with magnetic coupling).

User avatar
markrob
Posts: 1623
Joined: Wed Nov 28, 2007 1:14 am
Location: Philadelphia Area

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31571Unread post markrob
Thu Sep 25, 2014 8:56 pm

Hi,

Maybe a thin walled aluminum tube filled with some damping material. You want the high stiffness, with low mass, and low Q. The nice thing about a DIY design is that you can try things out and make changes to see if you can improve performance. So, go with your gut.

You might want to consider some sort of bench test setup to see if you are in the right ballpark before you start trying to do actual cutting. I mounted my head on a base and coupled the stylus holder to a ceramic stereo pickup. I used Trueaudio's RTA and signal generator to get some spectral plots of the head response, channel separation, distortion, and cutting velocity. I ended up using a ceramic pickup for the measurements because the magnetic pickups were seeing the field from the drive coil. In this case, the response is amplitude rather than velocity based (easy to convert from one to the other). You need a high impedance buffer (1 meg or better) to connect the pickup to a sound card input.

https://www.trueaudio.com/rta_abt1.htm

You can get a good idea of how the head will perform without wasting blanks and stylii. It won't completely simulate real world cutting since the mechanical load of the blank is not included, but its still useful.

Mark

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31572Unread post EpicenterBryan
Thu Sep 25, 2014 9:05 pm

opcode66 wrote:Why not print rods?
Good question. My first few attempts were to print the stylus holder and push rods in one flat printable part. And I printed domes to glue to the cones, with a 4mm square for the push rods to glue into. When playing with them rough, I discovered I could cause the perimeter of the square push-rod to de-laminate. And after reading about your project and what Fela was up to, I decided to go this other route and add feedback.
IMG_2924.jpg
This is what I went with, with a hex hole for pressing in a threaded nylon insert.
IMG_2921.JPG
You do not have the required permissions to view the files attached to this post.

User avatar
opcode66
Posts: 2700
Joined: Sun Nov 08, 2009 10:56 pm
Contact:

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31573Unread post opcode66
Thu Sep 25, 2014 9:30 pm

Yoy want to reduce the moving mass as much as possible to push resonany frequency higher. Moving mass is anything that moves. Not just the rods. Guessing you already know that. Super glue has been holding parts together effectively for me for now.
Cutting, Inventing & Innovating
Groove Graphics, VMS Halfnuts, MIDI Automation, Professional Stereo Feedback Cutterheads, and Pesto 1-D Cutterhead Clones
Cutterhead Repair: Recoiling, Cleaning, Cloning of Screws, Dampers & More
http://mantra.audio

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31574Unread post EpicenterBryan
Thu Sep 25, 2014 10:29 pm

markrob wrote: The nice thing about a DIY design is that you can try things out and make changes to see if you can improve performance. So, go with your gut.
Another batch of great inputs Mark. I'm going to use this project as a learning experience. The great thing about having a 3D printer is it's easy to make changes and print a new part. Now that I have something assembled and in my hand I'm already hearing problems just by handling it and paying attention. Like the cool little curved spacers... I printed them mostly hollow with a honeycomb internal fill (fast to print, low plastic usage, strong in compression). But tap on it and it sings like a banshie and is also heard coming from the main plastic body. The main boy is 100% in-filled, but it also rings when taped on. If the main body resonates then the stylus is going to transmit that through the torque tube.

This test bed may only be useful for making sure all the holes are in the right place....

I did recently receive some new printable Nylon for my 3D printer that I may try printing another main body. I tried printing push rods with it and thought it was too pliable for that use. But for a body it may (or not) help damp resonances. What the heck, let's see. Some times rigid is not so good if it rings like a tuning fork, right?

Ultimately the whole thing may need to be a 4 piece aluminum assembly, screwed, tig welded or machined from a block. But if I didn't get it to this point I would never have known.

One interesting thing your post also made me think about... if the goal is to get the feedback as close to the stylus because that is what we really want is to make sure it is moving they way we want, there is nothing that says the feedback transducer can't be the one closer to the stylus holder and the driver be the one on the outside. Hummmm? I love this whole process.

Oh, and until I has something in my hand it didn't cross my mind that this might fit mt Presto 8D (if I put an extra mounting hole in it). Duh.
IMG_2918-smaller.jpg
You do not have the required permissions to view the files attached to this post.

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31575Unread post EpicenterBryan
Thu Sep 25, 2014 10:45 pm

opcode66 wrote:Yoy want to reduce the moving mass as much as possible to push resonany frequency higher. Moving mass is anything that moves. Not just the rods. Guessing you already know that. Super glue has been holding parts together effectively for me for now.
I was worried about super glue. I've used it for years to mount wide angle lens adapters on my Flip Cameras. Holds up well, but when it gets shocked from the side, the whole lens crashes to the ground. I'm sure you are right and it would work really well for this application. Fear I guess on my part from my other uses. I get stuck in ruts. The last few repairs I have used J&B weld. But it has metal fibers and will suck toward any magnetic field, so I went with standard epoxy.
IMG_2925.jpg
You do not have the required permissions to view the files attached to this post.

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31607Unread post EpicenterBryan
Tue Sep 30, 2014 12:09 am

markrob wrote: You might try flipping the feedback sense driver and see if it has better shielding in one direction or the other when you do you blocked test.
I have some interesting test results I'll tabulate and post tomorrow about this. And an interesting twist - I decided to eliminate the unknowns related to a possible "microphone" effect with the feedback sensor (since it has a diaphragm) by removing the surround and inverted cone... Decided it was best to test as much of the magnetic pickup as possible without wondering about the audio linkage.

I used a recently acquired Phase Linear 400 amp, my 40 year old SAE preamp, a Behringer DSP8024 for sine wave generation, and an old school Data Precision 4 1/2 digit true RMS meter for data collection... Fun stuff. And yes, I forgot about the massive power supply reserve on the "Flame Linear" when I turned it off and moved cables around. POOF on one of the drivers. My bad. Glad I have spares.
IMG_2939.JPG
You do not have the required permissions to view the files attached to this post.

User avatar
EpicenterBryan
Posts: 738
Joined: Sun Aug 10, 2014 9:01 pm
Location: Eugene, OR USA

Re: V1.0 of Bryan's 3D printed head with feedback

Post: # 31630Unread post EpicenterBryan
Tue Sep 30, 2014 10:56 pm

OK, I have some test data and observations to share.

A big thanks to Markrob for his inputs on what to look for. I know Mark and Todd will be interested, and I hope some of the other trolls will find this of value.

The driver in the upper right is a stock - non modified driver. The one on the left one has been modified with the back port vent screen removed, a hole punched through the cone, and a threaded disc glued in the center. This can be used as a driver or a feedback sensor. One is used as a driver in all tests mentioned below. And one is used as a feedback sensor in test #1 only. The one in the lower center is a driver I have remove the cone and surround from for testing magnet coupling. This is used as the feedback sensor in test 2-4 below.
IMG_2936.JPG
As I mentioned earlier, I decided to modify a feedback sensor for test condition 2-4 to make sure I was measuring magnetic coupling only. In those conditions the driver and feedback sensor were not physically connected, and I wanted to make sure the sensor wasn't acting like a microphone and picking up sound transmission out the back port of the driver. For condition 1) this was not an issue since the sensor was mechanically coupled to the driver and any audio pickup would be in the noise floor comparatively speaking. But in test condition 2-4 there was no physical connection between the driver and sensor so the microphone effect could be an issue.

So here we go:

Test conditions:

1KHZ Sine Wave in all tests.
2.000 VAC RMS delivered to driver (8 ohm, 1/2 watt), all tests.
Same driver in all tests, with glued disc and hole in diaphragm as discussed earlier.

Test data:
1) Baseline as pictured earlier - Driver and Feedback sensor both with glued discs and 3mm steel push rod through both making mechanical connection.
Sensor output 0.01579 VAC RMS
2) Feedback sensor cone and surround removed, 3mm steel push rod through both but no mechanical connection to sensor.
Sensor output 0.00144 VAC RMS
3) Feedback sensor cone and surround removed, nylon push-rod, no physical connection to sensor.
Sensor output 0.00061 VAC RMS
4) Feedback sensor inverted otherwise like 3) above.
Sensor output 0.00056 VAC RMS


So what does this tell us?
The push rod needs to be nylon. The steel push-rod seams to magnetically couple the driver and feedback sensor. I did notice that the end of the rod had a slight magnet field when installed through both drivers - enough to hold onto a 3mm steel nut.
There is a slight advantage to reversing the feedback sensor, but I don't think it is worth it. It does move the coil farther from the driver and slightly reduces the magnetic pickup because of distance from magnetic source but also increases the push-rod length.

If we take test 1) and subtract test 2), we should have a rough estimate of what we might see by replacing the steel rod with a nylon one. That would be around 0.01435 VAC RMS with 1/2 watt drive. So that would be about 29mv feedback per watt drive. That looks reasonable. But it does mean the pickup wires should be twisted pair and shielded, and I will also opt to make the drive wires shielded as well.

What next?
A test needs to be done identical to 1) above as a new baseline but with a nylon push rod.
A second test should be done after wrapping the driver and feedback sensor with MuMetal to try to shield magnetic pickup.
Test leads should be all be shielded.


Example for configuration 4 above. Notice no physical connection - nylon - and the feedback sensor is inverted. What would normally be the cone is pointing away from the driver.
IMG_2941.JPG

Comments?

Bryan
You do not have the required permissions to view the files attached to this post.

Post Reply