Project "Groove Scribe"- A DYI stereo 45/45 head.

[EpicenterBryan]

After this check, I did another test to measure the phase shift. I used a great program called Smart 7. The result can be seen in the image.

IMG_3955.JPG

Ciuens, I'm looking at this phase plot and I'm a bit puzzled.
The phase plot seams to be inverted. But it also seams to be offset starting in the low frequencies by -90 degrees. Ideally, the phase should start at +90 degrees at low frequencies.

First, look into the polarity of the signal path from the feedback coil to preamp to your PC. As the low frequency response curve rises, the phase plot should go down. At the first resonant frequency the phase should be zero and start to go negative from there.

Next, reading some info on your program (Rational Acoustics Smaart 7), which is crazy expensive ... there is a phase function that lets you move the 0 reference where you want. I'm not saying it is correct to use it, but if the phase polarity turns out to be inverted, and after you fix it and it still starts at zero try moving 0 up to +90 and see if the chart starts to look correct. There is also a "Coherence" function that you could turn on to see if the measurement data is stable and if it can be trusted. It's a good thing to do until you get a good feel for what you are looking at. After that, turn it off. A high reading (near 100%) says the readings can be trusted, low % readings may indicate a problem like not applying the system delay measurements correctly.

Oh, that brings up another thing - be sure to read up on how to do a system delay measurement and apply it. In your manual - look for "Delay Locator / Delay Tracking". It could be run on the output to input on the PC to see what delays are being generated there and it could be applied to compensate. It could also go through the whole signal chain.

Some things to think about.

If you have time to put a signal flow chart together that would help us as well.

Bryan

[EpicenterBryan]

That said, there are several freeware or shareware "VST" EQ plugins that allow one to choose between linear-phase and quasi-analog (phase-lagging) filtering. Or, if you still have a 32-bit Windows machine going, try to turn up a copy of good old "Cool Edit Pro" before it became "Adobe Audition" - its "FFT Filter" function that allows you to freely draw your curve has a very "analog-ish" phase response, and does a rather good iRIAA on A/D transfers made through an analog RIAA preamp.

Chris, I'm thrilled you found us here on LatheTrolls as a new(old)bee. I say that jokingly as I have seen your other posts. If someone else hasn't said it... Welcome!

Thanks so much for posting about the free VST plug-ins that can do this. I'm sure many people will be interested in knowing that. Also, many of us actually DO have old versions of Cool Edit Pro (like me). That's really good to know. For me though, a full analog path can be done so I'll go that way so I don't have to think too much. Only so much processing power left you know.

Bryan

[EpicenterBryan]

Wow.
We are on to it.
In this test, I used the "minimalist" Spring / Diagram with the carbon fiber rod. I shortened the rod to 32.5mm and put a notch at about 27mm from the driver coil for the Phono Cartridge stylus. Notice I still have the "Ciuens hard drive Mu-metal shield" in place and a spacer. Also notice no support for the end of the rod.

IMG_3993.JPG

Here you see the notch cut in the shaft for the phono cartridge stylus.

IMG_3994.JPG

And now... for the data that will confuse, amaze, excite and make you question reality (and my measurements). Behold:

IMG_0080.JPG

Now, ignore the big phase change at the end. I still have the RC network in there that hits 30 degrees of positive phase shift at 10khz through 20khz. I'll take that out on another test.

The big thing is why does this phase plot start at -90 and not +90? I scratched my head, I checked my test setup and could not explain it ...until now.

There is no longer a spring in a spring mass equation that normally defines how this driver should work. There is only a dampener, which is kind of the opposite of a spring. Imagine pushing on a spring - When get to your maximum push, the spring has stored the energy and as you start pushing less, the spring now starts pushing on you - kind of like having a tail wind when you are flying from here to there.

Now think about pushing on a damper - it resists your movement, and once you peak your movement and try to go the other direction - it doesn't help you at all. In fact it's like having a head wind you have to fight against.

So, I do think the fact that the phase starts in the wrong direction is actually correct. But it gets even better: Notice the phase changes stay within 145 degrees to about where my extra +30 degree shift from the RC comes into play. The way I'm seeing this is that feedback could be applied all the way out to 10K and be stable. Perhaps further if I pull the C out of the RC boost in this test.

Now the fact that this minimalist spring has an even higher resonance than the other one (which I need to test again after shortening the rod like on this one)... probably has more to do with 3D printing inconsistencies than anything. I have mentioned before this Nylon material is very tricky to print, and tends to clog up a nozzle, spit steam if not dry enough and all that. So it's possible the full diaphragm version may have had some void or layer bonding issues internal. From the outside the print looked good - but I had 4 failures for one good looking print.

I think this phase data is for real. I'm going to shorten the rod on the other (full diaphragm) version, cut the stylus slot at about the same spot and see if we get a similar result.

The other take away from this is that if this phase response is for real for the modified driver, then it looks like introducing some springs here and there - like a V spring attached to each drive link to keep the driver from going out of ideal angle (causing voice coil rub) then we may have phase margin to play with when everything is connected into a final mechanical system.

I'm not thrilled to see the second resonance, but I'm thrilled to see a surprisingly minimal phase change - well within the ability of a feedback loop to fix. Again, only if this data is correct. I think it is.

Hey Mark, what do you think about all this?

Bryan

[markrob]

Hi,

Still looks to me like you have a polarity inversion in your signal path. If you reverse the wires going to the head drive coil, the phase will flip 180 degrees and the phase shift will become +90 at low frequencies. I can see any way a +6db/oct rising response would yield a negative phase shift if you have a minimum phase system.

At the end of the day, you still have a spring mass system here. Extreme overdamping still results in second order system, the poles just move from the complex plane to the real axis.

Mark

[EpicenterBryan]

If you reverse the wires going to the head drive coil, the phase will flip 180 degrees and the phase shift will become +90 at low frequencies.

It does. Below is a shot with the driver reversed. Nothing else changed.
And notice the phase is going positive heading toward resonance like the other plot. If it was a phase reversal problem, wouldn't the phase go down as we approach resonance like it did when a pure spring was in there?

IMG_0081.JPG

Bryan

[markrob]

Hi,

That looks a bit better. I'm still not sure that I'm seeing and damping effect. The Q of the primary and secondary resonances still seem to be very much present. I would expect them to be flattened quite a bit if the damping was working. The fact that your primary resonance is so high indicates that you have a very stiff system given the mass you have to sling around. Maybe the damping is really just adding stiffness.

There is quite a bit of crazy phase stuff going on at the upper frequencies. I'm not sure this is better than your earlier plot where you were able to close the loop. If you remove the RC network you added in front of the driver, does the rising phase go away? That would make sense as that network is a lead-lag.

Mark

[EpicenterBryan]

If you remove the RC network you added in front of the driver, does the rising phase go away?

So here is an update.
RC network removed. Gone back to the full diaphragm, carbon fiber rod shortened and stylus slot about the same as last plot but on this diaphragm and coil.

It's interesting to note that the response has changed since first testing of this configuration. I think it may be that the epoxy has fully hardened. The second plot is made with me just slightly touching the carbon fiber. Same result if I touch the end or the side anywhere on the rod. It's much more stable. I can also offer that I noticed the voice coil on this one is not perfectly parallel to the surface of the diaphragm and carbon fiber disc. It looks like I didn't get all the glue off from a previous mounting...

IMG_0086.JPG

IMG_0087.JPG

There are a few things to look at and experiment with. First, the outer area of my diaphragm design is 1mm thick but is printed with 100% in-fill. My layers are set at 0.25mm in height (4 layers for 1mm). Since it looks like we may be quite a bit too stiff, perhaps one or two inner layers can be less dense and still bond well enough. Also, we have discussed the speed of sound through materials and some experimentation with another light weight material but with faster sound propagation is in order.

We have a brave Canadian brother who has stepped up to the plate and will do some experimentation with some items I'm sending him. Yes, it's our very own Bennett!

Bahndahn will play with printer setting and other tweaks to try to get this Nylon material to print consistently. And hopefully come up with something that works and is easier for everyone to duplicate for their own project. He will also mess around with some magnesium material I'm sending which might be useful for rod segments or other components in this or other head projects. There may be special material processing lessons to be learned, or other insights to be shared - so I'm really excited about this!

One last set of inputs:
I have a set of Wayne's boards on the way, and hope to stuff them this weekend. If I have time, I want to use his preamp board to double check the HF response on this Groove Scribe driver. If it works out, watch for an update. Otherwise I may go silent for a bit on this subject but I'm sure Bennett will have some inputs when his care package arrives with hard to find goodies from America!

Bryan

[Snug Music]

Hi Bryan, I have my driver indeed changed (instead of the metal diaphragm I now have a PVC membrane), it may be that have thus suffered the highs? Sounds more middle with distortion!
Have made a pic from cut under microscope...
Can anyone help me?
Thx...Scotty (newBee)

[Ciuens]

Hello Scotty, some things you have to change to be able to increase the high frequencies.
1- decrease the size of the voice coil spring.
2 - replacing the rod with one that is rigid.

Looking at the picture of the microscope .... is low depth in the groove.
Another tip to get boost high frequencies is to use extreme equalization.

Ciuens

[Snug Music]

Ok, thank you .. ciuens, .. I'll take the next week to make the changes! Wishes nice and relaxing weekend. ..

Thx greetings...
Scotty

[opcode66]

However, no amount of EQ'ing or Feedback will adjust for a total lack of response inherent in the actual transducer design.

There are at least two other more critical things that must be done to increase high frequency response that have not been cited thus far.

I see no center line in the grooves. Were they cut with a chipped stylus? Grooves should appear to be three lines: two for the walls and one for the center cut by the very tip of the stylus. Center lines are not apparent at all. Also, the groove are being cut way too shallow. Finally, I may as well be looking at the disc with my naked eye. That level of magnification is insufficient to really be able to see the grooves and read them. You wouldn't be able to see high frequency data at this magnification even if you had it.

[Snug Music]

Ok, ..'m going to watch my Stylus under the microscope. And'm going to make a better image from the cut. Thank opcode ... you also nice weekend. Best regards. Scotty

[Ciuens]

There are at least two other more critical things that must be done to increase high frequency response that have not been cited thus far.
.

And what are those things that were not mentioned?

[Ciuens]

Ciuens hey , sorry to ask ... but the magnets feedback without isolation from each other could not cause interference or read errors on feedback ?

The insulation is being done with distance between the magnets and the copper plates.
To ensure isolation, I am now again using MU metal (Ciuens Shield, hahaha), but we will always have interference between the magnets and the coils, what matters is the difference between the gain of the feedback and the interference of the voice coil. I believe that a difference of at least 40db are sufficient for the feedback function properly.

Ciuens

[Ciuens]

First, I made double magnetic isolation with copper plates, as I show in pictures.

Hey Ciuens,
Copper is fine for radio frequency (RF) shielding, but for magnetic fields below about 100 kHz, you want to go back to your Mu-Metal idea. I don't think that copper is doing anything for you. I think you may be getting better isolation because your feedback coils are further from the driver and picking up less magnetic field from the driver.

But, that increased distance may be an issue now that we know about the propagation delay.

Read this section from Wikipedia:

https://en.wikipedia.org/wiki/Electromagnetic_shielding#Magnetic_shielding

Also, are you using raw pink / white noise? No riaa or iraa involved at all right?

Bryan

Hey Bryan, my friend ...
I am with you, copper does not insulate as well as I thought, but .... I'm now using in combination with MU metal, hahaha
I am using the pink noise generated by the Smaart 7 without IRIAA. I am also using the pink noise generated by a program called Soundcard Scope.
I modified the rod for non-magnetic metal (Bronze) and seems to have worked well.
My biggest problem is still the phase shift, I can not close the loop properly. I am having the same problem you have shown the video of Caruso Massimo
http://www.lathetrolls.com/viewtopic.php?f=15&t=5462&start=40#p35650

[opcode66]

Any conductive metal will absorb flux from an alternating EM field. A copper strip fashioned into a large sort of coil on a round pcb is what shields the electronics of a technics 1200 from em emitting from the coils of the bldc motor, FYI. Also, copper is the very stiff that is prducing the EM field to begin with. The same thing woks on reverse just fine. What is in all electroc generators to turn flipping magnetic field into current, copper.

I'll let you know of any of you identify the other two requirements correctly.

[Ciuens]

After this check, I did another test to measure the phase shift. I used a great program called Smart 7. The result can be seen in the image.

IMG_3955.JPG

Ciuens, I'm looking at this phase plot and I'm a bit puzzled.
The phase plot seams to be inverted. But it also seams to be offset starting in the low frequencies by -90 degrees. Ideally, the phase should start at +90 degrees at low frequencies.

First, look into the polarity of the signal path from the feedback coil to preamp to your PC. As the low frequency response curve rises, the phase plot should go down. At the first resonant frequency the phase should be zero and start to go negative from there.

Next, reading some info on your program (Rational Acoustics Smaart 7), which is crazy expensive ... there is a phase function that lets you move the 0 reference where you want. I'm not saying it is correct to use it, but if the phase polarity turns out to be inverted, and after you fix it and it still starts at zero try moving 0 up to +90 and see if the chart starts to look correct. There is also a "Coherence" function that you could turn on to see if the measurement data is stable and if it can be trusted. It's a good thing to do until you get a good feel for what you are looking at. After that, turn it off. A high reading (near 100%) says the readings can be trusted, low % readings may indicate a problem like not applying the system delay measurements correctly.

Oh, that brings up another thing - be sure to read up on how to do a system delay measurement and apply it. In your manual - look for "Delay Locator / Delay Tracking". It could be run on the output to input on the PC to see what delays are being generated there and it could be applied to compensate. It could also go through the whole signal chain.

Some things to think about.

If you have time to put a signal flow chart together that would help us as well.

Bryan

Hello Again Bryan.
I am studying the software, thanks for the information.
This is the signal flow that I am currently using.

SINAL.jpg

Ciuens

[Snug Music]

Upsi, ... I could locate the fault today, the diaphragm has come off the coil! Do I have to probably take another glue ...
And the needle is blunted front!
I'll take other peace!

[Snug Music]

Hi there,.. I have seen this. There is a (Orthophone reading system) is used as a feedback system. Sitting directly in height where to meet the driver. Souri used this system.
Best greeting...
Scotty

[Ciuens]

I have tested using this configuration, but did not like. the magnetic isolation is very good, but greatly increases the secondary resonance. I returned to the configuration using the coil wound on the push rod

Ciuens


Enviado do meu iPhone usando Tapatalk