JE-1D Version 356

[dubcutter89]

approx. 800Hz turnover, right! (..was looking just at the graph without thinking...)

[jjwharris]

Given the relatively low impedance of this wind, I'm wondering if you should try filling the bobbin with a higher gauge (one or two numbers maybe) wire and see where you stand.

The current coil are wired in 0.25mm, which seems to have the best results so far. I have experimented (Pre-DATS? B.D?) with adding an extra 25 turns on top of what I usually wind, and the top end and volume did suffer. As mentioned this was before the DATs so it was all done by ear - "100 winds sounds good, how about 75? how about 125?" etc...

I've also tried filling a bobbin with .2mm, which got me to 6ohms, on the DATS but the top end and volume/excursion suffered dramatically; also while baking the .2mm set of coils with 1A of current, they spools melted.

I'll have a search for some ~0.225mm 31AWG wire and give it another go...

I've finally cracked the source of the hum; It was a combination of the DC Power supply feeding the ADAU1701 evaluation board, and the ADAU1701 board itself. I replaced the DC supply with a regulated 5VDC feed from one of the windings from the toroidal transformer, but as soon as I connected the ADAU1701 there was still an increased noise floor.

Do you have any recommendations for DSP boards Mark? I have been using a cheapo one from aliexpress;



It's been helpful in the past, but I feel like these heads are getting to the point where they match each other pretty closely, so I think I will drive the amplifiers directly from the audio interface, without the DSP and hope that the three heads do have continuity.

But for now, need to get into the backlog of records I need to cut, but I'll keep in mind the 500hz turnover when winding the new coils.

I really need to set up a test lathe so I don't have to dismantle between runs.

Impedance ratings are still a bit of a mystery to me...I've come across;

Impedance at 1khz
Nominal Impedance
DC Resistance

Should I be trusting the DATs Re reading? When these go for sale, do I just supply a DATs readout?

I'm running (cheap) TDA7293 mono blocks which say they support 4-8ohm drivers and I haven't blown any with these low impedance coils. Is it by the nature of the iRIAA curve that the point of lowest impedance/frequency is generally the point of lowest amplification?

[jjwharris]

the heads being way way way too low impedance..

Are the coils wired in series or parallel? If parallel then try to do series and you should be right at 4 Ohms DC resistance which should be a little nicer to the amp...

Thank you for the kind words!

I have tried this with .2mm coils, the results were good, but as Mark mentioned above, the worry is that the the coils will see a different current.

I'm hoping to send one of these heads to a friend in NZ for testing with embossing, but after that I would also like to send one for testing stateside. If anyone is interested.

The cases do fit, but there is slight interference with the tops of the screws from the magnet clamps to the dampener, and the tops of the screws from the dampening holder to the front of the case.

I am happy with how it looks though. I'm on the lookout for knurled M2 screws, I can find 12mm, but I really need 15-17mm

[markrob]

Hi,

The head looks super!

You could spend a great deal of time playing with the windings, but if you are getting good results with your current wind, then I would think you are ok. Its basically a 4 ohm head. Most modern amps should be able to drive that fine. As far as the 500hz vs 800 hz, its not a big deal because you will EQ as needed to make the head flat. I think the DC resistance is the least important spec. I'm surprised that you would see such a radical performance drop by adding more turns and higher gauge within reason. It would be interesting to see if the DATS plot showed any parasitic effects other than increased DCR and inductance. It would also be interesting to see if you run the test with no damping to see the resonance on the plot. Might still be too small to show up. I don't have a better source for the DSP board.

Mark

[grooveguy]

Good advice from Mark! As for the DSP board, I've used those (probably from another source, but how different can they be?) and never had a noise problem. Make sure that the gain structure through your equalizer lashup is where it belongs. That chip has lots of dynamic range; what you describe sounds like you might be working way down toward the noise floor.

[jjwharris]

Just checking in...

I've made a few adjustments to the pole pieces and armature, this seems to have improved things. I've done the final order (which is the order after the last order which I said was final) to the laser cutter.

I was having issues with the threaded dampener holder being soldered to the armature, it's held on 3/5 heads, and I think it may be a case of over-torquing the stylus screw, but since I can't seem to find any M2 Thumbscrews that are long enough, I've decided it's best to have a rethink.

The armature is now being made in a single piece - this has a couple of advantages; mechanical strength and an ever so slight drop in moving mass, it's all mass close to the pivot point, so I doubt it will make a difference, but I'm going from an M4 threaded rod to a 3*3mm square, so the cross sectional area is a little smaller.

I have also shaved a mm or so from either end of the armature, which is at the extreme of the pivot, so should have some effect.

After making the adjustments manually, I have noticed that has made a positive change to the frequency response and volume.

The aluminum damper holder was originally threaded and glued in place, in the new order of parts I've had a brass damper holder made, this is so I can solder it to the armature.

---------------------

I hate doing this, as it's a recording of a screen, but I've since handed this cut record over to the customer so can't get a better recording - there will be something more definite in future.

But I put a 13khz brick-wall hi-cut when embossing this record and it definitely seems to be coming through.

312005763_5441139932668682_2418569265089883388_n.mp4

---------------------

The heads seem to be getting better and better, but they still seem to be lacking the consistency I am after. I've designed a few machining jigs which are being cut with this next lot of parts.

The other thing I'm considering is a test bench to align the pole pieces and tension the springs, it can be done by ear, but I feel that's a little too subjective.

I tried constructing something using a bone/skull tranducer, but the EMF from the coils is too great. I am wondering if a piezo pickup may be a better option.

------------------

I've picked up some piano wire (it's very nice stuff...Never thought I'd say that about music wire, but nevertheless I visited a harpsichord maker and he was generous enough to give me a short length!) and will be replacing my handmade die with a machined die. If this new die is successful, I will try to document how I've constructed it.

[markrob]

Hi,

Great work! I never had much luck with a piezo pickup (but don't let that stop you from trying). On the plus side its very high output. But, its also very high impedance, so its hard to shield from the stray fields. On top of that, the common pickups on the market today are not very Hi-Fi, so I've been skeptical about using them to make quantitative measurements. In your case, you really just want to compare to a golden sample, so its not critical that its accurate.

I'm working on a laser interferometer test jig, but I'm not quite there yet. I'll post if I get somewhere.

Interferometer.jpg

Mark

[jjwharris]

I'm going to give one of those cheap red piezo phono pickups a go... I've thought it could be expanded to make a nice pre-test for cutting records - gauge if there is too much over-excursion etc. But time will tell. I imagine that coupled with a DSP could give a fairly good preview of what the final cut will sound like.

How does the laser interferometer work?

It's been a long while since I've updated, but I'm in the process of manufacturing the final parts.

I've had a couple of interesting issues...

I was casting silicone dampers into shape, and fitting them to the cutting head, however they were doing strange things to the frequency response.

siliconedamperresponse.PNG

As you can see, it kind of killed everything above 5khz. I removed it and replaced it with the neoprene dampers I have been using previously and the issue went away. I'm looking towards cast urethane dampers for the final damper, I've reached out to a local plastic engineering company and I'm hoping they can help me with density, shape and thickness to tame the specific resonance of the head.

The brass damper soldered to the one piece armature worked well, it stayed in place and there is no chance of the set screw holder shearing off.

However, the frequency response was way off.

ArmatureWithSquareDamperHolder.png

I tweaked and tweaked it thinking it could be to do with tension of the torsion springs, or alignment of the pole pieces.

Eventually I decided to take the brass damper off, thinking it could have been the brass itself.

SquareArmatureWithNoDamper.png

This fixed it, except the resonance was clearly visible.

I tried my old parts - glued the old circular aluminum damper holder in place and screwed into that. - It went straight back to the original response I had before.

Out of frustration I cut the damper holder down into a square, to match the shape of the new armature set screw holder - It fixed it completely!

So there seems to be a weird connection between the shape of the piece the damper fixes to and the holder itself - unfortunately I have no photos of the final response, I've only dug the above responses from various PMs.


I'm pretty much done with tweaking things at this point, my only other consideration may be making the armature rectangular rather than square, to bring the pole pieces closer together and have more 'face contact' between the armature and the pole pieces. But I'm also quite keen to get this thing done and dusted...

[grooveguy]

You're making good progress, JJW, and taking the time to do that empirical testing along the way too! "Responsible R&D," good man!

[markrob]

Hi,

Sorry about the delay responding. I would give the piezo pickup a try. Remember this is natively a displacement sensor, so you will need to high pass filter (differentiator) the output if you want to correlate to a flat magnetic pickup response. The big issue here is that there are not currently high quality ceramic pickups on the market, so the response is not really categorized. Back in the day, the Microacoustics pickups had very good response.

http://www.roger-russell.com/sonopg/micropg.htm#cartridges

That said, if you intend to setup a test bench to do a QC check on you heads, I think that would work out well.

The laser interferometer allows you to measure very small excursions of the head movement without making physical contact and is immune to EMI from the drive coils. Several years ago, I came upon this article and I built up and tested this with great results.

Doppler Interferometer For Speaker Measurment.pdf

My current build is using a simpler approach that uses an inexpensive laser diode rather than a laser pointer in a feedback arrangement that eliminates all of the mirrors and beam splitters.

Reconstruction of Displacement Waveforms with a Single-Channel Laser-Diode Feedback Interferometer.pdf

The main issue with both approaches is that the recovered displacement waveform can't determine direction and needs some processing to reconstruct the motion. That's where I left off when I stopped working on the project. I do intend to get back to that at some point.

Mark

[jjwharris]

I'm tuning the coils at the moment. Interestingly 95 winds with .28mm wire gives me an impedance of 7ohm @ 1khz, with a turnover at 395hz adding a 1ohm series resistor would bring that up to 8ohm, and bring the turnover up to 500Hz, which I believe would be it into the correct ball park.

Just to confirm I am doing the math correctly...

R = 1.661ohm
L = 0.8561mH

1.661/(2*Pi*0.0008561) = 308.79Hz

1+1.661/(2*Pi*0.0008561) = 494.69Hz

Anyhow, Springs have arrived this week.

Untitled.png

The main components should be here next month.

I have a few different dampening materials to test, different shore hardness urethane, it's castable so I'm almost wondering if it could be worth experimenting with specific shapes.

Here's a couple of the latest tests;

https://drive.google.com/file/d/1ACtbZYNP61I6reb6CMkVmeEUUVx2R3pQ/view?usp=sharing

https://drive.google.com/file/d/1gZH-fKPSl3JLROz_35zTE9WKq_Aun0ME/view?usp=sharing

[jjwharris]

I've attached the new damper material, this is cut with my embossing needles, and one of the heads - I'm very happy with it and I do feel like I'm close to the end.

The coils are 95 turns at 0.32mm. for a total of 5ohm @ 1khz with a turnover frequency of 250Hz- Another element I'm not happy with, they're a bit untidy, and after my testing with the DATs unit the other day I have realized I need to make a coil winding rig that can get repeatable results, as the coil construction will affect the turnover..I'm hoping between that and losing the extra ohm I should be close.

In terms of frequency response I'm seeing 15-16khz with white noise, but I get some strange sounds when I cut this track without a 14khz cutoff - I'm thinking this could be the limitation of the stylus itself.

It's cut at 240 LPI at 45rpm on a 7", interesting you can see the HF drop off near the end of the song. I overcooked the bass a little too.

https://drive.google.com/file/d/1jr3wwo-4N-HsiZUx4LluC_-2tGzdLWj1/view?usp=sharing

I'm going to mill out some moulds for the dampers, my cases are a little smaller than the Presto 1-D, and the current damper is around 4.2mm thick, luckily the material is pourable, so I can gain 1-2mm back from the damper holder.

Damper construction is the other consideration - I've a feeling I'm seeing this increased response because the damper is free-floating, and not clamped, as it usually is. I need it fixed it space, but maybe I've been cranking the tightness too much.

Hopefully I have some diamond cutting stylus soon, which should pull back some fog on what the head is capable of.

[grooveguy]

Very commendable results, John. I think you really nailed this 1D re-engineering project. Congratulations!
I remember back to my Presto 1D days, and how top-end response could be extended from 10kHz to 15kHz simply by cutting the traditional "long shank" stylus down to "short shank" length. The long shank was de-riguer for cutting 78s, but the much lower deviation of microgroove techniques didn't require extra leverage, so it was a no-brainer tradeoff to nip-off some of the shank. Give it a try.

[jjwharris]

About 12 months ago when i first started testing these heads I did a few experiments backing the stylus off a bit each time, I found the best length was 8.1mm from the polepiece/end of the armature...I've yet to go back and check it, but it is on my to-do list.

Did a buuunch of cuts last night, trying to get a fairly broad spectrum of different material. I'm 95% ready to say it's done, but there is some distortion rearing it's head, and I think I know where/what it is and I think I will try to prototype a part to fix it, but it could be an extra cost, and I'm not sure if the payoff will be huge - that said, it's probably worth getting some quotes and comparing.

anyway, here's the test record;

https://drive.google.com/file/d/1OWBG1LYlGWyKccZpWDLu9KjfSGRZJCJQ/view?usp=sharing

Blue Monday - New Order
Bad Kids - The Black Lips
Real Slim Shady - Eminen
Louie Louie - The Kingsmen
Billie Jean - Michael Jackson
Money For Nothing - Fleetwood Mac (Lost heaps of material in the mono sum of this.)

The EQ before the cutting head is light de-essing around 6-7khz, Ozone 10 (izotope) for the mono summing and a high brickwall at 14khz. Tracks are all FLAC.

[diy_cutter]

Great job, a lot of work, head looks very professional .

Excellent sound, very balanced, sufficiently high level, low surface noise. I missed the highs of the Black Lips a little and when I listen with headphones there is a sense of overload or distortion. New Order and Eminem sounded very good (there are some distortions, but the track itself is built on overloaded sounds). Kingsmen - very worthy. Michael Jackson is also good, you can't hear any distortion, it lacks a bit of density, it sounds thin, probably due to mono mixing. Money for Nothing at the New Order level, of course, without stereo there is a loss of information, but this is natural.

I'm thinking about ordering one of your heads as we plan to restore our 1937 RCA overhead. We are too busy with our stereo heads and there is no time left for mono experiments .

[jjwharris]

Thanks for the feedback!

It's been great watching your progress with the stereo head!

The vocals on the Black Lips track seems to have a little distortion itself, but I agree, there does seem to be a sense of distortion.

I will be experimenting with laminated pole pieces like the original presto, originally having laminated pole pieces made was uneconomic, but after getting some prices for the quantity I need, it seems like they will be the way to go.

I also need to test these cutting heads with my GX5 cutting amp, at the moment they are powered through these TDA7293 based amplifiers; https://www.aliexpress.com/item/4000746305619.html

I need to work on getting some overload protection set up on the GX5 before I get there.

Mark, you were asking about the resonant frequency, it seems to be sitting at 4.2khz

Here is the undampened frequency response;

Resonance.PNG

I have removed some windings from the coil, and it seems to be sitting at 4ohm @ 1khz;

newheadresponse.PNG

DATs Response

Turnover is 250hz

I will be working on a coil winder that can give a repeatable coil every time, to try and ensure that turnover is in the same place each time place.

At the moment the armature is from 3mm mild steel, I would like to try experimenting making the armature from something thinner, I'm eyeing up 12 gauge steel, but keeping the same 3mm dimension across the pole piece face, meaning the armature is slightly rectangular rather than square, this slight decrease in weight could have an effect.

Another consideration has been this mention in the BBC type B paper

TypeBArmature.PNG

Working off the armature being solid (without the stylus holder) would make the thickness currently 3mm, which means there should be 2.7mm 'face-to-face' contact and a 1.5mm overlap of each end of the armature - from the photos I have seen of the presto, this isn't the case.

The next thought is that the stylus mounting hole decreases the thickness of the armature, so the 2d area would become 3*3mm = 9mm2, subtracting the stylus mounting hole (2.27mm2) would take us to; 2.24mm Thickness

So there should be an overlap of 2.016mm face to face and 1.12mm overlap of the pole pieces - both of these look entirely 'off-spec' to the original presto design. But maybe it is worth looking into.

[jjwharris]

I'll be updating soon with an assembly video, but these heads will be available for sale next month.

I'm waiting on a few other parts, along with boring stuff like damper material and potting compound.

I thought I'd show this coil connection, I like it as the wires come in the back and the PCB seals off the back entirely. The connection of the coils has always been a bugbear for me throughout this process, but I think this is the tidiest way to do it.

Heads will be wound to 8ohm at 1khz. It's quite irregular for these heads to blow the 1A MCB I am using for protection.

I'll be sending these for testing to a couple of guys in the states, if you're interested, feel free to get in contact.

I still want to provide a youtube assembly video of the entire thing as a repair reference.

The goal has always been linearity over all the cutting heads, but I'm beginning to think that in the analog realm this is unlikely to happen. I will see how the new precision parts perform, but if there is enough of a variance between the heads it is likely I will be selling them at different price points/grades.

[jjwharris]

https://drive.google.com/file/d/1_9zvdD25TrZhboYGzT6rgKBp85EimEsV/view?usp=sharing

Sorry, here's a couple of test tracks - tone at the start is 1khz @ 7 cm/s

Other two tracks are with one of my prototype heads which was hitting around 10khz, other track is one of the better prototypes.

Tracks are embossed, test tone is from the Ultimate Analog Test LP

[grooveguy]

For an embossed track, that's really pretty darned good. Better results than I've ever had. Good work!

[jjwharris]

Parts are all finished and here; the laminated pole pieces and new magnetic yokes seem to have changed the magnetic circuit quite considerably, to me the head sounds cleaner, but the impedance has increased from 8ohm to 10ohm, that's using the exact same coils which were in the prototype.

I'm potting the final coils in Elmotherm® 073-1010 which requires a baking temperature of 140-160C, this means I cannot use the acrylic bobbins as they melt into a mess. I've been experimenting with Micarta bobbins, which seem to be the best option at this point, I need to finesse my method for baking and winding a little more, but with the same number of turns (85 per coil), same gauge wire (0.315mm2, I'm assuming, my calipers are showing .25mm diameter) and same diameter coil, I'm now seeing a 6ohm impedance - it's a bit of a speed bump, but I think the end is in sight and after the boxes and manual are finished, I'm hoping to have them available for purchase at the end of the month.

JE1-DBlueInnards.PNG

The innards are the pieces that I'm most proud of, I'll be posting full assembly and photos soon, the case is the same as the one posted above, except the fixing screws have changed a little - all just for aesthetics really.

https://drive.google.com/file/d/1M0bePJ_l7UW7k_qAnc_ewH9ds1DqxQvZ/view?usp=sharing

Here's a dirty embossing sample, I haven't changed much as I would prefer to get these coils cemented before I start messing about too much.

Pink Noise Response is;

JE1-D Response.PNG

Corrective/Cutting EQ is this;

JE1-D Cutting EQ.PNG

I don't run an iRIAA curve, as I've found it preferable to do it all in one hit with a corrective EQ + iRIAA

Amplifier is (as always) is this UNISAN cheapo. I have yet to test these heads with the Power Amp I have.

UnisanAmplifier.PNG