JE-1D Version 356

[jjwharris]

As a side note, I've been measuring the weight of my heads as they travel across the records - It seems some of my rails are out of parallel with the platter - Interestingly enough, the machine I have the most issues with skipping at the outside of 10/12" records has the most consistent weight across the record. I'm wondering if it could be beneficial to have a slight 'lift' in the carriage towards the centre of the record.

This in theory would apply more weight at the outside of the record, which would produce a groove that is more easily tracked, while decreasing the top end, which is going to be more abundant at the outside.

[grooveguy]

Wow, like the 'diameter equalizers' of the pre-microgroove era. If that's for real it indicates that 'penetration' may be inversely proportional to groove velocity. If I had a Ph.D. I might be able to write an equation for that... for all the good it would do. Yeah, maybe because of stylus geometry too, who knows. Further testing for sure!

I like your idea of the tortional-mode spring too. Fairchild heads used that arrangement, I think Grampian too maybe.

One other question. It looks as if you stack a number of neodymium magnets "in series" to bridge the gap between the side pieces that carry the field down to your gaps. I've been told that this lashup actually increases the pull over a single magnet, but the guy who told me didn't have a scientific basis for it. I would have thought that putting magnets "in parallel" would deliver the greatest field, but then you are putting like poles together in that case and probably really reducing the field from the pair. What's your understanding? I was going to try to mill-out the "U" of an old Presto 1D and put neos in there.

[jjwharris]

Hey sorry!

I keep typing out replies and forgetting to hit reply!

Rather than penetration, I'm wondering if it's that the outer grooves are 'throwing' the playback stylus out of the groove with the increased distance they are travelling.

I've been playing with embossing angles a little more, and I'm now embossing at 84° - I can dial the weight back by quite a bit, 45g instead of 65-75g.

I've finally got the packaging completed for my stylus, so I will try send you some soon. Post from NZ will have two options - $~5NZD letter post, and $~20 parcel post.

Parcel post would be insured, whereas letter post isn't.

The torsional spring arrangement is based on what I've gathered from looking at the BBC Grampian docs - I actually have a lot of parts for the Grampian design, but at the point in time I was looking at it, it was too difficult to machine the body of the head. I should have a reasonable CNC together for aluminum soon, so it's on the cards - like most of my projects...

Funny you mention stacking magnets! I read the opposite so ordered longer rod magnets!

The 'magnet' components consists of laser cut steel plate that's tack-welded together, which carries the magnetic field down to the pole pieces.

My plan was to use laser cut steel plates to capture the dampening material - as soon as I installed them I realised my mistake - that I was shorting the magnetic field - It's on of those things that I've been super conscious of the entire design and build, but somehow slipped past.

I have aluminum replacements being made this week - The dampening component is one of the last tweaks I need to make - that and revisiting the omega springs.

[grooveguy]

All good learning... and progress; keep at it! And I like the embossing angle you're using now, closer to vertical. Plus, I don't know the exact geometry of your stylus, but would think that the closer to vertical that it is, the smaller the area actually in contact with the surface of the blank. So, yes, lower force is required, and maybe high frequencies do better as well.

I take it that you are embossing on polycarbonate, which I think is probably the better surface. Some blanks for us experimenters are being distributed now, in polycarbonate and in other plastics. Proponents of the alternatives claim a lower noise floor, which I can readily believe if the material is softer than poly-c. After all, we embossers are simply deforming the surface, so a hard material should have less "springback." If we were cutting a groove in lacquer, or thermally deforming the material, like a vinyl pressing, then that's a different story altogether.

The old Soundscriber dictating machines used a thin blank of some plastic material that remained unchanged from the 1940s forward. Not sure what that stuff was, but seemed pretty hard.

Furthermore, I have reservations about heating a blank for embossing. It may lower the noise floor, as the material is temporarily softer, but while it is softened by heat, I would think that it would spring back more readily. That is, continuing to heat an embossed groove might tend to encourage high frequencies (short wavelengths, actually) to sort of flow back together.

All food for thought only, I don't have a Ph.D. in Materials Science. By the way, I think regular (uninsured) post might be the best way to send. I haven't lost an international letter in decades.

[jjwharris]

I don't heat my blanks either - never really saw any sense in it...Polycarbonate does seem to have a thermal memory - I've been running a few EQ tests and saving on blanks by blasting a embossed blank with a gas torch - the grooves pop right out and you're left with a smooth (fairly warped, and noisy) blank. It's enough to re-use to adjust an EQ though.

I did some experiments dying record blanks. Synthetic dyes seem to create a bit of a noise floor, I assume the dye is sitting on the surface rather than impregnating the polycarbonate.

Part of that experiment was embossing a record and then dying it, and as you described, the higher frequencies vanished from recording. I believe the water had to be heated to around 80C.

However, It did make some cool looking records.

The springback of polycarbonate was why I went with a greater than 90 degree stylus tip, I have the feeling the groove geometry also springs back when embossing. I ground the stylus wider until surface noise became a problem, but they do seem to track better than the standard 90 degree angle - going the other direction makes the groove even harder to track.

A friend told me Peter King was pretty adamant about keeping the room at a certain temperature, it makes me wonder if it was more to do with the dampening system though.

I do get a quieter groove and more HF leaning back, but tracking suffers.

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

I've implemented omega shaped springs (1mm piano wire, formed in a die). They do seem to perform better. I had the die from previous attempts, but run into issues with the die pinching the spring, or snapping the spring. I made some adjustments and they are forming fine now.

REALLY pushing the EQ I get a fairly full response with the embossing needle, but clipping becomes an issue. I have an assortment of magnet wire arriving soon I hope, so there may be some adjustments I can make with the coil.

There's a lot of tinkering around the edges at the moment, but cases arrive on Monday, so something will be ready soon.

I think previously I may have been over-dampening the assembly, as the damper design is now like the presto and floating in free air, with a piece to hold it in place. It will be interesting to see if there is an effect clamping the damper with 4 screws as opposed to securing it between the two nuts.

There's a photo of the dyed blank, and a photo of 3/4 colours of polycarbonate I have arriving Monday...

[grooveguy]

Good to see continued experimentation and progress. And you may be right about over-damping; Presto relied pretty much on damping against the mass of the Prestoflex® block (which I think was pretty much just Neoprene), with it held lightly at just the one point to keep the damper assembly from unscrewing from the stylus holder. The old 1D was exemplary among that type of head, and may actually have had some science behind its design, perhaps a true "custom-tuned" assembly with lots of equations modeling it as a low-pass filter or something. I certainly am not that smart and depend a lot on trial and error, just lucking-in to acceptable results sometimes. But at least in our experiments today we are fortunate to have inexpensive instrumentation that helps guide us in the right direction. Looking forward to trying one of your styli for sure.

[jjwharris]

What was the thickness of the Prestoflex® block?

I have been experimenting with 2-5mm thickness. Isolating the mass of the damper definitely seems to make a difference to output.

I've finally overcome one of the biggest obstacles...Previously I was attaching the M4 thread to the armature with a VERY light tap from an arc welder...I then tried a MIG....then tried an alibaba 'pulse' welder...

I've managed to make these with simple brazing.

I have been trying to film the machining process, and I'm hoping to release a 'how to build' video on youtube soon...

I was planning on nickel plating the components myself, but I think at this point I may just got to someone local.

[grooveguy]

Prestoflex was about 1/8 inch thick (3.2mm). Beautiful work on those armatures, you're good! That would be a quite involved YouTube video, I would think, well beyond the skill level of most of us.

[jjwharris]

I've been meaning to start making them for awhile, but I've pretty much got a microscope over the lathe and a go-pro on an arm...I'll cut the footage together and try to give some sort of explaination.

All my cases arrived today, so I should have 'finished' pictures soon - I'm still trying to find a good way to test the performance of the heads, I thought of using a benchtop test with an upside down cartridge - I believe this is how markrob has done his pre-cutting cutting tests.

I'm winding a few coils today and taking notes to see if I can make any improvements on my current formula.

One of the tests is going to be wiring the coils in parallel, the theory being that it would double the current rating of the coils, while decreasing the distance of each coil/line of flux to the armature, which should create a more sensitive head...

Which leaves me to ask why the presto coils would be wound in series in the first place; is it because 16/500ohm was a more common amplifier output at the time?

I do remember in trade school, talking about running power mains in parallel and the effects that would have on loading. I don't think we ever really got a clear defined answer, but one theory was that as the wire resistance would never be completely equal, there would always be more current running through one wire, which could increase the temperature of the coil, increasing the resistance and cause more current to run through the other wire, see-sawing back and forth...

Do remember this was in my younger days, so a class of 15-20 year olds...

[markrob]

Hi,

I think you can run the coils in parallel if you like, but that would tend to reduce the impedance of the head. You would still need to make sure the coil polarity is correct. The 1C was available with an 8 ohm winding, so I'm not sure there is any advantage to doing it that way. Its also a bit harder to wire as compared with series configuration. Since the head is really a current responding device (F=BLi), having the coils in series insures that each coil sees the same current. In parallel, that might not be the case. The 1D and 1C were meant to have a series resistor installed to get the L/R time constant such that the RIAA 500hz turnover point is followed. You don't need to do that as you will correct the response with EQ externally. One interesting thing to try would be to run each coil via its own amplifier. That was not practical back in the day before stereo was the norm. The advantage of this approach is that you could null out any imbalance in the system due to variances in the two coils or the magnetic push-pull circuit. That would reduce 2nd and even order harmonic distortion. Not sure if the reduction is worth the hassle, but its a thought.

Mark

[jjwharris]

Thanks Mark, that makes a lot of sense. I think it was close to what I was thinking.

I've set up an oscilloscope to calculate impedance. It might be overkill, but since I will be selling these cutting heads, I would prefer not to kill someone's amplifier and be able to give a accurate reading.

I'm testing the coils at 1khz, I seem to be getting higher than expected impedances, 2.3ohm is giving me 12ohm impedance at 1khz.

I often see speakers rated at "4/6/8 ohm @ 1khz" but I imagine the more accurate way to measure impedance would be to take a range of measurements and average them all.

[markrob]

Hi,

I've been using the DATS analyzer from Parts Express. Not too expensive and gives great results. I previously owned the older version and it was a bit flakey. This version is much nicer and more stable. Not sure if you can get this in your neck of the woods, but its worth the price. If you are going to produce the head, it will make production testing much easier.

https://www.parts-express.com/Dayton-Audio-DATS-V3-Computer-Based-Audio-Component-Test-System-390-807

[jjwharris]

Thanks for the heads up on that - I'm hesitant to spend more money because a lots been going out lately! But that does look like a worthy investment...I was looking at an impedance meter on aliexpress and what you have suggested looks like it would be a much better tool suited for the job.

The results I am getting with the oscilloscope are interesting...

297228171_1073834056853425_3124744400320894829_n.png

This is the impedance shape I am expecting...

This is what I seem to be measuring - it seems like the test rig is doing it's job. I do need to check a few things, but from what I can see, I'm looking at a 8-9 ohm impedance.

I'm unsure if this checks out - I've been asking around a few amp/speaker guys I know locally and asking how they measure impedance, the most common answer is "You guess." or if they want to get into specifics, it seems to be that it DC resistance is 70% of impedance.

The graph below is the measurement of a head that I have equipped with 2 coils in parallel (Measuring 1.6ohmDC total, or 3.2ohmDC each)

Capture.PNG

From the logic of the people I've spoken to, that should give me an impedance of 2.28ohm (ish)

I've always assumed I've been underwinding my coils, but if the measurement I am getting is correct, I think I should be able to squeeze out a bit more volume from this head.

I think the next step will be ordering the Dayton Impedance tester, while waiting for that to arrive I will wind some coils to (my measured) 4ohm and 8ohm and see how they behave.

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

It should be noted that my impedance measuring setup is less than ideal... I'm working off voltage drops, and as I understand that doesn't take into account complex impedances and phase shifts.

I've attached a diagram of my test setup

setup.png

Maybe it's completely wrong! Will wait and see. Oscilloscopes do my head in.

[markrob]

Hi,

How are you calculating the impedance based on the measure of V1 and V2? Are you taking phase into account?

Typically speaker impedance is measured at 1Khz. But that's just a reference point. That's why the impedance curve is a much better measure. Gives you more details.

[jjwharris]

I'm using the method described here;
https://assets.testequity.com/te1/Documents/pdf/applications/oscilloscope-measure-impedence-capacitance-guide.pdf

Impedance = V2 / (V1-V2) * R

I don' think this takes into account phase shift.

----

I have ordered a DATS V3 analyser - Shipping was $80USD from Parts Express, but I found it on Amazon for the same price with a more reasonably shipping rate to NZ.

I think that will be the way to clear up any confusion and get the most solid results.

[markrob]

Hi,

That looks right to me. Should give you the magnitude, which is what is mostly used in speaker specs. Just won't tell you how reactive the load is. The link you provided shows you how to determine that, but I don't think you need that info.

The DATS analyzer should make it much easier to obtain the impedance curve. On a normal speaker, you can clearly see the free air resonance. On my DIY moving coil heads, I can see the system resonance. I can see how much adding ferrofluid damps the resonance which is useful.

I'm glad you didn't have to pay such a crazy shipping price. I hope you fine it worth the price of admission. You should find some other uses for it as well.

Mark

[jjwharris]

I'm running into slight problems getting the cases on - they look great, but on the 5 sets of aluminium parts I've had made up, things are just a little too snug...

My choice of hex socket screws have also taken up a lot of space, so I will be ordering more.

The really stressful part is the damper connection - I must have made a measurement error somewhere along the way as there's just not enough gap to fit the dampening system with my current screws and get the case on. I've got some 'flatter' screws on the way which will hopefully means I can clear it.

I've been winding more and more coils..the winner (for my setup at least) seems to be 85 winds on each bobbin, .315mm magnet wire, with a DC Resistance of 1.8ohm.

image (3).png

That response is with 4 filters; two high pass filters (1225hz and 500hz -10dB each) a high boost at 4.5k (10dB) and a high cut at 14k, essentially a close approximation of the iRIAA filter.

I'll see if this carries over with the 3 coils I have baking at the moment.

I'm still on the fence with the omega springs...I'll swap them out for the straight piano wire I was using earlier to see if I can notice any loss in response/output.

[jjwharris]

The DATS V3 arrived today - 10 minutes into using it and it seems like it's worth the cost.

VERY easy to use.

I've been struggling with some serious hum from the amp, and this points to the heads being way way way too low impedance..The thing is, they do sound great. I've been trying different wire thickness and turns and the coils I'm using I'm most happy with - except for the hum.

They're pretty cheap mono-block poweramps, powered by toroidal transformer so I was looking at trying to find a schematic to see if I can replace the smoothing capacitors on the amp board. after seeing the heads are a ~1ohm impedance, I think the first option is to rewind them.

Keep in mind the only mono head I have access to is an old RCA-MI4887, so up until now, it's all been by ear.

4ohmspeaker.PNG

DATS Reading of a 'known' 4ohm speaker

NewCoil1.1ohm.PNG

DATS Reading of the current coil that cut "abstractlathe.mp3"

There's a bit of surface noise on the lathe cut. I've still been playing with geometry and shapes - I had a conversation with Eposlab and we've come to similar conclusions with tip shape etc. His setup can get him to 100degrees, whereas I can only hit 97.5 - which I found to be the best shape, a little sharper and it lead to worse tracking, a little higher and the surface noise increased.

The back angle of the stylus seems to play the biggest part, I can't seem to get as a quiet groove with the head going towards being vertical, but as the head gets close to vertical, less weight can be applied, the stylus tracks much better, and the content can be louder. It's a bit of a trade-off. That said, I think there's an element of versatility with a sharper back angle, as you can lean the head back for a quieter groove, or bring it closer to vertical for louder content and more surface noise - I guess ambient music would be cut with the lean back, whereas louder/rock records could be cut with the head closer to vertical.

[dubcutter89]

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...

Also the plot looks really nice to me - a moving iron cutter looks different than a moving coil speaker.
The almost straight rising impedence shows nicely how the coil inductance kicks in, turnover frequency is around 200Hz so should be similair to 78rpm records from back then...
From my deeper understanding of how moving iron cutters work this (straight impedance rise with inductance) is part of getting the head flat velocity responding....

Keep up the good work!
Cheers
Lukas

[markrob]

Hi,

The curve looks very nice. As Lukas stated, its a pure inductance in series with a resistance. From the plot, I see R = 1.041 ohms and the inductance is 0.2007 mh. That yields a turnover of about 800hz, which is a high as compared with the 500 hz RIAA standard. The Presto heads were wound such that you needed to add a series resistor to adjust the turnover. However adding resistance in this case would move the turnover higher, which you don't want. 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.

Some years ago, I measured a Presto 1C that was factory wound for 8 ohms (the 1D never came as an 8 ohm wind). I found these parameters:

Wire Diameter: 0.0115" (29 or 30 Ga)
Coil Wire length: 8'2"

This yields about about .82 ohms per coil or 1.64 ohms total in series.

I never ran an impedance curve on the 1C as I didn't have the DATS tester at that time. But from your measurements, that looks like it would be in the right range at 1Khz. You are currently a bit under 4 ohms at 1Khz.

That might help you with your wind.

Mark