You have the equation incorrect, in fact 1/Re is only to the first power.
But it's only an approximation anyway, I already mentioned that D.B. Keele has an excellent paper with more accurate results.
The JBL 2226 is just on the limit of what the simple T/S equation predicts reasonably accurately.
A more efficient version will need the full catastrophe.
Best wishes
David
Benefit of the "sandwich" former is that the heat from the VC doesn't have to pass thru the former. Lower thermal resistance.
It's not been promoted but in occurs to me that there is also some benefit from increased adhesive bond area, uses both sides of the former.
Benefits of Carbon Fibre are many
.Carbon fibre is stiffer and has better tensile so it should keep the VC nice and round and bonded into one piece.
All at lower density than fibreglass or aluminium.
Doesn't soften or creep like plastic, low coefficient of thermal expansion to hold tolerances.
Doesn't deteriorate at elevated temperatures,
And thermal conductivity better than usual composite or plastic.
Well, this is only a PR blurb but it's based on a JAES article that presumably was peer reviewed, so it's a bit odd.
The SLMM is actually iron powder in a non conductive binder.
I think the touted non conductivity misses the point.
I suspect that the real reason for any improvement is simply that the saturation level is lower than the rest of the circuit.
Saturated pole tips in other words, exactly as you discuss below.
I didn't mention this previously because I didn't want to bias the discussion, so I am pleased to see you raise it spontaneously.
I have seen this idea proposed a few times, by A. Novak in the JAES and a DIY poster who I can't recall just now.
Prima facie I would expect the optimum operation point for the NdFeB to be less than 2T but the DIY poster had an FEMM sim that exceeded that.
Need to study this more to build some intuition.
But ATC use a lower conductivity to reduce the eddy currents. Hmm.
Yes! One of my ideas is a neat way to do this.
Current polyimide substantially exceeds epoxy. Expensive of course.
No problem, have a safe journey, and preferably fun.
Best wishes
David
Last edited:
Depends on whether you are talking power or voltage sensitivity
And yes, ignoring radiation resistance (et al) only works up to a certain point. Good luck.
Last edited:
Hope you are home safely.
Do you have any experience with plated copper Faraday sleeves on the pole piece?
Specifically, I wish to put the copper into a slot and I know it is difficult to electroplate into a recess.
Option is to try electroless copper, as used in plated thru hole PCBs, but maybe not suitable for substantial thickness.
I can cast it I suppose but that seems an awkward option.
Best wishes
David
If the aim is to reduce flux modulation and the increasing inductance, there are several ways of doing this.
A Cu shorting ring is probably best done as a turned sleeve and shrunk or glued onto a shelf on the pole. If you make it longer than the coil, it will have a significant cooling effect on the coil.
If you intend to saturate the pole & top plate, the Cu sleeve/ring provides little extra benefit for flux modulation.
I'm not sure Cu plating has much, if any, effect.
All my Jurassic trials were with 8" units 25mm & 32 mm poles so may not be applicable to your big units.
You need to dream up some way to measure Acoustic THD & compression at high levels to check that what you do is actually beneficial.
I see flux modulation as a separate issue from inductance, even if both often fixed with the same cure.
I take "flux modulation" to refer to the 2nd Harmonic variation.
3rd is a different problem.
But I have wondered if there is some deeper connection... hmm.
I want to do an annulus (or several), more like a washer rather than a sleeve.
The JBL 1500 (link to cross section pic. in post#13) actually does use washers.
I have an improved version but want more flexibility on the shape.
Actually, an induction heater would be a neat way to melt the copper onto the pole recesses, probably quite feasible.
JBL tech note #22 discusses this a bit, it's on their site.
But a recent JBL patent by Alex Voishvillo, (also easy to find) discusses the benefits of the combination of sleeves, annuli, use the aluminium frame as an eddy current component, all at once.
I think I can do even better, even looks patentable, AFAIK.
Same principles, don't be modest
Best wishes
David
OK, It's measurement time!
I don't have a Bruel and Kjaer microphone, or a fortune to spend on one, what do you recommend?
I am inclined to a USB microphone for simplicity of set-up, no phantom power to worry about.
Some of the early electret capsules had inbuilt FETs with nasty circuits, I wonder if this is fixed in the UMIK from miniDSP?
Best wishes
David
The software recommendation is ARTA. Les Watt's (ex Shure) uses the paid version for his own commercial mike making business.
https://www.facebook.com/LMWattsTechnology/
WARNING! I have not used ARTA myself but rely on my own DOS programmes. DOS machines are now extinct.
Microphones for response are easy. A Behringer ECM8000 but get it calibrated properly (WARNING! There are some really dodgy "calibrators" around.)
Distortion is more difficult. The least is a Panasonic WM61a converted to 3 wire. Eric Benjamin on Yahoo MicBuilders has measurements of 3 wire vs 2 wire.
Different vintages of Behringer vary. There have been at least 3 different circuits used in the ECM8000 and some of these were 2 wire.
Mine certainly has more THD than a DPA 2011 but I haven't opened it up.
If what Eric shows in his measurements of 3 wire WM61a are unacceptable for what you want, I would look for 1/2" B&K capsules (4133, 4191) on eBay and make my own preamp. The machining and other stuff is not trivial.
Earthworks use selected WM61a. I'm not sure what they use now as it is discontinued.
I doubt if any other maker of cheap "measurement" mikes do more for THD than convert WM61a to 3 wire.
I'm sorry if this isn't all that helpful.
Last edited:
ARTA is fine, reasonably priced and the developer is helpful.
But my first preference has moved to REW, it's even better priced (free) and this seems to have produced a very active users/developer community.
Likewise, the ECM8000 is reasonably priced and exactly what I had planned to use with ARTA.
But REW has some nice support for the UMIK-1 and there's also the simplification that it's a USB microphone so no need for phantom power or a pre-amp.
Yes, this is my main concern, I assume the UMIK is a WM61A or equivalent but don't know the circuitry.
I know you are active in the Mic forums so I had hoped you had some information.
Thanks for the ideas, I will ask around, maybe start a new thread.
Best wishes
David
As I said, I haven't used ARTA in anger but I have tried REW. My impression is that it's Mickey Mouse and orientated to those who want to convince themselves their investment in digital EQ is justified.
Much of the room EQ stuff on the forum is distinctly naive ... but I'm obviously biased towards stuff that might be used for serious R&D and Production Testing.
There's a free trial version of ARTA that you can compare with REW.
If UMIK-1 came out in the last 2 yrs, it is unlikely to use WM61a as this was declared EOL 2 yrs ago.
The significance of WM61a is that it has the most consistent response of the inexpensive 1/4" omni capsules.
Original ECM8000 used it and Behringer consistency went down dramatically when they switched to a lesser item. We have this from JLI who supplied it to Behringer (and others) in the early days.
The MicBuilders community are still searching for a good, consistent small omni capsule to replace WM61a ... especially for measurement.
I have some ideas to improve THD performance of simple WM61a based mikes inspired by Stefano Simonelli's work
https://groups.yahoo.com/neo/groups/micbuilders/conversations/messages/25146
https://groups.yahoo.com/neo/groups/micbuilders/files/ss%20files/ You have to join.
.. but they need to be tried out in real life. (No. I'm not using his circuit but a very different one.)
Even without ARTA experience you're ahead of me, I haven't used either yet.
What did you find MM about REW? (other than the naivety on the forum, which is only to be expected of course)
It's not the money, I had hoped to save time and not learn software that I don't eventually commit to.
But it's educational in any case, so I probably should try it, looks like pretty impressive work and maybe I'll use it.
I had read about the ECM8000 consistency problems but didn't realize there was still no WM61A replacement that was fully satisfactory.
The UMIK-1 is individually calibrated, and I can have the calibration rechecked by an independent lab so perhaps the consistency is not such a problem?
B&Kjaer is over $1000, even on discount...
Thanks for the information from MicBuilders
Best wishes
David
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.