Let's Design and Build a (mostly) Digital Theremin!

Lookin' sweet André!  My preference is the "simple box on an angle" look.  Very neat layout inside too, placing the opening on the bottom rather than the top precludes stress on the cabling when servicing it, and unwanted cable interaction when closed.

You might experiment a bit with the mounting of your coils.  The windings being placed immediately next to a lot of wood might be hurting their Q a bit?  Even a cm or so of spacing might make a difference.  You can check this with the librarian, or with a frequency counter placed somewhat near either antenna.  It might be nothing too, I don't mean to alarm you over this.

The LED brightness is dependent on their efficiency - you might be able to lower it by increasing the resistance of three current setting resistors on the tuner driver board.  The analog brightness range provided by these ICs isn't the widest, and there is a weird overlap between the two current modes.

André: would it be OK if I post your pictures on the D-Lev web site?

Hi Eric,
Thank you for your kind words!
I didn't realize that the wood could have an influence on Q.
I tried to measure the frequencies but my old frequency counter is not sesitive enough or I should use a different coil.

Anyhow, moving the coils is not difficult. The inside height of the box is 60 mm, so I can space the coil equally 1 cm from top, bottom and side.
Could there be a problem if the pitch coil is too close to the Prozor DACs? I will be at 1 cm also, instead of 2 cm now.

For the LED brightness I will try 6.8k resistor value, which is the maximum specified on the data sheet (page 14).
6.8k gives 3 mA
3.9k gives 5 mA
So, with 4.7k, we should have around 4 mA.
Since the current is given by the resistor value, what is the adjustment principle by the LED parameter on the Diplay screen? PWM?

Of course, I will be honoured if you post my pictures on the D-Lev site!
These pics are not of very good quality.
I will post better ones when I move the coils... next year!

"I didn't realize that the wood could have an influence on Q." - André

I haven't done an exhaustive analysis of this, but noticed years ago when performing my hand capacitance experiments that a wooden stick against the antenna influenced the frequency in a way that a plastic stick didn't at all.  

"Anyhow, moving the coils is not difficult. The inside height of the box is 60 mm, so I can space the coil equally 1 cm from top, bottom and side."

Sounds good.

"Could there be a problem if the pitch coil is too close to the Prozor DACs? I will be at 1 cm also, instead of 2 cm now."

I would worry more about placing the coils close to significant metal and / or operating circuitry than wood.  But it is all relative, and experimentation here gives the final answers.

Since fields drop off as the inverse of the square of the distance, even small separations can help.  The moisture content of the wood is likely a factor.  Not hollering at you at all, and it's hard to know how much the wood might be influencing things (most likely not a lot) but I try to discourage folks from disassembling the coil / AFE boxes in the kit unless there's an overriding reason to do so, as the boxes guarantee some minimal distancing for the coil, and the AFE is placed off to the side where it services both the hot and cold ends of the coil (something of a compromise position).

I just noticed that both coil wires seem to be coming out of one end of the coil?  That is likely more of a problem than any proximity to wood.  It would be best to run the low impedance drive wire outside the coil and away from the high impedance hot sense end of the coil.  Think of the coil itself as an antenna, with 0% sensitivity the cold end and 100% at the hot end, and with a gradient along the winding length.  You can run your finger along the coil and see this effect on the tuner.  Running the cold wire so close to the hot end of the coil is almost certainly reducing some of the absolute sensitivity.

"For the LED brightness I will try 6.8k resistor value, which is the maximum specified on the data sheet (page 14)."

Let me know how this works for you.  The current value tries to balance dark room viewing vs. bright stage viewing, but the dynamic range isn't the best.

"Since the current is given by the resistor value, what is the adjustment principle by the LED parameter on the Diplay screen? PWM?"

The brightness knob sends codes to the analog drive circuitry in the driver ICs to control the overall brightness.  PWM is used for the grayscaling between LEDs.

"I will post better ones when I move the coils... next year!"

OK, I'll wait until then, thanks!

"I just noticed that both coil wires seem to be coming out of one end of the coil? ... It would be best to run the low impedance drive wire outside the coil and away from the high impedance hot sense end of the coil."

The drive wire goes to the end of the coil which is close to the AFE.
The sense wire comes from the farther end of the coil, through the coil tube to the sense input.
So the drive wire is inside the tube but not inside the coil.
The sense wire does go through the coil.

If I understand well, the drive wire should go to the farther end of the coil, outside of the coil and at some distance (2 cm?)
The sense wire should come from the closest end a go directly to the AFE.

Am I right ?

"Am I right ?"  - André

Yes!  

It goes like this (very sorry for not providing any guidance here earlier):

1. Orient the hot sense end of the coil so it is closest to the antenna, the cold drive end farthest away.
2. Keep the drive wire outside of and at least 1cm away from the coil; this spacing matters more the closer it runs near the hot end of the coil due to the increased sensitivity there.
3. Locate and orient the AFE so it accommodates the above, and isn't super close nor too far away from the hot end of the coil.

Thanks Eric, this is very precise and clear.
I will position the coils about 10mm away fom the wood and I will run the drive wire in the corner between the top plate and the side, i.e. about 14 mm from the coil.

Photos and first results, beginning of next year.

Is there any open build project with D-Lev cabinet with pipe antennas based on standard D-Lev kit?

"I will position the coils about 10mm away fom the wood and I will run the drive wire in the corner between the top plate and the side, i.e. about 14 mm from the coil." -  André

Sounds good!  I must say that my advice is based less on precise experimentation and more on the intuitive sense one develops after dealing with this stuff for more than a decade + an abundance of caution.  Anything that loads the antenna with C without contributing to the hand signal will reduce the sensitivity.  Can the D-Lev function OK with reduced SNR?  Generally yes, but Theremins are accidents waiting to happen, so if some degradations aren't too difficult to fix, it makes sense to address them.

"Is there any open build project with D-Lev cabinet with pipe antennas based on standard D-Lev kit?" - Buggins

Not that I know of.  Perhaps someone will take this on.  I'm overdue for constructing another pipe type cabinet early this year and will post the results.

Eric, I have equal confidence in your experience and in your intuition !

RuoShui 4080 LCR Meter Review

My cheapo resonance based LC meter started acting up, so after looking around at the various offerings I pulled the trigger on the $120 RuoShui 4080 LCR meter:

I bought this meter to measure very small capacitances and it seems to resolve well to sub 1pF.  Also to measure inductances in the range 1mH to 50mH, and it does that well too.

Some observations after having used it for a bit:
- I really like the internal lithium battery + USB-C charging and wish my DMM had this instead of a leaky 9V.  When charging the screen shows "In Charging (200mA)" plus the battery level icon.  It would be nice if it displayed the charge level % numerically plus an ETA to full, but I suppose the icon is sufficient.  It appears you can use the meter while it is charging.
- I suppose I prefer this hand held form factor over a bench type instrument.
- It comes with short alligator clip test leads which have the typically too slippery inside vinyl boots, which makes opening the jaws more of a challenge than it should be.
- The screen fonts are a nice looking and clear sans-serif, some images show this as ugly serif.
- The screen info is somewhat more extensive and arranged somewhat differently than the images show.
- Pressing the "SET" button takes you to a single page that shows the Model, Serial, and software version.  Here you can set the language (Chinese | English), AutoOff (OFF | 5 | 15 | 30 | 60 min), Bright (backlight in 20% increments), PowerOn (Last | Default), Beep (Off | Key+Notice | Only Notice), Colors (White | Yellow | Cyan | Coral) - Yellow is almost unreadable, Cyan and Coral seem useful, I prefer White.
- Included is a shorting bar for calibration.  You stick it in and a long press of the "NULL" button starts the "Short" calibration, which takes about 45 seconds.  You remove it and a second long press starts the "Open" calibration, which also takes about 45 seconds.  A countdown from 10 to 0 on the screen shows the progress.  The meter automatically senses the shorting bar's presence or absence, so you can do the calibration in either order.
- Since the calibration takes a while to do, I was hoping it would be preserved through a power cycle.  It seems if you do SET | PowerOn | Last then it will be preserved along with all of your other test settings, otherwise you're starting over every time.  Of course you'll want to recalibrate if you switch between the direct insertion slots and any test leads.
- The direct insertion slots are actually 4-wire: the drive is on the bottom metal lip and the sense is on the top metal lip, and inserting a component lead shorts them together.
- You can buy inexpensive shielded 4-wire "Kelvin" test leads, but I think I'll be OK with the direct insertion slots and the short 2-wire leads that were provided with the meter.
- The LCR mode can be selected as AUTO, during which the meter tries to figure out what type (L | C | R) the component is.  Auto mode switching can be fooled if the test frequency is too low (e.g. it can confuse a larger inductor for a resistor) or too high (e.g. it can confuse a 10meg resistor for a capacitor).
- The modes have "Series" and "Parallel" sub modes which are selected with the AUTO/SER/PAL" button.  For inductance series mode seems most useful as the inductance value stays fairly constant with test frequency.  For resistance parallel mode seems most useful / constant.  For capacitance this doesn't seem to matter?
- The FREQ (100 | 120 | 1k | 10k | 40k | 100k Hz) selects the test sine frequency.  Going to 100kHz is more of a premium feature, and allows you to see how components behave over a more extended range.
- For the RANGE, "AUTO" seems to work the best in most cases.  For some reason the range display uses the Ohm symbol regardless of what is being measured.
- Pressing the HOLD button instantly freezes the displayed values, a long press shows running min, max, and average values at the bottom of the screen.
- You can use the up/down/left/right arrow buttons to select and change most of the measurement settings.  A bit confusingly, up/down are used to select, left/right are used to dec/inc the selected value.
- I couldn't find a use for the ENTER key?
- Pressing the NULL button offsets the displayed values and shows a delta icon on the screen, a long press initiates the calibration procedure.
- The SPEED (FAST | MED | SLOW) button controls screen updating and presumably trades response time for averaging time, so slower should give more consistent measurements.
- The LEVEL (300 | 600 | 1000 mV) button sets the RMS (1000mV = 2.5Vpp confirmed with my scope) test voltage, I suppose this should be set to max for the best SNR if not doing in-circuit testing.
- For some reason there is no BIAS button (though there is a suspicious gap on the front panel here) but you can reach the setting via the arrow keys which introduces a DC bias (0 | 100 | 300 | 600 mV) on the test leads.
- There is a dedicated capacitor button for electrolytics.  This sets the level to a fixed 600mV RMS and the offset to 1V, so the lowest part of the test sine wave is a bit above ground.
- Pressing the X/D/Q/theta/ESR button selects the secondary measurement below the main measurement.  ESR is probably the most useful, and I believe D is simply the inverse of Q.  I suppose one could compare differently constructed components of the same value to see which might have a higher Q, but D, Q, and theta aren't otherwise very meaningful without a continuously variable test frequency and knowing the SRF.
- The manual seems more useful than most, with charts and tables showing the accuracy of the various measurements and modes.  I didn't mess with any of the data logging features nor the included software.
- If I could redesign the user interface, I'd make the two columns of buttons align and correspond with the two columns of display data - as it is you have to hunt around for the correct button to press because the orderings are quite different, so I tend to stick to the arrow keys for navigation.  And I'd give it a dedicated BIAS button.  But this is nit-picking, I'm quite happy with it the way it is - it's a good value.