I’ve been studying the past couple of months some ways to improve the field coil performance and in a way to make my tools better, to sharpen them. My goal is to get the most flux density in gap for a given current in field coil.
I tried many magnetic circuits meant to take advantage of the shape of a field coil because being a solenoid it will work better having a length per diameter ratio above unity. Works better means a more uniform field in the core.
I managed to get a simulated result of 1.5T in a 1mm wide and 12mm high gap using standard 1010 steel and a field coil consuming 7.231 Watts!
The field coil is made with 1 mm EC wire so it will be able to handle a lot more current. You will get more flux density this way but the curve will get more and more bell-shaped.
As you can see for 1 Amp you get a very nice increase in flux density. Its not as linear but the difference is small.
There is a balance, hard to achieve, between linearity and maximizing flux density. And this has to do with the magnetic circuit and saturating the pole plates. I usually want to saturate them right at the gap. This way it’s less prone to modulation.
In placing the saturation area, the central pole piece and the top plate geometries have a large role. Always remember that the thin parts of steel saturate faster and zones with transitions from one dimension to another or sharp angles are also prone to saturation.
Below you can see a motor with a linear flux density curve over a 20 mm travel. The central pole piece is of the same diameter as before. You can see that making the top plate thicker killed the saturation around the gap. To bring it back, we must use a larger diameter for a voice coil. But more on that in a future article.
Do you remember BOSS Slow Gear pedal? If your a guitarist you most likely do or at least you’ve heard of it. It was a great pedal sold from 1979 to 1982 and it was made in Japan. The pedal would cut the attack of your notes giving a swelling sound. It god famous for making the guitar sound kinda like a violin.
I always liked that effect and i even made a clone a few years back. It is based on a 2SK30 JFET and it was a pain getting these transistors. It was a lot of fun though and i though i should make a Project Ryu swell effect pedal and so LAGGER was born!
Recently i worked on a few projects with LM13600/LM13700, one of them is a nice noise gate / compressor unit which i will present at a later date, and i really like the VCAs that can be built with these chips.
To cut the attack of a note and then swell the volume basically we need a triggered fade in effect. This means that we need to control our VCA with a rising voltage using what i call a ramp generator.
Below you can see the block diagram of the Lagger:
The input is fed into an ADC channel to be rectified and averaged in order to detect when a note is played. Once it is detected, the ramp generator is triggered and provides the control voltage for the first VCA.
Since LM13600/LM13700 is a dual amplifier the second one is configured as a VCA with manually set control voltage. In the picture below you can see how the circuit works. The top signal is the input signal, the middle signal is the output of the ramp generator and the bottom signal is the trigger.
There is a problem with using the ramp generator circuit this way. The capacitor is discharged too quickly when the trigger is interrupted and this causes an audible thump noise when trigger goes off. Looking below at the schematic we can see the discharge current goes through CE junction of Q1.
We can lower this current by inserting a resistor between ground and Q1’s emitter but in our specific application that will cause an offset and the output will not be totally silent in absence of input signal.
Another way to solve the problem is by paralleling a capacitor with R3 (Q2’s emitter resistor) This will cause a fade out effect and eliminate the thump noise.
Below you can find the schematic for the Lagger:
U5 shows as TL071 but you need an opamp with higher output current sink capability. Something like HA17358 with 50mA capability is good:
Trigger for the ramp generated is created when the microcontroller detects a signal from guitar. In my last article i have explained a way to rectify and average an analog signal using ADC and software. If the input level is higher than a set threshold level then ramp generator is triggered.
In the first units the middle pot was used to set a sustain period but that was changed to sensitivity control as it proved to be much more helpful.
J1 is a push-button which will generate an interrupt for the microcontroller and provide a true bypass via the SPDT relay.
You will notice some unusual supply voltages. For example the microcontroller’s Vdd is set to GND and Vss to -5V. This is done in order to provide correct trigger levels and avoid using other active components to shift the level.
Below you can see the PCB for the unit:
Here are some pictures with Project Ryu Lagger:
Here is a short video with the unit in action:
I will be supplying the hex file for the PIC18f1320 microcontroller in my next newsletters so if you want to built the unit and your not a subscriber yet please use the top right form to subscribe.
Also in my newsletter you will find offer for kits and complete units for those who don’t do well with electronics.
In a home recording environment equipment often doesn’t come in a large variety because of either limited budget or limited space… or both. I am presenting here an balanced attenuator which comes in between fixed gain preamp and recording device/soundcard.
The purpose of this device is to adjust the level and monitor it to prevent saturation of the next stage in the recording chain. It uses a L-pad followed by a balanced buffer stage. The attenuation steps are 0dB, -3dB, -6dB, -9dB and -12dB.
The meter section uses a microcontroller with a 10 bit ADC. It monitors both polarities of the signal and detects the peak within a frame of 1000 samples. ADCs samples the signal every 12us.
In the schematics above J7 will be used later on future revisions to indicate symmetry in the balanced signal.
J3 will connect the attenuation selector.
There are certain modifications i did the initial schematic tho. R7, R8 must be of greater value. At least 100k. With 100k you will get a -0.8dB signal at 0 dB setting.
The circuit is pretty straightforward, you can use any quad opamp chip for U3 as long at it operates from a 9V single supply. I do recommend a FET input opamp if R7 and R8 >= 100k as they tend to have lower noise than bipolar with high input impedance.
U2 i used a LM324 as it allows operation at 5V single supply. U2:C is used to bias the ADC inputs at 2.5V.
I will explain how to rectify the signal with a PIC in a future project but you will find the hex file for this one at the end of this article.
Here are the PCB drawings:
pcb top layer
pcb bottom layer
pcb top silk
Warning: NONE of the PCB images are mirrored!
Here is how the PCBs turned out:
I used a rotary switch to select the attenuation levels. For 0dB you can just omit R15. Use POT2 to calibrate 0dB on your meter. I usually set it to 0dBV.
After building the first unit i made some measurements. I used 5% tolerance resistors so i wanted to see if the attenuation levels are correct. Here are the results:
-6dB and -12dB settings are about 1dB off and it seems -3dB setting is also a bit off. For -12dB i soldered a 47k resistor in parallel with the 6k8 one and for -6dB i soldered one 100k resistor in parallel with the 22k. For -3dB i soldered a 470k resistor in parallel with the 47k one. Results were much better:
I will be making kits available for this device with PCBs and programmed microcontroller as well as fully built units. Please subscribe to receive more details about this offer in the next newsletter.
Recently i have finished a pair of field coil mid-bass drivers. The driver has a Fs of 50Hz and a frequency response up to 7kHz. It is a dual voice coil design features a 160mm diameter motor with a max of 1.4T in the gap, Ryu spider, non-treated paper cone, triple roll cloth surround and bullet shaped phase plug.
The voice coil construction was presented in an earlier article here. Inside the gap set at around 1T flux density, the voice coil inductance was measured at 0.52mH @1kHz and 0.15mH@10kHz.
The high compliance of the Ryu spider was measured at around 3.4mm/N for this mid-bass and makes the compliance of the cloth surround the determinant factor in the system’s total compliance value.
Below you can see the impedance measurement and frequency response of these mid-bass drivers. The frequency response is measured with the voice coils connected in parallel. you can see a nice constant 6db/octave rise in response. You can use this characteristic when designing the loudspeaker if you consider baffle step or horn loading.
I have tested the unit and played Isao Tomita’s Snowflakes are Dancing and The Police Synchronicity. The sound is thin with the voice coils connected in parallel and no baffle but with the voice coils in series even without a baffle you can reach a good tonal balance. This indicates the unit can be used successfully in an Open Baffle system or in a horn loaded enclosure with the voice coils in parallel.
For a while, we have been discussing on diyAudio the influence of the spider. Ideally we want a low mass, low radiating area compliance for this. Also a very high stiffness towards side-to-side movement to avoid rocking modes.
Work on Project Ryu continues with a new model. This one will be a mid-bass by design but as always i will try to push the limits in bandwidth and sensitivity. The new cones are a bit on the heavy side weighting 18 grams but i think it will balance out very well in the mass controlled region.
The 12 inch cone has a curved profile and concentric rings to reduce breakup effects.
Voice coil was designed using my Excel spreadsheet tool. You can find a link to download it at the end of the article. Double coil, each with estimated 8 ohm impedance allowing for 4-8-16 ohm operation.