Can someone model a sub for me? ported 18 inch "Johnny sub" (willing to tip)
- By jesburger2
- Subwoofers
- 19 Replies
Attachments
One DAC with TDA1541A and tube output
For sale one DAC withTDA1541A and tube output.
Is built with audio grade components and it sound very good.
The cost is 450 euro+ 25 shipping in EU only.
For more details, please send me a PM.
Regards,
Vlad
Is built with audio grade components and it sound very good.
The cost is 450 euro+ 25 shipping in EU only.
For more details, please send me a PM.
Regards,
Vlad
TDA1541 tube output stage?
- By JacobGreen
- Tubes / Valves
- 19 Replies
Hi,
A few years ago I was building a music streamer based on a pc and TDA1541 board.
Changing the original switching power supply to linear one made a great improvement.
Splitting to two power supply - one for the DAC and one for the computer helped a lot!
Upgrading the OPAMP of the DAC to a higher quality ones mad a great improvement.
The sound is great, but a little "too compressed" as opposed to my DIY turntable. And here goes my question: do you think that building a tube-output stage instead of the om-amps can help? I rely love the sound of the TDA1541, and, yes, I know it is a very old cheep. 🙂 I'm building a new streamer, based on a Raspberry Pi, and there will be much less wires in the new version... Is it time for tube output stage?
Any advice or direction to a proper kit / schematics will be highly appreciated.
A few years ago I was building a music streamer based on a pc and TDA1541 board.
Changing the original switching power supply to linear one made a great improvement.
Splitting to two power supply - one for the DAC and one for the computer helped a lot!
Upgrading the OPAMP of the DAC to a higher quality ones mad a great improvement.
The sound is great, but a little "too compressed" as opposed to my DIY turntable. And here goes my question: do you think that building a tube-output stage instead of the om-amps can help? I rely love the sound of the TDA1541, and, yes, I know it is a very old cheep. 🙂 I'm building a new streamer, based on a Raspberry Pi, and there will be much less wires in the new version... Is it time for tube output stage?
Any advice or direction to a proper kit / schematics will be highly appreciated.
One Driver - Two Tweeters
- By Databass
- Full Range
- 22 Replies
"One Driver - Two Tweeters"
A few years ago, when I didn't really know what I was doing, I decided to build a replica of a DJ monitor I had owned in my youth.
The plans called for 150 litre enclosure housing a 15" woofer and two KSN 1025's (Yeah, I know)
So I choose an Eminence Delta-15LF-C
https://www.ltt-versand.de/en/sound...eminence-delta-15-lfc-15-speaker-500-w-4-ohms
topping out at 3200Hz and a pair of Monacor MPT-025's rolling in at 3000Hz
https://www.ltt-versand.de/en/sound...-tweeters/20312/monacor-mpt-025-piezo-tweeter
Seemed like a sensible choice at the time.
Now, I have a few questions but I will start with this one.
When it came to wiring this set-up, I had no diagram and the local PA "guru" of the group advised me to
"wire the tweeters in series.... Parallel to the cone"
So I did.... But ever since then every configuration of one driver and two tweeters I have seen are not wired this way.
Is my system wired correctly.....?
I am currently re-building this box but obviously because of the cut-outs I cannot really change the tweeters.
I am reading lots of conflicting information online about resistor values and crossover circuits for piezos but again,
all for just one driver and one tweeter.
I would like to quieten these screamers down a few dB's..... Would anyone have the proper schematic or some sound advice for this
seemingly unusual configuration……?
A few years ago, when I didn't really know what I was doing, I decided to build a replica of a DJ monitor I had owned in my youth.
The plans called for 150 litre enclosure housing a 15" woofer and two KSN 1025's (Yeah, I know)
So I choose an Eminence Delta-15LF-C
https://www.ltt-versand.de/en/sound...eminence-delta-15-lfc-15-speaker-500-w-4-ohms
topping out at 3200Hz and a pair of Monacor MPT-025's rolling in at 3000Hz
https://www.ltt-versand.de/en/sound...-tweeters/20312/monacor-mpt-025-piezo-tweeter
Seemed like a sensible choice at the time.
Now, I have a few questions but I will start with this one.
When it came to wiring this set-up, I had no diagram and the local PA "guru" of the group advised me to
"wire the tweeters in series.... Parallel to the cone"
So I did.... But ever since then every configuration of one driver and two tweeters I have seen are not wired this way.
Is my system wired correctly.....?
I am currently re-building this box but obviously because of the cut-outs I cannot really change the tweeters.
I am reading lots of conflicting information online about resistor values and crossover circuits for piezos but again,
all for just one driver and one tweeter.
I would like to quieten these screamers down a few dB's..... Would anyone have the proper schematic or some sound advice for this
seemingly unusual configuration……?
3-Way with Coaxial Driver & Horn + Tube Amp (Quad II, PreAmp)
- By Stonesilver
- Multi-Way
- 7 Replies
Hello everyone,
Having realized my dream of designing and building a 3-Way, Coaxial Compression Driver, and Horn speaker capable of playing down to 300Hz by 112.5dB, paired with a Closed Bass 15” (101 dB), I now have the opportunity to take my hobby more seriously. Previously, I was building DIY speakers based on blueprints from the internet.
I recently acquired funds, allowing me to delve deeper into the world of amplifiers. Despite having zero knowledge about amps initially, I've gradually gained understanding. My preference is to drive these speakers with a tube amplifier. I have an DSP + Active Crossover + Preamp+DAC Allinone Solution from the German high-end horn manufacturer Hornsolution with a Nadja Board.
Through some coincidences, I obtained a pair of Quad II with original tubes, together with a DIYTube Preamp (1 Phono (mm), 4 Line 1 Tape In). My plan is to implement a passive crossover for the coaxial compression driver, enabling me to drive them with one tube amp. I intend to connect the amp to the Hornsolution Nadja, and have it only cut the frequencies below 300Hz, and use the Tube PreAmp for the CD Player/TT attached to the Nadja.
Now, my questions:
A) For the bass, which will receive a low pass filter for the crossover, I'm uncertain about the right amp. Should I go for a Tube or Class A or D with more wattage?
B ) I've attached a picture of the amp. Can you confirm if the amp is wired to 8 ohms? What's your initial impression of its condition?
C) The preamp was inherited, and the person doesn't know much about it. Can someone assess the quality/value of the preamp? Was it worth the 300€?
D) What are your overall thoughts on my plan? Does it make sense?
Thank you all, looking forward to your thoughts.
Having realized my dream of designing and building a 3-Way, Coaxial Compression Driver, and Horn speaker capable of playing down to 300Hz by 112.5dB, paired with a Closed Bass 15” (101 dB), I now have the opportunity to take my hobby more seriously. Previously, I was building DIY speakers based on blueprints from the internet.
I recently acquired funds, allowing me to delve deeper into the world of amplifiers. Despite having zero knowledge about amps initially, I've gradually gained understanding. My preference is to drive these speakers with a tube amplifier. I have an DSP + Active Crossover + Preamp+DAC Allinone Solution from the German high-end horn manufacturer Hornsolution with a Nadja Board.
Through some coincidences, I obtained a pair of Quad II with original tubes, together with a DIYTube Preamp (1 Phono (mm), 4 Line 1 Tape In). My plan is to implement a passive crossover for the coaxial compression driver, enabling me to drive them with one tube amp. I intend to connect the amp to the Hornsolution Nadja, and have it only cut the frequencies below 300Hz, and use the Tube PreAmp for the CD Player/TT attached to the Nadja.
Now, my questions:
A) For the bass, which will receive a low pass filter for the crossover, I'm uncertain about the right amp. Should I go for a Tube or Class A or D with more wattage?
B ) I've attached a picture of the amp. Can you confirm if the amp is wired to 8 ohms? What's your initial impression of its condition?
C) The preamp was inherited, and the person doesn't know much about it. Can someone assess the quality/value of the preamp? Was it worth the 300€?
D) What are your overall thoughts on my plan? Does it make sense?
Thank you all, looking forward to your thoughts.
Help me identify feedback loop
- By Kyleag89
- Analog Line Level
- 4 Replies
Can someone tell me which resistors in this diagram are responsible for setting the gain? This is a LPF using t072 opamps. It would be great to know exactly which caps and resistors are setting the lpf. I have been struggling to read opamp diagrams and translate them to what I'm looking at on the PCB itself.
Bassinga 'Flatwounds' desktop TL template and tutorial
- Full Range
- 6 Replies
Hello all
My name is Randy. I have been floating around the forum for a while. Recently, I started work on a pair of desktop speakers using drivers without proper specs. Only relevant info supplied were power, dimensions and a response down to 32hz. I asked for help on the forum, and a number of folks helped me work out some viable options with enclosures to try these out with. One of the directions was a TL of around 20 odd litres. Along the discussions, I ran various iterations of the model in CAD to visualise what the work file would look like. During the course, I noticed that I had created a template to allow these quick changes and its implications. This template would not have been possible without folks taking an interest and educating freely and being generous with their time and expertise. I found this to be the case with both commercial and enthusiast members of this forum
Allow me to share this template back to everyone. I'll be glad if this is of use to you to either easily make boxes for messing around or develop to your commercial boxes. The template will be able to assist you to create not only sophisticated speaker boxes but anything really. I will be using this for making a tube power amplifier chassis too
I call this particular implementation of the stepped tunnels the "Bassinga Flatwounds". Please mention this name with your work, so people may treat my misconceptions kindly 🙂
The system
This is based around a 3D model using FreeCAD. The template allows easy manipulation of all fields to a novice with CAD to alter required areas. Base model is for a 6.5" Pioneer coaxial driver that I am set to trail. The model can be scaled down and 3D printed or CNC'ed for tiny format speakers. My print bed is 40 x 40 x 40 cm, so I have the real estate to print my full box should I wish. This does take a while even with a 1 mm nozzle and fast settings, as well as material costs. This template will help solve these issues and I will show a tutorial in a future instalment once I am ready to proceed with my desktop TL build
For the moment, I am curious to sees how it prints. I have another pair of unknown 3" full range speakers with the only info being 30wrms and 8 ohms printed on the back. This will be a test for the comprehensive tuning abilities that my CAD template offers pre- and post-printing. This is for another future instalment
Template features
Full or semi prints
Model is converted to a building jig that ends up as decorative trim for boxes too large to print whole
Simple build process that can used already prepared flat material like finished floor planks as well as foam and foam core and the usual ply and MDF. Just about any suitable material can be used
The jig takes care of all the assembly work and removes the need for making wooden box edges look good
Some glue, a screwdriver and a drill are all the tools needed to assemble the box to result in a commercial level finish
Speaker workshops can adjust to their requirements and sell kits (please mention cab name)
Adjustments can be made to the template as well as to already made boxes, the tunnels can be inserted with all sorts of things to change tapers and such as well as stuffing due to the easy access nature inherent in the design
In the next instalment, I will show how to navigate the template and carry out changes. Do not hesitate to point out any mistakes. If any CAD sharks want to assist, go for it 😀
The zip file contains the FreeCAD 3D template
Let's keep this thread to the construction and assembly discussions and the audio discussions to the original thread now renamed and linked below
https://www.diyaudio.com/community/threads/desktop-tl-template-audio-discussion.396170/
Thanks and regards
Randy Bassinga
My name is Randy. I have been floating around the forum for a while. Recently, I started work on a pair of desktop speakers using drivers without proper specs. Only relevant info supplied were power, dimensions and a response down to 32hz. I asked for help on the forum, and a number of folks helped me work out some viable options with enclosures to try these out with. One of the directions was a TL of around 20 odd litres. Along the discussions, I ran various iterations of the model in CAD to visualise what the work file would look like. During the course, I noticed that I had created a template to allow these quick changes and its implications. This template would not have been possible without folks taking an interest and educating freely and being generous with their time and expertise. I found this to be the case with both commercial and enthusiast members of this forum
Allow me to share this template back to everyone. I'll be glad if this is of use to you to either easily make boxes for messing around or develop to your commercial boxes. The template will be able to assist you to create not only sophisticated speaker boxes but anything really. I will be using this for making a tube power amplifier chassis too
I call this particular implementation of the stepped tunnels the "Bassinga Flatwounds". Please mention this name with your work, so people may treat my misconceptions kindly 🙂
The system
This is based around a 3D model using FreeCAD. The template allows easy manipulation of all fields to a novice with CAD to alter required areas. Base model is for a 6.5" Pioneer coaxial driver that I am set to trail. The model can be scaled down and 3D printed or CNC'ed for tiny format speakers. My print bed is 40 x 40 x 40 cm, so I have the real estate to print my full box should I wish. This does take a while even with a 1 mm nozzle and fast settings, as well as material costs. This template will help solve these issues and I will show a tutorial in a future instalment once I am ready to proceed with my desktop TL build
For the moment, I am curious to sees how it prints. I have another pair of unknown 3" full range speakers with the only info being 30wrms and 8 ohms printed on the back. This will be a test for the comprehensive tuning abilities that my CAD template offers pre- and post-printing. This is for another future instalment
Template features
Full or semi prints
Model is converted to a building jig that ends up as decorative trim for boxes too large to print whole
Simple build process that can used already prepared flat material like finished floor planks as well as foam and foam core and the usual ply and MDF. Just about any suitable material can be used
The jig takes care of all the assembly work and removes the need for making wooden box edges look good
Some glue, a screwdriver and a drill are all the tools needed to assemble the box to result in a commercial level finish
Speaker workshops can adjust to their requirements and sell kits (please mention cab name)
Adjustments can be made to the template as well as to already made boxes, the tunnels can be inserted with all sorts of things to change tapers and such as well as stuffing due to the easy access nature inherent in the design
In the next instalment, I will show how to navigate the template and carry out changes. Do not hesitate to point out any mistakes. If any CAD sharks want to assist, go for it 😀
The zip file contains the FreeCAD 3D template
Let's keep this thread to the construction and assembly discussions and the audio discussions to the original thread now renamed and linked below
https://www.diyaudio.com/community/threads/desktop-tl-template-audio-discussion.396170/
Thanks and regards
Randy Bassinga
cheap VU meter board and noise
Hello everyone, simple question, would this kind of build generate distortion or noise in the signal which feeds the (class D) amplifier (can dirty noise creep back from the vu meter board, which has a small amplifier, and a noisy 12v SMPS)?
My basic knowledge of electricity tells me "no", but I would like an expert advice from people more versed in electronics and signal processing 🙂
My basic knowledge of electricity tells me "no", but I would like an expert advice from people more versed in electronics and signal processing 🙂
Sure WONDOM APM2 DSP
Hi, I’ve just purchased an APM2 board to provide DSP functions between a preamp and power amp going in an active loudspeaker. I’ve worked on several projects before with SigmaStudio and Sure / Dayton ADAU1701 based boards before with success however the APM2 doesn’t seem to be behaving as I would expect. Picture shows test setup
With power connected I hear the input signal on the speaker from AD0 connection (also from AD1). I expected this as the board ships with a program already loaded, howeverPOTS d the on’t do anything which suggests the default program on the APM2 is not working.
With power off and the ICP programmer disconnected I still hear audio through AD0 connection – even though the board is not powered up. I wouldn't expect to hear anything as the unit isn't powered. (The audio is actually louder when not powered)
When programming Sigma studio indicates correct download of program to the board and the LED flashes on the AMP2 - however there no change in function.
Anyone know what's going on here?
With power connected I hear the input signal on the speaker from AD0 connection (also from AD1). I expected this as the board ships with a program already loaded, howeverPOTS d the on’t do anything which suggests the default program on the APM2 is not working.
With power off and the ICP programmer disconnected I still hear audio through AD0 connection – even though the board is not powered up. I wouldn't expect to hear anything as the unit isn't powered. (The audio is actually louder when not powered)
When programming Sigma studio indicates correct download of program to the board and the LED flashes on the AMP2 - however there no change in function.
Anyone know what's going on here?
Digital tape counter for R-R deck
- By Leoniv
- Analogue Source
- 20 Replies
The digital tape counter is designed for installation in reel-to-reel tape recorders. Many tape recorders have mechanical counters that are best replaced with electronic ones.
Counter features:
Tape move sensor
The transport of the tape recorder must have a rotating roller that is constantly in contact with the tape. The roller diameter can be from 30 to 60 mm, but you can add another range of diameters to the source code of the program. Reflective optocouplers are used as roller rotation sensors. Small size optocouplers are well suited, for example, CNB1302, ON2170, KTIR0711S, KTIR0811S, KTIR0821DS, ITR8307, TPR-105F, QRE1113. An chopper is installed on the roller, which should have 5 blades. Another number of blades is possible, but this number must be entered in the source code of the program. Two optocouplers are used, which allows you to determine the direction of rotation. Optocouplers are installed at an angle equal to half the angular size of one blade. As a result, signals shifted by 90 degrees are generated at the output of the optocouplers (quadrature signals).
Pulse shaper
The optocoupler pulse shaper circuit is taken from a Studer A820 tape recorder. The optocoupler signal is first amplified by a common-base transistor, then fed to a comparator (Counter_rev6_sch.pdf). The goal is to reduce the voltage swing at the output of the photodetector in order to speed up the recharge of parasitic capacitances. Such a scheme improves the shape of the pulses at a high repetition rate, which is observed in the rewind mode. To estimate the required frequency range, it is necessary to calculate the maximum speed of rotation of the roller. When rewinding, the tape speed can reach 10 m/s. With a roller diameter of about 40 mm, its rotational speed will be about 80 revolutions per second (4800 rpm). Considering the number of chopper blades, the maximum signal frequency from the output of each optocoupler will be about 400 Hz.
Quadrature decoder
In the simplest case, you can count the blades that pass over the sensors. This will be a 1X quadrature decoder. When one blade of the chopper passes over the sensors, 4 states of the output signals of the sensors are changed. It is advisable to use all these states to increase the accuracy of the sensor. This will be a 4X quadrature decoder. To implement such a decoder, the ability of microcontrollers to generate interrupts on any change in the state of port pins is suitable. This feature is available in the ATmega48/88/168/328 line of AVR microcontrollers. Decoding can be implemented in an interrupt. The easiest way to calculate the direction is by analyzing the previous state of one of the signals.
An important feature - the count +1 and -1 must occur at the same point. Otherwise, roller jitter near the switch point will cause an infinite increment or decrement. The text of the 4X quadrature decoder is below:
Optical sensors do not have a bounce, which happens in mechanical contacts. They can vibrate for a while, closing and opening. Optical sensors also have certain anomalies, but their nature is different. Near the switching point, the light flux changes smoothly, so the output is a signal with a fairly low slew rate. To form a digital signal with short edges, a certain threshold element is required. Any analog signal has noise and interference. Interference can have a variety of causes, ranging from electrical pickups and supply voltage ripples, ending with external illumination of variable intensity and mechanical vibration of the chopper, leading to modulation of the light flux. When the output voltage of the photodetector overcomes the threshold value, due to signal fluctuations at the output of the threshold element, a whole series of switching can occur, which will lead to an erroneous count. To prevent this from happening, the threshold element must have hysteresis. The logic inputs of the microcontroller also have hysteresis, but its value is small. It is better to use an external threshold element, for which the hysteresis value can be set arbitrarily. The dual comparator LM393 is well suited for this purpose.
Software processing
Each change in the state of the optical sensors triggers an interrupt. There will be 20 events per revolution of an chopper with 5 blades. In the interrupt handler, these events are counted in a small 8-bit counter. Periodically, the main program polls this counter and adds the count to the large 32-bit Events counter.
Based on the Events variable, the metrage value is calculated:
Metrage = Events * Diameter * PI / EV_PER_REV
Calculations are carried out in integer mathematics. To prevent error, the Metrage value is calculated in micrometers. The roller Diameter value is not a constant, but is calculated during the calibration process.
Based on the Metrage variable, the real time value for the selected tape speed is calculated:
Time = Metrage / Speed
Tape speed 9/19 or 19/38 is selected using an external 9/19 logic signal.
Timer1 based meter is used to calculate actual tape speed. The meter uses a capture interrupt - one of the optical sensor signal is applied to the ICP pin. When the meter starts, the software timer starts with the interval M_PRE_TM. If during this time no pulse comes from the sensor output, the speed is considered zero. The minimum measurable speed with these constant values is approximately 5 cm/s. When a pulse arrives from the optical sensor, the timer value is captured into the CapBeg variable, the measuring interval M_RUN_TM is started, and the counting of pulses begins. When the interval ends, the pulses are counted up to an integer number of roller revolutions (otherwise the mechanical inaccuracy of the chopper will lead to a measurement error). Time M_FIN_TM is allocated for the completion of the calculation. If during this time the required number of pulses does not arrive, the speed is considered zero. Based on the last pulse, the counter value is captured into the CapEnd variable and the duration of one roller revolution is calculated:
TurnTime = T_timer * (CapEnd - CapBeg) / Turns
Then, based on this duration and the diameter of the roller, the tape speed is calculated:
Speed = Diameter * PI / TurnTime
The roller diameter calibration procedure uses speed meter data. In order not to include the old roller diameter (which may have a large error) in the calculations, the measured speed value is not used. Instead, the duration of one roller revolution is used. To improve the accuracy of the calibration, the measurement is carried out CAL_CYCLES times, and the total time TimeSum is calculated. Then the diameter of the roller is calculated:
D = (TimeSum * SPEED_19) / (PI * CAL_CYCLES / T_timer)
The duration of the calibration is approximately 14 seconds, the measurement is carried out 10 times, each of them lasts 2 revolutions of the roller. If the calculated value of the roller diameter is not within the range of 30 to 70 mm, then a calibration error occurs. After calibration, the roller diameter is displayed with micron accuracy.
An unexpected problem
Overflow handling caused unexpected difficulties. The counter supports several different tape speeds, real time readings at each speed will be different. It also supports the display of metrage. The information capacity of the display in metrage mode is ±9999.99 meters, and in real time mode it is ±9 hours, 59 minutes, 59.9 seconds (±35999.9 seconds). At a speed of 38.1 cm/s this corresponds to a metrage of approximately ±13716 m, at a speed of 19.05 cm/s it is ±6858 m, at a speed of 9.525 cm/s it is ±3429 m. Obviously, when switching speed or switching metrage-time, when switching back, the value should return to the previous one. This means that the counter overflow must occur at the same point for all modes. But how can this be achieved?
At first I decided to limit the metrage range to ±2999.9 m, which is sufficient for any size reels with any tape thickness. But then the real time range was at a speed of 9.525 cm / s ± 8 hours, 44 minutes, 56 seconds (± 31496 seconds), at a speed of 19.05 cm / s ± 4 hours, 22 minutes, 28 seconds (± 15748 seconds), at a speed 38.1 cm/s ±2 hours, 11 minutes, 14 seconds (±7874 seconds). When indicating the time, the counter began to overflow at these rather unexpected points, and at different speeds at different ones. It looked a little eccentric.
Since the time indication mode is the main one, I decided to make the overflow point for this mode a round number. For different speeds, this point will be ±8.00.00.0, ±4.00.00.0, ±2.00.00.0. In the metrage measurement mode, the number, on the contrary, became ugly ± 2743.19.
On overflow, the maximum positive number should turn into the maximum negative number. Here, too, there were all sorts of miracles associated with rounding numbers. I even had to write a counter model on a PC in order to debug all situations related to its overflow.
How the tough guys solve this problem, I could not find out. With the accumulation of experience, instead of facilitating design, the understanding came that making any, even a simple thing, is very difficult.
Counter features:
- Real time display: hours.minutes.seconds (selectable tenths of a second)
- Tape length indication (selectable in decimeters or centimeters)
- Indication of actual tape speed (selectable in hundredths of cm/s or thousandths of cm/s)
- Roller diameter calibration for precise measurements (from 30 to 70 mm)
- Three speeds supported (9, 19, 38 or others)
- Two independent counters (for example, total time and lap time)
- Automatic saving of all values in EEPROM when power is turned off
- LED display 6 or 7 digits (selectable)
- Menu controlled via button or encoder (selectable)
- Port for the implementation of the locator (no software support yet)
Tape move sensor
The transport of the tape recorder must have a rotating roller that is constantly in contact with the tape. The roller diameter can be from 30 to 60 mm, but you can add another range of diameters to the source code of the program. Reflective optocouplers are used as roller rotation sensors. Small size optocouplers are well suited, for example, CNB1302, ON2170, KTIR0711S, KTIR0811S, KTIR0821DS, ITR8307, TPR-105F, QRE1113. An chopper is installed on the roller, which should have 5 blades. Another number of blades is possible, but this number must be entered in the source code of the program. Two optocouplers are used, which allows you to determine the direction of rotation. Optocouplers are installed at an angle equal to half the angular size of one blade. As a result, signals shifted by 90 degrees are generated at the output of the optocouplers (quadrature signals).
Pulse shaper
The optocoupler pulse shaper circuit is taken from a Studer A820 tape recorder. The optocoupler signal is first amplified by a common-base transistor, then fed to a comparator (Counter_rev6_sch.pdf). The goal is to reduce the voltage swing at the output of the photodetector in order to speed up the recharge of parasitic capacitances. Such a scheme improves the shape of the pulses at a high repetition rate, which is observed in the rewind mode. To estimate the required frequency range, it is necessary to calculate the maximum speed of rotation of the roller. When rewinding, the tape speed can reach 10 m/s. With a roller diameter of about 40 mm, its rotational speed will be about 80 revolutions per second (4800 rpm). Considering the number of chopper blades, the maximum signal frequency from the output of each optocoupler will be about 400 Hz.
Quadrature decoder
In the simplest case, you can count the blades that pass over the sensors. This will be a 1X quadrature decoder. When one blade of the chopper passes over the sensors, 4 states of the output signals of the sensors are changed. It is advisable to use all these states to increase the accuracy of the sensor. This will be a 4X quadrature decoder. To implement such a decoder, the ability of microcontrollers to generate interrupts on any change in the state of port pins is suitable. This feature is available in the ATmega48/88/168/328 line of AVR microcontrollers. Decoding can be implemented in an interrupt. The easiest way to calculate the direction is by analyzing the previous state of one of the signals.
An important feature - the count +1 and -1 must occur at the same point. Otherwise, roller jitter near the switch point will cause an infinite increment or decrement. The text of the 4X quadrature decoder is below:
Code:
#pragma vector = PCINT0_vect
__interrupt void Opto_Handler(void)
{
static bool preF1 = 0;
bool nowF1 = TCount::Pin_F1;
bool nowF2 = TCount::Pin_F2;
(preF1 == nowF2)? TCount::Count-- : TCount::Count++;
preF1 = nowF1;
}Optical sensors do not have a bounce, which happens in mechanical contacts. They can vibrate for a while, closing and opening. Optical sensors also have certain anomalies, but their nature is different. Near the switching point, the light flux changes smoothly, so the output is a signal with a fairly low slew rate. To form a digital signal with short edges, a certain threshold element is required. Any analog signal has noise and interference. Interference can have a variety of causes, ranging from electrical pickups and supply voltage ripples, ending with external illumination of variable intensity and mechanical vibration of the chopper, leading to modulation of the light flux. When the output voltage of the photodetector overcomes the threshold value, due to signal fluctuations at the output of the threshold element, a whole series of switching can occur, which will lead to an erroneous count. To prevent this from happening, the threshold element must have hysteresis. The logic inputs of the microcontroller also have hysteresis, but its value is small. It is better to use an external threshold element, for which the hysteresis value can be set arbitrarily. The dual comparator LM393 is well suited for this purpose.
Software processing
Each change in the state of the optical sensors triggers an interrupt. There will be 20 events per revolution of an chopper with 5 blades. In the interrupt handler, these events are counted in a small 8-bit counter. Periodically, the main program polls this counter and adds the count to the large 32-bit Events counter.
Based on the Events variable, the metrage value is calculated:
Metrage = Events * Diameter * PI / EV_PER_REV
Calculations are carried out in integer mathematics. To prevent error, the Metrage value is calculated in micrometers. The roller Diameter value is not a constant, but is calculated during the calibration process.
Based on the Metrage variable, the real time value for the selected tape speed is calculated:
Time = Metrage / Speed
Tape speed 9/19 or 19/38 is selected using an external 9/19 logic signal.
Timer1 based meter is used to calculate actual tape speed. The meter uses a capture interrupt - one of the optical sensor signal is applied to the ICP pin. When the meter starts, the software timer starts with the interval M_PRE_TM. If during this time no pulse comes from the sensor output, the speed is considered zero. The minimum measurable speed with these constant values is approximately 5 cm/s. When a pulse arrives from the optical sensor, the timer value is captured into the CapBeg variable, the measuring interval M_RUN_TM is started, and the counting of pulses begins. When the interval ends, the pulses are counted up to an integer number of roller revolutions (otherwise the mechanical inaccuracy of the chopper will lead to a measurement error). Time M_FIN_TM is allocated for the completion of the calculation. If during this time the required number of pulses does not arrive, the speed is considered zero. Based on the last pulse, the counter value is captured into the CapEnd variable and the duration of one roller revolution is calculated:
TurnTime = T_timer * (CapEnd - CapBeg) / Turns
Then, based on this duration and the diameter of the roller, the tape speed is calculated:
Speed = Diameter * PI / TurnTime
The roller diameter calibration procedure uses speed meter data. In order not to include the old roller diameter (which may have a large error) in the calculations, the measured speed value is not used. Instead, the duration of one roller revolution is used. To improve the accuracy of the calibration, the measurement is carried out CAL_CYCLES times, and the total time TimeSum is calculated. Then the diameter of the roller is calculated:
D = (TimeSum * SPEED_19) / (PI * CAL_CYCLES / T_timer)
The duration of the calibration is approximately 14 seconds, the measurement is carried out 10 times, each of them lasts 2 revolutions of the roller. If the calculated value of the roller diameter is not within the range of 30 to 70 mm, then a calibration error occurs. After calibration, the roller diameter is displayed with micron accuracy.
An unexpected problem
Overflow handling caused unexpected difficulties. The counter supports several different tape speeds, real time readings at each speed will be different. It also supports the display of metrage. The information capacity of the display in metrage mode is ±9999.99 meters, and in real time mode it is ±9 hours, 59 minutes, 59.9 seconds (±35999.9 seconds). At a speed of 38.1 cm/s this corresponds to a metrage of approximately ±13716 m, at a speed of 19.05 cm/s it is ±6858 m, at a speed of 9.525 cm/s it is ±3429 m. Obviously, when switching speed or switching metrage-time, when switching back, the value should return to the previous one. This means that the counter overflow must occur at the same point for all modes. But how can this be achieved?
At first I decided to limit the metrage range to ±2999.9 m, which is sufficient for any size reels with any tape thickness. But then the real time range was at a speed of 9.525 cm / s ± 8 hours, 44 minutes, 56 seconds (± 31496 seconds), at a speed of 19.05 cm / s ± 4 hours, 22 minutes, 28 seconds (± 15748 seconds), at a speed 38.1 cm/s ±2 hours, 11 minutes, 14 seconds (±7874 seconds). When indicating the time, the counter began to overflow at these rather unexpected points, and at different speeds at different ones. It looked a little eccentric.
Since the time indication mode is the main one, I decided to make the overflow point for this mode a round number. For different speeds, this point will be ±8.00.00.0, ±4.00.00.0, ±2.00.00.0. In the metrage measurement mode, the number, on the contrary, became ugly ± 2743.19.
On overflow, the maximum positive number should turn into the maximum negative number. Here, too, there were all sorts of miracles associated with rounding numbers. I even had to write a counter model on a PC in order to debug all situations related to its overflow.
How the tough guys solve this problem, I could not find out. With the accumulation of experience, instead of facilitating design, the understanding came that making any, even a simple thing, is very difficult.
For Sale Deluxe large aluminium chassis plus heatsinks (Maybe UK only)
- By passive420
- Swap Meet
- 5 Replies
Attachments
Help with JVC RC-Q50 boombox repair - how does the HOSIDEN HSC1466 AC socket work? (schematic/datasheet available)
- By scragnoth
- Everything Else
- 1 Replies
Hi, I tried to repair the year-long issue with my father's JVC RC-Q50 boombox (FM radio, CD, cassette). It has a loose contact inside, the D-type power plug had to be in a specific angle, otherwise the boombox wouldn't turn on / shut itself off, back to standby. I used to bend the plug up by stacking cassette cases next to the boombox, but it was annoying since the slightest bump would cut the power. I found nothing obviously wrong with the power socket after I got the PCB with it out of the boombox. I de-soldered the power socket, it's part number is HSC1466, made by HOSIDEN, I disassembled the part itself and then the mishap happened. I heard two small pieces hitting the tiled floor, I looked everywhere but I only rediscovered the mechanical switch / lever. At first I wasn't even sure how that lever had to be put back inside the power socket. How I re-installed it at the end looks similar to the photo of the part inside the data sheet.
The lever between the two contacts for the D-type power plug will be pushed down by the power cable. The pins look a bit weird, I assumed that the contact would be better if I would bend them apart a bit, maybe that was the whole fault after all.
The lever in it's upright position, after I separated the two halves of the power socket.
When the lever is pushed down by the power cable, the three pins will be bend down.
The solder points of the three pins from before, back view of the power socket.
Screenshot from the PDF file of the part's data sheet / schematic which I found on www.datasheet.hk
After I found out how to correctly install the lever, I soldered the AC socket back in and tried it. I got no power at all. The lever / switch energizes the standby power circuit as far as I understand, since I am missing a part the circuit wouldn't be energized anymore.
Unfortunately I don't understand what should happen to the three contacts, I think they should be connected somehow as soon as the lever is pushed down by the power cord. Can someone tell me by looking at the schematic? I also cannot order a direct replacement of the part, it seems that it is nowhere available anymore, I remember that my father had this particular boombox since the early 90s tho. It would be nice to be able to repair it.
Thank you in advance.
The lever between the two contacts for the D-type power plug will be pushed down by the power cable. The pins look a bit weird, I assumed that the contact would be better if I would bend them apart a bit, maybe that was the whole fault after all.
The lever in it's upright position, after I separated the two halves of the power socket.
When the lever is pushed down by the power cable, the three pins will be bend down.
The solder points of the three pins from before, back view of the power socket.
Screenshot from the PDF file of the part's data sheet / schematic which I found on www.datasheet.hk
After I found out how to correctly install the lever, I soldered the AC socket back in and tried it. I got no power at all. The lever / switch energizes the standby power circuit as far as I understand, since I am missing a part the circuit wouldn't be energized anymore.
Unfortunately I don't understand what should happen to the three contacts, I think they should be connected somehow as soon as the lever is pushed down by the power cord. Can someone tell me by looking at the schematic? I also cannot order a direct replacement of the part, it seems that it is nowhere available anymore, I remember that my father had this particular boombox since the early 90s tho. It would be nice to be able to repair it.
Thank you in advance.
Zaerix E81CC
- By mdr30
- Tubes / Valves
- 24 Replies
Wonder which factory manufactured this tube for Zaerix. It has a transparent disc with a strange pattern on the upper mica and is marked E74 on the glass. Four seams on top. The inner structure is askew, sloppily mounted or the result of bad handling. Doesn't sound bad, piqued my curiosity.
Attachments
what to do about this awkward listening space?
- Multi-Way
- 22 Replies
Any ideas on how to make this basement room work? One side is a solid wall and the other is open. It is currently just drywall. The room is intended as a social space for my kids to bring friends to, and for watching movies as a family on a projector we just bought (Benq gp500. My god the image quality is great!). There will be a thick carpet, and a large squishy L-sofa. I can probably manage some things on the walls to deal with reflections but it wont be extensive. As much as I love audio, the family function is more important to me so I just need to figure out how to manage the compromise.
Now that i have a house again, I'm can resume building speakers, and I can do pretty much anything. I have some conventional box speakers I want to get finished but I've wanted to do an OB along the lines of linkwitz for a long time and I hope that will be soon.
I'm under the impression that OB interacts less with the room but I wonder if this is beyond the scope of what it can improve on. Should I look at this path or constant directivity or some other format, or should I accept that this a fools errand and make do.
I'll eventually get around to all sorts of builds but It would be best to get something working well in this room first to keep everyone happy. Would love to hear some thoughts.
Now that i have a house again, I'm can resume building speakers, and I can do pretty much anything. I have some conventional box speakers I want to get finished but I've wanted to do an OB along the lines of linkwitz for a long time and I hope that will be soon.
I'm under the impression that OB interacts less with the room but I wonder if this is beyond the scope of what it can improve on. Should I look at this path or constant directivity or some other format, or should I accept that this a fools errand and make do.
I'll eventually get around to all sorts of builds but It would be best to get something working well in this room first to keep everyone happy. Would love to hear some thoughts.
X-Wing open baffle - 2 x 15"
This weekend I had a chance to listen to a nice 2x15" open baffle, that had a large curved front horn and a conical extension on the back. It sounded great and looked very cool. Therefore I would like to put together a test baffle inspired by it.
It would be really easy to build. What I do not like on a H frame are the parallel walls, here, only the top/bottom (not shown) would be parallel.
My question is, would this work? Would the baffle width be counted including both sides of the wings or would the bass frequencies ignore the inner triangle and just bend to the back?
It would be really easy to build. What I do not like on a H frame are the parallel walls, here, only the top/bottom (not shown) would be parallel.
My question is, would this work? Would the baffle width be counted including both sides of the wings or would the bass frequencies ignore the inner triangle and just bend to the back?
Help with Parallel notch filter for small full-range speaker
- By Chasmo57
- Full Range
- 16 Replies
Attachments
PT2258 volume controller distorts signal. What am I doing wrong?
- By jeroen94704
- Analog Line Level
- 2 Replies
I'm trying to build a volume controller for a line level signal using the PT2258, which is a dedicated 6-channel volume controller IC. This is the reference circuit from the datasheet:
I built this circuit on a breadboard, and while it's functional there is a weird distortion of the signal. Here's a screenshot of what a sinewave (~300Hz) looks like coming in (yellow) and going out (blue), with the attenuation set to 0 dB:
Note that both signals are DC coupled, so it's clear the outgoing signal gets a DC offset from the PT2258. It also looks like the outgoing signal gets clipped at the bottom somehow.
In any case, I don't understand what I'm doing wrong here. I've rechecked the circuit I built several times and checked all the connections are good. So if anyone has any ideas what's going on here I'd love to hear about it.
Thanks!
I built this circuit on a breadboard, and while it's functional there is a weird distortion of the signal. Here's a screenshot of what a sinewave (~300Hz) looks like coming in (yellow) and going out (blue), with the attenuation set to 0 dB:
Note that both signals are DC coupled, so it's clear the outgoing signal gets a DC offset from the PT2258. It also looks like the outgoing signal gets clipped at the bottom somehow.
In any case, I don't understand what I'm doing wrong here. I've rechecked the circuit I built several times and checked all the connections are good. So if anyone has any ideas what's going on here I'd love to hear about it.
Thanks!
Crossover design why resistor has jumper on both sides
- By Kamranbits
- Multi-Way
- 21 Replies
Attachments
TA-242 Smart amplifier
- Swap Meet
- 0 Replies
I have a Smart TA-242 amplifier
DH19c boards I believe they are the same boards used on the DH 500 but with a lower power supply.
One channel is not working, no fuses are blown I believe the output gets are open. Need to be replaced.
strong amplifier!!!!
I'm moving I can't take it with me and I have no time for repairs. I don't want to throw it out.
Make me an offer. drdanrss@gmail.com
DH19c boards I believe they are the same boards used on the DH 500 but with a lower power supply.
One channel is not working, no fuses are blown I believe the output gets are open. Need to be replaced.
strong amplifier!!!!
I'm moving I can't take it with me and I have no time for repairs. I don't want to throw it out.
Make me an offer. drdanrss@gmail.com