electronics engineering notes

Fixing Bose Quietcomfort 35

2019-06-07T08:31:00.000-07:00

Short version:
Bose Quietcomfort won't turn on - open it and check if a plastic piece sitting on the switch is intact.
Bose Quietcomfort won't enter pairing mode - it's possible that the switch on the PCB can't reach position responsible for entering pairing mode. Open the headphone and check if moving the switch directly helps, if so, continue reading.

Headphone won't turn on when the switch is moved.
The problem was this small plastic part which is visible from outside as a switch and which moves the actual switch soldered on the PCB. To fix it you will have to open the right headphone. Start with gently removing the ear cushion and the fabric inside, then undo three screws:

Without screws aluminium cover can be removed. The actual switch is highlighted:

From the other side of the aluminium cover you see a plastic part which should sit firmly on the switch. In this case it is broken:

Find two pieces of the broken part and glue it together, I used two compound epoxy glue.

It's not pretty but it holds well.

Bose Quietcomfort won't enter pairing mode.

The mechanical tolerances of the switch, this plastic piece and aluminium cover are so tight, that it can happen that the headphone will not enter pairing mode. You move the switch but it neither enters pairing mode nor removes paired devices. The reason is, that the oval hole in the aluminium cover is to small and the switch will not contact in the "bluetooth" position. I extended this hole. It was around 8,9mm long but I extended it by approximately 0,75mm towards the bluetooth logo. I used a round file for this:

It is barely visible that the hole is longer:

But this bigger hole allows reaching the final position of the switch.

Fixing VOX AD60 VT guitar combo

2019-05-19T07:59:00.000-07:00

The VOX AD60 VT blown the main fuse when turned on. I received it with a comment that someone has already tried repairing it.
After opening the housing it was obvious that someone has already played with transistors in amplifier stage.

For test purpose I used two power supplies with current limitation to supply the power stage. I connected them to points marked with yellow boxes on the circuit diagram below. After turning power supplies on, I observed that without input signal both power transistor (Q13, Q16 - blue ovals) are turned on completely. My power supplies entered constant current mode. I removed both power transistors (Q13, Q16 - blue ovals) and turned supply again, this time I observed that R87 (green box) overheated. I measured voltages across R84 and R79 (blue boxes) and realized, that Q12 and Q15 (green ovals) are fully turned on, while they shouldn't. While inspecting pairs of transistors Q5-Q6 and Q7-Q8 I noticed that in each pair there are different types of transistor, while they should be actually the same. I swapped Q6 and Q8 as they should be according to the circuit diagram and after that the amplifier (still without power transistors) draw approx. 20mA. I soldered power transistors back and tested the amplifier. It works well.

The problem was, someone was trying to fix this amplifier or to test transistors but mixed them up while putting them back in the circuit. This caused a shoot through current, the power transistors survived that but the fuse tripped.

On the very beginning, when I observed that there is high current flowing from both positive and negative supply rail, I thought it is something with phase inverter, and not with upper or lower power stage. I did once mixed up transistor while fixing a power stage and after that I can say, I learned my lesson and this time was able to find the problem quite fast.

EMC Troubleshooting speaker relay matrix

2019-04-03T06:25:00.003-07:00

There is relay a matrix designed to switch between active and passive speakers in a music shop. There are actually two matrices, each in a different showroom. It has been observed that one matrix switches speakers almost inaudible while the other makes a quiet but annoying click. Here is a screenshot of a signal recorded at the output of the speaker relay matrix. This is matrix switching line signal for active speakers so it was possible to record it directly:

Recorded distortion on the output of the matrix, at the beginning and ant the end there are disturbances which I believe are 4 bytes digital transmissions coupled on the audio signal.

Because of the location and surrounding equipment only one rack with a matrix can be opened, namely the one which generates audible disturbances, it looks like this:

Wiring arrangement in the matrix

The matrix consists of daisy-chained modules, each module is switching one input signal to 1-of-8 output jack sockets. A single module can be simplified to the following:

Single module of the relay matrix

In reality a single module is more complicated (I assume GND is switched too) but it describes the idea. A complete matrix is built in two rows and wired like this:

Current wiring of 10 modules in the matrix (outputs not shown because they are shielded on the other side of the panel)

See how many times digital signal cable is crossing and running parallel to analog signal wiring. I believe there is magnetic coupling between digital and analog lines. To make it worse, both are wired without shield, such arrangement is actually prone to cross-talks and requires longer cables. I suggested changing arrangement to something like this:

Improved wiring of 10 modules in the matrix

In the improved wiring concept there are shorter cables, and there is no unnecessary zigzag. I believe this will remove unwanted disturbances on the switched analog signal. Rearranging is possible because every module has its own configurable address, I am not even sure what exact type of the digital transmission is used. Maybe daisy-chain order can be changed even without readdressing modules. I will update this post when it's done.

Fixing Dell 2400MP projector with power supply PA-4331-1D-LF

2019-03-14T01:20:00.000-07:00

Power, temp and lamp control lights are continuously on when projector is turned supplied. Nothing changes after pressing power button. In the user manual there is no description of such a combination of control lights.

Opening the projector:

There are silk screen descriptions on the PCB so it was easy to identify and measure voltages generated by the power supply PCB, the 5V net looked like this:

Which is mean that the power supply is not able to regulate voltage. It looks like a single cycle of the flyback is causing overvoltage and it leads to power supply shutdown and another try in couple of milliseconds when voltage drops to 4V. Visual inspection of the C206 5V output filtering capacitor reveals that the capacitor is swollen.

This capacitor has been replaced with a new low ESR capacitor but the power supply didn't start either. This particular power supply turns out to be very sensitive to correct filtering capacitors selection. After replacing the old capacitor with 3 electrolytic hybrid ultra low ESR capacitors the projector started working normally:

The C602 capacitor value is not critical (as long as it is at least 3300uF) but low ESR is.

I suppose that high temperature in the projector accelerated electrolytic capacitor's wear-out. Most of devices I repair is supplied by a flyback power supply and most of flyback power supplies suffer from issues with output filtering capacitor which usually works in high temperature, high current and high frequency.

Seeing problems with electrolytic capacitors another time I decided to buy a meter UT612 RLC meter:

I like price-quality ratio of UNI-T devices but looking at specs reveals, that the higher frequency you use for a measurement the lower capacitor values can be tested. In case of the projector with a capacitor of 3300uF I was able to measure ESR with highest frequency of 120Hz. For a flyback power supply working with tens of kHz this measurement should be performed with at least 10kHz to reflect real conditions capacitor has to deal with.

Fixing tube amplifier Suprem Concort 80

2019-03-11T05:41:00.002-07:00

The amplifier came with a note that it doesn't turn on. Control lamps remain dark after flipping the switch. The reason for that was damaged switch. It made a click noise but there was no connection. Establishing reliable connection requires force to applied on the lever in side direction, then the control lamps lit but there was no signal coming through the amplifier. After flipping the standby (Vollbetrieb) switch, one could hear that the power stage was working because it wasn't complete silence, it was typical quiet noise of a working tube amplifier.
For this amplifier there is no circuit diagram available, but there is one very similar (Superem Luxor 40 https://elektrotanya.com/suprem_luxor_40_sch.pdf/download.html ) which except for number of inputs and tubes reference designation is the same. I checked heater resistance of all tubes but they all appeared fine.
Keeping in mind that tube amplifiers require connected speaker I measured all supply voltages including bias and they were correct.
I observed that one of the ECC83 tubes look different, see the middle tube on the picture:

See milky deposit in the top part of the tube which lost vacuum, it should be uniform, shiny surface.

In a normal tube there is shiny surface on the inner part of the glass, above a getter. (see https://en.wikipedia.org/wiki/Getter ), the middle tube on the picture has no shiny surface but a milky deposit. This is a sign that the tube has lost vacuum, and that particular tube drives the end stage of the amplifier. This explains why the amplifier was mute. I have no vacuum tubes in my workshop so I shuffled other tubes in the way that there was no signal for vibrato but all input channels should work. And it works. It sounds pleasant with a bass guitar. I ordered replacement tube and will build it back together soon.

Fixing Allen & Heath Zed 420 mixer

2019-03-06T00:30:00.000-08:00

The mixer was turned on 24 hours a day for around two years. After that time once its LEDs were not as bright as in the past and the signal was not getting through the analog channels. The mixer was returned from the service with a note, that power supply was damaged and that a price tag for the power supply makes it unprofitable to repair. After that, the owner brought the mixer to me. I observed that output filtering capacitors, right behind rectifying diodes are swollen. It was clear indicator that their lifespan was exceeded. See swollen capacitors, especially the one marked as (2):

I replaced those with identical capacity and voltage, low ESR capacitors and it worked again.

Fixing acoustic research Phantom Sub 12D

2019-03-04T03:08:00.000-08:00

In the subwoofer there is one externally available fuse and two built in fuses which protect symmetrical low voltage supply for the analog path and the amplifier.
The problem was that internal fuses blew immediately after turning power supply on.

This is a D-Class amplifier and there are some descriptions on the web that in this case output filtering capacitors C16 and C17 should be replaced what the owner already did but fuses kept on blowing. I received the Amplifier without speaker housing and took the amplifier PCB out:

Type code of key components on the PCB are removed but it turns out that this PCB is exactly the same as the one used in JBL SCS150SI, SCS160SI and SCS180.6S subwoofers.
The circuit diagram and service manual are available online.

With the circuit diagram available I noticed, that the owner replaced C16 and C17 as they were damaged (you can see corrosion on the amplifier PCB resulting from leaking capacitor's electrolyte) but there were more components requiring replacement, in this case it was Q10, Q11 and Q5. I assume that output filtering capacitor C16, failed to short circuit because of electrolyte leakage what caused Q10 and Q11 to work with a short circuit as a load, this damaged that MOSFETs. Q11 was failed open and D10 failed as a short circuit between gate and drain what caused Q5 to fail. After replacing all three mentioned transistors, the amplifier started to work.

Fixing Ebike battery, resetting OZ890

2019-02-27T01:20:00.002-08:00

My Ebike's range dropped significantly. I took the battery apart and and found 7 pouch Lithium-Polymer cells and a BMS printed circuit board based on the OZ890. Every cell consists actually of two smaller cells connected in parallel so the battery architecture can be described as 7S2P.

Firstly I noticed that there is one cell with unusually fast voltage drop, when the cell voltage dropped to 3,2V the BMS turned the complete battery of to prevent damage. Removing kapton foil revealed a crack on the battery pouch:

Replacing complete battery is not an option, 6 of 7 cells are still fine. I wasn't able to find exactly the same replacement cell so I decided to cut the corners and built a cell combination which will fit into the place of the damaged one. 6 pieces of 18605 cell fit perfectly and if they are connected 2S3P then their voltage matches original li-po cells and their capacity is even higher than rest of the cells. I ordered 6 new 18605 cells already welded in the desired shape and put it instead of the damaged cell.

My strategy was to fix the battery by replacing damaged component and I believe that as long the new cell has got the right voltage and at least the same capacity (ideally a bit higher) then it has to work.

OZ890 BMS had to be reset to accept the battery. In this case one of the middle cells were replaced, it means, during the complete operation OZ890 was powered up continuously and it detected 0V on one of the cell which was replaced. I wasn't able to reset the OZ890 with original o2micro USB I2C interface, but disconnecting the BMS from the batterypack completely did the trick. After reset the BMS allowed normal charge and discharge procedure. This means, that there is no Coulomb counting in the BMS and that there is voltage supervision only. I don't expect this trick to work with more sophisticated BMSes. Fixing this battery wasn't all that difficult. It has been couple of month since I did it and the battery still serves its purpose.

Fixing Rocktile GB-15 bass guitar combo

2019-02-26T00:13:00.002-08:00

I bought a damaged super-cheap bass combo for practicing at home. Description provided by the previous owner mentioned that signal overdrives and that it sounds unpleasant.

After turning this on for the first time I noticed that:

there is unwanted overdrive/distortion
distortion changes over time
after couple of second there is no sound

As always I checked supply voltages first - they were fine. Power stage built around TDA2030 worked too so I concentrated on the preamplifier. Replacing opamp didn't help so I had to understand how the circuit works. There was no circuit diagram available so I reverse-engineered it from the PCB:

After connecting a signal generator to the amplifier I noticed that it operates normally for a while after touching pin number 2 or 3 of the 4558 operational amplifier. Using oscilloscope I observed that voltage on the pin number 2 of the 4558 is slowly rising and when it saturates the circuit doesn't amplify anymore:

I knew that this has got to do with an input bias current of the operational amplifier. In one of my favorites books "The Art of Electronics" I read that input bias current problems become visible some time after turning the circuit on, when impacted capacitors charge, for low input bias current amplifier it can be even minutes after turning a circuit on.

According to circuit diagram every input pin of the operational amplifier has got a DC path designed for input bias current. I measured R7 resistance on its pins and it appeared to be fine but by measuring resistance between pin 3 of the 4558 operational amplifier socket I noticed that the R7 is not there. I inspected the PCB and discovered that there is a crack on the PCB between the R7 and GND. On the movie below you will see (sound is important for that movie).

After fixing that PCB crack with a drop of soldering tin the amplifier started working

One more note here is the fact, that the input jack socket shorts input to GND when there is no plug inserted. It is so to avoid noise when there is no guitar connected. But to connect a signal generator to the PCB you have to be aware about that and to plug something in to open the input socket.

Fixing DENON DCD-510AE CD player

2019-02-24T09:51:00.000-08:00

While visiting friends I heard their complaints that they have a CD player which keeps on resetting. It wasn't able to eject the CD (there was one inside). On the display it was continuously saying "READING". From time to time this player was able to start playback but it ceased when pressed "next". I was told that the CD module was already replaced twice with no result. My first thought was that it has got to do something with supply voltage. Circuit diagram is easy to find on https://www.manualslib.com/manual/1115299/Denon-Dcd-510ae.html#manual .
I started by checking 5V and 8V net, see sample trace below. I also observed (you have to believe me because I have no waveforms for that) that 8V was dropping to lowest values when a skip track or eject button was pressed. Spinning the CD loaded the 8V and 5V net significantly too.

Blue trace i 8V and yellow is 5V.

5V and 8V are both generated by linear regulators and it is obvious that 8V is overloaded so much that 5V falls below reset threshold. I inspected C914, IC93 and C916, they appear fine. I measured current consumption on 8V (approx. 1A max) and supply transformer voltage. According to my crude calculations C914 value is way too low, and I believe that the power supply is calculated without margin. I don't have 22000uF capacitor and I didn't want to order one but I realized that I have a 230V to 5V 1A switching mode power supply and keeping in mind that the main load on the 8V net is the 5V regulator I decided to add an additional power supply to provide 5V from another supply.
I removed IC94 and connected additional power supply (see yellow rectangle) parallel to C917. This new power supply is connected in parallel to the primary winding of the main transformer of the CD player T901

It's not ideal, I could try experimenting with the original power supply and low ESR high capacity alternatives for the C914 but I had an extra switching mode power supply lying around so I used it. It works just fine.

Ni-Cd battery deep discharge protection circuit

2018-05-08T11:39:00.000-07:00

I have a LED lamp in my workshop which I supply from a recycled 12V 3Ah Ni-Cd battery. To protect the battery from a deep-discharge I built a simple circuit from what I found around. The circuit is built into the lamp and starts to beep when the supply voltage drops to 10V. The beep means that the battery has to be replaced.

The potentiometer PR has to be adjusted: Just set the desired voltage at which the circuit has to beep and then find a position of the PR at which the circuit starts to beep. The TL431 is a shunt voltage regulator and will pull down the pin 1 of the CD4093 when the voltage on its pin 1 is higher than 2.5V. One gate of the CD4093 forms an oscillator running at audible frequency. 3 unused gates have their inputs tied to GND. Piezzo is a of a passive type, it doesn't have built in generator.

The circuit has been built a piece of prototype PCB.

Voltrcraft LSP-1403 / Maplin N27GG / Manson SSP-7080 / Extech 382260

2016-09-12T12:05:00.000-07:00

I bought a Voltcraft LSP-1403 power supply on ebay. I intentionally choose the broken one so I could challenge myself and try fixing it. This power supply is a product of Manson but it's sold under at least three other brands which I named in the title of this note.

Voltcraft LSP-1403, Maplin N27GG, Manson SSP-7080 and Extech 382260

The power supply was turning on, display lit and showed reasonable numbers which could be altered by turning knobs, but it didn't provide any output voltage.

Block diagram of the LSP-1403 power supply

One look at the PCB shown that Q2, U4 R26, R27 and D9 were burned component. I replaced those damaged components and added a light bulb in series, to prevent a small explosion in case I overlook any other damaged/shorted component. Additionally, all the time the power supply was connected to mains over an isolating transformer, this way I was able to connect an oscilloscope to whatever point in the circuit I wanted.

Part of the circuit diagram of the main power supply

After putting this all together power supply provided correct output voltage but after loading it a light bulb lit up which was a sign that something is going wrong with the circuit.

I measured voltage on a shunt resistor R28 ant it turned out to look like this:

Voltage on R28 before fixing the power supply (low output current)

Voltage on R28 before fixing the power supply (higher output current)

Which is wrong because of too high current value and too steep rise. The UC3843 controller was detecting this high current and was turning the MOSFET off to prevent damage, this was the reason the power supply wasn't able to provide demanded output current. Such a steep current rise in the primary winding of a transformer is not possible because of primary winding inductance. My first guess was a damaged transformer (shorted primary winding) but with another transformer, from a similar power supply symptoms were exactly the same.

My second guess was that the snubber circuit is damaged. I measured all snubber components for an obvious damage but they all seemed to be fine. I added two shunts just to see if the current observed on the low side of the primary winding is flowing through the snubber circuit. or through the primary winding. I measured that there is no such a high current in any of the upper shunts. I started scrutinizing the (single sided) PCB for a short but there was none.

The solution, the MOSFET's radiator is connected to GND over a 10nF capacitor (a green capacitor on the circuit diagram above). The isolating layer between MOSFET case and its radiator was shorted and this 10nF was connected to the drain of the MOSFET. The capacitor charged to almost 400V over primary winding and when FET turned on it had to discharge this capacitor first what caused heavy current spikes.

After isolating the MOSFET from the radiator power supply works correctly, voltage on a shunt during normal operation looks like this:

Voltage on R28 after fixing the power supply

Lesson learned: mica is not always isolating, before starting with wild theories check very basics. I did some research on google and it seems that there are often problems at the primary side of the main power supply.

I do have circuit diagram of this power supply but I am not supposed to publish it. However, I will answer all questions to components.

Fixing РКСБ-104 - Voltcraft HS-036 Geiger counter (replacement for ИЖЦ-3 LCD display)

2016-05-16T04:53:00.001-07:00

Geiger counter RKSB-104 was produced in Belarus and exported under Voltcraft HS-036 brand to Germany. I bought one damaged counter on ebay. The defect is typical for this model: LCD display showing a huge black stain, everything else worked fine but the previous owner has removed the display.
I tried to purchase a new ИЖЦ-3 display but I failed to find a supplier, I decided then to reverse engineer the PCB and figure out if any modern LCD can be fitted into this device. I used this circuit diagram: http://c4r0.elektroda.eu/_hv/xrays/count2_sch.jpg

50	49	48	47	46	45	44	43	42	41	40	39	38	37	36	35	34	33	32	31	30	29	28	27	26
R1	BP	R1	G1	F1	A1	B1	G2	F2	A2	B2	G3	F3	A3	B3	G4	F4	A4	B4	BP	BP	BP	BP	BP	BP
Original RKSB-104 display pinout
BP	BP	E1	D1	C1	BP	E2	D2	C2	BP	E3	D3	C3	BP	E4	D4	C4	BP	BP	BP	R2	BP	R3	BP	BP
1	2	3	4	5	6	7	8	9	10	11	12	13	14	15	16	17	18	19	20	21	22	23	24	25

RKSB-104 LCD pinout, BP is a backplane, R1, R2 and R3 are some symbols which I couldn't identify based on a circuit diagram.

I suppose that DE119 LCD can be used to fix other similar devices for example for example РКС-107 or Bella (DBG-01N)

I prepared a small PCB which is an adapter between a original PCB and a new display

Adapter PCB circuit diagram

Adapter PCB layout

New LCD mounted on the adapter PCB

Display before fixing

The new display and adapter mounted on the counter PCB

The counter with a new display

Voltcraft VC840 RS232 interface circuit

2015-05-10T02:35:00.001-07:00

Q0 is a photo-transistor which receives IR light from a multimeter. If Q0 is illuminated then current flows through R1 and turns Q1 and Q2 on and the output R will be pulled down to positive voltage, this means mark will be transmitted. If Q0 is not illuminated then there will be negative voltage at the output R so space will be transmitted.

NTC temperature sensor connected over microphone input

2015-02-16T12:01:00.002-08:00

Simple measurement

The idea is to measure resistance by connecting it to smartphone, to the microphone port to be precise. The simplest implementation would be something like this:

R1 and R2 form a voltage divider which dumps a signal from headphone output while microphone input records and measures amplitude. If one of resistors is known then second value can be found by comparing known headphone output amplitude with that recorded on a microphone input. C1 and C2 are capacitors which are always at the output and the input of a sound card. C1 is cutting off DC component so, without C1 there would be constant current flow through a headphones, even when playing back silence. The unnamed resistor to VCC is providing DC supply current for a microphone, C2 is not letting this DC current flow into ADC converter.

However

This circuit has got a huge disadvantage: it requires calibration with every smartphone you would like to connect it to. Additionally, I am not sure if headphone output doesn't change it's amplitude with temperature. Calibration seems to be not elegant solution so I though of using both headphone outputs and conducting a ratiometric measurement. Assuming that both headphone channels (left and right) provide the same amplitude if driven with the same signal, the influence of absolute amplitude could be eliminated from the equation.

Ratiometric measurement

||here will be a circuit diagram||

In this circuit there are two voltage dividers, R1 and R2 is dumping a signal from a left channel while R3 and R4 from a right channel. Microphone input measures sum of both divided voltages. Measurement is conducted by driving the left channel with a sinus signal, recording input amplitude, turning left channel off while the right on, recording input amplitude and doing calculations on recorded amplitudes.

Three of four resistors are known, the fourth is measured. To simplify calculations it is assumed that R2 and R4 resistors are the same while R1 and R3 are reference resistance and measured resistance.

to be continued...

Stimulus signal

For some measurement I needed a specially prepared .wav file with sinus wave interleaving between right and left channel. Octave manage to prepare that file but can't manage playback under windows 7 so I will have to use Audacity...

bps = 16; % bits per sample
sps = 48000; % sample rate [samples/s]
freq = 1000; % frequency of the tone [Hz]
nsecs = 2; % number of seconds of the audio file
interleaving = 1; % duration of "pack" on a channel [s]

nsamples = sps*nsecs;
interleavingSamplesNum = interleaving * sps;

time = linspace(0, nsecs, nsamples);
wave = sin(time*2*pi*freq)';

window1 = repmat([ones(1,interleavingSamplesNum), zeros(1,interleavingSamplesNum)],1,(nsamples/(2*interleavingSamplesNum)));
window1 = window1';

window2 = repmat([zeros(1,interleavingSamplesNum), ones(1,interleavingSamplesNum)],1,(nsamples/(2*interleavingSamplesNum)));
window2 = window2';

wavwrite([wave.*window1, wave.*window2], sps, bps, 'stimulus.wav')

temperatures=[];

for i=1:10
%playback prepared file and record microphone input simultaneously
i
system("play.exe stimulus.wav -t waveaudio", [], "async");
recordCommand=sprintf("SoundRecorder.exe /FILE recorded%d.wav /DURATION 0000:00:02",i);
system(recordCommand);
recordedFilename=sprintf("recorded%d.wav",i);
%recordedFilename=sprintf("recorded.wav");
y=wavread(recordedFilename);
deleteComand=sprintf("del recorded%d.wav", i);
system(deleteComand);
a=(sum(y(20000:30000).^2)/(30000-20000))^(1/2);
b=(sum(y(60000:70000).^2)/(70000-60000))^(1/2);
ratioResults=a/b;
resistanceNTC=ratioResults*10000;
temperature=ntc(resistanceNTC)-273;
temperatures=[temperatures, temperature];
%ratioResults=[ratioResults a/b]
endfor

I'd like to measure resistance value in circuit:

if three of four resistors are know, and L-OUT and R-OUT can be driven with prepared waveform while MIC-IN records amplitude, than the fourth resistor value can be calculated. I prepared a trivial protype to check if such measurement is really possible on my PC. Octave generates a wave file, Audacity plays it back and records microphone input, Octave analyses collected data...
Of course, there is many limitations like AC coupling of mic input which forms actually a high pass filter, limited amplitude of microphone input signal, asymmetry between channels, clever estimating parameters of recorded sinus wave... but, let's give it a try. This is my prototype, four resistor can be put into DIL socket to ease changing values. One jack connector is for microphone input, another for headphones stereo output.

The measurement 1

I set up the prototype with following values

R2=R4=1k

R1=10k

R3=22k

Connected my oscilloscope to find out that the cheap Chinese USB audio card:

is actually providing around 2.23V DC into headphones output. This fact is not a problem for my measurement but can damage headphones. Anyway, here comes a waveform recorded by microphone input. The envelope of "packs" of sinus is surprising. I will do some further measurements to understand if DAC is not giving constant output or if microphone input is so bad quality. Probably these are some transitions because of microphone input capacitor. Next step will be to analyse recorded values to see if measured amplitude corresponds to expected amplitude...

Signal recorded by microphone input when stimulus signal has been played back.

Temperature logged in my room overnight. Celsius degrees, x axis are samples, not seconds.

Bicycle speedometer using hub dynamo to supply and as a speed signal

2015-02-06T14:03:00.002-08:00

Idea

So I got one of this cool hub dynamos. I really like that it's quiet and efficient way of generating energy for bicycle lamps. I considered adding USB charger to my bicycle to be able to supply my mobile phone while cycling but this is not trivial topic because of low output frequencies from dynamo. Another problem is relatively low output voltage, 6V is to low to use LDO reliable, under low speed condition output voltage may be lower than 5V so eventually a step-up step-down converter would be useful. Additionally rectifying low voltages is kind of tricky, I mean you can do this easily but to do this efficiently is already a challenge because of immanent power loses in diodes (0.7V drop at 6 V input is already more than 10% loss).

As a byproduct of such consideration I drawn a simple circuit which uses LCD display and tiny micro to track frequency, which is directly proportional to the bicycle speed. I like the idea of using dynamo output voltage as a supply and a signal at the same time. You will never need to replace battery, and you don't need additional sensor on the wheel.

Circuit description

The speedometer is supplied from a hub dynamo. D1, C7 and C8 form a half-wave rectifier. C4 is ESD protection and D2 is a Zener diode which should be short-term overvoltage protection. IC1 is an LDO voltage regulator, carfuly chosen to have low quiescent current, C5 is output capacitor of the voltage regulator. Hub dynamo produces pseud-sinusoid signal with a frequency which rises with speed. At low speeds at big wheels frequency can go as low as 5Hz, and C7 and C8 have to store energy for the device.

Input waveform when supplied from signal generator.

Signal at SPEED node. Voltage limited to the value safe for microcontroler digital input. Not very "digital" but thanks to microcontrolers Schmitt input it's not a problem. Circuit supplied from signal generator.

Waveform shows signal at SPEED node. Circuit supplied from a real hub dynamo (Shimano DH-3N31-NT).

Measurement
Because of required low power consumption micro has to run with 32[kHz] clock. The micro is actually measuring period, not frequency so division is needed. Fixed point arithmetic is used with assumptions that the speedometer shows integer part of the speed with proper rounding up and precision not worse than 0.1[km/h].

to be continued
(verification of fixed point arithmetic)

Prototype

First prototype supplied from signal generator

First prototype supplied from a real hub dynamo driven by a DC motor

Software

https://github.com/ArekKocowicz/spedo

Currently there is a github repository where you can download software and compiled.hex file. However this software doesn't show speed yet. It shows period in milliseconds before rising edges on SPEED signal. Stay tuned, new revisions are coming.

PCB

I designed a simple one-sided PCB using Cadsoft Eagle. Designing PCB with one layer only (the red top layer are wire connections which will be soldered manually) required changing association between mirocontroler pins and LCD segments. This is "layout friendly" now, software will have to compensate for this. Eagle files and gerber production files are packed along with software and can be obtained from github repository.

Todos

Add buttons

Add possibility to set up wheel circumference

Add captive touch sensor instead of mechanic switches

Add LED to back light the display while driving

Print a housing

Find a neat way of connecting the speedometer to the bicycle (electric and mechanical solution)

Housing trials
I've drawn a simplest housing ever, it consists of one part, PCB will be put in and sealed with epoxy. There is rounded part to fix the housing on the bicycle handlebar and fixing area to secure it with cable binders. Just click on "play" button to see it.

Your browser does not support iframes.

I didn't realize that this shape is actually not printable, the result can be seen here:

My buddy convinced me that using cable binders is not perfect way of holding stuff on the bicycle. Second housing will be fixed with an o-ring. I was thinking how I could make o-ring less visible while keeping model printable, but I changed strategy, if I can't make something unvisible, I will intentionally make it visible.

Your browser does not support iframes.

This is freshly printed housing with an LCD put in to if dimensions are right

Fixing Ravensburger Tiptoi

2014-06-15T09:01:00.002-07:00

Thanks to my college from work I had a chance to look at apparently popular toy called tiptoy. Basically it is a ~~RFID~~ infrared camera combined with a mp3 player. There are books with ~~embedded RFID tags~~ small points printed over the regular pictures. Those points are almost invisible to a human but tiptoy recognizes them and plays back some records when its tip touches tags in a book so a child gets interactive description of what is in the book. Different images or areas of images have different points pattern.
A typical problem of the tiptoy is that its speaker is not completely child proof and after some time the toy becomes mute. This particular toy worked with a headphones normal so my friend decided to try changing a speaker what solved his problem. Here come some pictures he took:

Take batteries out, unscrew four screws.

Take speaker out and desolder it.

Both speakers - the original one and the new not installed yet one.

This is close-up how the speaker is mounted.

You can ensure that the speaker caused a problem by measuring its resistance.

This is measured resistance of the new speaker, close to 8 Ohms is correct.

This is resistance of the damaged speaker. 2.3MOhms is almost open circuit - a speaker is damaged.

Brushless motor from RC model as a CNC spindle - 3D printed

2014-05-13T09:27:00.005-07:00

I do have a small CNC milling machine at home, I build to be able to prepare housings for my electronics projects. The machine has a working area of around 200x300x70mm and a typical 43mm tool holder. A spindle which I used at the beginning was 550W Kress FM 6955 without any speed regulation. I tried to supply it from phase controller but stable regulation range was just to narrow. Kress is a great tool but is extremely loud, I was thinking about building something smaller what would fit 43mm holder to be able to change tool quickly.

I don't expect it to be fast and powerful, the main criteria is to have a quiet machine to be able to work at home.

At the same time my great college from work got himself a Prusa 3D printer, I see he has got a lot of enthusiasm for printing and is happy to challenge his machine so I designed in a Freecad two holders for 608 bearings around which I build my tiny CNC spindle. Following pictures show the designing process.

This are screenshots from a FreeCAD:

And this are two holders as came from 3D printer. Note square nuts which are faced with a surface of "ears".

This is ER8 collet with a 8x100 roller and one 608 bearing already pressed on.

The construction is hold by three pieces of M4 screw threads, there is a brushless RC motor attached to the round aluminium plate which serves a s heatsink for the motor too.

I don't have access to lathe so I had to improvise a coupling out of a piece of elastic hose which is supposed to be a fuel hose in a garden lawnmower. I can recommend this type of hoses because of their incredible flexibility. Soon I will be able to tell how does it withstand higher temperatures because the motor is heating significantly.

Pros:

-the construction is much quieter than the "big" spindle and is strong enough to mile with 1x4mm bit in a plywood

-scaleability - you can easily mount a bigger and/or different motor

-easy accessible parts - ER8 collet costs around 10 Euro if ordered from China (ebay), motor around 20 plus 10 for ESC to drive it, rest of components is maybe another 10 Euros and can be obtained in a regular hardware store, usually DIY people have some sort of power supply so I will not put it on a bill.

Cons:

-bearing is getting warm when rotating, I didn't do long run tests but I have some concerns, BLA - material used for 3D printed parts has got a glass transition temperature of around 65 oC so there is not much headroom without forced cooling. Another possibility is to search bearing with a lower friction, I will update this page as soon I will get any test results.

Opel Astra H sport button

2014-05-11T08:10:00.001-07:00

In my Astra H there is a button which controls a park assist radar, some other versions of Astra H have a "sport" button there which changes a gas pedal, suspension (if car is equipped with IDS) and traction control settings.
I wanted to change the middle console module for a one with a "sport" button but I couldn't find any information if the middle console module in both variants (with parking radar and with "sport") has the same circuit - is this only a matter of a sign on a button and activating the right option in cars ECU or it is needed to change button module to be able to activate "sport" button?
I bought a middle module with "sport" and did a short investigation, here is the disassembled module:

Here is connector with pin numbers which I assumed:

And here is a reverse-engineered circuit diagram:

Safety related features in a car have additional diagnostics. Sport button is safety related because of influence on cars traction and for this reason there is additional 1k resistor across the switch so ECU can detect if the button connected correctly. If button is missing then most probably ECU will switch off this potentially dangerous sport mode. Door lock button and hazard lights can't directly cause a harm and they don't have this additional 1k resistor.

I didn't took the regular module without sport button yet but I will put here a circuit diagram and circuit comparison when I do this.

Fixing Voltcraft 8202 sweep/function generator

2014-02-05T11:16:00.003-08:00

Recently I ordered from ebay a defected Voltcraft 8202 function generator. It powered up properly and showed on the display a frequency. It wasn't completely damaged but there was no output signal, no matter what amplitude and offset I set there was 0 Volts on the output.

I assumed that the problem is somewhere in output stage, this sound like a typical damage for a signal generator which once had its output overloaded. After opening the generator up I traced signal backwards from output connector, circuit diagram of output AB-class amplifier area looks like this:

This is actually really popular topology, signal (point A) is fed to the positive input of the differential amplifier built around Q12 and Q13, Q14 drives Q15 and Q16 push-pull output stage while D6 and D7 provides bias voltage for Q15 and Q16. Signal from the point between R59 and R60 is fed back to negative input of Q12 and Q13 amplifier - this negative feedback improves linearity. Between point B and the output BNC connector there is additional switchable 20dB voltage divider which I didn't show on the circuit.

Multimeter set to a diode test showed that Q14 was damaged, additionally R59 and R60 were overheated, one of them was opened, another looked bad but was still conducting. R57 in the negative feedback has drifted because of overload and was around 30K instead of 6.8K. I replaced Q14 with BC557, put new R59, R60 and R57 but there still was no output signal.

It took me a while to notice that the original 2N3906 has got different pinout than the BC557 which I used, actually pin numbers are the same: emitter is the pin number 1 but take a look at this screenshots from datasheets:

2N3906

BC557

The order in which pins are numbered is different. After rotating the BC557 generator started to work again. I enjoyed watching sinusoidal and triangle signals on a screen of my oscilloscope. I didn't enjoyed watching square waves because their edges are really slow, slew rate of around 1V/us. It looks acceptable at 100kHz, but at 1MHz it is not a square wave anymore.
Another observation is about PCB quality, it was enough to solder one point twice to tear copper of the PCB:

I don't mind because not very often I look at this PCB but it says something about overall product quality.

Fixing Fender Hot Rod Deluxe PR-246 (solder cracking)

2014-01-26T11:46:00.000-08:00

Fender Hot Rod Delux is a 40 Watt hybrid guitar amplifier. This particular one had a problem of a cracking noise. There was a normal sound coming out of the speaker but with annoying full volume cracks, while playing cracks happened more often then in a standstill.

The owner told me that he wanted to by a new set of vacuum tubes because he noticed that he can cause this disturbance by touching the left most 12AX7A tube. I took a look at the tube and noticed that it is quite difficult to take it out because tube sockets are fitted very tight.

After opening a back of the implement I took a look at a PCB and noticed that there actually was a solder crack on the pin of tube socket. The crack is visible in the center of this picture:

I heated this broken solder joint with soldering iron, added some new solder and problem was fixed. As a precaution I reworked the same way all pins of this tube socket because if one solder joint was broken than another one was just waiting to crack.

Lesson learned:

1. Be careful with replacing tubes - they are fragile and rest of an amplifier is too.

2. PCB holes diameter for this tube socket are in my opinion a little bit to big giving pins possibility to cause mechanical tension while replacing tubes. Sockets have no mechanical support except for solder joints - shame on you Leo Fender!

Fixing Honeywell Rondostat

2014-01-08T10:56:00.005-08:00

In a scrap parts box I found a Honeywell Rondostat - a radiator thermostats which allow you to define different time zones and different temperatures for them, it is even aware of a day of a week and can control temperature accordingly. The thermostat I found was where it was for a reason, actually two: after putting batteries it turned out that a metal spring which connects batteries in series is loose and doesn't provide right electrical contact. This shiny metal part here:

It was a bit tricky to get it out:

But it was possible without demolishing whole device:

I bent this metal part so it was a bit tighter put it back, plugged two AA batteries inside and the Rondostat started to work immediately. New problem was that a adaptation process in which the thermostat learns what is possible valve operation range ended with an error. As it turned out it was a problem with a built-in gear. I pulled out the gear by levering three snaps, for two snaps I used stripes of EC card which I sicked into holes which are actually there for detaching gear:

Third snap is under control PCB which is easier accessible:

Root-cause of the gear adaptation problem was this little gear - metal axle was pushed into plastic part too deep:

With a pair of tweezers I pulled the axle out and put it back in but not that deep like it was before. Last step was to assembly it all back together. If you try it on your own, be aware that the gear with two metal parts has to be placed on top. This metal parts make gear visible for rotation sensor on PCB.

Further steps for putting it back are fairly easy and Rondostat works now.

This is PCB with NEC microcontroller, red connector is for a DC motor connected to the gear.

Honeywell Rondostat doesn't look like a new one, from a layer of dust I suspect it is more than one year old but mechanics is still OK. I didn't see any wear-out, it makes reliable impression. One problem with axle moving into plastic part might be planned obsolescence but even if - It is comparatively easy to fix it. I know that this thermostat was in a scrap box more than two days and surprisingly its clock was almost correct, turns out to have some super-cap or battery backed up RTC.

Last point is that I suppose that Rondostat does not like rechargeable batteries because of their lower voltage, I will try to test it deeper maybe my rechargeable batteries were already flat.

Fixing Squier Champ 15 type: PR-408

2013-12-30T11:33:00.000-08:00

My friends Squier Champ 15 had a problem - it didn't produce any sound. Red LED was illuminated but there wasn't even a hum from the speaker. My first guess was that a power stage was damaged. After opening up the Champ 15 reveled its uncomplicated construction:

One-sided PCB with two double operational amplifiers and TDA2030A monolitic amplifier. I searched for any sight of thermal events and measured supply voltage which turned out to be correct -20V/+20V. After feeding input 1 with audio signal from a laptop I measured with an oscilloscope that there is a correct signal on pin 1 which is input of the TDA2030A (it means that a preamp is working fine). There was amplified signal on the output of the TDA2030A to, so I moved on to connection between amplifier, speaker, and two connectors for headphones and a external speaker.
Plugging in headphones should disconnect the built-in speaker to make the combo silent, in case of this amplifier the problem was that one terminal of the speaker was disconnected because of defective headphones connector. Pair of normally shorted pins which should be shorted when nothing is plugged in was disconnected. I hard-wired them to check if this was it. Amplifier started to work.

I didn't want to replace defective connector so I've just fixed it. I desoldered and took out an elastic metal part which wasn't switching reliable enough, bent it so it was tighter, put it back and voila! Squier Champ 15 is working now.

This is mentioned elastic metal part, I am holding it with tweezers.

Metal part is on its way to be fixed:

Metal spring after bending:

crude power supply with LM741 2N3055

2013-11-28T11:58:00.001-08:00

Recently I had chance to look inside vintage power supply of my friend. It has got settable voltage and current limit so it's just typical, quite old workbench power supply (I don't dare to call it "laboratory")

After opening housing I've seen one-sided PCB with all components neatly described. The owner of the power supply told me that "current limitation doesn't work below 5V" and I was curious why it was so and if we could do anything with it. I could find damaged component quickly (there is not too much parts inside) but I started drawing circuit diagram to have clear overview.

As you can see this couldn't be simpler. According to circuit diagram current limitation should work just normal - when voltage across shunt resistor rises to around 0.7V, the transistor starts to conduct and pulls base of BC337 to lower voltage so it closes BC337 and in consequence closes 2N3055 to. This should be independent of voltage set-point.
I figured out that BC337 was damaged and replaced it with the new one. (OK, actually there were two transistors in Darlington-like configuration and I just omitted one). Power supply worked almost perfect, the only one problem was that current limitation (it worked below 5V to) smallest possible current limit setting was 1.4A when I expected it to be around 0.7V/1.5Ohm=0.46A. I just assumed that this two parallel power resistors which form shunt resistance (two white resistors below) are really 3 Ohms each like on the visible description on PCB. It turned out that they are 1Ohm each what gives 0.5Ohm and explains that current limitation can't be set to smaller value than around 0.7/0.5Ohm so around 1.4A. Never trust such a descriptions, measure it first!

By the way, here are some hints if you would like to build this power supply.
1. LM741 is not rail-to rail operational amplifier and this causes that the lowest possible to set voltage is around 5V, possible workaround would be to change operational amplifier to the one which can pull its output almost to the negative supply rail like LM321 for instance.
2. Current limit like on the circuit works actually from zero to around half of the potentiometer position, if you turn the knob further there will be no result. To eliminate that dead movement of potentiometer you could put another 470R in series with current pot, on top of it
3. By placing larger values of shunt resistor you get possibility to set lower current limit, I would prefer to put around 5.6Ohm as a shunt so I could set around 125mA as a lowest possible current limit.

Sachs Pentasport gearhub with two cables - fixing blocked extreme gears

2013-08-22T11:34:00.002-07:00

If your Sachs Pentasport blocks extreme gears (you can switch only between 2 to 4, gear 1 and 5 can't be chosen because you are afraid of braking the lever) then:
1. Make sure that you've disassembled, cleaned with white spirit, lubricated and assembled gear correctly.
2. Make sure that you are adjusting the lever correctly (starting from gear 4 which means all cables are in their "longest" position and than adjusting length).
3. Check if you've placed sun gear correctly, in my case it was the root of all evil. Befor mounting a wheel to bicycle it seemed to work but after riding for a while gearhub resisted against switching gears 1 and 5, I didn't want to break plastic lever so I've started to investigae what might be the reason. Sun gear can be easily flipped (two side are different but it seems to make no difference but it does make a huge difference).

The part 25a on this drawing is the mentioned gear which can cause problems if flipped.