CFL Electronic Ballast 01

CFL stands for "Conmpact Fluorescent Lamp".
How do these new low energy consuming bulbs light-up? With electronic ballasts!

1. Description
I salvaged this electronic ballast from a CFL (before I knew there was mercury in the bulb...).
I didn't write down the power, but because of the size (board is 55 mm in diameter) it was a relatively large one, likely for 1000 lumen (220 V, E27 socket).

The board has the following marks:
BS-03
GH372

Two pictures:

2. Electronic Ballasts
"Electronic ballasts contain a small circuit board with a bridge rectifier, a filter capacitor and usually two switching transistors. The incoming AC current is first rectified to DC, then converted to high frequency AC by the transistors, connected as a resonant series DC to AC inverter. The resulting high frequency is applied to the lamp tube." (source) 

I have found the following documents regarding the history, functioning and design of electronic ballast of interest:
- Electronic Fluorescent Lamp Ballast, application note from STMicroelectronics,
- Electronic Lamp Ballast Design, also an application note from ON Semiconductor.

Unfortunately, my knowledge of electronics does not allow me (at least for now!) to be able to follow them.

There are some sites involved in tinkering/reverse-engineering these type of components:
- PAVOUK,
- SHADDAK
- Nuno Suceda.


3. Schematics and Components
I have tried to sketch the circuit in QUCS, as usual in this series, but I am not even able to clearly identify the components, so this will have to come later...

Toaster Controller A0201D

1. Description
I found few days ago a toaster on a street. I normally do not pick this type of gadgets because I feel that somebody else may use them for better - but this time I got the electronics.

Because I only got the control board, I do not know the model: it was a toaster for two-slices, with independent selectors and three operation buttons: "reheat", "cancel" and "defrost".

For each slice-heater there is a control, with two boards each: for the timer/controller and for the buttons. 

Pictures of the control board and buttons:

 

 

The boards are 80*50 mm and 105*20 mm.

They are connected by 5 cables, and they have the following marks:
KT-223-MAIN for the main board,
and
KT-223-KR or KT-223-KL for the button boards (right & left).

The main board has two additional connector sockets: one with two pins, which is for the relay activating the heating AC side, one with three pins, which is the DC power supply for the board.


2. ICs for Toasters
The IC at the core of this board is labelled as:
A0201D
TN1430G3C+

I have not managed to get the datasheet of the IC A0201D - this is probably related to an intelectual property issue, at least this is what I asume.

This youtube upload is titled A0201D, and clearly talks about a toaster (model Aurora AU 153). It also mentions IC PT8A2511. Unfortunately it's ... in Russian!

This other webpage has the European certificates and tests on a toaster. The certificate mentions two suppliers for the control board, and the use of A0201D for certain models from one manufacturer, and PT8A2511 for the same models made by the other manufacturer.

Here, somebody asks for the equivalent subtitute of the A021D IC on a Kenwood TTP112 toaster (no reply was provided). Here some German guys were fighting with another toaster using this same IC (in German!). And there are also some Hungarians having fun/troubles with the toaster and the IC, and quoting pages in Chinese!

(In this Babel Tower I'm only missing the Finns!)

The A0201D chip is not unheard of: it is possible to find it on sales pages (f.eg: here or here), but no way to find the datasheet...

At least, for what seems to be an alternative IC, the PT8A2511, a datasheet is readily available (see here). As per the Hungarian blog mentioned above: "The Russians dare to believe and 99.99% certainty that the IC is the same as the PT8A2511".

In this page, which may be a/the manufacturer of the A0201D, the supply voltage is mentioned to be 3.5-5.5 V (other sources mention 12 V for the control board of toasters).

So, all in all, I have managed to collect the following:

- The "EMC Emission Test Report" for an electric toaster in whose control board, for some models, is installed the A0201D IC, as well as the PT8A2511. (The document is skimmed of what is no relevant here). The following scheme is of special interest:

- The PT8A2511 IC datasheet. This is a "CMOS LSI chip designed for toaster", presented in DIP8 package and working at 3.5-4.5 V. It includes few more information besides this scheme:

- The GA5210PH IC datasheet. This is also a DIP8 IC for toasters. The document is mostly in Chinese, mentions a standard range of 9-12 V, and includes the following scheme (among others):

Despite the apparent differences, all schematics share the pin utilization of the IC:
PIN 1 : DEFROST
PIN 2 : REHEAT
PIN 3 : RELAY / OUTPUT
PIN 4 : VCC
PIN 5 : RX
PIN 6 : OSC / RC
PIN 7 : CX
PIN 8 : GND

As per the PT8A2511 datasheet, the PIN operation is as follows:

3. Schematics

The circuit is modeled in QUCS, and can be found here.
A snapshot of the schematics:

Remarks:
1. The IC A0201D is not a sub-circuit, it's only a "box" without functionality. On this regard the schematics is "useless".
2. The lights at the button board are not LEDs (or at least their resistance is equal in both directions...)
3. The power supply (6 V) is just a figure: as mentioned above, I'm not sure of the required voltage.
4. The switches at the button board are actually of the push-button type (momentary, they remain open unless they are kept pushed).


4. Operation
The board is powered from connections S3_3 (VCC) and S3_1 (GND).
The coil is powered through S2_2, but only activated if the transistor T1 is closed, whose base is controlled by  PIN3.
Pushing puttons DR1 or DR2 closes the circuit through ground to PIN1 (a negative pulse). Similarly, pushing puttons RR1 or RR2 closes the ciruit through ground to PIN2. These negative pulses, according to the previous table, activates the DEFROST or REHEAT timing, resp. Once the function is activated, PIN3 activates the AC relay.
Once PIN1 or PIN2 are activated, they generate a positive voltage which keeps on the lights in the button board (DEF and REH).
If the main board is powered, the POW light is on through the direct VCC-GND circuit, regardless of the position of CR1/CR2.
If CR1 or CR2 are closed, in addition, a branch is closed through the base of the transistor Q1. Whether the voltage at the base is enough to overcome the transistor resistance will depend on specific values, but if it can, then it will also activate the coil which controls the heating of the slices.
VCC to PIN4 is constant.
The variable resistors VR1&2 control voltage at PIN6, and then the frequency of the oscillating system and the duration of the pulses at PIN2 and PIN3.
PIN5 and PIN7 are likely starters for the time counter controlled by the oscillator.

SCART board

1 Description
This is a SCART board from an old DVD player (if I'm not mistaken). The board size is 80*25 mm.

Some pictures:

The board has several marks:
LFM 200492-0001
76VCA
94V-0

Besides the female SCART connector it has two male pin sockets: one for 8 pins/cables, the other for 5.

2 Simulation
The circuit is modeled in QUCS, and can be found here.
A snapshot of the schematics:

There are a number of parts which seem to be ferrite beads: they are black cylinders, unmarked, the PCB labels them as FBNNN, and it makes sense that there are filters for signal cleaning. Their properties are unknown to me and I have simulated each of them as a subcircuit, with the model and properties mentioned here for one example. The impedances of all them is in the range of 0.5 ohm, as measured by my DMM.
There are also some SMD parts whose identification is unclear:
- ZDNNN, which I assume are Zener diodes protecting the line in case of voltage peaks (all are marked E29 except one (ZD135) which is H38).
- CNNN, which I asume are capacitors, but of unknown capacitance.

3 SCART
See here for information on the connections. The standard, EN 50049, is here.

There are several SCART connectors (at least three as per this source, in Spanish).

Pins 1, 2, 3 and 6 are for the audio (L and R, IN and OUT); the Zener protects lines 1 and 3 which are both OUT. This side goes to pin connector RB203.

All video signals are protected by Zener diodes, and go to pin connector RB202.

As per the pins used on the board, this is a type 1 connector. Pins (7, 11, 15) form the RGB channels IN, and pin 19 for composite video OUT; pins 8 and 16 control the communication acording to voltage levels (see here).
Voltage levels are (source here):
- Audio: 0.5 Vrms
- Video RGB: 0.7 Vrms
- pin 8:
* 0 - 2 V = no signal, or internal bypass
* 4.5 - 7 V (nominal 6 V) = widescreen (16:9) signal; (5-8 V as per this source).
* 9.5 - 12 V (nominal 12 V) = normal (4:3) signal
- pin 16:
* 0 - 0.4 V = composite
* 1 - 3 V (nominal 1 V) = RGB
- pin 19: 1 V   

All impedances from teh SCART female connector to the internal pin connectors are low (below 1 ohm as per my DMM).
 

Power Source 01

1 Description
This is a small board (56+60 mm), whose origin I do not recall.

It seems a sort of noise filter: the main capacitors are as per EN132400, which seems to be a standard for capacitors in line filtering applications, so that the amount of radio/electromagnetic interferences generated by the equipment, and transmitted back to the supply line, is minimized. Two sources of information on this subjec: one and two.

Two pictures of the board:

It shows a mark: "KX 1011-553".

(The switch seems to be broken: it is closed at all times).

2 Simulation

The circuit is modeled in QUCS, and can be found here.
A snapshot of the schematics:

 Mmmm ... I do not know how does this work:
- The main noise filtering capacitors seem to be placed following a X (C1) and Y (C2) arrangements.
- I have simulated the bobine as a transformer, but with a weird layout of primary and secondary.
- As I see it, both LEDS should be on.
- I cannot test it because my oscilloscope is max 30 V.
... 

Another RCA Board from another Old TV Set

The economic crisis is behind, it seems: it's time to change TVs and leave the old units in the streets!

1 Description
A small breadboard (100*33 mm) salvaged from another old TV set.

It features three RCA female sockets (the standard yellow, white, red) and a 3.5 mm jack female socket (black).

See another page of this blog with a similar unit here.

The board is labelled as from Phillips and has the following marks:

"

96977-4

3139 127 27491

MOD SB.PNL-HP-S

2 40     ARO

D4201

"

And on the back:

"

3139 123 5384.1 WK105

"
A Google search has produced no meaningful result for these marks.

A couple of pictures:

2 Simulation

The circuit is modeled in QUCS, and can be found here.
A snapshot of the schematics:

 4 Results
Composite Video (yellow RCA jack)There is nothing on this side, other than a diode: if it were a Zener (but I do not see any mark for that) it would protect the next stage from peak voltages; if it is a standard diode I do not understand its function...

Audio (input, red and white RCA jacks)
The audio channels are identical with a low-pass RC filter with extra components. As per the simulation, the cut-off frequency (70% of peak incoming voltage) is around 2.3 MHz, which is in line with what is mentioned in my previus RCA blog. (The result is the same for the simple RC circuit calculated on a spreadsheet).

Here also there a resistor voltage divider at the audio signal income, whose function is unclear to me.

Pin 9 acts as detector: when there is no jack the path 8-9 is closed with continuity; when the jack is introduced the circuit is open. The same happens with pins 3 and 6, but only pin 9 is connected to the rest of the board. (Because of this, when the red jack is in, the resistence between pins 8-9 is 150 ohm).

Audio (output, 3.5 mm black jack)
The jack is a standard 3.5 mm TRS.
This part of the board is also symmetrical: J1 is ground, J2 and J3 are the audio channels, J4-J8 and  J5-J7 are detectors, J6 and J9 are unused pins.
If there is no jack, none of the internal contacts shows continuity; when the jack is in, there is continuity between pins 4-8 and 5-7.
The circuit is a low-pass filter with a cut-off frequency (70% of peak incoming voltage) of less than 0.3 MHz - a bit on the low side as per my previous blog on RCA boards.

However, the innings of this part of the circuit are unclear to me: why an electrolytic capacitor is used (aren't they suppossed to be used only with positive voltages?)

Rectifier 220/10 V

1 Description
This is a small salvaged rectifier - I do not have a clue about its source, any small home apparel could be!

A couple of pictures:

On the secondary it features a rectifier bridge with an electrolytic capacitor, no more fancy features! The diodes, 1N401, have a nominal capacity of 1A, so that's the nominal intensity of the machine. For an output voltage about 10 V, that's approx. nominal 10 w. The capacitor is rated at 16 V, so little margin.

Actual measurements with no load:

V_rms Primary: 225 V

V_rms Secondary: 8.2 V

V_dc output: 11.2 V

2 Simulation
The circuit, modeled in QUCS, is here. 

A snapshot of the circuit, with some results:

 (I have adjusted the transformer ratio to fit the measured Vcc).

3 RMS
Crest Factor = (V_peak/V_rms) and for a sine wave, the crest factor is sqrt(2). Then, if Vrms is 225, Vpeak is 315 V, as in the model. 
My DMM has read 8.2 Vrms on the secondary, while the simulation shows a sine wave with a peak-to-peak amplitude of (-0.79,+11.8) V.
The Vrms of this voltage is:
Vrms = (11.8+.79)/2 + (11.8+.79)/2/sqrt(2) = 6.3 + 4.5 = 10.8 V
So, what's wrong?
It is possible that my DMM (Promax FP-2b) is deriving the rms value from average or peak values, which might be not correct for a dc-shifted wave.
Not clear... will have to check with the scope if the secondary is behaving as shown in the model.

Nope.
I have checked with my USB oscilloscope and the results at the secondary are:
8.2 Vrms
23.5 Vpp
18 mVdc
50 Hz
And at the load probe:
11.0 Vcc

Weird secondary...

4 Inverting the Rectifier 
What would happen if the voltage source is placed at the output?

For a DC source nothing: the diodes and the capacitor would block it.

And an AC source should blow the electrolytic capacitor.

5 Removing the Capacitor
If the capacitor is removed, the output shows a superimposed AC wave of 2.9 V of amplitude. 

RCA board from old TV set

1 Description
This is a small breadboard (78*36 mm) salvaged from an old TV set (cathodic rays!) that I found in a street.

It features three RCA female sockets (the standard white, red, yellow) plus a 3.5 mm jack female socket (black).

For the internal connections of the RCA signals, it has two separated sockets with 3 and 2 pins (for white+red, and yellow jacks, resp.; pin spacing is 2.5 mm). There are no further connectors for the part of the circuit rlated to the 3.5 mm jack, although the components are there.

On the board one can read:

"

2129

KABINLERDE KULLANILACAK (which seems to be Turkish for "USE IN CABIN")

11SB118-1

230104

"

A couple of pictures:

2 Simulation
The circuit, modeled in QUCS, is here.
(The voltage sources are not part of the board, of course).
A snapshot of the schematics:

3 Connectors
On RCA connectors see here.
With regard to the 3.5 mm 9 pin female connector, I have not looked for the model or checked in detail the connections, but I found of interest the information here. The socket is likely case C of the following image (source):

Are the sockets for input or output?
For the RCA, I think they are input:
- Protection: why would otherwise place Zener diodes here rather than on the previous stage?
- Filtering: as an input circuit the yellow RCA side (composite video) is an LC low-pass filter; the red and white RCA jacks (audio channels) are RC low-pass filters.
For the 3.5 mm jack, my assumption is that it is audio output (no technical argument: it's just the usual situation).

4 Results

Composite Video (yellow RCA jack)
For composite video see here and here.
(Why is this "composite video" and not "component video"? Not totally sure, but one: the colors of the connectors; two: the symmetry of circuits for red and white).
The resonance frequency of the LC circuit is 15.9 MHz.
The simulation of this part of the model is here.
As the sub-circuit has no resistor a short occurs at the resonance frequency - which is identified in the model results with a peak of more than 100 V at approx. 16 MHz.
(In reality there are resistances all around, f.eg I have measured about 0.6 ohm at the inductor.)
The introduction of the resistor and the Zener diode in the simulation has the result of reducing the voltage peak (from 160 to 16 V) and lowering the resonance frequency (from 16 to 7 MHz). Beyond that, the total capacitive impedance grows quickly and signals are filtered out.

Audio (white & red RCA jacks)
These sub-circuits are RC, acting as a low-pass filter, and that's shown in the output of the simulation:

However, the RC simple circuit whose behavior I have calculated with a spreadsheet has a much flatter pass branch:

What happens here ... ?
Tachán! Simulate with logarithmic steps and you get:

Still the numbers do not add up: QUCS says that at 1MHz there is 1 V, while I get 1 V at 5.9 MHz. Interestingly, the reason is related to the Zener diode: remove it and 1 V is shown at 5.8 MHz. I do not know... the Zener has a rupture voltage of 5.1 V on one side and is grounded on the other!

There a resistor voltage divider at the audio signal income: for what if both sides are directly grounded at their other ends?
No answer for this yet...

Why has the designer selected a cut-off frequency in the range of 1 MHz for the audio, and 7 MHz for the video?
According to this source, a TV signal requires 4 MHz of bandwidth, which at the end are transformed into 6 MHz. The audio is added centered at 5.75 MHz with a bandwidth of 0.5 MHz.
If this is so, the frequency carrier has been removed prior to this stage, and only the signal is left: 6 MHz for the video and 0.5 MHz for the audio. 

And why using an RC circuit for one and LC for the other?
No answer for this yet...