Refinement of the amplifier "Radio engineering U-101"
So, let's begin!
1.
Source of power.
Power supply circuits may vary slightly!
To get a decent output power, you need to have a decent power supply. Let's use the gift of the transformer manufacturers: the entire secondary winding is made with one thick wire (0.8 mm in my opinion). Therefore, it is quite possible to switch the power supply of a powerful rectifier VD5…VD 8 from pins 4 - 4* to 3 - 3*, which will increase the voltage from +/-26V to +/-31V. At the same time, a low-current rectifier VD 1…VD 4 becomes unnecessary and is removed along with the wires, and its storage capacitors C2 and C7 are connected in parallel with the corresponding capacitors of a powerful rectifier. But all connections with contacts 5,6 and 9,10 should be preserved.
Then the witchcraft begins.
1. We determine the geometric mean between the ground terminals of the capacitors C2, C3, C4 and C7, C8, C9 on the board of the filter capacitors, clean and tin it. We assign this point as the main common point of the entire amplifier.
2. From it we start up 2 thick wires to the minuses of the output connectors.
3. From it we start up common wires to the UM and UE.
4. From it we start up 2 wires to contacts 6 and 6 * of the transformer, removing the jumper between them. At the same time, we remove the connection between the rectifier board and the case.
5. We organize the connection of the common wire with the case at the input connectors of the amplifier.
6. And check that there are no more contacts of the common wire with the case anywhere else.
And finally, we connect a 0.047x630V capacitor in parallel with the primary winding of the transformer to suppress impulse noise from the mains. 
After removing VT 6 and VT 7, installing a jumper, replacing R 10 with a diode D 7 and shorting R 15, the D 7-VT 5-R 11 circuit turns into a diode stabilizer for the current source on VT 8, which is already powered by the swinging transistor VT 10 To reduce non-linear distortion, the swinging transistor VT 10 must be high-voltage, powerful and with a high gain. just meets these requirements, so we replace the original transistor with a more suitable one. The classical scheme. Only the R 42 resistor breaks the idyll. It is soldered from the print side into the notch of the printed conductor near the VT 2 collector. The introduction of this resistor increases the stability of the entire PA and allows you to get rid of compensating capacitors C4, C5, C9, C10, as well as resistors R 20 ,R 21. Side effects introductions of R 42 will appear when listening.
For normal operation of an electrolytic capacitor, it needs a charging potential of 0.6V, but it is not on the C3 plate. Therefore, there should be a non-polar capacitor here, limiting the bandwidth around 5 Hz. Hence the nominal value of 22 microns NP.
The setup is normal: connect an ammeter to the power break and set the idle current to about 40 mA. Then restore contact and start working.
color removed components are highlighted.Numbering corresponds to standard schemes e.
3. Preamplifier .Djvu 60 kb
Chip DA 1 is included in the preamp solely to match the piezoceramic pickup. I think that now this is no longer relevant, but it adds noise, and therefore we boldly throw out the DA 1 chip along with all the strapping and throw a jumper using the freed holes on the printed circuit board.
Newly introduced or changed components and jumpers are highlighted in red, and blue color removed components are highlighted. The numbering corresponds to the regular scheme.
This figure shows the DA1 chip and the components that should be removed along with it on the U5 ULF-P board.
Next, we more accurately adjust the loudness circuits to the volume control. Then we expand the bandwidth of the DA 2.1 and DA 3.1 amplifiers both in HF and LF and adjust the parameters of the tone block. To return the supply voltage of the DA 2 and DA 3 microcircuits to an acceptable direction, it is necessary to correct R 47 and R 48.
The preamp contains trimmer resistors R 24 and R 26 to adjust the gain of the entire amplifier. Setting conditions: at the input - 0.5V 1kHz; volume control - to the maximum; at the output - 14V without load, set with resistors R 24 and R 26.
rednewly introduced or changed components and jumpers are highlighted in color, and blue color removed components are highlighted. The numbering corresponds to the regular scheme.
This diagram shows the ULF-P refinement circuit, the DA1 chip is not shown.
5. Corrector UPZ-15.
To date, all known moving magnet magnetic pickups operate with a correction capacitance of 470pF. Accordingly, the capacitance of capacitors C1 and C2 is changed to 470pF. 6.Payment of inputs.
To expand the bandwidth down from 20 to 7 Hz, you can increase the capacitance of capacitors C4, C5, C14, C15 to 0.33 microns. This is at the end of the work as the cheeks puff up.
This treatise was compiled on 3.06.09.
/ nivaga /
or by mail, address attached.
The product of the Latvian industry Radiotehnika U-101-stereo (later, Radiotehnika U-7101) was a welcome purchase for any music lover in the mid-eighties. A complete set of Radiotehnika equipment consisted of at least four blocks - an amplifier, a tuner, a cassette deck, and a vinyl player. There might have been something else, but I didn't get it.
Some time ago, I found myself face to face with a Radiotehnika U-101-stereo amplifier, a Radiotehnika M-201-stereo cassette deck and a pair of Romantika 25AC speakers. There was a lot of time, there was nothing to do, next to the dream of a music lover of the mid-eighties, there were cassettes with recordings of The Beatles and Al Bano & Romina Power. It was decided to listen to Felicita and Let it be, but no such luck. The cassette deck did not spin the cassettes, and the amplifier gave such a background that it was scary for the speakers.
With the cassette deck, everything was decided quite simply - a little liquid lubricant, a bottle of cologne and a check of vodka brought the old woman to her senses. Here is a small photo report:
Just pour alcohol and oil on top of everything, well, glue the cracked plywood case together. This is of course not for long, because. and the gears were brought up and the belts stretched out
With an amplifier, in principle, everything is also quite simple. All the salt is in the electrolytes 🙂 As it turned out after a five-minute study of the problem through Google, it is enough to replace a couple of electrolytes in the RF unit and it is possible to replace the electrolytes in high. Here is a small photo report:
Since I didn’t remember exactly which pair of electrolytes to change in the RF unit (such a small shielded box on a cold contact plugged into the main board), I had to replace everything. Similarly with electrolytes on high. Everything was aggravated by the fact that I did not have a multimeter, and a soldering iron too. I had to buy everything in the same place where I came for electrolytes. DIN 5 pin and TRS 3.5mm connectors were also bought just in case.
As a result, everything about everything took about 40 minutes of work and the dream of a music lover of the mid-eighties sang first with the voice of Al Bano, and then with the Moby synthesizer, taking the signal from a mobile phone.
It is soldered, disassembled and assembled quite easily, I soldered with a decent Chinese 100W soldering iron. All parts are available and common, high - six pieces 50V 2000uF, low pair 6.3V 50uF, pair 10V 20uF and pair 50V 2uF. You just need to keep in mind that the tracks from the RF unit board peel off easily and naturally, and you need to solder carefully so as not to tear anything. Otherwise, you will have to “duplicate” the tracks with electrolyte legs.
Yes, I almost forgot, the amplifier circuit:
- (PDF, 100KB)
- (PDF, 100KB)
A short story about the repair of the Radiotekhnika U-101 stereo amplifier, the replacement of UMZCH modules with a power amplification circuit with TDA7250, preventive maintenance, experiments with output transistors TIP142 + TIP147, BDW93 + BDW94.
Brought short list basic preventive maintenance when repairing an old factory-made amplifier, many useful nuances have been clarified.
The stages of assembly and adjustment of a power amplifier based on the TDA7250 microcircuit are described. I'll tell you how I encountered the effect of overexcitation of a home-made UMZCH (noise, buzz, overheating and burnout of the output transistors) and how the solution was found.
I will share the bitter experience of using non-original TIP series transistors, I will show in the photo the differences between the originals and clones of unknown origin.
I have long planned to repair this old low-frequency power amplifier, considering the circuit on the LM3886 as a replacement for the old UMZCH modules or the design on the TDA7250 that has been tested in the work.
An additional impetus to this now was the desire to find and solve the problem that arose when repeating the ULF on the TDA7250 chip from one of the readers of my site, Andrei Vladimirovich. Thus, the fate with the choice of the ULF circuit to replace the old modules in the Radio Engineering amplifier was decided!
It turns out interesting story with investigation, many useful information and a successful finale with a demonstration of the amplifier!)
Amplifier Radio engineering U-101 stereo
First, I will briefly talk about the Soviet audio power amplifier "Radiotehnika U-101 stereo" (Radiotehnika U-101 stereo). One of the copies of such an UMZCH turned out to be in my possession (condition 4/5):
Rice. 1. Low frequency power amplifier - Radio engineering U-101 stereo.
Main specifications amplifier:
- Number of channels - 2 (stereo);
- Output power (nominal) - 20 W;
- Output load resistance - 4 ohms, 6 ohms, 8 ohms, 16 ohms;
- Reproducible frequency range - 20...20000 Hz;
- Power consumption from the mains 220V - 80 W;
- Rated pickup input voltage - 2 mV;
- Rated input voltage uni/tuner/playback - 200 mV;
- Signal/noise ratio (weighted, at Pout=50mW) - 83 dB;
- THD - no more than 0.3%;
- Case dimensions - 430x330x80 mm;
- Weight - 10 kg.
Useful features:
- Electronic input selector;
- Volume control + stereo balance control;
- Tone control (HF + LF);
- Loudness on/off;
- Pickup input;
- Headphone output;
- Output power indicator (separate channel indication);
- Protection of output stages against short circuit (short circuit) at the output;
- Protection of acoustic systems (AS) from direct voltage entering the UMZCH output;
- Overheat protection.
Inside, the amplifier is assembled in blocks, which contributes to its convenient repair and even replacing some of them with similar ones from another ULF or with home-made ones.
Below is a photo with the internal device of the amplifier (electrolytic capacitors have already been replaced):
Rice. 2. Appearance Amplifier Radio engineering U-101 inside (after minor repairs).
To understand further actions with the amplifier, I will give here its schematic diagram:
Rice. 3. Schematic diagram of the main blocks and their connections in the radio engineering U-101 stereo power amplifier.
Rice. 4. Schematic diagram of the ULF-50-8 power amplification module.
Rice. 5. Schematic diagram of the module for amplifying the signal from the pickup UP3-15.
Preventive work
Highly common cause inoperability of the old Soviet-made UMZCH - failure of electrolytic capacitors in the power supply. As a rule, they are large cylindrical aluminum cans, each with a capacity of approximately 2000 microfarads.
Rice. 6. Old electrolytic capacitors in the Radio Engineering U-101 amplifier.
In my copy of the Radio Engineering U-101 amplifier, six electrolytic capacitors were also previously installed in the power filtering circuit (see the diagram in Figure 3 - the U3 module).
Four of them (Figure 6 on the right) were in the rectifier to power the output power amplification stages, and the remaining two (Figure 6 on the left) were in the rectifier to power the input stages of the power amplifier, as well as for the preamplifier (U5 ULF-P).
Instead of 4 electrolytes at 2000uF to power the output stages of the UMZCH (U3 Rectifier board - C3, C4, C8, C9), I installed 2 pieces of 4700uF at 50V - this should be enough for this amplifier. And instead of 2 electrolytes to power the rest of the amplifier nodes (C2 and C7) - 2 pieces of 2200 microfarads at 63V, which I found in my stock.
In addition, it was decided to replace all remaining electrolytic capacitors in the preamplifier module, as well as in the indication, switching, protection circuits and in the UMZCH modules.
Three non-polar electrolytic capacitors with a capacity of 5 μF (C9, C10, C23) were installed on the board of the U5 ULF-P module - I could not find such electrolytes, and therefore I replaced them with pairs of back-to-back connected ones (plus to plus, and minuses to the circuit) polar electrolytic capacitors with a capacity of 10 microfarads.
Rice. 7. Scheme for replacing a non-polar electrolytic capacitor with two back-to-back polar ones.
On the board of the electronic input switch (U2) there are two more electrolytic capacitors in the power stabilization circuit of this unit - I did not change them, the switch is working properly.
- Installing or replacing a power fuse (especially if there was a jumper, "beetle");
- Inspection of all conductors (especially on the power supply of the UMZCH output stages) for damage (melted insulation, etc.);
- Replacement of electrolytic capacitors on the U3 rectifier board;
- Replacement of electrolytic capacitors on other boards (optional, judge by the sound and operation of the modules);
- Replacing dual variable resistors (volume, balance, treble, bass), if there is a crack or loss of sound during adjustment;
- Cleaning the insides from dust and debris + external cosmetic cleaning.
New UMZCH module based on TDA7250 + Darlington transistors
The first repair of Radio Engineering U-101 was carried out due to the inoperability of one of the amplification channels - the output transistors in the ULF-50-8 module burned out. The cause of this problem, as practice later showed, could be electrolytic capacitors that had lost their capacity, because of which the voltage balance in the rectifier arms was unstable (the difference was more than 5V).
I replaced the capacitors, the transistors in the output UMZCH installed serviceable ones and the amplifier continued to work further. The tracks on the getinax board began to fall off after soldering, the board itself deformed a little over time, some transistors were completely soldered to the board and connected to the broken tracks using pieces of MGTF wire in fluoroplast insulation.
Over time, one of the UMZCH channels burned out again - either due to overload, or due to some already burnt transistors in other amplification circuits of this module. The amplifier has been sent to rest.
Now I decided to revive it and completely throw away the old modules with power amplifiers, replacing them with a homemade scarf with a two-channel UMZCH. As a replacement option, I chose the amplifier circuit I used earlier based on the TDA7250 chip + Darlington transistors.
As output transistors, I considered the following options:
- KT825 + KT827 (powerful, reliable, slightly complicated mounting on the radiator);
- TIP142 + TIP147 (haven't tried these transistors yet, simple mount).
As a result, I nevertheless settled on completely different pairs of composite transistors, I will talk about this later. And now, I will write in detail how I made and assembled a printed circuit board for this UMZCH circuit.
The publication on the amplifier contains a schematic diagram, its description and a set of printed circuit boards from visitors to my site and those who have already assembled and launched this module.
For installation in the Radiotekhnika U-101 amplifier, I decided to make a printed circuit board from Alexander - it is compact and designed to connect output transistors using insulated conductors.
Thus, it will be possible to fix the transistors on the amplifier radiator in any convenient way and at any distance, the printed circuit board with parts can be placed vertically or horizontally.
Production of a printed circuit board for UMZCH using the LUT method
Manufacturing process printed circuit board for an amplifier with track layout from Alexander, I described in a separate article -.
There, the manufacturing process of this printed circuit board is shown in great detail, useful nuances and recommendations are painted.
Details for power amplifier circuit
To assemble the AF power amplifier circuit (a link to the article with a circuit diagram is given in one of the previous sections), the following were purchased: a TDA7250 chip, a set of TIP142 + TIP147 transistors, as well as powerful ceramic resistors (though 0.1 Ohm, nominal value 0.15 Ohm was not available).
I took the rest of the parts for the circuit from old stocks, but I still had to make some adjustments, since I didn’t have the right amount of some parts with exactly the same ratings.
Here is a list of the changed denominations that I decided to use:
| Detail | Denomination on the diagram |
Denomination used me |
Note |
| Capacitor | 100 pF | 82 pF | C13, C14 |
| - | 150 pF | 68pF + 82pF | replacement by a pair connected in parallel |
| - | 100uF / 50-63V | 220uF / 63V | C3, C4 - nutrition |
| - | 100uF / 50-63V | 150uF / 63V | C1, C2 - chains feedback(OS) |
| Resistor | 33 ohm | 56 ohm | R20-R23 - protection circuit circuits (SZ) |
| - | 1.5 kOhm | 1.6 kΩ | R2, R3 - OS circuits |
| - | 390 ohm | 360 ohm | R12-R15 - transistor control |
| - | 0.15 ohm | 0.1 ohm | R16-R19 - SZ detectors |
Electrolytic capacitors for power supply can be set to 150-470 microfarads, but in feedback circuits it is still better not to go far beyond 100-150 microfarads.
I did not find 150pF capacitors in the amount of 8 pieces, so I decided to assemble approximately the same capacitances from two 68 + 82 (pF) capacitors connected in parallel, that is, I will solder two capacitors together in the board in place of the 150pF capacitor.
Powerful resistors that are involved in the current detection circuit for the quiescent current stabilization circuit and protection of the output stages should preferably be set within 0.1 - 0.18 (Ohm).
Increasing the resistance of these resistors will lower the protection circuit's trip bar (reducing the maximum output power), while decreasing it will raise it (increasing the maximum output power, but be careful).
Rice. 8. Printed circuit board and a set of parts for assembling the UMZCH on the TDA7250 chip.
On the body of the TDA7250 chip, "MALAYSIA" is indicated. In the comments to the article describing the amplifier circuit, I gave a photo of two microcircuits that are used in my home-made UMZCH "Phoenix-P400".
As you can see, all the TDA7250 microcircuits I use have different markings, and at the same time, they all showed themselves well in work.
Rice. 9. Appearance of the purchased TDA7250 chip with the inscription MALAYSIA.
Two pairs of TIP142+TIP147 transistors were purchased for about $1.4 each.
Rice. 10. TIP142 + TIP147 transistors I bought.
I immediately noticed that TIP142 is significantly different from TIP147 in many ways. outward signs Let's see how they perform in action...
Winding homemade inductors
To assemble from the necessary parts, only inductors are still missing - they will have to be made independently. They should be wound on a mandrel with a diameter of about 10mm, for this purpose I used a metal screwdriver holder rod.
At first I thought to fasten the turns with fusible silicone, but then I decided to use a different material - an elastic thin thread, which I already used in the manufacture of a contour inductor for a homemade tube radio.
Also, to wind 40 turns, you will need a piece of enameled copper wire with a diameter of 0.8-1mm and sufficient length (did not measure, since the wire is in abundance). To fix the beginning of the wire on the frame and after winding the first layer, electrical tape is also useful.
The coil will be wound in two layers - 20 turns each.
Rice. 11. Preparation for winding the inductor, the necessary materials.
We start winding by fixing the beginning of the conductor with an insulating tape, we also tie a thread to the conductor and tie several knots so that the thread does not come loose during operation.
Rice. 12. We start winding the inductor, fixing the conductor on the frame.
We wind the first layer of the coil turn-to-turn, after each turn we wrap the conductor once with a thread with a slight stretch. As a result, the thread will look like a seam along the entire coil. the coils will hold tightly and together.
Rice. 13. The first layer of the inductor is ready and fixed with a thread.
As you can see, one side of the coil along the rod-frame holds well, but the opposite side can "walk" a little, which in turn will interfere with the winding of the second layer on top of the first.
To eliminate this problem, it is enough to use construction tape - cut a strip along the height of the coil and wrap the first layer of turns, pressing the tape tightly to the turns.
Rice. 14. Insulating the first layer of the inductor with construction tape.
Now you can start winding the second layer over the insulated turns of the first layer. Similarly to the first layer, we tie a thread to the beginning of the winding of a new layer and fix each of the turns with it during the winding process.
At the end of the winding, we connect the beginning of the thread from the first layer with the end of the thread from the second layer and leave a piece 30 cm long.
We unwind the tape and remove the frame from the coil. We thread the remaining piece of thread inside the coil and with a stretch we wrap the coil with it from the inside out about two times, we connect the end of the thread with the remaining thread from the previous stages.
Here's what should happen:
Rice. 15. Fixing the layers of a homemade inductor with an elastic thread.
We also fix the coil on the opposite side. We tie the remaining ends of the thread into several knots and cut off, leaving about 15 mm in length. After that, we take a lighter or a match and melt the remaining ends of the threads up to the knot. Be careful not to melt the knot itself or the ligament will be damaged.
Rice. 16. Fixing the ends of the knot of threads with a flame of fire.
One coil is ready, another one is made in the same way.
Mounting and soldering of electronic components
When starting the installation, the first thing I want to do is solder the microcircuit into the board, but do not rush, before that you still need to solder two jumpers that go under the microcircuit into the board.
Rice. 17. Two jumpers on the printed circuit board for the chip.
All conductors must be of large cross section, since a small current will flow through them at a high ULF output power. Here, bundles of conductors from non-working switching power supplies (from personal computers and servers).
Rice. 18. Insulated colored conductors of large cross section from computer switching power supplies.
Using colored conductors, I decided to give them the following assignments:
- Blue - to the Base of transistors;
- Orange - to the collectors of transistors;
- Red - to the Emitters of transistors;
- Black - earth;
- Green - ULF outputs;
- Red - plus power;
- Gray - minus power.
Thus, when experimenting with transistors, I will have almost no room for error - to confuse B-C-E connection or apply plus power to minus shawls.
Rice. 19. Low-frequency power amplifier board on the TDA7250 chip, assembled.
After soldering, it is advisable to clean the side with the tracks from the remaining rosin and wipe it with a cotton swab dipped in a solvent.
Rice. 20. View of the finished bass amplifier board from the side of the tracks.
Checking the health of the output transistors
First, I decided to check the operation of the assembled amplifier with powerful composite transistors KT825 + KT827. But before that, I considered it necessary to check all the transistors in stock using a universal tester for electronic components on a microcontroller.
Similar testers can be ordered from local online stores or from the Chinese for less than $8 per set.
Rice. 21. Indications of a universal microcontroller tester when checking the transistor KT825 (P-N-P).
Rice. 22. Checking the composite transistor KT827 (N-P-N), instrument readings.
The tester correctly recognizes the transistors, and also determines that a diode is connected inside them between K and E.
Similar checks were performed for the TIP142, TIP147 transistors I bought.
Rice. 23. Checking the health of the TIP142 (N-P-N) transistor using an electronic component tester.
Rice. 24. Checking the health of the TIP147 transistor (P-N-P).
For some reason, the tester did not detect the presence of an internal diode in these transistors. In addition, the hFE readings (albeit not showing exactly, but still) for 147 and 142 differ by almost 2 times, which is a bit strange when comparing the difference in readings for 825 and 827.
I thought that it would not hurt to check all the transistors with a tester in the dialing mode.
Rice. 25. Preparing for the continuity of transistors with a multimeter.
All the results and readings of the multimeter in the continuity mode (resistance measurement up to 2K + sound signal at low resistance) are given in the plate:
| Transistor | B+ K- | B- C+ | B+ E- | B- E+ | K+ E- | K-E+ |
| KT825 (PNP) | ? | 693 | ? | 837 | 536 | ? |
| KT827 (NPN) | 667 | ? | 989 | ? | ? | 535 |
| TIP147 (PNP) | ? | 737 | ? | 921 | 599 | ? |
| TIP142 (NPN) | 762 | ? | 1374 | ? | ? | 716 |
Note: the symbol "?" indications are indicated on the screen of the multimeter, when 1 was displayed on the left, this means that the measurement limit in the continuity mode has been exceeded (resistance is more than 2K) or the current does not flow at all (break).
The K-E legs ring in one of the directions, since inside between them, all the Darlington transistors under consideration have protective diodes installed.
But if you switch the multimeter to the resistance measurement mode of 20K, then the B-E legs, when changing the probes, will show different resistance in places (4-7 kOhm each), the reason for this is the resistors installed inside between the B-E, also parallel to one of the resistors can also include a diode.
Each of these transistors contains a small circuit that contains:
- Two transistors (one medium and one high power);
- Two resistors;
- Powerful diode between K-E;
- Some composite transistors can have another diode installed - between B-E first transistor in parallel with one of the resistors.
It is not for nothing that such transistors are called "composite", since they consist of several electronic components connected to each other.
Rice. 26. Schematic diagrams composite Darlington transistors - TIP142 and TIP147 (from the datasheet).
Also, to check the key mode of transistors, you can assemble a small circuit with an LED, I talked about it in the main publication with the ULF circuit on the TDA7250.
First use and security measures
It's time to check the assembled scheme in action. So, I prepared the KT825 + KT827 transistors - I found the fastening elements for connecting the conductors to the collector:
Rice. 27. Fastening elements for connection to the collectors of transistors KT825, KT827 in the TO-3 package.
I will take the power for the circuit directly from the radio engineering U-101 stereo amplifier, for this it will be necessary to disconnect the old power amplification blocks from the circuit. AT this case I am interested in the supply voltage that goes to the output stages of the UMZCH, the conductors are quite thick and are connected on the left side through the terminals.
By measuring the voltage between ground ( general scheme) and the power terminals of the UMZCH scarf received values of approximately 26V in each arm.
Rice. 28. Measurement of the supply voltage of the output stages UMZCH Radio engineering U-101.
I disconnected the old and faulty power amplification boards, and wrapped the remaining connectors with electrical tape so that during operation they would not short to ground or other working components of the amplifier somewhere.
Rice. 29. The purpose of the connectors connected to the UMZCH boards in the Radio Engineering U-101 stereo amplifier.
In order to protect the module of the new homemade power amplifier from burning out the components in case of any errors, it was decided to supply power to it through powerful lamps with a spiral inside.
Walking through one store in the lighting department, I found powerful miniature incandescent lamps for a voltage of 12V and a power of 35W!
By turning on three such lamps in series, they will shine at full brightness when a voltage of 36V is applied to them. The resistance of the spiral of each of these lamps is approximately 0.29 Ohm.
I will put a bunch of 3 of these lamps into the gap of each of the power lines (positive and negative) of the amplifier, this will protect against explosions of transistors, from melting the insulation of conductors and other troubles in the process of experiments.
Rice. 30. Powerful incandescent lamps for 12V 35W.
It was necessary to somehow figure out how to connect them, since I did not have cartridges available, and the legs are made of very durable metal which cannot be soldered.
I decided to get out of the situation in the following way:
Rice. 31. Connecting incandescent lamps with bare copper conductors.
Every three lamps are connected bare copper conductors extracted from a twisted pair cable (UTP Cat-5). From the extreme conclusions of each of the extreme lamps, I made small ears from the wire - I will solder the power wires to them.
Since this composite lamp is designed for a voltage of 36V, then in case of some kind of malfunction or failure of the transistors, a maximum of 26V will go to this set of lamps, they will not shine at full brightness, and this is good.
I tried to power one of these lamps from a 6V battery - it shines quite brightly even at this voltage, and heats up to a temperature of more than 60 degrees in just a few seconds.
I connected the volume control to the input of the handkerchief of the self-made bass amplifier - a dual variable resistor of 47 KΩ, set the control knob to the minimum volume for a start. I will give a signal from a smartphone, the volume in operating system Android set the average.
As for the first turn on, I decided to connect the first speaker that came to hand to the output, for safety I turned it on through a 470 Ohm resistor (so that the speaker does not burn out when a constant supply voltage is applied to it).
I just connected a 470 ohm resistor to another channel so that there was at least some load on the amplifier output. This is how the test installation looks like for the first inclusion of the self-made UMZCH module:
Rice. 32. The amplifier is ready for the first inclusion with additional measures security.
Transistors placed at some distance from each other. because if they collide with bodies (collectors), then it will turn out short circuit on power lines (26V + 26V = 52V).
Turned on the Radio Engineering U-101 amplifier (the circuit was powered from it), turned on the playback of a music track on the smartphone, added volume variable resistor- the amplifier sang! One channel works and it's already good.
Turned off the power, switched the speaker to another channel, turned it on - there was a click and silence in the speaker ... Turned off the power, set the multimeter to measure DC voltage (up to 200V), turned on the amplifier and measured what was happening at the output of this amplification channel on the scarf - and there 26V , supply voltage!
If I had not connected a 470 ohm resistor in series with the speaker, I would have to say goodbye to it. Since the lamps in the power circuits do not glow, this means that only one of the transistors is open, you need to look for the cause.
I turned off the power, rang the transistors of the problematic amplification channel with the tester - they are intact. I decided to check if there was any debris under the board and if there were any extra connections on the board itself - in just a minute I found a short circuit between the tracks, which appeared in the process of soldering an adjacent electronic component.
Rice. 33. Random erroneous connection on the board, which arose during the process of soldering parts.
But everything worked out, the microcircuit and transistors remained intact, after eliminating this short circuit between the tracks, the amplifier began to sing properly in two channels.
After making sure that the circuit is working properly, I connected the Radiotehnika S-30 speakers to it directly and checked the sound at medium and high volumes - the sound is excellent, there is enough power to swing the speakers at 8 ohms almost to a dangerous limit.
I want to note that the KT825 and KT827 transistors were connected for tests without radiators, even in this form the amplifier worked for literally 40-50 seconds at high volume until the transistors started to heat up to 50 degrees, then turned off the circuit so that they cool down.
I decided to measure the quiescent current of the output transistors, turned on the multimeter in the current measurement mode (up to 10A, also switched the red probe to the corresponding socket) - 0.11A or 110mA, about the same value as in my homemade UMZCH Phoenix P-400 on the same microcircuits and transistors.
Rice. 34. Measurement of the quiescent current of the output transistors of a homemade low-frequency power amplifier.
Attention! After completing measurements with a high current multimeter, do not forget to switch the plug of the red probe to the previous socket (for measuring low current, resistance, etc.), because in this form, when trying to measure the supply voltage or some other value in the working circuit, a short circuit will occur through the internal shunt (low-resistance resistor) of the multimeter.
The voltage at the Bases of transistors in rest mode is 1.2V each.
I filmed a short video with the operation of the amplifier at low volume and with transistors without radiators:
The song playing in the demo is: Frozen Style - I See in Your Eyes.
Transistors TIP142 + TIP147 and ULF self-excitation
It works well with Soviet transistors 825 + 827 ULF, it's time to check the operation of the transistors that I plan to put in the amplifier, since they are much easier to attach to radiators (than the same CTs in the TO-3 package) - these are TIP142 and TIP147, they are shown close-up in figure 10.
I soldered new transistors to the conductors, just in case I connected the speakers to the ULF outputs through 470 Ohm resistors. I turned on the power of the amplifier, so far I have not applied a signal to the input - a whistle and a rumble are heard in one of the channels, and silence in the second.
I felt the transistors with my fingers - in one of the channels (the one that makes noise), the transistors very quickly heated up to high temperature. Turned off the circuit, waited for the TIPs to cool down, turned on the power and gave a signal - both channels are playing.
Interestingly, with the use of KT825 + KT827 there was no such effect, in the no-signal mode the transistors are barely warm, it is possible that TIP142 and TIP147 were caught with a very high gain or fake ones.
Nevertheless, I decided to cut several tracks and conduct a couple of experiments that might show the cause of generation in this amplification channel:
- Throw the ground that goes to the feedback circuits;
- Bring out the RC feedback chain that goes close to the rest of the components.
Rice. 35. Experiments to find the cause of excitation of the amplification channel.
I cut the necessary tracks, soldered the conductor and the RC circuit (100K + 30pF) from the side of the printed connections, turned on the amplifier - nothing has changed.))
So the reason is somewhere else. I tried to spread the conductors with transistors over a greater distance - the noise decreased a little, gave an input signal and gave volume and ... the lights in the power circuit lit up ... on New Year's Eve.)
The TIP142 burned out, the voltage skewed in the microcircuit controller, and thus, along with the burned out transistor, the TIP147 also completely opened, but it survived ... and this is largely due to the incandescent lamps that shone brightly all 6 pieces. I put 825 + 827 in the burnt out channel - it works, the microcircuit is intact!
I decided to take a closer look at these TIP142s, on the left of each of the pairs in the photo these transistors are shown in comparison with TIP147, and below is a drawing of the case and lining of these transistors from the official STMicroelectronics datasheet.
Rice. 36. Comparison of the TIP142 transistors I bought (it looks like a fake) and TIP147 (original).
Noticed differences between these strange TIP142s and TIP147s:
- Screw hole for fastening - smaller diameter;
- The finish of the legs is very "cheap" shiny, it is not the same as on TIP147 and most parts;
- The ST logo and inscriptions are very different in quality;
- Two of the three indented circles are under the hole, not above it as in the datasheet;
- The shape of the lining is in the form of a simple rectangle, not curly;
- The legs on the sides should be straight, and they have ledges.
To all this, one can also add that the resistances during the dialing of B-K and B-E differ by almost 2 times, I already wrote about this above.
The next day I went to the market for new transistors, bought two pairs of BDW93C + BDW94C in the TO-220 package for the experiment, managed to find one original TIP142 and still took another suspicious TIP142 for testing.
Rice. 37. TIP142 - original and fake, transistors BDW93 and BDW94.
Checking these transistors (in continuity mode, with a signal) showed the following picture:
| Transistor | B+ K- | B- C+ | B+ E- | B- E+ | K+ E- | K-E+ |
| BDW94C (PNP) | ? | 774 | ? | 920 | 596 | ? |
| BDW93C (NPN) | 730 | ? | 1062 | ? | ? | 561 |
| TIP142 (original) | 764 | ? | 870 | ? | ? | 615 |
| TIP142 (not original) | 758 | ? | 1365 | ? | ? | 722 |
As you can see, the original TIP142 does not have such big difference in measurement readings transitions B-K and B-E. Readings for transistors of the BDWxx series - there is a slight variation, but everything seems to be in order.
First of all, I decided to test the BDW93 and BDW94 in the case, and since the linings of their cases are quite small, I installed these transistors on small radiators removed from the board of some non-working old monitor with a CRT tube.
Rice. 38. Test of the amplifier with transistors BDW93 + BDW94 in one channel and KT825 + KT827 in the other.
The amplifier sang right away, no overheating was observed and everything worked well.
I connected the original TIP142 and TIP147 to the problem channel, applied power - the same hum and overexcitation. I decided to solder the transistors directly into the printed circuit board from the side of the tracks, without conductors, it is likely that the conductors in combination with these transistors create pickups here.
Rice. 39. TDA7250 amplifier board with TIP142 and TIP147 transistors soldered into it.
I turned on the amplifier in this form - there was silence in the speakers, the transistors were warm, gave a signal and both channels worked, though I didn’t turn it on at high volume, since here it’s better to put radiators on the transistors.
I shortened the conductors by half, left pieces 8-9 cm long to be enough to connect the transistors fixed to the radiators, applied power - everything is fine, there is no overexcitation, there is no abnormal heating, two channels work.
Rice. 40. Connecting transistors to the board with shortened conductors, test.
After that, instead of the original TIP142, I installed the one with a strange case - it also works. You can mount transistors on a heatsink and then conduct a full-scale test at high output power.
Conclusion: when repeating such ULFs, try to make the conductors to the transistors as short as possible, do not twist them together into a bundle!
Perhaps these fake transistors will show themselves well, I only have one original 142nd, in other cases the sellers offered me exactly non-originals, so I still have to use one non-original, let's see ...
Mounting transistors TIP142, TIP147 and connecting the UMZCH module
Before attaching the transistors to the radiator, it was necessary to remove the old UMZCH modules from the amplifier. To do this, you need to unscrew the three screws that attach the radiator to the amplifier case, and then you can conveniently unscrew the handkerchiefs with transistors.
Rice. 41. We unscrew the old UMZCH modules of the Radio Engineering U-101 amplifier from the radiator.
The transistors of the UMZCH modules are screwed by collectors in pairs to separate cooling pads made of thick metal, which is also soldered to the scarves with legs pressed into it.
These metal pads are glued with some kind of sticky adhesive to the radiator through an insulating film (not mica). From the disassembled structure, it can be seen that these linings did not fit very tightly to the radiator, voids were formed between the film and the glue, which is probably not in the best way affected the cooling of the output transistors.
I decided to place the scarf of the new power amplifier vertically - it is compact and its height allows it to be done in the Radio Engineering U-101 amplifier case. I immediately estimated the length of the conductors to the transistors and then shortened them to the desired value.
The surface of the radiator, on which the transistors will be mounted, was cleaned of adhesive residue with a cotton swab and ethyl alcohol.
Rice. 42. Layout of the printed circuit board in relation to the radiator for transistors.
I decided to fasten the transistors with the same screws that were used to fasten the metal pads with the old modules to the radiator.
The diameter of these screws turned out to be slightly larger than the diameter of the holes in the TIP147 transistors, and what can we say about the non-original TIP142. This problem was solved with a round diamond file.
Rice. 43. Diamond file for adjusting the hole diameter of the TIP series transistors.
Each pad of the TIP series transistor in this case is connected to the collector, so these components need to be screwed to the heatsink only through insulating thermally conductive gaskets. I removed such pads from non-working switching power supplies.
Rice. 44. Rubber thermal pads, TIP142+TIP147 transistors, screws and heatsink.
The transistors were soldered to the conductors coming from the UMZCH module, the joints were insulated with heat shrink.
Rice. 45. Transistors are mounted on a radiator and connected to the UMZCH module.
To connect the new UMZCH module to the power and ULF outputs, Radio Engineering first thought to use four-pin MOLEX connectors from computer block power supply, but then I found an easier way in which everything is almost ready - use connectors from old UMZCH modules.
Rice. 46. Connectors for connecting the conductors of the Radio Engineering U-101 amplifier to the printed circuit board.
To install these flat connectors in my homemade UMZCH board, you will need to slightly correct the holes leading to the power supply and outputs of the two channels.
I solved this problem with the help of a jigsaw: I slightly drilled holes in the board so that a jigsaw file fit into them, threaded it, clamped it and sawed out the necessary elongated holes. After that, I soldered the connectors to the printed circuit board without any problems, sparing a lot of solder for this so that they held well.
Rice. 47. Installation of power connectors in the LF power amplifier module.
When installing the radiator in its place, do not forget about one interesting component - the sensor temperature system protection, it must also be installed in its place.
Here, one junction of the KT315V transistor acts as a temperature sensor (see the diagram in Figure 3, module U6 - transistor VT5).
Rice. 48. KT315V transistor as a temperature sensor of the amplifier thermal protection system.
I decided to fasten the printed circuit board with the components using one strong connection, consisting of a long screw and tubes. Additional support for the scarf is provided by thick conductors that are soldered to the transistors.
Rice. 49. Mounting unit of the printed circuit board to the heatsink of the power amplifier.
Here's what the mount looks like:
Rice. 50. The UMZCH module board is securely attached to the radiator.
I connected the already assembled module to all conductors:
- Three power connectors (ground, plus and minus);
- Two connectors from the protection board;
- I soldered the outputs of the preamplifier to the UMZCH input (two common inputs rang soldered together).
I connected the power cord to the amplifier, inserted the conductors from the speakers into the output jacks and, just in case, connected it through a 470 Ohm resistor, you never know what.
For convenient signaling, I decided to use the front input jack of the amplifier called "play", for this I set the "copier inputs" switch knob to the "2-> 1" position, and the "INPUT SELECTOR" knob to the "2" position.
The pinout of the Soviet DIN-5 signal connector in this amplifier is as follows: if you look at the connector (socket) from the front with the key located at the bottom, then the middle contact on top is the common one, the two contacts on the right are inputs, the remaining two contacts are not used.
Rice. 51. Signal supply to the amplifier Radio engineering U-101 stereo, positions of the input switches.
I turned on the power of the amplifier, started playing a song on my smartphone, started turning the volume knob of the amplifier - it works! I added volume so that the signal level on the output power indicator was visible - the sound disappeared, one channel on the indicator glows completely up to red, immediately turned off the amplifier.
I thought it might be a false alarm of the protection (perhaps it will need to be adjusted), turned it on again - the filled segment of one channel on the indicator immediately lights up on the indicator maximum level, which is typical, there was no click of the relay when turned on.
When switching on again for a short time, I measured the voltage at the outputs of the channels - one of the channels had 26V, which is why the protection worked. The vertebra of the transistors showed that the TIP142 (not the original) was out of order, it K-E conclusions in both directions they ring with a resistance of about 5 ohms, it is broken.
There was a possibility that he would drag the chip into the trash can, but no, everything worked out. Since the speakers are connected through 470 ohm resistors, I thought maybe a load with such a large resistance somehow affected this situation ...
I decided to take a chance and connect the speakers directly, I replaced the burned-out TIP142 with the remaining new non-original one, let's see what happens, in any case, I already know that the protection in the amplifier is working properly.
Turned on the power, gave about 20% volume - it plays, waited a bit and increased the volume level to about 60% - the sound disappeared, the protection worked and turned off the speakers, the output power indicator showed with its "off-scale" that the problem was again with the same channel, quickly turned off the power .
I rang all the transistors - the non-original TIP142 burned out.
Rice. 52. Failed non-original TIP142 transistors where they belong.
I don’t have any serviceable TIP142 left (although the second channel with the original one works fine), no one has the originals in the market yet, ordering from the online store and explaining to managers how the transistor I need should look like will take time, but I want to already to finish everything, so there were already adventures ...
Of course, you can tinker for several hours and install KT825 + KT827 on the radiator, but I still have transistors of the BDWxx series - I'll try them in action.
Mounting transistors BDW93, BDW94
Mounting these transistors is a little more difficult - you will have to drill new holes in the radiator, and also take care that the mounting screw is not connected to the transistor lining.
I used insulating washers and pieces of cambric for this purpose, which will be put on the screw and isolate it from the inner ring of the transistor lining.
Rice. 53. Elements of isolated fastening for transistors BDW93, BDW94.
I marked the holes on the radiator and drilled them with a drill with a diameter of 2.5 mm, then I cut the thread with a tap for a 3 mm screw. If there was no tap, I would drill holes with a large drill and use longer screws with nuts.
Rice. 54. Preparing holes in the heatsink for mounting transistors in the TO-220 package.
To isolate the transistor pads (collectors) from the heatsink, I also used rubber thermal pads, only smaller, just under the TO-220 case.
Rice. 55. Preparing to install transistors on a radiator through rubber thermal pads.
I installed a radiator with the UMZCH module on the TDA7250 in the amplifier case, connected all the connectors, and soldered the input. Turned on the power and gave a signal from the smartphone - it's playing!
I added the volume by about 60% - everything is fine too. I added a signal level so that the output power indicators show full load (with red marks) - the speakers are literally bursting with power, everything plays and no problems.
Rice. 56. Transistors BDW93 and BDW94 in the output stages of the new module UMZCH amplifier Radio engineering U-101.
I drove this modernized design for about 20 minutes at high volume - the radiators are a little warm, the sound is quite good, it feels like there is still a power reserve, but the speakers did not fire.
In conclusion, you can slightly correct the display of output power levels on the gas discharge indicator by changing the sliders of the resistors R4 and R5 (the circuit in Figure 3 is the U8 module).
Below are photos of the insides of the amplifier with a top and bottom view (clickable photos):
Rice. 57. Photo of the upgraded Radiotehnika U-101 stereo amplifier (top).
Rice. 58. Photo of the Radiotehnika U-101 stereo amplifier with the new UMZCH module (bottom).
Conclusion
The task of restoring and modernizing the "Radiotehnika U-101 stereo" amplifier has been completed! I thought that it would do without adventures, but they turned out to be enough. Got interesting experience, which may be useful in the future not only to me, but also to those who read this article.
At the end of the article, I mounted a small video demonstration of the operation of the amplifier with the top cover removed. Filmed on a smartphone, from a high sound level, the microphone of the smartphone began to distort what was happening, however, this is enough for the demonstration.
Attention! About halfway through the video will increase the playback level in the amplifier, lower the volume in your video player.
