DIY AUDIO BLOG

TDA2050 - High -end version. The Gainclone Rival.

2012-02-23T03:05:00.000-08:00

The Gainclone amplifier is well know amplifier among the audio fans. It's based on the National Semiconductor's LM1875 chip. This article shows an alternative design using the TDA2050 by ST.
The original implementation of the chip can be found inside the datasheet:
http://www.datasheetcatalog.org/datasheet/stmicroelectronics/1461.pdf

The development of the new topology started as a group project in which many people participated. The original forum thread can be found here(in Bulgarian):

http://www.bgaudioclub.org/showthread.php?535-TDA2050-%D0%B7%D0%B0%D0%B2%D1%8A%D1%80%D1%88%D0%B5%D0%BD

The amplifier includes a delay circuit that keeps the speakers disconnected for few seconds while the amplifier settles to its nominal working condition. This will prevent the amplifier from introducing the unpleasant pops in the speaker during power up.

Here is the complete schematic for this project:

As you can see the NFB loop was modified.

Here is the original schematic suggested bu ST:

Comparing the two schematics we can see the differences. The modified circuit has the following advantages:
1. Eliminating the electrolyte capacitor in the NFB loop and using a non electrolytic one. This results in a lower overall distortion and sound improvement.
2. Using a low value resistor in the NFB loop (R2) we introduce a much lower noise in the signal path.
3. Using two identical capacitors at the input and NFB loop will result in cancellation of the any kind of distortions introduced by the capacitors due to the differential structure on the amplifier.
4. No DC gain. This results in no DC offset at the output.

To prevent the circuit form unwanted roll off at low frequencies, the NFB was further modified to prevent DC gain, but allow low frequencies to be amplified with enough gain.

The speaker delay circuit was designed by a fellow forum mate Bis and was adapted to suit the needs for this project. It is a simple but efficient. It uses a standard 555 timer. The time delay is set with the value of C2 capacitor. The relay used in the schematic is a one pole 10A power relay. I recommend Omron G2R series relays here. But you can use any other brand you like. The schematic shows a low cost version with relay by Goodsky:
http://store.comet.bg/bg/Catalogue/Product/5180/?dl=2122

This is how the PCB looks like:

Few words about the grounding. You will notice that the signal ground and the power ground are separated. Each ground should be connected with its own ground wire directly to the center of the power capacitors. (pin1 and pin 3 of X2 terminal). The 12V supply should be coming from a separate low power transformer.

The original PCB artwork is available for download HERE and as usual the design is free for personal usage. The commercial usage is of course not allowed.

THANK YOU FOR READING THIS ARTICLE

SSM2017 Line Preamplifier - High Quality Design

2012-01-20T07:39:00.000-08:00

This page will show you how to use the SSM2017 as a line preamplifier. With some clever designing a very high quality preamp could be build at affordable price.
For those of you unfamiliar with the chip, here is the datasheet:
http://www.analog.com/static/imported-files/data_sheets_obsolete/139484016SSM2017.pdf

Now some few words about this device. It has been designed as a microphone preamp with selectable gain. It requires only one external component - the gain set resistor. As a microphone preamp it was designed to work in harsh environment, long cables, signal pollution, noise etc. Despite that the SSM2017 is able to achieve extremely low noise performance and high gain. This makes it perfect for a normal line preamp with selectable gain.

Now the SSM2017 is obsolete. Few replacements are available from few manufacturers.

AD's SSM2019:
http://www.analog.com/static/imported-files/data_sheets/SSM2019.pdf

TI's INA217
www.ti.com/lit/ds/symlink/ina217.pdf

THAT's THAT1510/THAT1512
LINK

THAT's IC's have a superior noise performance at low gains so those are preferable. They are a drop in replacement to SSM2017/SSM2019 but a different values of the gain set resistor must be used because of the different input stage of THAT's IC's.

Whichever IC you chose to use, a high quality line preamp is the result. Another interesting thing is the REF pin. It allows us to design a servo. This way we can null all the DC offset at the output and remove any capacitors from the audio path. For this purpose we need a low cost FET input operational amplifier. TL071 is the right choice. The servo is not in the signal path so the servo opmap quality is not critical.

Here is the schematic for this project:

The PSU is prety straightforward. Two adjustable vregs and filter caps. Nothing fancy here. Use good quality industrial capacitors (Panasonic FC, NCC KY series, Rubycon etc.).

Now for the stage itself. The power supply goes trough an individual RC filtering to each IC. This helps reducing noise even further. Again good quality parts are used. All resistors are metal film to help keeping noise low.

The SSM output is forced into class A operation using a JFET cascodes ad the outputs. This increases the overall performance of the stage. A better sound is the result as well as better driving ability reducing the interconnect cables negative effects.
It is important to match the loading JFET's so that the two SSM's work under same conditions. Because the line stage is meant to drive high impedance loads, no high currents are needed. That is why the 2N5484 FET was used. It has a typical Idss of 1-5mA. For our application a value of 2-3mA is more than enough. It is important for the upper JFET to have a higher Idss rating. That is why the 2N5486 JFET was used in this place. It has a typical Idss rating of 8-20mA so no matching is needed here.

The whole project is assembled on a single layer PCB including the rectifier. The original PCB artwork is available for download here:
http://dox.bg/files/dw?a=5a603ec49c

Now some few words about matching JFET's by Idss. This is actually a very simple job. All you need is a voltage source and a multimeter. Here is the testing schematic:

Simply tie together the gate and the source pins and apply voltage to the drain pin. Connect the positive node of the ammeter to the gate and the source and the negative node to ground. The JFET will saturate at its own Idss. Please note that the Idss value is temperature dependable. Let the JFET to settle for about 10min and then note its Idss. Repeat the procedure until you find two closely matching devices.

Attention: This design was provided for personal usage only and is FREE for such purpose. Commercial usage is not allowed!

THANK YOU FOR READING THIS ARTICLE !!!

Diamond Buffer Headphone Amplifier

2011-11-27T12:00:00.000-08:00

Here is a truly classical implementation of the diamond buffer. A headphone amplifier.
There isn't much to be said about this topology. It's pretty popular and is used in variety of applications. One application is buffering an opamp's output so that it can drive low impedance high capacitance loads like a pair of headphones. Here is a schematic showing the diamond buffer as a current booster for an opamp:

The gain is set by the resistors R5 and R2. If you are using NE5534 the C6 capacitor is a must when gain below 3 is set. For more info on this read the 5534 opamp datasheet.

The buffer is biased in pure class A at about 300mA. That will give a plenty of headroom for plenty of voltage swing.

This amp will output a substantial amount of power so it must be used with care. Always switch it on with a volume pot set to minimum. It's very easy to damage you hearing so be warned!

The PCB artwork is available for download HERE. You need two PCB's for a stereo application.

**********THANK YOU FOR READING THIS ARTICLE**************

TPA6120 Headphone Amplifier - COMPOSITE TOPOLOGY

2011-10-20T14:22:00.000-07:00

Yet another TAP6120 based headphone amplifier you'd say. Well maybe you are right. This one started as a group project involving myself and few other guys forum mates. So here it is - the UBIQUITOUS TPA6120 amp.
The TPA6120 chip is a rather strange beast. Looking at the TI's portfolio of headphone amplifiers you would not find any other amp similar to this. Especially if you look at the slew rate. So why is that? The answer is rather simple. The TPA6120 chip is actually a re-branded high speed line driver that had failed to meet the requirements. This is a common practice among the manufacturers. Many so called "audio opamps" are actually the same thing - lower grade opamps that had failed to meet the criteria.

However failing to comply with the high speed line driver requirements does not make the TPA6120 chip a useless junk. Believe it or not it's an excellent performer. Many people believe that it can outperform much more expensive headphone amplifiers. I don't want to put any thoughts on this mater though.

Some facts about the TPA6120 chip
- as it's a high speed IC it requires a careful PCB layout. Any parasitic capacitance may cause the amp to go unstable;
- the chip is rather hard to run alone. The simple truth is that you may destroy your headphones if you don't put any attention to this. The chip will have a severe DC offset at the output depending on the volume pot setting. This can go up to several volts. You wont find this in the datasheet so BE CAREFUL!

This being said I've decided to go safe. This design uses an input opamp based stage. It's not an ordinary two stage project though. This design is based on the so called COMPOSITE opamp design. This approach uses the best of both CFB and VFB design worlds. More info on this matter can be found here:
http://focus.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=sboa002&fileType=pdf&track=no

I went for the non inverting design.

So my input stage is based on a precision opamp. Few candidates here:
OPA2132 - http://focus.ti.com/general/docs/lit/getliterature.tsp?literatureNumber=sbos054a
LT1057 - http://cds.linear.com/docs/Datasheet/10578fc.pdf
LT1215 - http://cds.linear.com/docs/Datasheet/12156fb.pdf

I've decided to use opamps in DIP package for an easy experimenting. However you could use just about anything you find suitable and use a DIP->SOIC adapters. Just remember that you need a PRECISION opamp in this stage. This would help keeping the DC offset low.

Few words about the project realization. This one is completed on a two-sided PCB using smd components and TH metalization. A special attention was paid on the bottom ground plane to minimize the parasitic capacitance. More info on this matter can be found in the TPA6120's datasheet:

www.ti.com/lit/ds/symlink/tpa6120a2.pdf

"A ground plane should be used on the board to provide a low inductive ground connection. Having a ground plane underneath traces adds capacitance, so care must be taken when laying out the ground plane on the underside of the board (assuming a 2-layer board). The ground plane is necessary on the bottom for therma reasons. However, certain areas of the ground plane should be left unfilled. The area underneath the device where the PowerPAD is soldered down should remain, but there should be no ground plane underneath any of the input and output pins. This places capacitance directly on those pins and leads to oscillation problems. The underside ground plane should remain unfilled until it crosses the device side of the input resistors and the output series resistor.

The power supply regulation is complete on the board itself near the TPA chip. This requires a +/-18 to +/-21VDC prefiltered.

The complete schematic for this project can be downloaded HERE.

Here is how the PCB looks like:

Since it was the first batch of PCB's I had some problems with the silkscreen printing. However it's not a big issue at the moment. In order to keep the return paths as short as possible I was forced to use jumpers. This could be easily solved with a 4-layer PCB. These come at a price though.

WARNING: This project was brought to the DIY community for free. It is intended for personal DIY needs only. Commercial usage of this project is not allowed!

+++T H A N K Y O U F O R R E A D I N G T H I S A R T I C L E+++

ECC85 RIAA Preamplifier

2011-09-19T13:52:00.000-07:00

This idea came to me when a friend from the forum asked for a simple schematic based on the ECC85 tube.
He had plenty of these around and wanted to try them out. So I sat and and made some calculations. Since we had a 300V B+ supply I had to make calculations for this supply voltage. The preamp itself is a pure classic. Two stages separated by the RIAA EQ network. The RIAA parts were calculated using this tool:

http://www.kabusa.com/riaa.htm

I have used it many times. It's always spot on. Of course when calculating RIAA equalization some parts of the schematic must be taken into account. Since the output impedance of the common cathode stage is pretty high it can not be neglected. The output impedance is effectively in series with R1(from the link), altering it's value. This of course leads to an incorrect RIAA equalization. The other part altering R1 value is the next stage input resistor. This must be taken into account too.

Of course before calculating the RIAA equalization network I needed to set the operating conditions of the two gain stages. Those stages are identical. Both sections are biased at 175V anode voltage, -2V grid, 6mA anode current. This setup gives about 28 of voltage gain using a 20K anode resistor. Here is the bias point:

The total gain of the preamplifier is about 37.9dB. This will give you around 400mV output from a cartridge giving around 5mV of output.

The complete schematic is available for download HERE.

Here is the predicted nonlinearity of the RIAA equalization for this project (+/-0.3dB):

***************THANK YOU FOR READING THIS ARTICLE***************

DIY USB Isolator Made Easy

2011-09-12T12:55:00.000-07:00

Are you into high quality PC audio? Do you have one of these nice USB DAC's that pull all the good musical stuff out of your computer? Then you might need to check this article out. You can vary well improve your USB DAC performance by isolating it from all the garbage coming from the noisy(electronically) PC.

What I'm about to show in this article is a way to build you own USB isolator and improve your USB audio. The isolator itsel is build around the ADUM4160 chip from Analog Devices. The complete datasheet is here:
http://www.analog.com/en/interface/digital-isolators/adum4160/products/product.html#prod_cross_sell

Since the schematic for this project is rather simple, a single layer printed circuit board can be made. Of course a double layer is always better option.

This is how the whole thing looks:

The project also makes use of one high quality LDO regulator. In this case I'm using Linear Technology's LT1763 fixed voltage (5V). I've used this one before in some other projects and I'm quite happy with it. The datasheet is available here:

http://cds.linear.com/docs/Datasheet/1763fg.pdf

The board allows you to use external regulated PSU just in case you don't trust the LT's chip. This is the complete schematic for this project:

The circuit is pretty straightforward but here is some explanation. JP1 switches the speed - open=low speed(1.5Mbps), close=full speed(12Mbps).

JP2 function is explained here:
The ADuM4160 has an option to delay application of the upstream pull-up under control of the peripheral. This function is controlled by the PIN input. In this application, the PIN input is jumpered high so that the upstream pull-up is applied as soon as peripheral power is applied. In other applications, it can be connected to a GPIO pin of a controller, a fixed delay circuit can be applied, or it can be connected as shown in this circuit. It is the designer’s choice how to use this functionality.

I personally use id closed. JP3 allows you to switch between the power supplies of the second half of the isolator. I recommend using the build in regulator. In this case the board need 10-15VDC. A simple DC adapter will do the job.

The PCB artwork in 1:1 scale pdf file and component placement is available for download HERE.

The PCB can be easily made using a laser printer and an iron. I encourage you to try this simple upgrade, as you won't be disappointed.

************THANK YOU FOR READING THIS ARTICLE************

DIR9001+TDA1541A Classic NOS DAC

2011-09-09T04:21:00.000-07:00

The whole thing started as a simple idea - to design a simple S/PDIF input DAC for my PC. So I turned to the classical chipset - DIR9001 + TDA1541A in NOS mode. I just wanted to keep thing simple and at low cost(compact PCB). So eventually the thing grew up to a group project with many people sharing opinions on the mater.
The whole discussion thread is here(Bulgarian language):

http://www.bgaudioclub.org/forum/viewtopic.php?t=22012

Since some people suggested to make it possible to use TDA1541(non A) DAC chip the project was altered this way. The interesting thing about TDA1541(non A) is that it allows a master clock to be fed directly to pin 4. This one accepts 11.2896MHz clock frequency. In TDA1541A non such thing is available.

From TDA1541 datasheet:
"A separate system clock input (SCK) is provided for accurate, jitter-free timing of the analogue outputs AOL and AOR."

So the PCB was designed to allow this mode, should a TDA1541(non A) was used. For some reason Philips had decided to turn that feature off in their later TDA1541A DAC chip. Here is how it should be implemented according to a Philips application note from 1985 :

The application is a bit misleading. The SAA7220 chip actually outputs 11.2896MHz at pin 9.

As for the analog section. It's of course a pure classic. Opamp based I/V converter, followed by an active low-pass filter. Nothing fancy here. The project uses one double opamp (NE5532) per channel. Of course a DC null correction is used at the DAC chip outputs. Despite that a output coupling capacitor is paced on the PCB.

Here are some pictures of the PCB:

Almost completed project forum mate Stanislav

Higher end implementation from forum mate Nikolay Kolev

Here is a set of measurements taken by user pid_58 (thnaks mate ;) ) using EUM 0404 sound card.

The final version of the schematic is available for download HERE.

WARNING- this project is provided to the DIY community for free and is intended for personal usage ONLY. Therefore a commercial usage is NOT allowed without author's exclusive permission!!!

*****************THANK YOU FOR READING THIS ARTICLE*****************

Low Jitter Clock For CD Player - Upgrade Module

2011-09-07T08:32:00.000-07:00

Do you have one of those expensive CD players that use a simple quartz oscillator to provide the clock? Anyways, the clock upgrade is no doubt the most important upgrade to any CD player. Otherwise you will never hear that nice smooth analog like sounding from your gear.

What you need is a low phase noise clock source to feed the CD player.
Searching the web will come out with many ways to upgrade your CD player. Some of those upgrades come at substantial costs. Many sources simply don't provide enough data to prove those high costs in my opinion.

So i searched and searched for a nice clock project to upgrade mu trusty Arcam Alpha 5 CD player. This one was already upgraded with a nice new TDA1541A S1 DAC chip so I was impatient to reveal its true potential.

So finally I came across a nice article about The Flea Clock from Pink Fish Media. An very interesting project indeed. What that project aims at is in reality an ultra low noise PSU to power up a Tent Labs oscillator. So my project was based on that one from PFM forum.

I just had to change some things to fit my personal needs.
First - the oscillator was unable to drive any substantial capacitive loads, so I needed to buffer the outputs. Buffering is an easy task. A simple inverter would do the job. In this project I'm using two 74VCH1G04 ic's to provide me two outputs.
Second - i wanted a oscillator with real measured parameters showing the phase noise and jitter performance since I had no way(equipment) to measure the jitter myself. Few manufacturers that in My opinion are suitable for the task:

Fox Electronics HC-73 series oscillators:
http://www.foxonline.com/pdfs/FXO_HC73.pdf

EuroQuartz XO-91 series oscillators:
http://www.euroquartz.co.uk/Portals/0/xo91.pdf

Silicon Labs Si510 series oscillators:
http://www.silabs.com/pages/DownloadDoc.aspx?FILEURL=Support%20Documents/TechnicalDocs/si510-11.pdf&src=DocumentationWebPart

In my case I'm using 11.2896MHz oscillator to feed the digital filter's XIN pin(pin 11 on SAA7220 chip). This is the most basic way to feed the new clock to your CD player. Simply remove the quartz form the player and the two caps and the resistor. Here is how I did it:

A more advanced reclocking to your player can done using a flip-flop to reclock the I2S bus before fitting it to the TDA1541A chip. Here is how its done:

This method has one major disadvantage. It uses one IC to process all three wires of the I2S bus. A better way is to use three IC's but it's cost and space consuming.

Possibly the best way is to feed the clock separately to the digital filter(SAA7220) and the DAC chip (TDA1541A). The DAC chip itself requires half the frequency provided to the digital filter. So we need to divide the clock coming from the module. This can be done again using a flip-flop:

In my opinion this is the better sounding reclock. Anyways, a wide field for experiments is available.

So this is the final result. Here is the clock module PCB fitted into my Arcam Alpha 5 CD player

Ultra low jitter clock fitted into Arcam Alpha 5 CD player

The module itself is powered by a preregulated PSU using a LM317 vreg providing 18VDC. I'm using a small transformer to power the whole thing up. This is the better way to do it. One can still use the already available supply rails from the CD player but 18VDC is the minimum voltage required in order the module to work.

A complete schematic for this module is available HERE.

WARNING - This design is provided to the DIY community for free and is intended for personal use only. Therefore the commercial usage is NOT allowed.

*************THANK YOU FOR READING THIS ARTICLE**************

The V.S. REF RIAA Phono Preamplifier - SSM2017 Based Project

2011-09-06T16:59:00.000-07:00

This project started as a experiment after reading the Solidophono article at TNT-Audio website. I was interested in the SSM2017 opamp so I've decided to give it a try.

On my project I've decided to go for a bit different topology using a split RIAA equalization. All my previous projects were created using a traditional passive equalization so I was exited to see the results.

The whole preamp is made of two stages. The first one is built around the SSM2017 amplifier from Analog Devices.
http://www.analog.com/static/imported-files/data_sheets_obsolete/139484016SSM2017.pdf
One could use the modern substitutes - SSM2019 or INA217. The gain of the first stage was set at 60dB! The second stage uses the OPA627 opamp from Texas Instruments. One could experiment with a variety of opamps here. The gain was set to 2. I used a jumper to change the second stage gain to unity in case no further amplification was needed.

The complete schematic for this project can be downloaded HERE.

Some pictures of the finished PCB. The project is using two separate aluminum boxes. One for the PSU and one for the preamp PCB itself.

Almost finished PCB.

The finished PCB inside a nice aluminum box.

The four voltage regulators are placed on the PCB. I'm using a small 20VA toroidal transformer to power up the whole thing. Here is the PSU section:

As you can see, this is a serious power supply. Such configuration could be seen in some expensive power amplifiers. Such a PSU has a very high ripple suppression and since the current demands are no big issue in RIAA preamps, it's affordable and easy to build as well.

So how does it sound?
Well, I'm not going to describe my own creations as my judgement could be a bit subjective. However I could say this preamp has some very good dynamics with good transient attack.

I encourage you to try this project and see for yourself.

WARNING: The project is provided to the DIY community for free. As such it can only be used for personal needs and NOT for commercial usage.

THANKS FOR READING THIS ARTICLE.