Audio Amp Component Selection

[schmitt trigger]

Do you know if the engine is 1200 or 1100 CC?
I'm eagerly awaiting for your photos. My friend's radio (which I talked before) had a defective vibrator. He threw the thing away!!! Can you believe it? Nowadays, it would be a real collectible.

 

[vne147]

Do you know if the engine is 1200 or 1100 CC?
I'm eagerly awaiting for your photos. My friend's radio (which I talked before) had a defective vibrator. He threw the thing away!!! Can you believe it? Nowadays, it would be a real collectible.

1200 cc's and 36 horses of pure fury baby! Enough to smoke any lawnmower out there.

I'm going over my friend's house tomorrow. I'll get some pictures then. Is there anything specific you want pictures of or just some overall shots of the radio?

 

[schmitt trigger]

Just the radio, and the instrument cluster if you may (if you can actually call that an instrument cluster).

My friend's engine was 1100 cc. The speedometer went all the way to 80 Km/hr, which you could only be achieved full throttle on a steep downhill.

 

[vne147]

Can someone double check me on this?

I want to use this amplifier to achieve a desired power output of 5W to a 4 Ω speaker. After looking through the data sheet I think that I'll need a 13V supply that can source at least 600 mA. Can someone double check this for me please? Here's the data sheet.

**broken link removed**

Also, I need to convert ~6V to 13V to be used as the supply. I'm thinking of using this boost converter.

**broken link removed**

The data sheet says it's capable of supplying up to 1.5 A but the application circuit example shows an output of 28V/175mA so I'm a little confused.

Thanks.

 

[audioguru]

The TDA2003 datasheet has a graph that shows its distortion with a 14.4V supply and a 4 ohm load. It clips at about 4.2W and at 5W its distortion is 0.3% and is becoming audible.
If the supply is only 13V then it clips at only 3.5W.

The 34063A can switch the low input voltage to the inductor at up to 1.5A but the output current is much less because the voltage is stepped up. It cannot manufacture additional Watts.

 

[schmitt trigger]

Yes indeed, you need a boost converter.

My suggestion on this one, since you are not familiar with the design of switchmode power supplies, is to use a Simple Switcher device from National Semiconductor.

They have an on-line simulator that works amazingly well. You only enter the input/output parameters, it will suggest devices from its extensive lineup, and it will calculate the required component values and provide all sorts of plots and data.

 

[vne147]

The TDA2003 datasheet has a graph that shows its distortion with a 14.4V supply and a 4 ohm load. It clips at about 4.2W and at 5W its distortion is 0.3% and is becoming audible.
If the supply is only 13V then it clips at only 3.5W.

The 34063A can switch the low input voltage to the inductor at up to 1.5A but the output current is much less because the voltage is stepped up. It cannot manufacture additional Watts.

Audioguru,

I understand that power is conserved. It's just that I have a VERY basic understanding at best of boost converter theory.

I think I understand your comments about the clipping at 3.5W with a 13 V supply. I also think based on your comments that if I use a higher supply voltage say 17V, I can achieve 5W output with much lower distortion. Is this a correct statement?

If that's true, then there would be a problem because with a 17V supply, the maximum output of the amp could go to ~ 8.5 W with a 4 Ω load which would likely blow my 5W speaker. Could I protect my speaker by adding an extra resistor at the input (resistor labeled Rp in attached schematic)?

I would just have to make sure that the resistor value I pick will ensure that the power output never exceeds 5W even when the volume pot is turned all the way to it's minimum resistance. How would I go about determining an appropriate resistor value?

Yes indeed, you need a boost converter.

My suggestion on this one, since you are not familiar with the design of switchmode power supplies, is to use a Simple Switcher device from National Semiconductor.

They have an on-line simulator that works amazingly well. You only enter the input/output parameters, it will suggest devices from its extensive lineup, and it will calculate the required component values and provide all sorts of plots and data.

Schmitt Trigger,

I went to the National Semiconductor website and used the tool you suggested. It looks like the LM2588 is going to do what I need. What do you think? The only downside is as best I can tell Mouser does not carry any of these parts and they are my preferred supplier. They are in Dallas and I'm in Houston so shipping is usually cheap and I almost always get my order the next day even though I only pay for normal UPS ground. I also wasn't successful at finding any equivalent ICs made by other manufacturers that Mouser carries. Oh well, you can't have everything I guess.

 

[schmitt trigger]

Indeed, the LM2588-ADJ will work for your requirement.

I also live in Texas, and sometimes have used Mouser for the reasons that you mention, but if you order from DigiKey, they do carry National Semi's line and you avoid paying the state sales tax.

 

[audioguru]

I understand that power is conserved. It's just that I have a VERY basic understanding at best of boost converter theory.

1) Power out must equal power in, minus some losses.
6V at 3A (18W) will be stepped up to 12V at about 1.2A (14.4W) plus 3.6W of heat.
A 5W amplifier is about 50% efficient so it uses a max power of 10W from its power supply. The extra 5W makes heat in the amplifier.

I think I understand your comments about the clipping at 3.5W with a 13 V supply. I also think based on your comments that if I use a higher supply voltage say 17V, I can achieve 5W output with much lower distortion. Is this a correct statement?

An amplifier has low distortion until its output runs out of supply voltage and clips which produces severe distortion (the audio waveform is "clipped" into a square-wave). Unfortunately all amplifier manufacturers exaggerate the output power by spec'ing 10% distortion where the volume control is turned up too high so the output is a badly clipped square-wave that sounds awful.

In the datasheet the graph for power with a 14.4V supply and into 4 ohms with 10% distortion shows 6W. But the graph for distortion with a 14.4V supply and into 4 ohms shows the distortion suddenly climbing (clipping) at only 4.2W. So the ratio is 4.2W/6W= 0.7.
It shows 8.5W with a 17V supply and a 4 ohm load with 10% distortion so it clips at 8.5W x 0.7= 5.95W.

If that's true, then there would be a problem because with a 17V supply, the maximum output of the amp could go to ~ 8.5 W with a 4 Ω load which would likely blow my 5W speaker. Could I protect my speaker by adding an extra resistor at the input (resistor labeled Rp in attached schematic)?

A resistor in series with a speaker makes it sound boomy because the resistor destroys the damping of resonances provided by the extremely low output impedance (0.04 ohms or less) of an amplifier.

We don't know if the speaker is rated in Watts or in Whats. We don't know if the speaker can survive 5W for 1ms or for 1 hour. Music is never at full volume all the time, its average power is probably 1/10th its peak power.
I had a speaker that was marked "5W". It was tiny (4") and I fed it from my 70W amplifier for years. It was rated at 40 Whats.

 

[vne147]

I’m having some trouble moving forward with my design. My difficulty is coming mostly in the form of determining component values for the switching converter portion of my circuit but I have a few other questions too. I downloaded the switchers made simple software from National Semiconductor and also tried their online tools but for the life of me I still couldn’t figure it out. Here are my questions:

1. The current plan is to connect the voltage going into the switching converter directly to the car’s battery, however this voltage will not be constant. The battery voltage of the car varies between 6 and 7.5V depending on engine RPM. Is it necessary and/or beneficial to place a voltage regulator in between the vehicle battery and the switching converter so that the converter sees a nice stable voltage at its input?
   
   
2. I’m unsure, because of my inability to use National’s software, what component values to use for the switching converter. I attached a schematic of my preliminary circuit. The components shown with red labels are the ones I’m unsure of.
   
   
3. It would also be helpful to get some feedback on the circuit itself. Is there anything you see that’s outright wrong? Are the other component values I’ve chosen (the ones without red labels) appropriate? Any comment you have would be helpful.

Here are links to the data sheets for all the parts I’m using:

LM2588:

https://www.electro-tech-online.com/custompdfs/2009/11/LM2588.pdf

LM317:

https://www.electro-tech-online.com/custompdfs/2009/11/LM317.pdf

TDA2003:

http://www.datasheetcatalog.org/datasheet/stmicroelectronics/1449.pdf

As always thanks in advance for your help.

P.S. Schmitt Trigger, sorry I’ve been busy with work but I haven’t forgotten about you. I have some pictures of the old VW radio I’m going to post tonight.

 

[audioguru]

Get rid of the LM317 voltage regulator because the switching converter regulates its output voltage. Set it for 15V then the TDA2003 amplifier will produce 5W into 4 ohms.

 

[vne147]

In the datasheet the graph for power with a 14.4V supply and into 4 ohms with 10% distortion shows 6W. But the graph for distortion with a 14.4V supply and into 4 ohms shows the distortion suddenly climbing (clipping) at only 4.2W. So the ratio is 4.2W/6W= 0.7.
It shows 8.5W with a 17V supply and a 4 ohm load with 10% distortion so it clips at 8.5W x 0.7= 5.95W.

The power output vs supply voltage graph shows 6.5W at 15V. Based on your explaination above, wouldn't a 15V supply clip at 4.5W?

Also, the ouput of the LM2588 is regulated but according to the datasheet on page 21 it does have some voltage ripple at what looks like about 100-200 mV. I sthat not going to affect the amp? Do I not need to worry about that?

Thanks.

 

[audioguru]

Based on your explaination above, wouldn't a 15V supply clip at 4.5W?
The ratio from no distortion to 10% clipping distortion is 0.7 times. So 6.5W at 10% distortion with a 15V supply means that the amplifier begins clipping at 4.55W (simplified calculations). The difference in volume between 9W and 4.5W is small. You will not hear the difference between 4.55W and 4.5W.

Also, the ouput of the LM2588 is regulated but according to the datasheet on page 21 it does have some voltage ripple at what looks like about 100-200 mV. I sthat not going to affect the amp? Do I not need to worry about that?

Amplifiers have a power supply hum rejection spec where the ripple is attenuated. It should be very low at the output of the amplifier.

 

[schmitt trigger]

1. The current plan is to connect the voltage going into the switching converter directly to the car’s battery, however this voltage will not be constant. The battery voltage of the car varies between 6 and 7.5V depending on engine RPM. Is it necessary and/or beneficial to place a voltage regulator in between the vehicle battery and the switching converter so that the converter sees a nice stable voltage at its input?
   
   
2. I’m unsure, because of my inability to use National’s software, what component values to use for the switching converter. I attached a schematic of my preliminary circuit. The components shown with red labels are the ones I’m unsure of.
   
   
   /LIST]
   
   
   P.S. Schmitt Trigger, sorry I’ve been busy with work but I haven’t forgotten about you. I have some pictures of the old VW radio I’m going to post tonight.

1. 
   
   -The Nat Semi's software WILL PROVIDE you with both an annotated schematic and a Bill of Materials which lists the value AND full description of each component.
   See the attached file of an example circuit which I had run previously.
   
   -Also as Audioguru mentioned, you DO NOT require an additional linear regulator after the switching regulator. Yes, there will be some ripple, but it will be at least 15 to 20 times higher than the highest audible frequency. Very straigthforward to filter, if required.

 

[audioguru]

The switching converter does not need its input voltage regulated.

 

[vne147]

The ratio from no distortion to 10% clipping distortion is 0.7 times. So 6.5W at 10% distortion with a 15V supply means that the amplifier begins clipping at 4.55W (simplified calculations). The difference in volume between 9W and 4.5W is small. You will not hear the difference between 4.55W and 4.5W.

Amplifiers have a power supply hum rejection spec where the ripple is attenuated. It should be very low at the output of the amplifier.

Audioguru,

Thanks for the advice. I'll ditch the LM317, change the supply voltage to 15V, and also forget about placing a voltage regulator at the LM2588 input.

-The Nat Semi's software WILL PROVIDE you with both an annotated schematic and a Bill of Materials which lists the value AND full description of each component.
See the attached file of an example circuit which I had run previously.

Schmitt Trigger,

I'm sure that the software will provide me with the info you stated, but I just can’t figure out how to work it. The problem I keep running into is I can't figure out how to load the adjustable version of the LM2588. Whenever I select the LM2588 the schematic and component values are for the 12V version and I can figure out how to change that. Maybe I'm slower than the average bear but I’ve seriously spent hours trying to figure it out. I know it's asking a lot but if you have time would you mind seeing if you can run the software for the following conditions:

Device: LM2588-ADJ
Input Voltage: 5 - 8V
Output Voltage: 15V
Max output current: 800 mA

For the current requirement I looked at the data sheet for the TDA2003 and found that for a 15V supply across a 4 Ω load the maximum power output would be 6.5W. Also from the data sheet the maximum power dissipation looks to be around 3W. So:

6.5W + 3W = 9.5W, I = P/V = 9.5W/15V = 633 mA.

I just bumped it up to 800 mA for a little buffer. Does that seem reasonable?

I would really appreciate your help with the software. Thanks in advance.

Also, as promised here are some pictures of the VW radio. There are actually pictures from two different radios here. The working radio is installed in the car and my friend didn't want to take it out because he said it is a major pain in the a$$ to get it back in. However, he has a second non-working radio that he used for parts. It is almost complete and is just missing one of the tube amplifiers. I hope it satisfies your curiosity. Enjoy.

And last but not least:

The plan seems too complex and somewhat wrong for the task.

If it will stay with one speaker to keep originality, the MP3 source can have both earphone or better line-out channels merged and fed to the radio volume potentiometer with a selector switch.
Is that a vacuum tubes radio ? Is it 6 V ?

The 5V MP3 power source can be obtained from the 6V VW electrical with a couple of 1A Si diodes in series to bring it to around 5V. A 7805 is not a good choice. Perhaps a 2950-05 if you really want regulation.

Externet,

I know you made this reccomendation much earlier in the thread. I just wanted to let you know that I suggested the idea to my friend but he was much more receptive to a solution that didn’t require doing anything to the antique and expensive radio. Also, your solution was very simple and would have taken away my project. Anyway, just didn’t want you to think I was blowing you off. Thanks for the input.

 

[audioguru]

It will sound like an AM radio. Bad sound quality.

 

[schmitt trigger]

VNE147:
I already simulated it, and the solution is included on figure STEP 3, but I really want you to learn how to do it; because you may want to play around with the component selection, as some of the components you may have the need to substitute, and you can run the simulation again and see how the performance is affected.

So here it goes, in three easy steps:

On National's web site, on the WEBBENCH DESIGNER box, enter the apropriate voltage and current parameters see figure STEP 1. Click "START DESIGN".

If you have not decided upon your preferences, the next screen will ask you whether you want an HTML or Flash-based design. The Flash is much better, so select that.

Then a screen will appear with the recommended devices, see figure STEP 2. Highlight LM2588-ADJ and press "START DESIGN".

The computer will think for a few moments and Voila! you get screen shown in STEP 3. There it is, your solution. On the "Bill of Materials" you can scroll the component selections. In case you want to substitute. Click the "Select Alternate Part" Button, and the software will automatically recalculate everything and display that for you.

Also, in the schematic itself, if you point the cursor on any component (STEP 4), it will provide you with full details on that particular component.

It is really easy!

 

[vne147]

It will sound like an AM radio. Bad sound quality.

What is going to have poor sound quality? The circuit I'm about to build? I don't think my friend is much of an audiophile. Given the speaker I have to use, what can I do to improve the sound quality?

schmitt trigger,

I followed your instructions, sort of. I have to use the html version because I'm at work and we're restricted here. Anyway, I could have sworn I already did what you described but this time it worked. I mean it's pretty simple like you stated so I don't know where I went wrong before. Thanks!

 

[audioguru]

A CD sounds perfect.
An MP3 and an FM stereo radio sound very good.
A high quality cassette tape deck with Dolby noise reduction and high bias tape sounded pretty good 20 years ago.
But a lousy old AM radio sounds awful.

Most people can hear frequencies as high as 20kHz. But an AM radio rolls off frequencies above only 3.5kHz like a telephone. An AM radio also has high distortion and lots of noise.