Anyone interested in a digital amplifier project? - Page 14 - diyAudio
Go Back   Home > Forums > Amplifiers > Class D

Class D Switching Power Amplifiers and Power D/A conversion

Please consider donating to help us continue to serve you.

Ads on/off / Custom Title / More PMs / More album space / Advanced printing & mass image saving
Reply
 
Thread Tools Search this Thread
Old 18th December 2003, 07:09 PM   #131
diyAudio Member
 
Join Date: Sep 2002
Location: US
Quote:
Originally posted by dwk123

Wow - a total of only 28Ah? Thats much less than I would have expected. I forget exactly what I was looking at, but I was thinking along the lines of 2x 80Ah. Sounds like that might be unnecessary.
Yeah, they are really efficient. One charge last about a week listening, couple of hours a day.

Quote:

How many channels are you using - all 5 or only 4?
4 channels, 2 almost full range and 2 below 45Hz.

Quote:

Assuming you're right about the 0dB level being around -20 on the volume dial, I can see this working. I usually listen around -30 or so and so *might* get away with 24V, but maybe not - there might not be much headroom.
I had no time to verify, but I'd rather say somewhere between -12 - -18dB. One day I'll feed the i2s from the Pana directly to the crossover and can tell exactly.
I've tried 6V supply, and it worked great with high efficiency speakers - 2A3 tubes be afraid
  Reply With Quote
Old 18th December 2003, 08:07 PM   #132
diyAudio Member
 
Join Date: Aug 2002
Location: Wisconsin
Default Supply Levels and Batteries

The following relationship applies when using the power supply voltage for volume control:

Vcontrol=Vorig*2^([dB change]/6)

For example, for an original supply voltage of 42V, to get a -5dB adjustment:

42V*2^([-5dB]/6dB)=23.7V

Here's a chart showing the relationship for the SA-XR10, which has a max supply voltage of 42.2V:
(The 65% wattage adjustment factor is an approximation for max output duty cycle and output losses.)

Supply 42.2V
Load resistance 8ohms
.................................................. 65%factor
dB......Vsup.......Wp..........Wrms......Wrms
0........42.2........222.6......157.4......102.3
-1.......37.6........176.7......124.9......81.2
-2.......33.5........140.2......99.2........64.5
-3.......29.8........111.3......78.7........51.2
-4.......26.6........88.3........62.5........40.6
-5.......23.7........70.1........49.6........32.2
-6.......21.1........55.7........39.4........25.6
-7.......18.8........44.2........31.2........20.3
-8.......16.7........35.1........24.8........16.1
-9.......14.9........27.8........19.7........12.8
-10.....13.3........22.1........15.6........10.2
-11.....11.8........17.5........12.4........8.1
-12.....10.6........13.9........9.8..........6.4
-13.....9.4..........11.0........7.8..........5.1
-14.....8.4..........8.8..........6.2..........4.0
-15.....7.5..........7.0..........4.9..........3.2
-16.....6.6..........5.5..........3.9..........2.5
-17.....5.9..........4.4..........3.1..........2.0
-18.....5.3..........3.5..........2.5..........1.6
-19.....4.7..........2.8..........2.0..........1.3
-20.....4.2..........2.2..........1.5..........1.0

Here's what I actually measured on the SA-XR10:
(It matches the above calculations pretty well)

0dB 42.2V
-1dB 37.6V
-2dB 33.5V
-3dB 29.8V
-4dB 26.6V
-5dB 23.7V
-6dB 21.1V
-7dB 18.8V
-8dB 16.8V
-9dB 15.0V
-10dB 13.4V
-11dB 11.7V
-12dB 10.6V
-13dB 9.5V
-14dB 8.5V
-15dB 7.6V
-16dB 6.8V
-17dB 6.1V
-18dB 5.4V
-19dB 4.8V
-20dB 4.3V

Here's another chart showing a 48V supply (which is right at the rating limits of the parts):

Supply 48.0V
Load resistance 8ohms
.................................................. 65%factor
dB......Vsup.......Wp..........Wrms......Wrms
0........48.0........288.0......203.6......132.4
-1.......42.8........228.6......161.6......105.1
-2.......38.1........181.4......128.3......83.4
-3.......33.9........144.0......101.8......66.2
-4.......30.2........114.3......80.8........52.5
-5.......26.9........90.7........64.1........41.7
-6.......24.0........72.0........50.9........33.1
-7.......21.4........57.1........40.4........26.3
-8.......19.0........45.4........32.1........20.8
-9.......17.0........36.0........25.5........16.5
-10.....15.1........28.6........20.2........13.1
-11.....13.5........22.7........16.0........10.4
-12.....12.0........18.0........12.7........8.3
-13.....10.7........14.3........10.1........6.6
-14.....9.5..........11.3........8.0..........5.2
-15.....8.5..........9.0..........6.4..........4.1
-16.....7.6..........7.1..........5.1..........3.3
-17.....6.7..........5.7..........4.0..........2.6
-18.....6.0..........4.5..........3.2..........2.1
-19.....5.3..........3.6..........2.5..........1.6
-20.....4.8..........2.8..........2.0..........1.3
-21.....4.2..........2.3..........1.6..........1.0

(Note that only about 1dB additional volume is obtained by the extra six volts)

Using switched 12V batteries would give the following:
48V.......0dB
36V.......~-3dB
24V.......-6dB
12V.......12dB

******************

One of the concerns I had about using unregulated batteries for the output bridge supply was regarding a possible decrease in volume over time. Starting with 48V, the above chart shows that the voltage has to drop a full five volts before a 1dB decrease occurs. With 7Ah batteries, an amp could operate for just over an hour at maximum continuous power (one channel) before a 1dB drop occurred. Music statistically doesn't exceed 1/8 of full power. With this consideration, the amp could output full volume for over six hours before a 1dB drop occurred.

A 12V nominal SLA battery typically has about a 13V initial voltage. Within about a minute, at light to moderate loads, the voltage will drop to about 12V. Once this first drop occurs, the voltage will remain within a couple of tenths of volts for quite a long time, even with drastically changing loads. As long as there is sufficient capacitance in parallel with the battery, any of this voltage variation will occur at subsonic frequencies, and will only cause a small fraction of a dB in amplitude variation.

Once I realized the above I decided to abandon attempting electronic supply regulation.

I think this is a beautiful example of a brute force design. An SLA battery and parallel capacitor is an extremely stiff supply over the audio bandwidth. This is just what an open loop switched output section of a digital amp needs. Better still, the high efficiency of the switched output makes batteries a practical consideration.

******************

A final concern is the potential for crosstalk or other channel interaction when powering more than one channel from the same battery. Careful layout and decoupling could avoid this. Of course, using separate sets of batteries for each channel completely avoids this issue.

SLA battery pricing is somewhat proportional to the number of amp-hours you are buying. Because of this, I think it makes sense to buy separate batteries for each channel. (i.e. more smaller batteries vs. fewer larger ones.)

******************

A couple of implementation issues:

Directly connecting a battery to a parallel capacitance causes a very large in-rush current. Worst case, this can cause a capacitor to explode. Even if the capacitors survive the in-rush, they will be degraded over time.

One easy solution is to put a series power resistor between the battery and capacitor. Once the cap has been precharged through the resistor, a relay can be used to short out the resistor and directly connect the cap to the battery.

With a little more work, the precharge circuit could also provide a means to switch in a battery charger when the amp wasn't being used.

Regards,
Brian.
  Reply With Quote
Old 18th December 2003, 08:42 PM   #133
diyAudio Member
 
Join Date: Sep 2002
Location: US
Default Re: Supply Levels and Batteries

Quote:
Originally posted by Brian Brown

With this consideration, the amp could output full volume for over six hours before a 1dB drop occurred.
It is more like 10 hours for 4 channel for me, but my speakers are relatively light loads in the midband, and the sub channels rarely go above -30dB. The average level of classical music is also lower. With continuous recharging every day it is not a problem for me.

Quote:

As long as there is sufficient capacitance in parallel with the battery, any of this voltage variation will occur at subsonic frequencies, and will only cause a small fraction of a dB in amplitude variation.
.
.
A final concern is the potential for crosstalk or other channel interaction when powering more than one channel from the same battery.
None of them seems to has any audible effect in the current layout. I also could not measure any effect of it but my measurement tools are very limited.
  Reply With Quote
Old 26th December 2003, 11:42 PM   #134
diyAudio Member
 
Tube_Dude's Avatar
 
Join Date: Mar 2002
Location: Aveiro-Portugal
Default Sony STR-DB2000

Hi Folks!


To the people insterested in digital amplifiers

After my work in the Panasonic SA-XR10 (see posts # 77, 78, 79)

Here is my new baby(victim!!!) ... the Sony STR-DB2000.

Here is a pic of the front of the "beast"
Attached Images
File Type: jpg im000024.jpg (55.4 KB, 1149 views)
__________________
Jorge
  Reply With Quote
Old 26th December 2003, 11:47 PM   #135
diyAudio Member
 
Tube_Dude's Avatar
 
Join Date: Mar 2002
Location: Aveiro-Portugal
Default Sony STR-DB2000

Here is the inside of the "victim". Gone is the video board!...the phono and analog boards (anyway who wants a phono stage in a digital amplifier?)
Now is near a straight line digital amplifier
Attached Images
File Type: jpg im000025.jpg (63.8 KB, 1166 views)
__________________
Jorge
  Reply With Quote
Old 26th December 2003, 11:48 PM   #136
diyAudio Member
 
Tube_Dude's Avatar
 
Join Date: Mar 2002
Location: Aveiro-Portugal
Default Sony STR-DB2000

Here is the switching power supply!
Attached Images
File Type: jpg im000026.jpg (65.8 KB, 1512 views)
__________________
Jorge
  Reply With Quote
Old 26th December 2003, 11:50 PM   #137
diyAudio Member
 
Tube_Dude's Avatar
 
Join Date: Mar 2002
Location: Aveiro-Portugal
Default Sony STR-DB2000

And finaly the digital input board and DSP!
Attached Images
File Type: jpg im000027.jpg (68.2 KB, 1486 views)
__________________
Jorge
  Reply With Quote
Old 24th January 2004, 06:24 PM   #138
diyAudio Member
 
Join Date: Feb 2002
This thread has been dormant for a while - I hope that means everyone is busy with their projects, not that they've gone idle.

I have some 12V batteries on the way to experiment with battery power on my two Panasonic Equibit units - the direct digital interface experiments are on hold for a while.

I was browsing the TI site again in preparation for this, and saw that there is a new multi-channel modulator out (or one I hadn't been aware of) - the 5076. The very cool thing for the DIY-er is that the EVM (using the 5182 and providing 6-channels @100W) appears to be directly orderable from the TI website.

The EVM looks to be perfect for experimentation, and they claim a measured performance of 105dB S/N on the two main channels from the evm board (102 for the other 4 channels). It appears that there is a serial port and associated PC software for config etc. including tweaking the output bridge timing params which appears to be critical to realizing the potential of the chips.

At $500 it looks pretty good to me. If I weren't between jobs at the moment, I'd order one immediately.

Link:
http://focus.ti.com/docs/toolsw/fold...5182c6evm.html
  Reply With Quote
Old 24th January 2004, 09:28 PM   #139
diyAudio Member
 
Join Date: Aug 2002
Location: Wisconsin
Default Yep... Still busy...

I've been looking at that new EVM as well.

The one thing that I've always been disappointed with TI's digital amp EVMs is that they don't include a reference power supply, one of the biggest quality variables.

They've mentioned a couple of power supply application notes:

SLEA027:
Power supply recommendations for DVD receivers.
"Gives basic requirements and guidelines for selecting the optimal power supply for a non-feedback digital amplifier used in a DVD receiver."

SLEA028:
Power supply requirements for AV receivers.
"Describes power supply and system considerations when designing Class BD-mode amplifier without feedback, for example a system with TAS5076/5182 used in AV receivers."

I've been very interested in reading these, but they're still not available.

I'm still using the SA-XR10 Equibit board while I work on other parts of the project.

The SA-XR25 / SA-XR45 use TAS5036 / TAS5182 chipsets. I don't have one of these receivers, but I've reviewed their schematics. The SA-XR25 is less than half the cost of TI's development board. The TAS5076 appears to be pin compatible with the TAS5036. Hacking into one of these receivers might be something to consider as an alternative.

The new six channel TAS5076 modulator chip looks very interesting. It appears to have better specs than all but the TAS5015 (which doesn't have an internal PLL). Previously the TAS5012 had the 2nd best specs.

The TAS5015 / TAS5012 use external capacitors to set the H-bridge deadtime. The TAS5076 has a digital adjustment (ABD Delay parameter). I like the idea of being able to easily tweak this value. This could be especially useful when matching the amp to a specific speaker driver. The ABD Delay parameter could be tweaked for minimum distortion using MLS measurements.

The TAS5076 has an interchannel delay parameter so that the switching of the six different H-Bridge channels can be offset from each other. This improves performance when operating from a single supply by preventing crosstalk at switching frequencies. (It still doesn't prevent supply crosstalk from occuring at audio frequencies.) If separate supplies are used, then this feature isn't important.

The other feature the TAS5076 has that the TAS5015 / TAS5012 don't is programmable correction for DC offset due to asymmetry in the output stage. I'm taking a wild guess that this may be the main reason the TAS5076 has better specs than the TAS5012.

In the meantime I'm still plugging on a multiple TAS3013 processor board. Hopefully it won't be too much longer...

Regards,
Brian.
  Reply With Quote
Old 24th January 2004, 11:01 PM   #140
JohnW is offline JohnW  Hong Kong
diyAudio Member
 
JohnW's Avatar
 
Join Date: Jan 2004
Location: Discovery Bay, Prague, Paris...
I’ve been evaluating digital modulators from Zetex, TI & Sony. All have been externally Re-latched using an external low phase clock circuit.

Here’s a brief round up of the results, all Dynamic Range results are Awtd and are about –2dB worse then optimal due to poor phase noise of the LVDS RX / TX used on the common Eval board.

THD @ 0dB 997Hz measured differentially, IMD results are the 3rd order products (18KHz & 21KHZ) of the stimuli tones at 19KHz & 20KHz @ –6dB each: -

NFM = Noise Floor modulation

Zetex 1MHz HPWM, DR = 118dB, IMD –63dB! NFM Greater then 10dB! THD 0.0019%

Zetex 2MHz HPWM, DR= >124dB, IMD –85dB, NFM Less then 10dB, THD 0.00050%

Sony CXD9634 / M65817AFP, DR=116.7dB, IMD –95dB, NFM Less then 1dB, THD 0.00037%

TI TAS5015 DR=114.7dB, IMD –115dB, NFM Less then 1dB, THD 0.00020%

The Zetex has very poor Noise Floor Modulation & IMD results @ 1MHz.

The Sony’s a good all round modulator, having a SRC & gain control, and good stop band rejection.

The TI has brilliant IMD & good THD results, but poorer stop band rejection, with no Gain or SRC. I’m very impressed with this modulator – as it only operates at ~ 384KHz PWMFs (x8 Fs)

Both the Sony and the TI sound VERY good – without an external SRC, the Sony beats the TI, although the TI beats the Sony when it’s combined with a good SRC (the SRC side-steps the TI’s Stop Band image problems).

For some FFT plots of the TI, see my posting “Wanted circuit for Low Phase Noise 100MHz clock”
  Reply With Quote

Reply


Hide this!Advertise here!
Thread Tools Search this Thread
Search this Thread:

Advanced Search

Posting Rules
You may not post new threads
You may not post replies
You may not post attachments
You may not edit your posts

BB code is On
Smilies are On
[IMG] code is On
HTML code is Off
Trackbacks are Off
Pingbacks are Off
Refbacks are Off


Similar Threads
Thread Thread Starter Forum Replies Last Post
Noise measurement amplifier - interested? jackinnj Solid State 15 27th March 2008 10:36 PM
New digital XO project borges Multi-Way 2 8th January 2008 07:49 PM


New To Site? Need Help?

All times are GMT. The time now is 05:29 PM.


vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2014 DragonByte Technologies Ltd.
Copyright ©1999-2014 diyAudio

Content Relevant URLs by vBSEO 3.3.2