Optimizing TDA7294 Output - Page 34 - diyAudio
Go Back   Home > Forums > Amplifiers > Chip Amps

Chip Amps Amplifiers based on integrated circuits

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 17th February 2013, 11:27 PM   #331
bcmbob is offline bcmbob  United States
diyAudio Member
 
Join Date: Dec 2009
Location: Bay City, Michigan
I have a bunch of small caps from the MyRef FE tuning process and already started sorting them out. A lot of 22 - 47 nF Wima and CDE Mica.

I tried to fit some 100/50 (new but with short leads), but it would have taken too much fiddling. I'll order two more of the 35V with long leads to match what I have and shoot for 200 on the nose.
Attached Images
File Type: jpg 100s.jpg (560.9 KB, 241 views)
__________________
Bob M.
"Arrange Whatever Pieces Come Your Way."
  Reply With Quote
Old 19th February 2013, 01:35 AM   #332
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
For the NFB-shunt cap selection workout, you can install a pair of vertical pins made of bell wire pieces. The pins would need something like a "Z" shaped kink with a solder ball applied to the "Z" at the topside of the board, both for wider spacing and as a blockade so that we can't accidentally push the tracks loose (the "Z" and their little solder blobs supports the pins--the tracks don't). The prospect is very much like prefab pins, except that homebrew can give wider spacing. The pins added topside can allow stacking the caps sideways atop the board, and the main feature of that is plenty of room to work without so much labor between swaps.

Preferred signal caps are the usual contenders, with Nichicon FW, Panasonic FC, Sanyo, Ruby, "non-pet" Elite and Elna Cerafine. The most likely of those to do 100||100u or 220u||220u most easily is Nichicon FW because of its extended high range, but the Panasonic FC may deliver higher resolution and there is no way to tell in advance which might be better. The Elna Cerafine may be most charming, but any size larger than 1u always requires a treble bypass cap (so popular that the specific name for Cerafine||polyester is "the classic combo"). Easiest possible general purpose treble bypass cap, if there is such a thing, is Nichicon ES 0.47uF (and smaller sizes of Nichicon polar or bi-polar electrolytic caps). Second easiest is 22u and 10u little polyester dip caps--the little green things you can find at the Radio Shack and probably also *somewhere* on Mouser.

Directly affecting such a large selection labor is:
#1). Amp board power circuit caps versus midrange loudness (for level response).
#2). Rail to rail cap versus midrange resolution (decreased upper-midrange distortion, and cooler running).
#3). Feedback-Shunt resistor and cap versus bass resolution (see post#27)
#4). Input cap value (should begin with a known good 1uF).
#5). Feedback resistor current (too little is noisy for poor resolution, too much is audio masher for poor resolution, so the middle ground must be found).

Rather fascinating that feedback-shunt resistor and cap are chosen by voltage and bass; however, feedback resistor is chosen by current and imaging size of a single/each speaker (audible resolution micro-detail).

Needless to say, it would be good to install plausible values for amp board power caps and rail to rail cap before going on a small signal workout.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.

Last edited by danielwritesbac; 19th February 2013 at 01:38 AM.
  Reply With Quote
Old 19th February 2013, 05:55 PM   #333
bcmbob is offline bcmbob  United States
diyAudio Member
 
Join Date: Dec 2009
Location: Bay City, Michigan
Had a Murphy's Day yesterday, sorry.

At first I couldn't hear a lot of difference with the cap transplants. With mono sources I separated my bench speakers to 20" and stuck my head in between them.. I perceived a definite increase of highs and dynamics with the new caps in that position but could not hear as much when I backed off to about 4 feet.

Again, with the big speakers there was nothing that jumped out at first. Then I realized no matter the distance from the speakers it sounded like I had wax in my left ear. The modded amp is definitely brighter, clearer and more responsive - I'll guess ~ 10 - 20% or so. I think that's the "clearing the bottleneck" Daniel has been talking about. I plan to keep the other amp stock through a few more upgrades.

BTW, I used a great little program named REAPER to convert a couple tracks to mono. It's from the same folks that did Winamp and is worth a look/trial. It produced flac files that are playable on a Galaxy S II. Not a lot of volume but high quality.

Daniel, If you would please paint in the Z extensions you suggest. I looked back at the photo you posted when you received your board but couldn't decipher.

Also, I'm still using the stock input cap (i believe) and noticed you have an alternative. Any info?
Attached Images
File Type: jpg TDA_Big (1).jpg (118.0 KB, 199 views)
File Type: jpg DSC00902.JPG (223.5 KB, 185 views)
File Type: jpg IP Cap.jpg (98.0 KB, 186 views)
__________________
Bob M.
"Arrange Whatever Pieces Come Your Way."
  Reply With Quote
Old 19th February 2013, 10:48 PM   #334
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Yes, non-audiometric values are like ears full of wax. That weird pressure is foldback distortion of the bass from allowing more bass into the amp than it could pass through. If you didn't change the resistor values, then there is actually a lot of bottleneck still present. See again, post#27

True that TDA7294 can't do high end sound unless bass is allowed to pass through the feedback-shunt RC intact. Ways to resolve this particular bottleneck is:
Decrease input cap size (1u or smaller).
Increase NFB-Shunt cap size (greater than 199u).
Increase Feedback-Shunt resistor value (greater than 1.1k).

I'm after both increases.

Note1
Finishing power circuit tuning first, to make the job easier
Before proceeding, I'd again like to remind that using the rail to rail cap could make the small signal fine tuning a lot easier and the amp slightly cooler and cleaner so that it has more capacity for powerful clear output. We could proceed without that part, but I wouldn't know why it would be omitted.

Note2
The large size caps we're using for NFB-shunt cap probably necessitate a tiny companion--the treble bypass cap, which can prevent either mumbling or shout effect of treble roll-off. If there's any of that unpleasantness, do be sure to fix it with a cute little bypass cap in parallel to the larger cap. Even a rough approximation can be more pleasant than listening to an error near ear sensitivity peak.

Example with parallel pair of 220u
To do 220u||220u (440u) NFB-shunt cap, we'd need a feedback-shunt resistor of 1.5k. In that case, plausible feedback resistor values are then in a range from 39k to 56k.

Gain versus Dynamics
Given the resource of clean power, I'd certainly make a try for more fun dynamics by setting high gain. The 56k feedback resistor with 1.5k feedback-shunt resistor is in the ballpark for having your cake and eating it too. I would suggest to ballpark at this point, and then experiment from there.
You can try reducing the gain (47k feedback instead of 56k) to see if the soundfield may or may not get bigger, and determine and keep the option that you like best.
However. . .
I predict that with either the Galaxy S II or a normal computer, that you would most like the high gain option (38X) because a source that isn't straining is a good upgrade, since a source in good condition is more attractive to amplify. Assisting source quality with a high gain power amp is generally far more beneficial than the rather small negative consequences of high gain at the amp. A different way to do it is install a preamp, but I cannot determine if the loss of high gain amp or if the insertion loss of preamp is the correct compromise. Perhaps it is personal taste. Either option requires clean power.

Resistor swapping Labor
Finding the ballpark and then fine tuning from there, will take a lot of parts swapping. You can easily paste and remove resistors from trackside (without wearing out a desoldering iron), but caps are not so easy. Adding pins doesn't work for the resistors, but it can make for ease in cap swapping. . .

Cap Swapping Labor (add Pins to make it easy)
Before doing the pins and cap interviews, plausible resistor values would be good, and then. . .
Leftmost photo (in your post above), the spots marked 1uF and 47uF are both signal caps, there's gain on them and it will take some experimentation, such as "contest of 5 peers" interviews to find ideal capacitors for both locations. The "z extensions" are nothing other than homebrew pins that stick out the top and give you room to test drive at least 5 capacitor candidates (plus find them attractive bypass cap partners) quickly and without wearing out the PCB traces or desoldering iron. A touch of normal soldering iron will free capacitors from pins so much more quickly and in that way you can conduct interviews more easily. The pins can be removed and the "winning" caps can be installed normally after the "interviews" are finished. May I suggest current production high availability parts?

Cap quality control
In a 5 peers contest, you've already got the 1u polyester you've been using for input cap and your NFB-shunt cap is a 220u panasonic. That's good. To finish the contest, you'd find 4 more contenders for each location in comparison test. There's gain on these two caps, so that high end results necessitates a quality control. Contenders that do poorly at first, may do excellence if given a just right bypass cap as helper, so yes, there is a bit of labor involved.

Validity check
The shield function of the amplifier board power caps requires that the NFB-shunt cap isn't larger than the combined value of the amplifier board power caps (and preferably nfb-shunt cap isn't larger than 1 of the amplifier board power caps). At near maximum scale, a pair of 270uF for amplifier board power caps (one per rail) and a parallel pair of 270uF for a low loss 540uF NFB-shunt cap, gives us an answer that the feedback-shunt resistor can't be smaller value than 1.2K. We need 1.2k or larger values, and range of validity is 1.2k through 3.3k. If we're not inside that range, then we're not in the ballpark.

Guitar amp & Bi-amp
The 680R doesn't work unless we wanted a bottlenecked mid-fi or to decrease the input cap value as may be used in the mids&treble portion of a bi-amp system or guitar amp. Otherwise, the 680R is probably dreadful.

Band-aid fix
A DC Tracker may (or may not) provide a dodge that could let us use a 680R feedback-shunt resistor, but I would suggest to change the resistor value to something plausible instead of adding a complex and noisy circuit with caps of its own.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.

Last edited by danielwritesbac; 19th February 2013 at 11:07 PM.
  Reply With Quote
Old 19th February 2013, 11:05 PM   #335
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Where to start? With the feedback resistor and feedback-shunt resistor values that came with the kit, we're stuck between a guitar amp and a mid-fi. Those two values need at least doubled, preferably more.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.
  Reply With Quote
Old 20th February 2013, 01:20 AM   #336
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
For rapid-fire component comparisons, how about SIP breakaway sockets?
  Reply With Quote
Old 20th February 2013, 01:28 AM   #337
bcmbob is offline bcmbob  United States
diyAudio Member
 
Join Date: Dec 2009
Location: Bay City, Michigan
Good Idea - I have piles of those and will attempt to use them where possible. We used about 12 pair on the FE. There may some spots that need the added flexibility of Daniel's extensions as some of the small caps have 1/4" leads.
__________________
Bob M.
"Arrange Whatever Pieces Come Your Way."
  Reply With Quote
Old 21st February 2013, 05:25 AM   #338
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Well, there's two resistors that need attached firmly and with excellent quality fluxed solder connections:
Feedback resistor
Input load resistor
If either of those should happen to come loose, the amplifier may do something unexpected and a bit disturbing.


So, for this phase of experimentation with the hasty component swapping, I suggest to use speaker protection (monobloc version shown):
Click the image to open in full size.
High quality version shown in the photo and still costs less than speakers.
Either positive to positive or negative to negative orientation works.
If used with split rail amp, one could add optional 100K drainers.
You can put this in series to either speaker- OR speaker+.
Small size, low cost, 35v compact caps are used.
Location is at speaker jack or speaker.
Drive first some cable.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.

Last edited by danielwritesbac; 21st February 2013 at 05:29 AM.
  Reply With Quote
Old 21st February 2013, 05:33 AM   #339
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Default My components are fairly simple. :)

For the amplifier board component swapping, optimization, selection play. . .
Personally, all of the caps that I'd use for signal, do all happen to have long leads.
Little green 4.7n, 10n, 22n polyester dip caps have long leads
Audio electrolytic caps have long leads.
472 470p ceramic from the radio shack has long leads, and a ceramic cap of such a small value isn't for passing audio but rather removing some inductance sound issues from big bodied nfb-shunt caps.

Whoops! Okay, I was mistaken. AVX little yellow radio grade polyester box caps marked with a T * in very, very small values can be used for a big blast of airy treble. They're probably overkill, but they are usable. And, those do have short leads.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.
  Reply With Quote
Old 21st February 2013, 06:13 AM   #340
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Default Audiophile tomfoolery that happens to cost less. . .

Diy audiophile resistors:
There's, no need to buy high end resistors for feedback divider when you can make ordinary parts do the same tasks. The feedback divider controls open loop gain and therefore hearing difference is possible due to the vast gain that is applied to those resistors (only those). When I finally thought of it for long enough, there was plausible science in some (but not all) audiophile resistor selections.
On research, this is plausible:
Lower noise with higher current tolerances, low inductance feedback resistor, moving the feedback resistor onto the chip pins, slightly lower gain at RF and ultra low capacitance feedback-shunt resistor. With that information revealed, it is then doable with ordinary parts at low cost . . .

The carbon film resistors at Radio Shack are Xicon with the nice thick leads. Mouser also has Xicon and Koa carbon film.

1). Feedback resistor:
A paralleled pair of identical value 1/4 watt Xicon or Koa carbon film, with minimal pin length between the two identical resistors, (spiral twist the pins, flux, and solder so there's no gap between the resistors) makes a high end carbon resistor with low inductance leads for feedback resistor. This small and sturdy assembly can go trackside, right on the chip pins. For example 100k||100k makes 50k feedback resistor. The extremely low inductance is a good thing. Carbon comp, carbon film, metal comp or metal film can be used since the paralleling has enforced a quality control already. For feedback resistor, I think we'd like to assure zero loss at RF.

2). Feedback-Shunt resistor:
However, for feedback-shunt resistor, a solo 1/2w Xicon carbon film could be quite attractive. I wish I had ordered 1.33k, 1.5k, 1.68k 1.8k, 2.0k 1/2w carbon film from Mouser, for testing each. Although we want the HF loss and low capacitance typical of a single 1/2w carbon film resistor, we don't want that to happen inside the audio band, so the resistor must be of decent quality carbon or carbon film (Xicon or Koa), so that the miniscule loss happens way up at RF, effectively decreasing the gain on noise. It doesn't do much, but even a tiny step in the right direction is nice when so easily done. At feedback shunt, we need the lowest capacitance resistor, and a good brand of 1/2w carbon film works nicely. Other ultra low capacitance resistors typical for stopper use in tube amplifiers may work well--For feedback-Shunt resistor, I think we'd like to use stopper suitable resistors and avoid metal film.

Current vs resistor wattage:
At least we could say that the resulting sturdy feedback divider could handle double the current without degrading the signal. More audio, less noise = more amp power. So, if you're planning on some full blast fun, this sturdy, inexpensive, audiophile resistor prospect might make a tiny, yet appreciable, difference.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.

Last edited by danielwritesbac; 21st February 2013 at 06:40 AM.
  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
OPtimizing the VBE MUltiplier fglabach Solid State 41 17th January 2013 01:27 PM
Optimizing active crossover atledreier Multi-Way 27 19th August 2011 02:53 PM
optimizing the VBE multiplier hienrich Solid State 80 29th November 2010 10:25 PM
Interesting linestage optimizing ideas PixelPlay Tubes / Valves 5 23rd June 2008 10:18 PM
optimizing for efficiency Dwiel Solid State 13 10th May 2005 09:31 PM


New To Site? Need Help?

All times are GMT. The time now is 09:07 AM.


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