♫♪ My little cheap Circlophone© ♫♪ - Page 113 - diyAudio
Go Back   Home > Forums > Amplifiers > Solid State

Solid State Talk all about solid state amplification.

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 11th October 2012, 11:23 AM   #1121
Elvee is offline Elvee  Belgium
diyAudio Member
 
Elvee's Avatar
 
Join Date: Sep 2006
I see no other problem than the one already mentionned by Kean: the non-linearity of the low voltage zeners, both in resistance and capacitance.
BTW, D12 and 13 are superfluous
__________________
. .Circlophone your life !!!! . .
♫♪ My little cheap Circlophone© ♫♪
  Reply With Quote
Old 11th October 2012, 10:05 PM   #1122
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Thanks!
Quote:
Originally Posted by Elvee View Post
I see no other problem than the one already mentioned by Kean: the non-linearity of the low voltage zeners, both in resistance and capacitance.
Goal: Reduces computer & speaker damaging surges (and works in practice).
Since it is the RC value that needs investigated further, I just did some homework. Given a variety of ~3.3v glass bead zeners in the range, 1/5thw, 1/4w and 1.3w, a series pair of any measured close to the cap value of 82p, and I cross checked the DMM with real ceramic caps for confirmation. Probably 100p for simulator would work fine. But I do not know the loss figure or dynamic behavior.
No idea what the dynamic RC value would be.
Can you rough estimate the capacitance and resistance while the Circlophone is running from a single rail source, such as a computer?
Quote:
Originally Posted by Elvee View Post
BTW, D12 and D13 are superfluous
Goal: Reduces "offset bounce" problem (and works in practice).
Thank you for this excellent news that D12 and D13 aren't especially harmful. They are helping to make the offset trimmer superfluous in a greater number of cases because of 1/3rd volt farther away from misfires that cause ~250mv DC offset bounce. Boosting the misfire tolerances up to avoid installing the offset trimmer is more attractive for low components count builds. Simply put, adding D12 and D13 allowed me to remove the trimmer, remove the feedback kludge and return the NFB cap. . . and get unstuck.

P.S.
Some of what you're seeing makes the difference between a power amp's features versus an integrated amp's features--mitigating source problems and variances. Indeed if that can be done with just 2 transistors and a few simpler parts, then it can remain a low component count elegance.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.

Last edited by danielwritesbac; 11th October 2012 at 10:12 PM.
  Reply With Quote
Old 14th October 2012, 07:40 AM   #1123
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Default 2N5551, 2N5401 failures documented.

Please don't install OnSemi 2N5551.
The OnSemi 2N5551 from Sanyo's China fab plant (Digikey stock) may have HFE 30 or HFE that falls unexpected with current, plus sometimes half speed, and may fail in service. They're even worse than fakes. The same may or may not happen with Moto-China (today's Motorola), so those need checked too, although the usual problem with new Motorola parts is the less obvious issue of noise performance that varies.

Solution:
Fortunately, there are many brands of 2N5551, without a problem:
Philips, CentralSemi, KEC, Fairchild, NXP, and more, do still have normally functioning 2N5551's and 2N5401's (and other similar parts).

NXP website search also documents some substitutes, since the 2N5551 is from a large family of similar devices.
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.

Last edited by danielwritesbac; 14th October 2012 at 07:46 AM.
  Reply With Quote
Old 27th November 2012, 09:45 PM   #1124
FOXYE is offline FOXYE  Israel
diyAudio Member
 
Join Date: Feb 2009
How this amp sounds against the LM3886 and F5?

And what the total estimated cost of the amp with parts?
  Reply With Quote
Old 28th November 2012, 02:28 AM   #1125
diyAudio Member
 
terranigma's Avatar
 
Join Date: Feb 2007
Quote:
Originally Posted by FOXYE View Post
How this amp sounds against the LM3886 and F5?

And what the total estimated cost of the amp with parts?
I built some Gainclones including balanced mode bridged versions. My Circlophones surpassed all of them in terms of clarity, background noise level, bass, vocal performance and dynamics although my Gainclones were very good performers.

The cost is determined by output device selection (which varies) and PCB build skills in my opinion. (I assume you already have a Power Supply unit.)
  Reply With Quote
Old 28th November 2012, 04:07 AM   #1126
abetir is offline abetir  Philippines
diyAudio Member
 
Join Date: Nov 2010
Location: Easternmost Region V
..against the P3A, Leach, and the PWM200rms class D...I prefer the Circlophone.
but then again "comparison" is always subjective to one's taste...
Referring to DIY, cost is dependent on parts selection.
Attached Images
File Type: jpg DSC04364.jpg (351.2 KB, 532 views)
File Type: jpg TO_39_heatsink.jpg (195.5 KB, 783 views)
  Reply With Quote
Old 28th November 2012, 09:26 AM   #1127
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Quote:
Originally Posted by FOXYE View Post
How this amp sounds against the LM3886 and F5?
F5, I don't know. For TV use, I don't know. For music replay, the Circlophone can beat the LM3886.
Quote:
Originally Posted by FOXYE View Post
And what the total estimated cost of the amp with parts?
It could be about $8 USD more than LM3886. But, I think that a Circlophone is "more" amplifier.

P.S.
Circlophone build thread: Building Elvee's Circlophone: Documentation, Parts, Accessories, & beginner friendly
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.

Last edited by danielwritesbac; 28th November 2012 at 09:30 AM.
  Reply With Quote
Old 23rd December 2012, 04:23 PM   #1128
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Default Audiometric, Offset bounce, hammer-pillow-thumb, and current

I was looking at the offset bounce issue with a new perspective, and that is audiometrics. For offset bounce to occur, D10 and/or D11 switch on to the bass beat when not all of the bass can travel the 470u. Well, a ballpark audiometric value for C3 (same value that would keep the diodes off) is far too large to consider; however there are several alternatives.

Simulations of Circlophone consistently show some benefit to buffering the input in several ways. One day, not long ago, I was wondering about these added transistors that effectively serve as a pillow between the hammer and the thumb, and if instead we could reduce the size of the hammer by reducing the current a little bit, and then not need any added transistors.

Reducing the current can be done with R16, R19 to 1.2k or higher figures. It seems to answer both of the above questions; but, as usual, there is the new question if the fix is suitable?
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.
  Reply With Quote
Old 23rd December 2012, 04:49 PM   #1129
diyAudio Member
 
keantoken's Avatar
 
Join Date: Aug 2006
Location: Texas
Blog Entries: 2
Increasing feedback impedance is the opposite direction. The buffers provide a low source impedance to the LTP inputs which is why they improve performance.

Increasing R6 and R7 may work towards not needing buffers, but probably not by much.
  Reply With Quote
Old 24th December 2012, 07:49 PM   #1130
diyAudio Member
 
danielwritesbac's Avatar
 
Join Date: Nov 2007
Of course I want the buffer, but I'm doing a "back to basics" review for the dual purpose of possibly running without the buffer and to make the compensations a lot easier. There's something "off" with the tolerances and I don't want to cover that up without first discovering the source. After reading the whole thread again, I find a few things that take priority. First, the audiometrics are so powerfully off the mark that D10, D11 have been damaged by overcurrent surges on a few builds. This area is also the source of the offset bounce.

One possible option is to fix the audiometrics with 470u~1.5k vs 56k so that the NFB cap is then apparently big enough so that D10, D11 accidents are rare rather than frequent.

If audiometrics are not repaired, I we chould change to a triple chain of BAT85 for 0.9v, and even 3 of those makes a LOT less noise than one 1N4148. However, audio quality is not the only concern. This area breaks small diodes from over-current, and then it is difficult to figure out what went wrong. Eventually, not fixing the audiometrics then requires something very much like this: 4.7R~FR\\FR~FR\\FR, a series pair of quiet 0.45v soft fast silicon diode clippers series to 4.7R, giving the approximately the 0.9v tolerance that the circuit seems to need, in combination with enough current capacity to withstand error conditions, still a lot more quiet than 1n4148's, and the resistor is added to the clipper for allowing the NFB cap to compete with the diodes until the loss at cap roll off exceeds 4.7R (a lot lower pitch). In this combination method of postponing the diode switch-on, an undersize NFB cap (470u~470R) may be used without turning the bypass diodes into a problem.

Since we've already exploded some 100ma diodes, we might want to consider that BC556B has 100ma maximum surge rating, and that BC327-40, with 800ma surge tolerance might be a safer bet for longevity of the input pair.

I'd rather use audiometric sized feedback-shunt area (470u~1k5 vs 56k) removing error instead of using high current parts to withstand error. But of course we could do both, since the replacements are excellent quality low noise devices.

P.S.
Previously, we tried to automate D8, D9 by replacing them with a single transistor resulting in dull audio and compensations that quit working--the last time I tried this, there was strong HF distortion as if the feedback capacitor was removed. This needs tried again but a bit closer to the rail or with a different reference point.
Perhaps we need to automate R21 prior to any other mod? Elvee proposed using one of these ccs for replacing R21:
Click the image to open in full size.
At the time, I thought it was a complicated way to make an inconvenient resistor; but, now I think it could really help, since hitting a moving target of the rails bouncing with the speaker is a job for one of those ccs. Right? And are those 1n914's switching or are they simply on?
__________________
Tools, Models & Software for DIYClipNipper boostLM1875 TurboPowerful TDA7293 kitTDA7294 pt2pt ♦ My post has opinion.

Last edited by danielwritesbac; 24th December 2012 at 07:55 PM.
  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
Need help with OB design - cheap, cheap, cheap djn Multi-Way 38 30th May 2010 06:26 AM
Scosche SPL meter: Cheap find with potential, or cheap junk? theAnonymous1 Everything Else 5 11th October 2006 03:40 AM
Alpine cda-9831 vs. 9833 vs. 9835 cbecker33 Car Audio 3 10th December 2004 12:47 PM


New To Site? Need Help?

All times are GMT. The time now is 03:00 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