Go Back   Home > Forums > >
Home Forums Rules Articles diyAudio Store Blogs Gallery Wiki Register Donations FAQ Calendar Search Today's Posts Mark Forums Read

Solid State Talk all about solid state amplification.

Amplifier
Amplifier
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 10th September 2019, 06:51 AM   #11
StevenStanleyBayes is offline StevenStanleyBayes  Canada
diyAudio Member
 
Join Date: Jan 2012
Location: Ottawa
Again : I am happy with Co and Csi also because, for the new schematics, a lower value may be possible to be used. Also, I can add external capacitors, as I have, in a separate module.

Also, I have tried to make the direct schematics work : amplifier and two buffer, common collector transistors only. No common emitter transistor with a resistors, no diodes with resistors. This has been explained in other threads I have opened.

The only thing I had to use is Co which may or may not be necessary. Co, also, may or may not be necessary with other parts, mainly, transistors.

Quote:
Originally Posted by Mark Tillotson View Post
I'd be interested to see a graph of distortion against output level for 1kHz and 20kHz. <snip>
The only problems this circuit has is the jump. The rest is the same as with every other amplifier.

The jump creates an error of nonlinearity, or, distortion. The faster the jump the lower the error.

The document discusses the non linearity in space ( amplitude ) and time ( half period ). The greater the frequency, the higher the nonlinearity. However, when the jump is completed for t ( the goal is 250ns ), the relative error to the amplitude is :

( Asin(wt) / A ) * 100 %

This can be calculated for any frequency. The document shows a calculation for 20kHz. The nonlinearity error is 5% for 1% of the half period ( 25us ).
This is only in case the jump completes for 250ns.

As the document says, I do not have any equipment to measure this.

In regards to a possibility to increase Ceb with a capacitor from bases to emitters, I have tried this, but, I do not want to slow the charge of Ceb, on the contrary, I want to speed this up and prefer to use an output capacitor Co instead.

Yes, capacitor Ceb and Reb do slow the transistor. Reb limits the range and can be used with Darlingtons. Ceb does not help with the noise unless the value is huge ( 10uF as far as I remember ) and seems to introduce even more noise. I used the Ceb of the transistors with resistors from 10 Ohms to 1K and OPA134. 1K worked extremely well resulting at perfect sound. Of course, again, the voltage over the 1K resistor lowers the range.

Also, I have tried to avoid any resistors and capacitors. I would not put Co in case possible. I have not tried the new configuration without Co, but, with the old configuration ( 1% resistors and a resistor from the amplifier's output to ground ) was incredibly noisy. This is why a VERY LOW ESR Co was needed. 1uF did the job with the old schematics and I decided to use this on the new ( lower noise ) schematics just to be more sure. Did not do more experiments.

Co is 1uF, 100V, X7R. 100V X7R has lower ESR and ESL than 50V X7R or so I have seen.

Also, forgot to mention again : as mentioned in the document, another difference between this circuit and a standard amplifier is Co introduces a frequency dependent resistance ( impedance ) of 1 / jwCo which is 8 Ohms at 20kHz and 5.33 Ohms at 30kHz. With RL = 4 Ohms, the amplifier will see not 4 Ohms but 4 || 8 = 2.67 Ohms at 20kHz and 4 || 5.33 = 2.28 Ohms at 30kHz. The amplifier must be able to handle these resistances. The amplitude of these frequencies is normally very low with audio signals.

Of course, the introduction of a frequency dependent resistance introduces problems with linearity, which, hopefully, are negligible.
Quote:
Originally Posted by zjjwwa View Post
OK, ... so strike the resistors. <snip>
As mentioned, I am only interested in the direct circuit : amplifier and two common collector buffer transistors. Nothing else.

Otherwise, yes, there are very many possibilities as discussed in other thread I have posted. Also, I have mentioned the standard approach : amplifier, common emitter transistor with only Rc, two Schottky diodes ( can be omitted ), transistors.

There are many variations of this standard solution. There have been many updates to the document since last posted. Here is the document again :Amplifier – Google Drive

Quote:
Originally Posted by zjjwwa View Post
What about "not to be biased" ===> "almost not biased" ? <snip>
Although I am only interested in the direct circuit and not in any other, as mentioned in the previous reply, I decided to try to make your idea not so much resistors dependent, because, your idea is brilliant and is not so standard. I used a standard approach of replacing resistors with transistors and, although I have not had the possibility to check the circuit logic, I have decided to post, because your idea is brilliant. I also do not remember to have discussed your idea in previous posts.

The transistors will allow for a larger current flow when the output transistors are fully or close to fully open. During the jump, they are not so important as not so much current is needed.

Again, I have not checked the schematics, hopefully, I have not made a logical mistake.

EDIT: Collector resistors of T11 and T21 OK, but, not needed.
Attached Images
File Type: jpg Idea, Reverse Control.jpg (42.7 KB, 272 views)
  Reply With Quote
Old 11th September 2019, 07:47 AM   #12
cumbb is offline cumbb
diyAudio Member
 
cumbb's Avatar
 
Join Date: May 2005
Location: different/relative
Which speakers do you use? If not multiway, I would remove all the different caps. Would left the biggest only. Currently it sounds like a distorted mirror on the hustle and bustle;-)-; Multiway speakers don't show that.

I would never do use TO-3 transistors.
__________________
different/relative,-)
  Reply With Quote
Old 11th September 2019, 08:12 AM   #13
zjjwwa is offline zjjwwa  Poland
diyAudio Member
 
zjjwwa's Avatar
 
Join Date: Oct 2011
Location: Warsaw
Amplifier
Quote:
Originally Posted by StevenStanleyBayes View Post
Although I am only interested in the direct circuit and not in any other, as mentioned in the previous reply, I decided to try to make your idea not so much resistors dependent, because, your idea is brilliant and is not so standard. I used a standard approach of replacing resistors with transistors and, although I have not had the possibility to check the circuit logic, I have decided to post, because your idea is brilliant. I also do not remember to have discussed your idea in previous posts.

The transistors will allow for a larger current flow when the output transistors are fully or close to fully open. During the jump, they are not so important as not so much current is needed.

Again, I have not checked the schematics, hopefully, I have not made a logical mistake.
Interesting development. I think that I "feel" where this concept is intended to be going, but I am not sure about some details here.
a). What does 'Zener 0V' mean?
b). As soon as D1 starts to conduct, T11 switches on. Will it not fail? (I am afraid that at start of conduction of T11, jointly T11 and T12 will create an inter-rail conduction path, a "short", with no limit to the current, and this could result in failure of these transistors).
c). In this transistor-assisted-Schottky concept, that you are thinking about, would the condition be fulfilled / met, that the Schottky diodes are conducting constantly, all the time? Or would they have periods where the one or the other is inactive? My gut feeling tells me, that the Schottky diodes "should" be conducting all the time, but maybe you see some alternative mechanism, concept, solution here? What is the intent?

===

Was thinking about how "not to interfere" with your design goals (resistors, capacitors, et al). Was thinking about how to "force" a constant current flow through the Schottky diodes, irrespective of the fact that they are riding "on the signal", or "floating" on the signal and jumping all around across all possible values ranging from the plus supply rail to the minus supply rail. How do you make a "floating" CCS-CCD pair? I don't know. Not sure. I tried to envision such a "floating CCS-CCD pair" that would be capable of floating on the output of your op-amp, and (possibly/hopefully) not interfering with your circuit too much.

I came up with this stuff. A combination of two series connected CCS (top side) and CCD (bottom side). But in order so as not to antagonize them against each other (so that they do not "fight" against themselves), added a mirror. Result: hopefully a "floating" current source. Probably nonsense. But maybe? Just take a look. Should this floating CCS-CCD be powered from a separate DC power source, or can it be powered from the same Rails as your amplifier? Not sure. The safer bet would be to power it from something separate. But maybe it could also be powered from Your rails. Then You probably would "loose" some 2 volts of your maximum possible output swing. But my bigger concern is that if powered from "Your rails", then there is the risk of various "capacitance" or other problems likely to emerge. Whereas when powered from a totally "separate" power source, there is a higher chance that the floating CCS-CCD would not in any way "interfere" with your low-capacitance Op-Amp concept. Just thinking out loud, fishing for some alternative ideas.

How about conducting a "Simple and Stupid" experiment? Because if no practical value gained, then not worth all the talking. A 9V battery. In series with a 3k9 resistor. Voila! A KISS floating current force. Two Schottky diodes, ... and Rock and Roll. Hopefully.
Attached Images
File Type: jpg Non-Conflicting CCS-CCD Op-Amp.jpg (85.3 KB, 261 views)
File Type: jpg Simple and Brutal Test.jpg (63.8 KB, 258 views)
  Reply With Quote
Old 11th September 2019, 09:36 AM   #14
StevenStanleyBayes is offline StevenStanleyBayes  Canada
diyAudio Member
 
Join Date: Jan 2012
Location: Ottawa
Quote:
Originally Posted by zjjwwa View Post
Interesting development. I think that I "feel" where this concept is intended to be going, but I am not sure about some details here. <snip>
D1 and D2 are to ensure T11 and T21 can be fully closed. Not much of a deal. Zeners can close amplifiers which are away from the rails by more than 0.7V and a better closure of all amplifiers. I wrote Zenners OK. I meant D1 and D2 can be replaced with Zeners.

Here is the principle :

Assume the amplifier has an output of 0V. Then, T11 and T21 are open and their base resistors ( and collector resistors ) ensure to what extent they are open. Say somewhere in the half of the needed collector current. Exact values are not important. When the output of the amplifier goes high, T11 opens more, T21 closes more. T12 opens more, T22 closes more. T13 can get more current.

T13 and T14 are still governed by the Schotkys and the output of the amplifier. Their bases are 0.35V away from the amplifier. Just the same as resistors.

Of course, precautions must be taken for the currents not to be too high. This can be done by the base and collector (not shown) resistors of T11 and 21.

Need to think more for logic and values. But, generally, I replaced your resistors with transistors which open a lot when the output voltage of the amplifier is very high (or very low). They are opened normally (and, theoretically, equally) when the voltage of amplifier is 0V (in the middle).

Quote:
Originally Posted by zjjwwa View Post
Was thinking about how "not to interfere" with your design goals (resistors, capacitors, et al). <snip>
I do not have the possibility to look at this schematics closely now, but, looks like you have combined the standard approach of a common emitter transistor and your original circuit. Good think is you tried to replace the resistors with transistors too.

Quote:
Originally Posted by zjjwwa View Post
How about conducting a "Simple and Stupid" experiment? <snip>
Separate source of the Schottkys is much better. I am still worried of the resistor, though. Your original idea with a current source must be OK.

Quote:
Originally Posted by cumbb View Post
Which speakers do you use? If not multiway, I would remove all the different caps. Would left the biggest only. Currently it sounds like a distorted mirror on the hustle and bustle;-)-; Multiway speakers don't show that.

I would never do use TO-3 transistors.
The Co is necessary. Csi allow for Co to charge and discharge with maximum speed. The external capacitors help Co. They are not necessary but good. Power supply capacitors are good as they are. The more the better.

TO3 transistors are not a problem. 2N3055 and MJ2955 are too powerful for the job which is good because their B is high at 3.375A, but, bad because their parasitic capacitances may be high, although, this may not be true for the new versions. Old 2N3055 and MJ2955 have high parasitic capacitor values. New must have low parasitic capacitor values.

The speakers are normal, general purpose 8 Ohm, 50W, Toshiba tiny speakers. Everything is OK with the capacitors and the speakers. No effect.
  Reply With Quote
Old 11th September 2019, 12:37 PM   #15
kokoriantz is offline kokoriantz  Lebanon
diyAudio Member
 
Join Date: Dec 2015
Location: south east asia
Where is the schematic of your amplifier. Are you ashamed of posting it here?
Attached Images
File Type: jpg amplifier.JPG (29.6 KB, 248 views)
  Reply With Quote
Old 11th September 2019, 02:03 PM   #16
Nico Ras is offline Nico Ras  South Africa
diyAudio Member
 
Nico Ras's Avatar
 
Join Date: Nov 2005
Location: East Coast of South Africa
Quote:
Originally Posted by kokoriantz View Post
Where is the schematic of your amplifier. Are you ashamed of posting it here?
Are you to lazy to click on the link he supplied?
__________________
Kindest regards
Nico
  Reply With Quote
Old 11th September 2019, 02:53 PM   #17
kokoriantz is offline kokoriantz  Lebanon
diyAudio Member
 
Join Date: Dec 2015
Location: south east asia
Quote:
Originally Posted by Nico Ras View Post
Are you to lazy to click on the link he supplied?
I get the picture shown .
  Reply With Quote
Old 17th September 2019, 05:34 AM   #18
StevenStanleyBayes is offline StevenStanleyBayes  Canada
diyAudio Member
 
Join Date: Jan 2012
Location: Ottawa
Quote:
Originally Posted by kokoriantz View Post
Where is the schematic of your amplifier. Are you ashamed of posting it here?
The schematics is in the ADDENDUM of the document as per the link, posted and repeated on many occasions. Here is again : Amplifier – Google Drive
  Reply With Quote
Old 17th September 2019, 06:12 AM   #19
StevenStanleyBayes is offline StevenStanleyBayes  Canada
diyAudio Member
 
Join Date: Jan 2012
Location: Ottawa
Quote:
Originally Posted by kokoriantz View Post
I get the picture shown .
Sorry. Please, read this :

With this post, as well as, with many more, I upload a document on Google Drive and make the folder as well as the document " Public ". Anyone who goes there can download the document. The document contains a lot of explanations, formulae, schematics, etcetera. However, the main schematics is in the Addendum of the document. Addendum is a strange word for the end of the document where information is added. Addendum is a Latin word, but, is used in British English and the root of the word Addendum comes from the English word Add.

Although to add the main schematics is strange, I think the word Addendum means a chapter of the document where specific information is published. Thus, I, usually, publish complete information in the addendum : complete schematic of the device as well as pictures of the device. The rest of the document is explanation and, yes, there are explanation of the complete schematics, published in the addendum, but these are still part of the explanations.

In the explanations section, I, usually try to explain the logic of the electronics and not the mathematical descriptions and formulae unless these are necessary for the explanations, which is seldom the case as electronics is logic and not mathematics. This is why there is a specialty called Electronics Engineering and, another, called Mathematics. Electronics is usually taught in Technical Institutes, whereas Mathematics is taught in real Universities.

However, because of stupidity, some teachers who want to sound scientific and justify their titles and salaries have decided to teach mathematics in Technical Institutes which is a crime against humanity. There is almost no mathematics and whatever there is is very basic and simple and can be figured out in a week or two. Has to be mentioned as a result and not to require theorems to be proven, etcetera in classes of Higher Mathematics.

The biggest stupidity is students of Computer Science are also needed to study mathematics and the only thing they need is to add, subtract, multiply and, occasionally divide. True in binary, mainly and in Hexadecimal, but still.

In Automation theory and control systems, mathematics seems to be necessary, but, this is not true. Only seems. They still use simple end results such as imaginary, filter and system response, etcetera.

In terms of study and use of Mathematics in engineering, I claim to be a specialist, because this is what I have studied for 5 years and, thus, I know how wrong to study mathematics is. The biggest stupidity is to concentrate on integrals which are not used even in integrators.

This is why I strongly fight for electronics, as well as any other engineering, even control systems, without mathematics and only with logic. There are two types of logic : analogue and digital. Analogue logic is more like a process which happens for a give period, although, this period is, usually, accepted to be 0 because electrons travel with a speed of light ( nearly ). However, to accept the period to be 0 is wrong. Thus, even without capacitors and other delays, the current flies slowly with the speed of light and not immediately. Digital logic is self explanatory. There, processes happen immediately as the processes are not part of digital logic but of analogue and, yes, physically, digital systems work in an analogue way, but this is not important. The digital logic they perform is.

In summary : this is why I divide the documents in two :

1. Main Document : Logic : Explanation of the used electronics by logic without mathematics. I try to go as low as possible, such as, how a transistor and a diode work.

2. Addendum : The full schematics and pictures.

Thus,

1. Click here to go to the document : Amplifier – Google Drive

2. Click on the document to download the document to your hard disk or you may just open the document. The best is to use " Save As " and download the document elsewhere, say, on your desktop and, after use, delete the document or move the document wherever you want.

3. The document is written with OpenOffice for obvious reason ( free ), but, should be possible to be viewed with Microsoft Word for Windows 97 and newer. Once the document opens, press Page Down to reach the Addendum. The schematics is first in the Addendum and pictures are second. The schematics, in this case, shows only one of the two stereo channels. The other is identical. Basically, the main components of the simple schematics is one amplifiers, two transistors and one capacitor Co and the cousins of Co, Csi.

You can use the scroll bar, but, you may miss the schematics. Page Down is preferable. There are not so many pages. After each press on Page Down, wait for the page to load fully and include graphics.
  Reply With Quote
Old 17th September 2019, 06:29 AM   #20
StevenStanleyBayes is offline StevenStanleyBayes  Canada
diyAudio Member
 
Join Date: Jan 2012
Location: Ottawa
In case anyone is interested, because I have the bases of the transistors outside, I may try to put a capacitor from the bases to ground. Such a capacitor is not a good idea because the amplifier will slow down. Anything between the bases and the output of the OA is not welcome and also makes noise.

However, a capacitor with a value 10% of the Ccbe ( the lowest value during the jump ) may be OK and may also reduce Co. The problem is this value of Ccbe is not known and has to be measured. I cannot measure this and can only speculate. LM7171 can drive capacitive loads of 0.5nF without a problem and, probably 1nF and higher. Again Ccbe is not known and the said capacitor is added to Ccbe.

Another good news for LM7171 : I read now and before, the documentation states the open loop outut impedance as 15 Ohms and not 110 as I said in the document. I decided they put a 100 Ohm protection resistor because the maximum current when the output is shunt is 135mA continuously ( the voltage drops to 8V, though ). Looks lie the current protection may not have been done by simple output resistors.

Anyway, 15 Ohms open loop ( saturation ) output impedance of an amplifier sound excellent.

Also, in regards to the LM7171 input limit of 10V pin to pin, 10V is the RMS of 14.1V which is the maximum amplitude of a preamplifier supplied by + - 15V or 30V. Not a very scientific protection, but, for now, I do not need another. LM7171 seems to be very sturdy even without input protection. I have not been able to damage any as far as I remember and I have abused them a lot. However, damage is not the problem. The problem is electronics degrades with the abuse. Thus, an abuse amplifier may seem to be perfectly OK, but, the parameters may have changed after abuse. This is why I replaced them before this version, although, all of them seem to be perfect in terms of parameters too.

And in regards to capacitors and speakers concerns. I do not think 1uF capacitor in parallel of a speaker with a very tiny amount of inductance ( in case any at all ) may affect the performance. The filter would have a huge frequency and, hopefully, no signal can have such.

So far, there have not been any changes of the document after the previous post on the topic of document changes and updates.
  Reply With Quote

Reply


AmplifierHide 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

Forum Jump

Similar Threads
Thread Thread Starter Forum Replies Last Post
Questions about how to decrease the PASS A-40 amplifier into a class AB amplifier. novtango Pass Labs 6 22nd November 2018 04:36 PM
Hypex SMPS400 + 2x UCD180 wiring (DC error, Auto Amplifier Enable, Amplifier Standby) maxw Class D 48 26th May 2017 07:47 PM
Boss NXD3500 car amplifier hooked to an Onkyo 7.2ch home amplifier Evolver Class D 0 20th August 2016 01:57 AM
zenith console amplifier (a stereo mono amplifier :) ) dr_skoobie Tubes / Valves 1 14th February 2015 04:42 PM
Motional feedback amplifier based on Philips 40 wats amplifier 1970s sanbadgujar Solid State 14 13th October 2013 02:35 PM


New To Site? Need Help?

All times are GMT. The time now is 08:55 PM.


Search Engine Optimisation provided by DragonByte SEO (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Resources saved on this page: MySQL 14.29%
vBulletin Optimisation provided by vB Optimise (Pro) - vBulletin Mods & Addons Copyright © 2019 DragonByte Technologies Ltd.
Copyright ©1999-2019 diyAudio
Wiki