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.

My attempts at a design of a 3 stage amplifier
My attempts at a design of a 3 stage 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 13th June 2019, 07:07 AM   #551
edbarx is offline edbarx  Malta
diyAudio Member
 
Join Date: May 2018
Quote:
Originally Posted by mjona
It appears you are now wanting this as part of a professional entertainment system for continuous high power output. It seems this is more about demonstrating specifications than being objective and practical.
Here, that business sector is highly competitive and I am not in it. Anyone in it will certainly seek to get the most reputable and expensive products. I purchased a Wharfedale amplifier for the fact, these amplifiers have a flat frequency response, and the price of the S1500 was comfortably within my spending power.

As it is, the amplifier can be used, but I must always make sure the outputs are not shorted, and the 2 Ohm minimum output impedance is adhered to.

At home I do not have 2 Ohm speakers suitable for this amplifier output. Such speakers would require four 8 Ohm woofers connected in parallel. Additionally, such a speaker cabinet would require another four tweeters also connected in parallel. With so many drivers, a cabinet like that, would require a crossover filter, and the cabinet itself, must provide acoustic isolation from all drivers. To handle that kind of power, the woofers have to be at least 12 inches in diameter.

A speaker as described previously at home, will only result in any neighbour getting deeply irritable for the penetrating power this kind of speakers deliver. Furthermore, there is also the more serious reality of hearing damage, which powers like these in confined places, definitely cause.

I am not in the path of this kind of expenditure. What I am after, is a home system with two amplifiers that can be exchanged with a set of switches. For the most time, a chip amplifier that uses far less power will be used. Currently, I have one based off the chip TDA8563AQ.

Before implementing the circuit additions physically, I would like to ask readers of this thread whether a simulated -22.6mV DC offset is acceptable.

With the new additions the short circuit average current has been greatly reduced. Plotting I(out)*V(rail), and using the generated graph to estimate the short circuit power dissipation, the result came to be about 60W to 70W.
  Reply With Quote
Old 21st June 2019, 01:40 PM   #552
Mooly is offline Mooly  United Kingdom
diyAudio Moderator
 
Mooly's Avatar
 
Join Date: Sep 2007
My attempts at a design of a 3 stage amplifier
Not all that many years ago we used to reckon offset was acceptable if kept within -/+100mv... so yes, the simulated result is fine.
  Reply With Quote
Old 22nd June 2019, 12:38 AM   #553
mjona is offline mjona  New Zealand
diyAudio Member
 
Join Date: Jul 2007
Location: Upper Hutt
Your dc offset is fine. Re the constant 60W-70W dissipation I simulated your 72volt supplies circuit for dc operating point data without C10 which I assumed to be a 1 uF test capacitor which has been removed. Under static conditions the simulation suggests a short on the output will activate the protection and the dissipation would be negligible.

The situation changes with sine wave input 1 V at 1kHz where there is no dissipation in the driver power combination in the positive supply and although the dc offset is not significantly affected the transistors in the negative rail stay on.

The problem is the power device bases in the latter sit more than 0.6V lower than the negative voltage rail while those in the positive half sit at virtually the same voltage as the positive rail and with zero base emitter voltage drop this turns off.

You have to find some way of changing this so the driver and power devices in the negative rail turn off by reducing the base emitter voltages for the latter.
  Reply With Quote
Old 22nd June 2019, 01:15 AM   #554
mjona is offline mjona  New Zealand
diyAudio Member
 
Join Date: Jul 2007
Location: Upper Hutt
Default Rough Recast of Circuit

The attachment shows some ways to reduce the negative supply short circuit current with sine wave input. I'm not suggesting this is a complete solution - having seen this before I suggested rail fuses for protection. If you end up with an unacceptable compromise performance wise you might like to think about this option.

If you prefer to work on a solution I suggest working backwards from the driver stage to see if you can get complete cut off. The series resistor and diode in the negative rail contribute to the voltage drop on the power devices and have been removed increasing the emitter resistor values for the drivers and to some extent the common resistors to the output rail have moderating effects on the NPN power device vbe. I increased the 820R,s in the protection to 1k for the same reason.

Good luck with your endeavours.
Attached Files
File Type: asc 2019_06_09_final_amplifier_circuit-xy.asc (15.3 KB, 22 views)
  Reply With Quote
Old 22nd June 2019, 10:38 PM   #555
edbarx is offline edbarx  Malta
diyAudio Member
 
Join Date: May 2018
Careful inspection of the various DC voltages is indicating when there is a hard short circuit of the output, the input stage is losing control. This is indicated by the fact that, the tails' currents are no longer balanced notwithstanding of the presence of the current mirror. This affects the drive of the VAS's transistor causing it to stop conducting. As a consequence, the VAS's current source's current is driven towards the ground formed by the output extremely low resistance. The input stage's standing current is 3mA, when there is this kind of short circuit in the output, the tails' currents are 1.340mA and 1.606mA.

The cause of the VAS's transistor not conducting was insufficient base drive at the non-inverting input. Increasing the resistor values of the negative feedback chain to 18kOhm and 330 Ohm, restored the VAS to proper function. However, the quiescent current in the output stage was not restored to balance. Now, the problem has been shifted to the positive rail output devices.

Using simulation values for driver base and collector currents, current transfer ratios, betas, for the driver transistors are 177 and 100. Restoring the feedback chain resistances to 180R and 10K, the short circuit quiescent current imbalance settles itself to the rail transistors which are driven with the driver transistor having 177 as beta. I am attributing this is causing simulation difficulty. The original circuit driver transistors are MJE15033 and MJE15032G, there is no _Cordell ending in the transistors' part number. Searching online for these transistors, I found that parts with numbers MJE15033 and MJE15032 have equal gains of 50.

I think, something is wrong in the LTSpice's transistor database. How can complementary transistors have very different betas and yet they are called complementary.
  Reply With Quote
Old 23rd June 2019, 08:34 PM   #556
mjona is offline mjona  New Zealand
diyAudio Member
 
Join Date: Jul 2007
Location: Upper Hutt
In truth complementary transistors are representations of the elements used to dope the silicon wafer - Phosphorus and Aluminium which have different atomic numbers and different energy levels for the electrons that orbit them.

Put them in intimate contact the surplus of electrons in the Phosphorus will be absorbed by the deficiency of such in the Aluminium creating a depletion zone where there are no electrons and a difference in potential between the N and P areas.

The art is in manipulation of the doping levels to produce complementary devices that are good enough for general use. This situation applies to small signal transistors too. That is a brief description about an area that is well worth studying for greater insights. Some processes in nature are not reversible such as corrosion in metals by contaminants in the air or in liquid.

The previous simulation I sent to you does not work properly. I had one that worked on a short circuit but forgot to save before closing Spice down and in my haste to finish the reply I relied on my memory for the changes I had made - sorry about that.

The way to address the sort of imbalances in a circuit is to get the vbe biasing right so complementary transistors give equal collector current - the vbe bias voltages may differ to get this result.

What I have tried to do is rearrange the circuit so the voltage drops across R20 and R22 can be made closer to the supply rails so the conduction of the output stage on short circuit is minimal.

The protection sense voltage take off point had to be moved - I have changed some resistor values in this section was off the cuff.

The constant current bias resistor was increased to reduce the change the bias voltage on D1 and D2 and R1 was increased at the same time. Symmetry in both circuit halves does not always work e.g. with R7. This attempt was to get inside the ball and leave finding the best seats to you.

Due to a lack of time I used a mix of transistor .models from Cordell and the onboard SPICE library. The problem with my last simulation is the protection circuit limited the power output to senseless low level.

I have checked the one attached here does not do that. You can get this circuit sorted to your own satisfaction.
Attached Files
File Type: asc 2019_06_09_final_amplifier_circuit-xy.asc (15.0 KB, 9 views)
  Reply With Quote
Old 24th June 2019, 07:19 AM   #557
edbarx is offline edbarx  Malta
diyAudio Member
 
Join Date: May 2018
This is what I can get while keeping all existing circuitry unchanged except the negative feedback resistors. These had to be increased to bring the differential pair into balance. The new values are 330R instead of 180R and 18k instead of 10k.

The negative feeback network had to be forked into two sensing prongs with one at the output, and the other at the mid-potential of the VBE multiplier. To get such a midpoint two resistors, 47R, and two diodes, 1N4148, were added. A 33k resistor feeds the inverting input directly from the latter, so called, midpoint. Another, 39k, feeds the base from the output. Since, these two points are 'hypothetically' at the same varying voltage, calculation shows that to keep the previous amount of negative feedback, the 33k and 39k resistance must have a parallel value equal to 18k.

I have an uncomfortable feeling my changes, although they corrected the short circuit imbalance, are so unorthodox, that anyone who had been in electronics, may view them as problems, rather than a solution.

Needless to state, these changes have impacted the distortion figures negatively. The new distortion figures are:
a) 0.011% with 8R output
b) 0.042% with 4R output
c) 0.163% with 2R output

On hard short circuited output, 1 micro Ohm resistance, the quiescent currents are:
a) 15.95mA per output PNP transistor (2SA1943)
b) 10.51mA per output NPN transistor (2SC5200)

The output short circuit current is triangular with bases of around 26us and amplitude of 33A. The frequency is the simulated frequency of 1000Hz.

I am posting the circuit. I made a conscious effort not to change the original circuitry to enable myself to implement the new additions without having to rebuild the PCB from the very beginning.

I am attaching the circuit, a Fourier distortion chart with a 2 Ohm load, and the output on short circuit (1 micro ohm).
Attached Images
File Type: png 2019-06-24-091213_1366x718_scrot.png (99.1 KB, 118 views)
File Type: png 2019-06-24-091617_1366x718_scrot.png (50.2 KB, 121 views)
Attached Files
File Type: asc 202019_06_20_final_amplifier_circuit.asc (17.8 KB, 8 views)

Last edited by edbarx; 24th June 2019 at 07:26 AM.
  Reply With Quote
Old 24th June 2019, 09:13 AM   #558
edbarx is offline edbarx  Malta
diyAudio Member
 
Join Date: May 2018
Looking at the Fourier chart, one cannot fail to notice the third harmonic is quite strong although narrower than the second harmonic. Acoustically, the second harmonic may be acceptable as a distortion, but the third, many deem may sound unpleasant to the ear.

A drastic, but nevertheless easier way out of this, is to remove the short circuit protection altogether. The two pronged negative feedback network is sensing voltages where there is nothing of its business. The so called, midpoint potential of the VBE, should not be expected to move in synchrony with the output. At high current outputs this midpoint potential has to move above the output + 0.7V + some_extra_voltage. The latter extra voltage depends on the instantaneous current driven into the output.
  Reply With Quote
Old 24th June 2019, 09:43 AM   #559
mjona is offline mjona  New Zealand
diyAudio Member
 
Join Date: Jul 2007
Location: Upper Hutt
If you run the simulation I sent look at the visible traces for I (R21) and I (R24) which are very brief spikes which the output stage should be capable of handling. I did not adjust the standing current in this simulation so this is quite low and not equal for both halves so in that state it runs close to Class B the THD for sine wave input output of 50V peak. On a 8R load the peak current is 6A and when the output is dead shorted to ground this converts to a series of spikes up to 13A for 540uS between cycle swings I (L2). On 8R the 1kHz THD is about 0.02%.

The only high power CFP based power amplifier with short circuit protection that I have seen is one using Darlington power outputs which turns the output stage into a triple arrangement.

The protection circuit is far from simple and you don't want a major redesign after all the design effort you have put into this project.

When you consider the situation you have reached, 33A is a hazard to be reckoned with if it extends too long in terms of safe operating area for the combined output transistor set. I think it is likely to be a balancing act to reduce this without increasing THD which could be hard to swallow. In that case you would need to think whether rail fuses are a safer option and whether or not they compromise performance as much as the protection circuit does.

As I mentioned earlier your amplifier ought to be stable without that near short circuit for ac signals due to C10 - the 1uF capacitor loading the output. This will be spoiling your THD figures across the board
  Reply With Quote
Old 24th June 2019, 03:23 PM   #560
edbarx is offline edbarx  Malta
diyAudio Member
 
Join Date: May 2018
Earlier in this thread some posters mentioned the use of an input servo mechanism. This was mentioned as a measure to correct too high DC offsets. Although, the latest modifications do not suffer from overly high DC offsets, simulations are showing I can further minimize circuit modifications, by superposing a DC 64mV fed trough a 33k resistance to the non-inverting input. On hard output shortening the quiescent currents per output device is about 13mA. This solution does not negatively affect distortion and, more importantly, does not require the use of some exotic circuit modifications.

On a post earlier in this thread Mooly suggested an IC based input servo mechanism but I couldn't find it or a link to it. I remember it can easily get DC offsets down to nano volts.

The thing with the suggested servo is that it is a tried circuit unlike what I am doing. Sometimes, my thinking out of the box, makes think, I should have been in the arts, rather than technology.
  Reply With Quote

Reply


My attempts at a design of a 3 stage 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
Phono stage design considerations part 1: choosing 1st stage tube Joshua_G Tubes / Valves 215 26th August 2018 05:30 PM
Amplifier design: how to connect a stage with the next one? Bricolo Solid State 11 10th August 2018 07:34 AM
Pre-amplifier circuit not working despite attempts to fix it Rudi1441 Analog Line Level 5 18th July 2016 02:35 PM
LF Line Level Stage design to preceed class D amplifier undefinedza Analog Line Level 13 18th September 2014 08:26 AM
how to design mosfet amplifier stage? pzung Solid State 2 22nd January 2010 06:58 PM


New To Site? Need Help?

All times are GMT. The time now is 05:08 AM.


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