Hey, i also don t agree to throw them to the scrapeyards..;
Whatever the trademark, amps of these times don t
exist anymore , even brand new..
I did save at least a 200 from being reduced to metallic powders..
Compensation comparison
The curves are the ones of the original schematic
and the ones of the modified amp.
Contrary to the claim of a self designed "specialist " by there,(that only
manage to display his lack of knowledge in the discussed matter),
the models have nothing to do with the amp lack of stability, as the transfert
function of this amp is determined only by the compensation capacitors.
We can see that the original compensation doesn t work at all.
The system is not damped enough, therefore, there s an overshoot
in the frequency response along with an abrupt plunge of the phase.
This will translate in the time domain by a massive overshoot of the
step response, that is, instability...
In the original amp s closed loop gain, we can see that the bandwith
is 150KHZ , in line with the one measured by magazine Stereophile.
Modification was to remove the lead compensation cap (C4) and
increase the lag compensation cap (C2) to 47pF...
The curves are the ones of the original schematic
and the ones of the modified amp.
Contrary to the claim of a self designed "specialist " by there,(that only
manage to display his lack of knowledge in the discussed matter),
the models have nothing to do with the amp lack of stability, as the transfert
function of this amp is determined only by the compensation capacitors.
We can see that the original compensation doesn t work at all.
The system is not damped enough, therefore, there s an overshoot
in the frequency response along with an abrupt plunge of the phase.
This will translate in the time domain by a massive overshoot of the
step response, that is, instability...
In the original amp s closed loop gain, we can see that the bandwith
is 150KHZ , in line with the one measured by magazine Stereophile.
Modification was to remove the lead compensation cap (C4) and
increase the lag compensation cap (C2) to 47pF...
Last edited:
Why don't you post the full .asc file as you changed it, list all your changes, and the device models that you used. Drop the personal insults will you please or get out of the thread. You have had an attitude since you first claimed it was an unstable design and were corrected by the fact that there are (many) thousands of units out there that have worked just fine. Most failed when the outputs were shorted. This has been pointed out to you on more than one occasion here and I find your thread crapping annoying. You deliberately leave out information in an attempt to promote your ideas.
Start a new thread- What Wahab thinks is wrong with the Adcom because I see little value to your input here.
Your figures in the .zip files have no labels on the axes, you call this good engineering?
Start a new thread- What Wahab thinks is wrong with the Adcom because I see little value to your input here.
Your figures in the .zip files have no labels on the axes, you call this good engineering?
Last edited:
I dont use LT spice, so unless you use Simetrix, it won t
be of any utility..
About the curves, i think they are way enough, as any
first year school engineer would understand the meanings instantly.
It s calibrated linearly in degrees for phase and in Vout/Vin logarithmic
scale for the gain.
I think you have no trouble understanding that 1V is 0db,
10V is +20db and so on.
atached are the schematicS and sxsch files.
be of any utility..
About the curves, i think they are way enough, as any
first year school engineer would understand the meanings instantly.
It s calibrated linearly in degrees for phase and in Vout/Vin logarithmic
scale for the gain.
I think you have no trouble understanding that 1V is 0db,
10V is +20db and so on.
atached are the schematicS and sxsch files.
Last edited:
Deadamp
Having had about 2500 amps across my bench (5 this week) I can say for sure thousands of amps in the field are classified here as pure and unadulterated (place derogative term here). It is very seldom any trouble at all to get them to ring like a bell or simply break out into oscillation. And forget about driving a bad load. Does anyone here know about the "Power Cube" test? I tell you why it is not popular. Amps turn into smoke. Heck, when I plug my speaker cables into most (97%) amps they blow a fuse, go into protect, or catch on fire or are badly distorted when a signal is put through them. I consider most amps an insult to design and product quality which is why there have been so many short lived audio amplifier companies.
Most of these products are set out with the best of intentions but the actual result is not nearly so. Manufactures have a great variety of reasons to sell such items but the number one reason is to stay in business. Look for amplifier companies which have made ultra-reliable amps that last for decades and you might find amps worth owning that actually sound half way decent.
So in the in I must agree with wahab.
Having had about 2500 amps across my bench (5 this week) I can say for sure thousands of amps in the field are classified here as pure and unadulterated (place derogative term here). It is very seldom any trouble at all to get them to ring like a bell or simply break out into oscillation. And forget about driving a bad load. Does anyone here know about the "Power Cube" test? I tell you why it is not popular. Amps turn into smoke. Heck, when I plug my speaker cables into most (97%) amps they blow a fuse, go into protect, or catch on fire or are badly distorted when a signal is put through them. I consider most amps an insult to design and product quality which is why there have been so many short lived audio amplifier companies.
Most of these products are set out with the best of intentions but the actual result is not nearly so. Manufactures have a great variety of reasons to sell such items but the number one reason is to stay in business. Look for amplifier companies which have made ultra-reliable amps that last for decades and you might find amps worth owning that actually sound half way decent.
So in the in I must agree with wahab.
Chris, I'd appreciate it if you could move Fotios' and my posts concerning the MJE182 model to his thread so that we can continue the most pleasant conversation there:
http://www.diyaudio.com/forums/solid-state/163698-simulation-parameter-mje182-correct-wrong.html
Thank's - if it's not too much trouble.
http://www.diyaudio.com/forums/solid-state/163698-simulation-parameter-mje182-correct-wrong.html
Thank's - if it's not too much trouble.
wahab, it's easy enough to download the legally free ltspice and then you can talk to these other people with a lot of ambiguities removed. Surely that's of value if you consider the tone this thread has lately got.
Thanks, ikoflexer , i ve already have LT spice , although i
don t still use it as i find simetrix being much more user
friendly.
Anyway, the simulators display close values when it comes to
phase and gain measurements , so it s not a problem.
LTspice will display the same OL and CL behaviour with this
amp as simetrix , since the impulse response is primarly
influenced by the compensation network.
I m still waiting some phase/gain response simmed using LTspice,
as it will be insightful to compare the results of differents simulators.
Well, consider yourself lucky though, that you can easily get ltspice. In some remote part of this forum I once showed that some regulator was not even in simulation living up to the claims of the builder, let alone in real life. The response was "you're using an inferior simulator, ltspice! We, real engineers use orcad and we got these amazing results" They shut me up real good, except I had access to orcad as well and took the time to sim their circuit and indeed, very close result to ltspice. But then they could no longer blame the simulator. My point? Sometimes you have to accommodate the other guy and use with his own tools.
@PB, nothing against you or your claim, just thought I could help the discussion a bit.
@PB, nothing against you or your claim, just thought I could help the discussion a bit.
Arrogance at his best...
At least , i did produce some sims of this amp, while yo keep
feeding us with insults..
Where are your sims ?.
Stick to the technical debate, rather than
branding the others as incompetents people..
I would never be as irrespectful as you are with Sumaudio, and i highly
doubt that in real life you would dare to speak to people like in this forum...
We are in a public forum, and you are contradicted TECHNICALLY,
so if my technical demostration doesn t suit your ego , it s you who
must leave the debate...
Your real problem is that i exposed some blatant lacking in your knowledge,
and that you re unable to respond in a technical way...
By the way, good idea to ask andy_c, i m totally sure that he will find the same things about the phase/gain response of this amp.
If my sayings doesn t suit you, thre s an ignore button on this site.
Last edited:
infinia and I were making progress just fine, the model as I posted has no oscillation problems - you claimed it did without stating your changes.
We confirmed many days ago that the 33.2K is actually 33.2 ohms, that was obvious. You come in here claiming that it oscillates, in reality it does not, I try to ignore you, then we find that you have changed devices, are using a different simulator and do not disclose this right from the start. You are the one who is arrogant, and thinks you are a better designer who can instantly decide what is wrong with a design. Now, how many times do I have to ask you to leave, we were making progress but you have turned it into you trying to prove that your claim of instability still stands.
By the way, all I did was offer this, infinia did the sims you ask for, I don't answer demands especially from people like you. I didn't offer the sim to prove anything for you or anyone else.
You also don't seem to read my posts, I'm not going to ask andy_c for help, I'll just post the discussion on his new forum where I am already a member.
We confirmed many days ago that the 33.2K is actually 33.2 ohms, that was obvious. You come in here claiming that it oscillates, in reality it does not, I try to ignore you, then we find that you have changed devices, are using a different simulator and do not disclose this right from the start. You are the one who is arrogant, and thinks you are a better designer who can instantly decide what is wrong with a design. Now, how many times do I have to ask you to leave, we were making progress but you have turned it into you trying to prove that your claim of instability still stands.
By the way, all I did was offer this, infinia did the sims you ask for, I don't answer demands especially from people like you. I didn't offer the sim to prove anything for you or anyone else.
You also don't seem to read my posts, I'm not going to ask andy_c for help, I'll just post the discussion on his new forum where I am already a member.
Last edited:
I used the models you did publish, but since the simulator is different and that
you think it has an influence, i will try with the asc file you posted and
also with LT spice , although i m not familiar with this tool..
Hope we ll ultimately find a common ground for the sake
of the technical debate...
If anyone is interested, my gut feeling is that there are problems in this design as far as power/current goes under a short circuit condition. This is where I plan to put my emphasis. I will certainly take a look at stability however none of this has any priority since my amps are working just fine at the moment. I would also like to find output models closer to the actual devices.
I do welcome infinia's analysis on the stability side. I have already taken a quick look at it, and found ways to improve it, however I don't want to debate it.
I have also added a current mirror, and beta multiplier however I've not decided if I'm going to mod my amp so no point in posting it here if it will just add to the confusion. I might also do bootstrapped cascodes - we'll see.
I do welcome infinia's analysis on the stability side. I have already taken a quick look at it, and found ways to improve it, however I don't want to debate it.
I have also added a current mirror, and beta multiplier however I've not decided if I'm going to mod my amp so no point in posting it here if it will just add to the confusion. I might also do bootstrapped cascodes - we'll see.