Goldmund Mods, Improvements, Stability
- Thread starter keantoken
- Start date
Sorry about that Andrew...
I guess it is like the Fourth of July...
We celebrate it and those in the United Kingdom view it with a different point of view... Same day, different Meanings.
911 has dual meaning here in the states... it use to mean call for assistance or emergency, but now its meaning has many other connotations.
I guess it is like the Fourth of July...
We celebrate it and those in the United Kingdom view it with a different point of view... Same day, different Meanings.
911 has dual meaning here in the states... it use to mean call for assistance or emergency, but now its meaning has many other connotations.
no need for an apology. I'm sure I do all the time, but since I am speaking/writing in english, I am allowed. It's americanisms that are not allowed because this is decreed an english language Forum.
As for 911,
was that tragic date chosen because of the connotation to emergency services tel no? and that you Stateside folks swap the month with the day !
I don't think it's a good idea to mix signal ground and decoupling ground.
Sorry, I was a bit short.
Ground on the left:
The two 10uF caps might add some nasty current to the sensitive input ground plus the caps are a bit too far away from where they are needed plus this way there's a dead end going to the caps what is always bad and to be avoided.
A solution could be to place the 10uF caps directly next to the -/+90V connections and put another decoupling ground right next to it. This way the decoupling ground can't influence the sensitive signal ground plus you have the return right next to the +/-90V, can twist them together and thus have no big loop.
Power ground on the right:
The connectors are close to each other what is good. You could move the small 10nF caps much closer to the pins of the OP devices (where they are needed and have lower series resistance) and thicken the ground considerably.
C8,C9:
Why are the caps dissimilar?
Ground concept:
I suggest to have three grounds:
1. input and small signal ground
2. 90V decoupling ground
3. 80V power decoupling ground
What do you think?
Ground on the left:
The two 10uF caps might add some nasty current to the sensitive input ground plus the caps are a bit too far away from where they are needed plus this way there's a dead end going to the caps what is always bad and to be avoided.
A solution could be to place the 10uF caps directly next to the -/+90V connections and put another decoupling ground right next to it. This way the decoupling ground can't influence the sensitive signal ground plus you have the return right next to the +/-90V, can twist them together and thus have no big loop.
Power ground on the right:
The connectors are close to each other what is good. You could move the small 10nF caps much closer to the pins of the OP devices (where they are needed and have lower series resistance) and thicken the ground considerably.
C8,C9:
Why are the caps dissimilar?
Ground concept:
I suggest to have three grounds:
1. input and small signal ground
2. 90V decoupling ground
3. 80V power decoupling ground
What do you think?
If you ignore proper VAS ground loading you'll have not only ringing but every imaginable oscilating problems to solve with unnecessarily compensation. Really not needed at this point ...😀Already the circuit is shown to be unstable and prone to osculation around I meg. Layout will be tricky to avoid loss of speakers as it can make a bad design far worse. Alex wait for a signed off circuit, you are running away from time.
As it stands, the amp posted in 1`037 is seriously flawed. Let the boys fix it up first, then do a final layout.
Alex your abilities are exemplary but don't put too much effort until the design has been sanctioned. For the moment it is far too early.
As it stands, the amp posted in 1`037 is seriously flawed. Let the boys fix it up first, then do a final layout.
Alex your abilities are exemplary but don't put too much effort until the design has been sanctioned. For the moment it is far too early.
This was the prototype no 8 built early in 2007
It was a slow process to integrate such an amp that would be commercially appealing and at the right price as well as be MUCH BETTER THAN A GOLDMUND. The launch date was 12 March 2007 and it was done with the appropriate partying and sound industry guests.
It is again such time and RAS-ELD 2 which uses a mixed BJT/MOSFET configuration and undergoing severe testing for the next two months will be launched in February 2011 too distributors and in March 2011 to the public.
Because ELD1 will soon be redundant, I will offer it to DIY members to build.
Kind regards
Nico
It was a slow process to integrate such an amp that would be commercially appealing and at the right price as well as be MUCH BETTER THAN A GOLDMUND. The launch date was 12 March 2007 and it was done with the appropriate partying and sound industry guests.
It is again such time and RAS-ELD 2 which uses a mixed BJT/MOSFET configuration and undergoing severe testing for the next two months will be launched in February 2011 too distributors and in March 2011 to the public.
Because ELD1 will soon be redundant, I will offer it to DIY members to build.
Kind regards
Nico
Attachments
I think at least one or two have to be BUILT before we can definitively say that it is flawed... therefore we are laying out a BOARD to be tested, molested and played with to make it NOT be a rush to GB...
THIS IS NOT A Group BUY at this juncture... IT is a test board to iron out the winkles
Some of the 'wrinkles' can be ironed out before a board is made. I think (IMHO) that is what some of the more learned members are saying...
I agree, BUT
Both Keantoken and OS have "simulated" the board, and had NO such instability. 😉
SO, just how and with what models did he Simulate? Will he share his insight?😕
IF there is a problem then lets get it out in the open. I do not think there is a problem at this juncture... IMHO.🙂
This is a work in PROGRESS😛
Thanks
Both Keantoken and OS have "simulated" the board, and had NO such instability. 😉
SO, just how and with what models did he Simulate? Will he share his insight?😕
IF there is a problem then lets get it out in the open. I do not think there is a problem at this juncture... IMHO.🙂
This is a work in PROGRESS😛
Thanks
I think you are saying that Keantoken is ignoring Lazy Cat, is that correct?
I am not ignoring him, but I am going to let Keantoken address Lazy Cat's concerns.😉
I certainly HOPE and believe you are not implying that other here including me are "SOFT" on experience. If you are then say so. My SKIN is very thick. But I also have a FEW years of engineering experience. AND I shudder at the thought that you imply that maybe just maybe SOME are just Sales Engineering TYPES....😱 I think IF you go over there and spend a little time you will find some post that I and others made that will dispel any doubts you may have. This thread is NOT that thread.
Well, LET us BOTH wait and see what Keantoken has to say. I know I will.
Thanks🙂
Last edited:
I don't imply things, I come out and say them.
Time and again we've seen these 'design by committee' project stall or fail, for lack of a driven, decisive hand at the helm.
If you are the hard experience type, you grab the bull and build it, you don't wait around for permission or guidance.
I have NO years of engineering experience but if I was hot to build this amp (which I certainly am NOT), it most certainly would get done.
Yeah I hate to say I told you (guys) so....
But I did. We should have stayed much closer to the original design, but correct the errors and instabilities and make it safer etc... Now it became a different beast altogether.
I say: Go back to the roots. This design-by-committee is clearly not working (just as I suspected)
But I've voiced my opinion on this before and it wasn't welcome then, so it probably won't be now.
But I did. We should have stayed much closer to the original design, but correct the errors and instabilities and make it safer etc... Now it became a different beast altogether.
I say: Go back to the roots. This design-by-committee is clearly not working (just as I suspected)
But I've voiced my opinion on this before and it wasn't welcome then, so it probably won't be now.
You know MJL21193, I agree with you up to one point... THIS IS KEANTOKEN's Thread, he started it and I politely give deference to him because he did. SO "I" am willing to indulge the "design by committee" short-comings. That is my choice. OK?🙂
Now may I ask one question of you, and I aint hating on you. WHY are you HERE? I am sure it is to help. RIGHT?
Many others tried unsuccessfully to give the other thread leader🙄 a nudge, hint shove in the right direction from time to time. I got SIN binned for it.
I just want to build an AMP that sounds good to me and my family.
SO lets hold hands and work TOGETHER and make that same thing happen for EVERYBODY🙂
Impatience is not my brand of waiting.
TELL me what you think I can do to improve... I will listen. What did I do wrong? Remember this is Keantoken's Thread...
I am old enough to learn something.
THANKS