anatech said:
Joshua_G
You have brought out some good points from time to time, and presented them well. These last posts are not doing anyone any good and run against the general tone here.
If you have a problem, then please bring it up in a more insightful manner in an appropriate place. Try to avoid stirring things up. That action is not gaining you any respect here.
I do have an email button, you are welcome to use it. If you wish to rant, be assured I'll ignore you. If you want to talk about something that is bothering you, I'm game. I would much rather you get more out of being here than to be able to say to a friend "Hey! Look what I just posted! I really showed 'em". However, that is your choice.
If you want to be taken seriously, please calm down and post either information or questions relating to a technical topic.
-Chris
Edit:
Forgot the other "cop" symbol.
You are referring to a post of mine, which was deleted by one of the moderators, which contained only quotes of your words to me in another thread.
Well done!
Quoting your words is offensive!
syn08 said:
Well, here's a sample of brilliant ignorance:
Straight from one that made a good living designing simulators for Cadence Systems: I swear all situations are accounted for, including non-conventional topologies
So, you are saying that simulation software are perfect.
So you said.
How about when simulation contradict an actual circuit?
syn08 said:
Well, here's a sample of brilliant ignorance:
Straight from one that made a good living designing simulators for Cadence Systems: I swear all situations are accounted for, including non-conventional topologies
syn08
Most of the people here are using LTSpice it’s a freeware thing.
PSpice is of course a lot better and it also runs Mextram models.
So when I referred to his post I was thinking about LTSpice.
Straight from one........ ( are you talking about yourself?)
Cheers
Stinius
stinius said:
- Models are not part of the simulator, they are plugins.
- Models have nothing to do with the (quote) "topologies".
- All Spice incarnations are horribly similar in architecture to the original Berkeley (available as source code to anybody willing to get e.g. a YACC dose).
- Sssshhhh... Yessir
Which pretty much explains why I hate simulators syn08 said:
- Models are not part of the simulator, they are plugins.
- Models have nothing to do with the (quote) "topologies".
- All Spice incarnations are horribly similar in architecture to the original Berkeley (available as source code to anybody willing to get e.g. a YACC dose).
- Sssshhhh... Yessir
I thought so.
Hi stinius,
I have been trying to be crystal clear here. The "Edit" line in my post was light humor. Believe it or not, I am in a good mood. I guess I should ask you if there is something bothering you.
I stated the following :
Stinius, what point are you making here? I have been very clear I think, what is your point or concern, or are you just stirring things up? I would rather like to think you simply only read what you wanted to read. As I said to Joshua_G, there is a mail button down there. If you want to clear up a concern, please just send me mail and I'll try to address your concerns. Fair?
-Chris
That post was not supposed to be missed or ignored, as you have pointed out I did use the official cop symbol. That post was a direct request to Joshua_G to keep his comments on a technical level. The last couple posts he made were designed to stir things up and had exactly zero technical content. Much like in the Blowtorch thread.
I have been trying to be crystal clear here. The "Edit" line in my post was light humor. Believe it or not, I am in a good mood. I guess I should ask you if there is something bothering you.
You didn't read my post.
I stated the following :
This is the exact example of a good post that Joshua_G has made, and I did acknowledge. The next few posts after this were personal attacks, as I pointed out.
Stinius, what point are you making here? I have been very clear I think, what is your point or concern, or are you just stirring things up? I would rather like to think you simply only read what you wanted to read. As I said to Joshua_G, there is a mail button down there. If you want to clear up a concern, please just send me mail and I'll try to address your concerns. Fair?
-Chris
Hi Joshua_G,
I did not delete any posts at all. I was not even aware your post had been deleted. You didn't even have to courtesy to ask before making your comments. I would have been able to clear this up with you.
I will give you an honest, serious answer. No tricks, no making fun of anyone.
If a simulator does not agree with a physical circuit, then the results from the simulator program are wrong. It is that simple. Understand that a simulator and model make only approximations of how a component reacts in real life. The real circuit must follow the real laws of physics in every detail, no simplifications. Other members have commented that a simulator approximates the operation of a circuit and makes sweeping assumptions as well. So the result from a simulator must be viewed as a rough tool to guide you in the right direction, nothing more.
Joshua, I am concerned that you may be heading for bin time due to your constant attacks on other members. I've tried to point out that you need to avoid making personal comments and disruptive remarks on more than one occasion. If I were trying to give you a hard time, you would already have been binned. Instead, I have warned you often and perhaps that was my error. Perhaps I should have given you some cooling off time earlier.
I am going to say this clearly again. I will not deny that your attitude bothers me, and that is my primary problem with you. Your attitude is my only problem that I have with you. I have invited you to contact me if you feel like there are problems, and I would rather do that than take action against you. I normally would say this privately, but I do fear you may force one of the moderators to react. Personally, I don't think that is in your interest or anyone else's.
Okay everyone. I can see people starting to get into things other than Steve's amp. Those responsible are doing Steve a disservice, and all the other members who are interested in this amp design. So for the sake of everyone else, personal attacks will not be tolerated. Let's keep things on a technical level please, from here on in.
-Chris
I did not delete any posts at all. I was not even aware your post had been deleted. You didn't even have to courtesy to ask before making your comments. I would have been able to clear this up with you.
Here is a good example of a post not designed to move anything forward. You are simply making a personal attack.
Joshua. Did you think about this before you posted it?
I will give you an honest, serious answer. No tricks, no making fun of anyone.
If a simulator does not agree with a physical circuit, then the results from the simulator program are wrong. It is that simple. Understand that a simulator and model make only approximations of how a component reacts in real life. The real circuit must follow the real laws of physics in every detail, no simplifications. Other members have commented that a simulator approximates the operation of a circuit and makes sweeping assumptions as well. So the result from a simulator must be viewed as a rough tool to guide you in the right direction, nothing more.
Joshua, I am concerned that you may be heading for bin time due to your constant attacks on other members. I've tried to point out that you need to avoid making personal comments and disruptive remarks on more than one occasion. If I were trying to give you a hard time, you would already have been binned. Instead, I have warned you often and perhaps that was my error. Perhaps I should have given you some cooling off time earlier.
I am going to say this clearly again. I will not deny that your attitude bothers me, and that is my primary problem with you. Your attitude is my only problem that I have with you. I have invited you to contact me if you feel like there are problems, and I would rather do that than take action against you. I normally would say this privately, but I do fear you may force one of the moderators to react. Personally, I don't think that is in your interest or anyone else's.
Okay everyone. I can see people starting to get into things other than Steve's amp. Those responsible are doing Steve a disservice, and all the other members who are interested in this amp design. So for the sake of everyone else, personal attacks will not be tolerated. Let's keep things on a technical level please, from here on in.
-Chris
That leads me to a question.. I now have a jig like MJL's and
my LT ..all on the same desk.
Should we keep our discussions to a real world level?.. oops
I fried my zobel resistor
( LT just predicted oscillationwith ksa1381/ksc3503).
I just tinker with the sims , and I have found operation wise
the sim can go further than rough (LT predicts 7.12 mA -
real circuit 7.15mA) I don't have an AP ,so I have to
rely on my ears and LT to "give it to me rough"
This aspect created problems , different devices , whether
the krill would "switch" ...here I acknowledge the sim is
"rough" (I did get the KSA/C's to work with different CCS currents)
even through the sim said it would oscillate. BUT , where
the sim would actually run the circuit , it would also
run on the breadboard.
BTW .. 80-90 uA through the diode string with the KSA/C's
and 210-230 uA with the MJE340/50's. the MJE's run best
at 8.5 to 9 ma CCS and the ksa/c's 6-7ma.
AndrewT said:
Yes? And what are the major differences? Tell me how mine is catastrophically unworthy?
I think that too many here (you included Andrew, no offense) nit pick too much. Exactitude rules.
I played with that circuit after, reduced the 39K r's down to 24k and monkeyed with some other things and you know what? It looks exactly the same on my scope.
You fellows are completely missing the point entirely, go back and read my posts through again.
Right on , the values you use for the CCS resistors don't make
a bit of difference. You could even use led's and 150R , as long
as you get 7-10mA. I went through a whole "junkbox" of
used devices and could get them all to work with the krill.
ANY relative high gain device is good for the diamond , lower gain is best for Vbias/drivers.
you will never get a perfect match for p/n devices , hence the
trimmers in the latest schema. I used MUR120 diodes (best for njl/
mjl0281/0302), for the bias string.
Close enough is alright with the krill , a "window"(not exact)
of operating currents is where it works the best.
OS
MJL21193 said:
Steve, didn't include the resistors just to match the board, he included them for offset adjustments especially if using this as an OPS only, which is what you are doing. MJL it's best to read the thread and acquaint yourself with the details of the circuit before you build it. In audio, the devil is often in the details!
Anyway, according to yourself you were not trying to reach audio nirvana (although you disputed the .0005% THD claims of Steve) & as stated above you seem fulfilled. Cheers
Do you have any offset?
Edit: That's good technical info OS - so I wonder does MJL's circuit have bias in the correct range running in the string for the devices he's using? Maybe he can answer?
OS, when you say you could get a range of junk devices to work with the Krill, I'm wondering how optimal the circuit was operating? Our purpose here is to investigate this little circuit & it's a pity you don't have the measuring equipment to nail down some results as you've built the most variants.
Hi ostripper,
Hi Andrew,
It would seem that John's circuit is close enough and does illustrate the latitude you can have with a circuit. I don't know, but possibly different simulators would be invalid because they differ greatly from the one Steve uses? I think there is a valid question in there. You could extend that to the models used, and that is a valid point.
In future, maybe we should have a look at differences before judging them to invalidate something like a built circuit? Hey, the PCB isn't the same, in fact it's not even used. That has got to perform differently, yes? I'm just showing that we need to allow some latitude as long as the errors are not severe.
Hi jkeny,
Well, I do have the equipment required. I just do not have the exact PCBs, so there is nothing I could build here in Canada that would be valid. The PCB (as stated above) has the ability to greatly affect a circuits performance.
I think that we are investigating the circuit idea as implemented by various members. Taken with the knowledge that each one is a little different, we can accept measured performances. Then we look for commonalities between them to get to the average character for this circuit arrangement or idea.
-Chris
No. I think simulated results are useful to a point and worth discussing. Now, if a real world result contradicts what the simulator is showing, we had best either fix the sim or ignore it's results in that case. Arguing a working circuit doesn't because the simulator says it can't is rather silly in my view. That's more an opportunity to fix the simulator in my book.
Hi Andrew,
It would seem that John's circuit is close enough and does illustrate the latitude you can have with a circuit. I don't know, but possibly different simulators would be invalid because they differ greatly from the one Steve uses? I think there is a valid question in there. You could extend that to the models used, and that is a valid point.
In future, maybe we should have a look at differences before judging them to invalidate something like a built circuit? Hey, the PCB isn't the same, in fact it's not even used. That has got to perform differently, yes? I'm just showing that we need to allow some latitude as long as the errors are not severe.
Hi jkeny,
Well, I do have the equipment required. I just do not have the exact PCBs, so there is nothing I could build here in Canada that would be valid. The PCB (as stated above) has the ability to greatly affect a circuits performance.
I think that we are investigating the circuit idea as implemented by various members. Taken with the knowledge that each one is a little different, we can accept measured performances. Then we look for commonalities between them to get to the average character for this circuit arrangement or idea.
-Chris
With any 2 diodes and the MJE's he would have to be running
7+ ma , no way around it. With the various junkbox trannies
I subbed 56R, 68R,82R ..just read the voltage drop - ohms
law =, tried to get in the 7-8mA range. The red leds work
good too , 100-150R depending on led. You can place 22u
caps from the LED's/rails for a much "cleaner" CCS.
(all of the above with 40v rails)
the MJE's seemed to work best , as the bias trim value was
closest to the schema's specified value.
Sounded "tinny" with the ksa/c's for Vbias.
So , what is on the LT sim is the best (MJL's prototype) ,except I used
much lower gain drivers (MJE15032/33 instead of the 2sa/c
japanese devices)
I am going to order some chipamps and power the krill with
them, should be interesting.
Yes , very silly . What they were arguing was the distortion claim,
It always worked , just at .02 -.1 % ..with Andy c's kind help,
.01 was the norm.. which isn't bad for open loop.. sort of
"picky" as one would never hear that level to begin with.
closed loop it does as good as my best EF triple (.0015%)
OS
ostripper said:
Seem so, in my limited understanding.
I have garbage on the simulator with the MJE340,350 models I have - total krap. Nowhere near the actual state of things in the actual OPS, I've changed to my 2SA1381/C3503 models for results that don't stink (simulator only, not my construct).
OS, what are the MJE model you have?
Care to post them here?
- Status
- This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.