Krill - The Next Generation

[Steve Dunlap]

Ignore the 3 blank pages in the screen shot. I don't know why they are there. I am still having trouble with the third file. I will try again.

[Steve Dunlap]

It finally worked. I have no idea what was going on. Here is the netlist.

[andy_c]

Okay, let's backtrack a bit. The detailed simulation data you presented were contained in posts 4 through 14. Those are the data I'm trying to correlate. Then in post #39, you included the netlist. Is it correct to say that the netlist of post #39 is from the same simulation used to create the graphs in your posts that were between 4 and 14?

[Steve Dunlap]

No, it is not safe to say that. The first netlist I posted for you did not include the complete amp. I was trying to keep the component count down for the demo version. I didn't realise at the time that only 20 components were allowed. I had stated back in post 27:

I threw everything out and started over. Fortunately, parts are cheap in simulation. I even changed the brand of some of the transistors, so these numbers won't be exactly the same as the previous amp.

You should have everything you need in the files I posted today. Now everybody has the same information if they use those files. I promise I wont change anything now that I feel the design is finished.

[Steve Dunlap]

Would you like to see the detailed simulation data presented again on the finalized design?

[GK]

Originally Posted by Steve Dunlap

Referring back to my question in post 77,

Krill - The Next Generation

I am not trying to insult GK with my question. Perhaps I am misunderstanding what is being said by either GK or Andy. It seems that Andy is saying the circuit will only operate as I think at high frequencies and GK is saying it can only work at low frequencies. Since these appear to be opposing views, I simply ask which is correct. I could be interpreting on of you incorrectly. Is that the case?

We're talking about 2 different things here.

1) With a cap between the bases of the drivers, it is only at low frequencies (depending on the cap size) that a "dynamic" biasing circuit could maintain non-switching of the output devices. At higher frequencies a large cap ensures that the biasing voltage is pretty much fixed.

2) At high enough frequencies, without a bias resistor for the drivers of the double EF, the power output devices may not have enough time to fully turn off.

[andy_c]

Steve,

The PDF you posted, "50W DIY Amp Scxreen Shot .pdf" is missing a bunch of the circuitry, making it impossible to associate all the node numbers on the schematic with those of the netlist. See the capture below. The front end came out fine, but the front end is not really part of the discussion about data correlation at this point. The circuitry in between the front end and the output stage - shown on the left side of the diagram below - is missing.

[andy_c]

Originally Posted by Steve Dunlap

No, it is not safe to say that. The first netlist I posted for you did not include the complete amp.

I think it's preferable not to look at the complete amp for the purpose of correlating simulation results. Since the discussion thus far has centered on the behavior of the output stage anyway, adding the input stage doesn't really clarify anything. It just muddies the waters further, making it more difficult to construct a working simulation file from the netlist. For example, suppose you aren't able to post a legible schematic containing node numbers corresponding to the netlist. From the one you posted above, it looks like there's some technical difficulties in the capture that are causing garbled schematics. If these difficulties can't be overcome, I could generate the simulation from a netlist of just the output stage - without the schematic. If you still have the full simulation files corresponding to the original netlist you posted in post #39, that may be the simplest way around all this. Do you still have them? It shouldn't matter that the netlist isn't for the final design. In fact, the discussion on output stage behavior was about your already existing design anyway, not some as yet undetermined future design.

[Steve Dunlap]

PHP Code:

Were talking about 2 different things here. 


Thank you for that explanation.

[Steve Dunlap]

Andy,

I am attaching the screen shot with node numbers again. This time it looks to be complete. I posted the complete amp because that is what I thought most here (as DIYers) would want to see. This is not some some "as yet undetermined future design". This is still the same basic amp, both VGS and OPS, that I presented originally. It has been changed slightly for improved low impedance performance and drive so it can be used in a bridged design for 200W output at 8 ohms. That is the reason it is an inverting amp. This has been built by me in the past, and is a working product. I have taken advantage of the sim program I have now, but not then, to adjust some component values. I do still have the files from post 39. What exactly would you like me to do with them? Lets keep it clean here!