Amplifier Circuit Simulation Services

[Yamaha5000]

so migrate some of the CA-2010 circuit features to the CA-1000

Precisely the goal

 

[Yamaha5000]

I'm sure you appreciate the hours of work by @LesE to load the circuit into LTspice, he has also done the class A/AB switching in a neat way, good work.

I most certainly understand the tremendous effort that @LesE has put forward and the praise is off the chart!

 

[Yamaha5000]

use a smaller transformer gets you close

I have spoken to the Chief Engineer at Toroid and he feels comfortable that he can design and manufacture a Toroidal Transformer for this application. He needs the V&I load values that the transformer needs to provide.

Toroid Corporation

2020 Northwood Drive

Salisbury, MD 21801

 

[mbz]

No Sir, I like the idea. I did propose the idea to a couple Audio Engineers and didn't get much reaction from them. But that doesn't mean much.

I'm tempted to "pimp" the CA-1000 on the bench (refurbing) by adding the low noise regulator, do it on a breadboard for "proof of concept" probably add a switch so can switch between OEM and breadboard, maybe a couple of hours work with design, assembly,installation..

 

[Yamaha5000]

low noise regulated power supply based on that diyaudio thread I sent

Please send that link again.
We got hit with some major storm damage this Spring and it just about wiped us off the face of the Earth. I have been so busy rebuilding my home since then and just have not had the time to devote to this overall project. I have just started to get back into it over this past week.

 

[Yamaha5000]

I think this was the circuit you are speaking of?
There are 2,957 post to that thread!

 

[Yamaha5000]

@Yamaha5000 ... If you are still looking for this, I put together an LTSpice simulation for the CA-1000 power amplifier which seems to behave well using models for modern transistor replacements.

Is there something in particular that you would like to see from the simulation?

What changes did you find in performance with using different current model parts? How did you settle on selecting the ones' you used?

 

[Yamaha5000]

@Yamaha5000 ... If you are still looking for this, I put together an LTSpice simulation for the CA-1000 power amplifier which seems to behave well using models for modern transistor replacements.

Is there something in particular that you would like to see from the simulation?

The more I look at your drawing I am more and more impressed, so much great information with the listed V&I values!

One of my goals is to use the QA403 at every point of signal flow and document the results. Having the V&I values along with these results could be considered the ultimate troubleshooting tool?

 

[Yamaha5000]

@LesE Have you created any of the other boards in the CA-1000?
Have you created Gerber files from your design?

 

[Yamaha5000]

My current parts selections, maybe these should be reevaluated:

 

[LesE]

What changes did you find in performance with using different current model parts? How did you settle on selecting the ones' you used?

For Class A/B operation at rated power, the original design gives a THD of 0.022696%. Adding a current mirror to the input differential amp reduces the THD slightly to 0.011277%. A initial reconfiguration of the voltage amplifier to a Darlington configuration dropped the THD to 0.003198%. That's almost an order of magnitude.

I have transplanted the bias circuit from the CA-1010 into the CA-1000 design but more work is required. I'm not seeing any additional improvement at this point but I may have to simulate the CA-1010 in order to fine tune the modified circuit in the CA-1000.

 

[LesE]

Having the V&I values along with these results could be considered the ultimate troubleshooting tool?

Yes, that's primarily why I simulate these circuits. It can be a very powerful troubleshooting tool.

It's also very helpful when selecting modern replacement semiconductors. I prefer not to rely solely on transistor cross reference lists.

@LesE Have you created any of the other boards in the CA-1000?
Have you created Gerber files from your design?

This is my first look at a Yamaha amplifier which I created only after seeing your thread. For the most part, I focus on Pioneer products but I do occasionally stray.

I have created Gerber files for other projects but I'm generally not into creating new boards.

 

[LesE]

My current parts selections, maybe these should be reevaluated:

If you provide a list of the replacement parts used, I can plug them into the simulation to see what it looks like.

 

[LesE]

If your CA-1000 is currently running, it would be helpful to know the actual unregulated power supply voltages at the amplifier board. For class A operation, the schematic shows +/- 18VDC which is not quite sufficient to meet the THD specified for class A operation at 15W output. I suspect that the actual voltage is slightly higher.

 

[Yamaha5000]

If you provide a list of the replacement parts used, I can plug them into the simulation to see what it looks like.

See post 30

 

[Yamaha5000]

I have transplanted the bias circuit from the CA-1010 into the CA-1000 design but more work is required. I'm not seeing any additional improvement at this point but I may have to simulate the CA-1010 in order to fine tune the modified circuit in the CA-1000.

Beyond amazing!

 

[LesE]

See post 30

If you could make a list referencing the TR#'s on the schematic and the replacements used, that would make it a lot easier than trying to figure it out from the photo's.

 

[Yamaha5000]

If you could make a list referencing the TR#'s on the schematic and the replacements used, that would make it a lot easier than trying to figure it out from the photo's.

 

[LesE]

The good news is that your replacement transistors do work in this application and I was able to dial in the idle current.

The first thing I tend to look at is the frequency response looking for any signs of instability. With your choices, here is the predicted frequency response. Here you see an undesirable peak which may result in instability.



Here is the response using KSC2960 which is better behaved. There is evidence of increasing response above 10MHz but I haven't found any consequence of that so far.



Adding a current mirror to the input differential amplifier and improving the VAS provided further improvement. In this case, I used a KSC1845 and KSC3503 for the VAS.



Otherwise, it's normally good practice to use a complementary pair in situations such as TR607 and TR608. So if you use KSC1845 for TR607, then TR608 should be KSA992.

 

[Yamaha5000]

response using KSC2960 which is better behaved.

At what location?