2stageEF high performance class AB power amp / 200W8R / 400W4R

[kgrlee]

You can get an EF3 stable with your compensation scheme if you use a shunt network at the VAS output to emulate the EF2 driver parasitics, without reintroducing the associated distortions.

Evil! Truly EVIL

 

[keantoken]

Okay, but this amp already has the same shunt network, which is the driver parasitics except worse because they change with voltage (what is the total Cbc of 2 pairs of C5171/A1930? Near 500pF?). So unless you have a different argument, in this situation the only thing an EF2 provides over a triple is an extra portion of low-order harmonics, rather than lower high-order harmonics.

But I digress. I don't care about "improving" this design with an EF3. I'm just stating that it's not incompatible with an EF3.

 

[kgrlee]

Okay, but this amp already has the same shunt network, which is the driver parasitics except worse because they change with voltage (what is the total Cbc of 2 pairs of C5171/A1930? Near 500pF?). So unless you have a different argument, in this situation the only thing an EF2 provides over a triple is an extra portion of low-order harmonics, rather than lower high-order harmonics.

But I digress. I don't care about "improving" this design with an EF3. I'm just stating that it's not incompatible with an EF3.

We totally agree here.

I didn't realise C5171/A1930 was as bad as that.

If I was developing this into a commercial product, I might simplify it even further depending on whether I think anyone can measure any difference with SOA test gear.

 

[sajti]

(what is the total Cbc of 2 pairs of C5171/A1930? Near 500pF?)

It's less than 64pF. But of course it will change with the output voltage. Say it can be 3 times more when the Vcb is low.
The question is the capacitance of the output devices...

Sajti

 

[astx]

Don't forget: since post #119 we are using 2SC4793/2SA1837 instead of 2SC5171/2SA1930 due to the better and also needed higher SOA!

• 2SC5171: ft 200MHz, Cob 16pF
• 2SA1930: ft 200MHz, Cob 26pF

• 2SC4793: ft 100MHz, Cob 20pF
• 2SA1837: ft 70MHz, Cob 30pF

As output devices TTC5200 and TTA1943 are used:

• TTC5200: ft 30MHz, Cob of 145pF
• TTA1943: ft 30 MHz, Cob of 240pF

(Cob of 2SC5200 and 2SA1943 are higher: 200pF/360pF)

BR Toni

 

[sajti]

Don't forget: since post #119 we are using 2SC4793/2SA1837 instead of 2SC5171/2SA1930 due to the better and also needed higher SOA!

• 2SC5171: ft 200MHz, Cob 16pF
• 2SA1930: ft 200MHz, Cob 26pF

• 2SC4793: ft 100MHz, Cob 20pF
• 2SA1837: ft 70MHz, Cob 30pF

As output devices TTC5200 and TTA1943 are used:

• TTC5200: ft 30MHz, Cob of 145pF
• TTA1943: ft 30 MHz, Cob of 240pF

(Cob of 2SC5200 and 2SA1943 are higher: 200pF/360pF)

BR Toni

The 10pF between the 2 different drivers is not critical, I guess.

Sajti

 

[astx]

One of the things that allows this amp to be so stable with such little internal compensation is the relatively high parasitic C at the OPS input. This can be visualized as a large cap to ground after the total base stopper resistance. However this capacitance is a proverbial lightning rod for high order switching harmonics of the OPS.

On the note of consistency in stability, I think it's a good idea to use .12R emitter resistors in the OPS so they can be biased higher; otherwise they may be too slow in the crossover region and too fast outside of it, and you will oversize your compensation in order to compensate for the slow region. Unfortunately that makes an amp like this dissipate a lot of heat. A more unflattering option is to slow down the outputs at high current to match speed in the crossover region. I think this can be done by bridging the base and emitter resistor with a cap, say 3-5 times the Cbe of the transistor in question.

For C5200/A1943, Cbe is quite large. In addition to this, .47R emitter resistors mean they need to be biased low to get the null point (13mV per Re) - so OPS speed is probably bouncing up and down wildly while in operation. Moving to faster outputs would probably allow for more consistent stability and therefore lower compensation.

kgrlee, I make my own C5200/A1943 SPICE models which are included in the standard.txt file I posted way back. I matched Ft to the datasheet - why don't you use those models?

Bias is optimized for lowest crossover distortion so bias voltage at Re is tuned to about 21mV. If done so correctly the distortion residual looks very smooth.

On using 0.12R resistors you

• have to match your devices and resistors exactly and
• the risk for second breakdown especially on driving 4R loads is rising if one device conducts better than others, heats more ...

So I decided to use 0.47 due to above reasons and thinking about environment. With 0.12R it would be better to design/use a pure Class A power amplifier.

BR, Toni

 

[astx]

kgrlee, I make my own C5200/A1943 SPICE models which are included in the standard.txt file I posted way back. I matched Ft to the datasheet - why don't you use those models?

If kgrlee is using all files from post #109 he is also using your models. My fault: I have copied your models into the file "2SC5200_2SA1943.lib" and renamed them to 2SC5200 and 2SA1943. Please forgive me .

BR, Toni

 

[keantoken]

So it doesn't really have too much more than an EF3 would. I didn't realize the voltage influence was so dramatic. So Early voltage is the next source of high-order distortion loading of the VAS.

I put the suffixes on the models so they're traceable and anyone looking at the schematic knows that it's a custom model.

 

[sajti]

I made some simulation with different base stoppers in my triple EF.

There was no significant difference in the distortion, even with increase the base stoppers to 10ohms(output), 100ohms(driver) 1kohm(predriver). Currently I use 10ohm/22ohm/100ohm

Sajti

 

[manso]

I used Michael Kiwanuka's method of just moving R27 in Toni's original from earth to the output in a working and stable TPC version. Yes. I know there are better methods to get evil TMC stable. You may like to download Toni's model from post #109 which is what I used .. & play yourself.

I couldn't get it to oscillate with Toni's original values using the methods in my post #177. There was little change to THD or the 20kHz residual on what I simmed (which didn't oscillate)
Please excuse me if I didn't spend zillion hrs working on something which doesn't have any potential (or real) advantage over 'pure Cherry' ... and I believe to be evil.

And Yes. I know Toni has some later TMC versions which were stable. But its a lot of work doing full stability .. even in SPICE world. You might like to try both his new & old versions. If you can modify the models to show 'old unstable' and 'new stable', you will have improved the sum of human knowledge and deserve our adoration.

I'll post my own final model and the tools for my results in a bit.

Michael Kiwanuka knowingly or unknowingly is misguiding how TMC should be implemented for some obscure reason, this is obvious from his posts. If implemented exactly as Edmond recommends or how Bob Cordell explains in his book its simple and stable and works better than TPC.
It takes me less than 5 minutes to implement it on a blameless design and the design always benefits from it. TPC and TMC are different and should not be used by just simply changing the resistor takeoff point. He doesnt want to see it this way as if he did he would find a very good way of compensating a amp which he didnt invent. Your implementation in post 177 is incorrect, now try implementing as Bob Cordell has explained numerous times.
There are complete designs here on the forum using TMC that have been built and tested by hundreds of users, surely one of them would have complained if there were problems with the implementation of TMC.

 

[dadod]

Michael Kiwanuka knowingly or unknowingly is misguiding how TMC should be implemented for some obscure reason, this is obvious from his posts. If implemented exactly as Edmond recommends or how Bob Cordell explains in his book its simple and stable and works better than TPC.
It takes me less than 5 minutes to implement it on a blameless design and the design always benefits from it. TPC and TMC are different and should not be used by just simply changing the resistor takeoff point. He doesnt want to see it this way as if he did he would find a very good way of compensating a amp which he didnt invent. Your implementation in post 177 is incorrect, now try implementing as Bob Cordell has explained numerous times.
There are complete designs here on the forum using TMC that have been built and tested by hundreds of users, surely one of them would have complained if there were problems with the implementation of TMC.

Yes, all well explained.
By the way, I think that well implemeted TMC will chalenge pure Cherry, but without stability problems.

 

[manso]

Yes, all well explained.
By the way, I think that well implemeted TMC will chalenge pure Cherry, but without stability problems.

I wouldnt be surprised.
I admit I am biased against Cherrry because of the bad rap Ive heard from others but Im open minded and should try it myself before condemming it. Each of these implementations have their benefits but I like TMC the most as it targets the output stage distortion which is exactly where its needed the most.

 

[astx]

We totally agree here.

I didn't realise C5171/A1930 was as bad as that.

If I was developing this into a commercial product, I might simplify it even further depending on whether I think anyone can measure any difference with SOA test gear.

Dear kgrlee,

attached current amp using new drivers and with your cherry compensation changes. A quick check of loop gain shows we have less 30 degree phase margin and less as 10dB gain margin with new drivers. Would you please be so kind and try to adapt your cherry recommendations to this version?

Thx and BR, Toni

 

[astx]

Phase shift using different compensation Methods ...

Has anyone compared input versus output phase shift at 20kHz using

1. Miller
2. TPC
3. TMC
4. Cherry

Miller and TMC have about 1.5 degree, but TPC nearly no phase shift.
Cherry is TBD - but think same as Miller.


TPC seems to be the Winner... or we have simulation problems.



BR, Toni

 

[dadod]

Has anyone compared input versus output phase shift at 20kHz using

1. Miller
2. TPC
3. TMC
4. Cherry

Miller and TMC have about 1.5 degree, but TPC nearly no phase shift.
Cherry is TBD - but think same as Miller.


TPC seems to be the Winner... or we have simulation problems.



BR, Toni

You should not generalize it in that way. Each type of compensatio should be optimized(specialy TPC-TMC should not use the same C R values) and then compare.

 

[astx]

You should not generalize it in that way. Each type of compensatio should be optimized(specialy TPC-TMC should not use the same C R values) and then compare.

Dear dadod,

who said I am using same compensation values? You can't get the amplifier stable with same values for Miller, TPC and TMC!

Of course you must first optimally compensate the amplifier and afterwards measure AC response and phase!
And then we can see a small but simulateable and measureable difference in phase at 20kHz! Not very much and maybe completely ignoreable as every speaker and crossover does much more phase shift...

 

[AndrewT]

Does 1.5° @ 20kHz matter?
That equivalent to 0.21us (maybe?)

 

[dadod]

Dear dadod,

who said I am using same compensation values? You can't get the amplifier stable with same values for Miller, TPC and TMC!

Of course you must first optimally compensate the amplifier and afterwards measure AC response and phase!
And then we can see a small but simulateable and measureable difference in phase at 20kHz! Not very much and maybe completely ignoreable as every speaker and crossover does much more phase shift...

Sorry if you understood me in that way. I meant that your general statment does not say anything about particular configuration just type of the compensation and that could be misleading.
BR Damir

 

[astx]

Spectrum comparison

TMC compensated. New spectrum analysis.

Attached spectrum of oscillator output and amplifier output.
Not really a difference between fundamental and H2, H3.

BR, Toni