♫♪ My little cheap Circlophone© ♫♪

Edit:
Does it do wide range voltage, such as from 18v rails to 39v rails?
Yes, it works from +/-15 to +/-42V, the zeners do not actually regulate, they serve to keep the voltage within safe limits for the FETs.

If a very good matched pair of FETs is available, or if one of the 220ohm source resistor is made adjustable, the non-polar 4µ7 could be discarded.

Ok it seems, I should expect an answer like "it doesn't matter of how you add RC filter at Circlophone's output as principle"
Basically, the order in which series elements are connected does not matter, except in cases of high impedance levels or religious issues.
 
I think Daniel means that could we get similar affect like Elvee's new CFP implementation if using aka 700 Hfe LTP transistors rather than 250 Hfe ones. I may wrong but it sounded like me that.
BC560C, 50v, HFE=420~800 (Fairchild datasheet specification)
Your idea, and it is a good idea. I wonder about the more orthodox approach of fine tuning both resistors of the gain divider. This much gain is easily sufficient for 90w to 8R size amplifier and you might not need to use up as much gain if your source is capable of pushing hard enough without blaring or distorting itself.

Taking steps towards very stiff rails, such as high amperage transformer, thick cable and a sturdy schottky at D4, D5 (see previous discussions), plus the addition of a soft clip circuit, could make the 90w to 8R size amplifier compete with 150w to 8R size amplifier and require output devices suited for the 150w to 8R size amplifier, especially when soft clip circuit is added.

The only thing you need to ask yourself is if you would like to turn up the influence of your amplifier or if you would rather push it with a preamp. I think that both options are well worth an interview.
 
I am grateful to see the new CFP "Circlophone 200" but applying it confuses me. Since the compensations and simulation is very specific, I would like to see it specifically (re)tuned for easily available parts capable of the target voltage.
high output power on 8 ohm (~150W), with a high input sensitivity
Can it have a new simulation with specific compensation fine tuned for either Hitachi 2SD669a (available, 160v, 12pF), Or Toshiba 2SC3421 (available, 120v, 15pF) for phase inverter?, And also maybe with a conveniently matching complimentary partner for driver (since they are stocked together)?, And a widely available output device capable of target voltage and current?
 
I should quit weird topologies to focus upon rolling parts. Vastly more irritating.

IIRC, the tubeophiles started the trend with tube rolling, then came the IC rollers replacing the NE5532, with types like OPA2134, OPA627 et. al......and now transistor rollers....

not as easy to do as tubes, transistor rollers do it thru simulation.......

and why not? this is DIY, a DIY'er has lesser constraints than a commercial designer....

sorry for the OT.
 
I have made a quick physical test of the CFP version:

http://www.diyaudio.com/forums/solid-state/189599-my-little-cheap-circlophone-61.html#post2892032

264284d1328393804-my-little-cheap-circlophone-circlopcfp.gif


I first attempted a direct plug-in replacement of the two input transistors by two composites, including their 15K BE resistors, without changing anything else.
There was a small (200~300mV) HF oscillation on the output waveform.

I then added the 270p/100R specific compensation network, and the oscillation went away.
I didn't care to make the rest of the modifications, and connected it to a source and speaker: as I write, I am listening to it and everything seems OK :)
So, after all the predictions of the sim were not that bad.
I would not recommend such a careless attitude to anybody else, but it proves it is safe enough for others to begin experiments.

My "test-bed" used the standard recommended components, except for the OP devices which were BD249C instead of TIP3055
 
I'm merely suggesting that Elvee has the patience of a saint.

Maybe that's the one of the significant reasons of ability of design such topologies.

I have made a quick physical test of the CFP version:

http://www.diyaudio.com/forums/solid-state/189599-my-little-cheap-circlophone-61.html#post2892032


I would not recommend such a careless attitude to anybody else, but it proves it is safe enough for others to begin experiments.

My "test-bed" used the standard recommended components, except for the OP devices which were BD249C instead of TIP3055[/QUOTE]

Thank you for good news :happy2: I'm almost hopeless to find some Philips BDT85/87 for my project. Their specs despite their little case (TO220) has surprised me. TIP and BD series from TI wink at me for a while.
 
I'm merely suggesting that Elvee has the patience of a saint.
He does have the patience of a saint.
I have made a quick physical test of the CFP version. . . as I write, I am listening to it and everything seems OK :)
Congrats! How do you like its subjective performance? Are there any noticeable differences except for more gain?

I cannot find reliable models for these transistors; anyway, that is something that needs to be tested for real, a sim alone is not able to tell the complete story.
Here was the amount of data that I was able to retrieve:
Hitachi 2sd699 (12pF), 2sb649
.model 2SD669 NPN (IS=5p NF=1 BF=250 ISE=5p NE=1.5 IKF=3 VAF=150 RB=1 RC=0.25 RE=0.25 TF=1.14ns CJC=50p )
.model 2SB649 PNP (IS=5p NF=1 BF=250 ISE=10p NE=1.5 IKF=3 VAF=75 RB=1 RC=0.25 RE=0.25 TF=1.14ns CJC=50p )
Onsemi MJE344 (15pF but maybe too slow), MJE350
Toshiba 2sc3421 (15pF), 2sa1358
Toshiba 2sc5171 (16pF), 2sa1930

All of those are generally available, and you could mix-n-match if you wanted to. And, if you want samples, PM me with your mailing address.

P.S.
I did finally manage to acquire some of those 1980's Philips BD140's, and when on unregulated power (meaning exposed to our multiple daily power surges), perhaps I could use 33v rails. This isn't really more powerful than what I already had, but maybe there an audible difference to explore. With these considerations and more, I've been shopping for speaker efficiencies in the mid 90's. . .
 
Congrats! How do you like its subjective performance? Are there any noticeable differences except for more gain?
Since I have tin ears, my subjective assessments are essentially based on negatives :D
I am satisfied when I hear no noise, no crossover distortion, no dynamic artifacts and no bandwidth anomalies.

Here, I left the gain unchanged, I just wanted to make a reality check with the CFP.
I am confident it will work properly at higher closed loop gains


Here was the amount of data that I was able to retrieve:
Hitachi 2sd699 (12pF), 2sb649
.model 2SD669 NPN (IS=5p NF=1 BF=250 ISE=5p NE=1.5 IKF=3 VAF=150 RB=1 RC=0.25 RE=0.25 TF=1.14ns CJC=50p )
.model 2SB649 PNP (IS=5p NF=1 BF=250 ISE=10p NE=1.5 IKF=3 VAF=75 RB=1 RC=0.25 RE=0.25 TF=1.14ns CJC=50p )
Onsemi MJE344 (15pF but maybe too slow), MJE350
Toshiba 2sc3421 (15pF), 2sa1358
Toshiba 2sc5171 (16pF), 2sa1930

I had found similar models, but they are too simplified, and the CJC value amongst other things is suspicious.
 
I'm listening the setup below for a few hours, :hphones: it is just wonderful! I couldn't imagine 200kHz RCA BDY29's will perform so beautiful. Maybe it is the benefit of their outstanding SOAR.

Another thing is after realizing that 1N5819 is not a good thing as D7 according previous comments, my noise performance get closed to actual one now with using BAT85 as D7. Frankly, I hadn't any idea about the difference between them until those comments and nice graps by Elvee. This journey became educational for me as being as joyful same time.

C12: 180p
D7: BAT85
Q5-Q6: Transitron 2N5858 (200MHz, 12pF)
Q11-Q12: Harris BD140 (No datasheet found)
Q1-Q3-Q4: Philips BC560C (Hfe ~540)
Other small transistors: Philips BC557B, BC547B, 2SC2240
Power Transistors: RCA BDY29
Without C11 of course.
 
Last edited:
Originally Posted by Elvee
I cannot find reliable models for these transistors;

I've got these 2 models, I don't know how good they are; You might be able to use them.

Regards,
Ted

.MODEL 2SD669 NPN (IS=12.0986F NF=0.8124 BF=255.1933 IKF=0.7146 ISE=15.842F NE=1.1041 VAF=88.4468 RC=0.48 RB=1.9 IRB=0.1787 RBM=0.475 RE=12M VAR=24.146 IKR=2.3712 ISC=3.8496F NC=1.135 NR=0.8027 BR=9.478 CJC=53.91P VJC=0.5 MJC=0.418 CJE=59.31P VJE=0.5 MJE=0.418 TF=0.936N XTF=0.318 ITF=0.8168 VTF=3.5927 TR=53.5038N XCJC=0.618 FC=0.543 XTB=1.487 XTI=2.818 Vceo=120 Icrating=1.5 mfg=Hitachi)

.MODEL 2SB649 PNP (IS=56.5F NF=1 BF=220 VAF=197 IKF=0.6 ISE=32.4p NE=2 BR=4 NR=1 VAR=20 IKR=0.9 RE=0.479 RB=1.92 RC=0.192 XTB=1.5 CJE=202p VJE=1.1 MJE=0.5 CJC=65.1p VJC=0.3 MJC=0.3 TF=1.14N TR=789N Vceo=-120 Icrating=-1.5A mfg=Hitachi)
 
Here is the input system that I'm using right now.

I'm not sure if the zener protector (parallel with the 10k input load) is the right value for everyone? How much voltage does an ordinary CD player output? Well, the computer likes to output a lengthy, subharmonic noise during start up and power management cycling, causing "x-max screen saver" and the zener thing shown, reduces those problems, and it blocks some RF, since zeners are modestly capacitive.

This has Ken's soft clip adapter too. I've modified it to remove ringing by adding a 470R resistor series to the diodes. A 1k didn't work there (neither ringing nor soft clip), so I picked up the 470R from the desk and used it. A smaller value, like 330R, probably works similarly.
That soft clip adapter does soft clip as shown. Man, you can really crank it up. For reference, I have 31v rails and an effective 70w is no problem, but now it does that cleanly. The resistors on that divider are 150R with 330R but I hunted through a few with the ohmmeter till I found some that came up a bit short so that the total would be really close to 470R, since the divider has replaced R16.
 

Attachments

  • inputmods.gif
    inputmods.gif
    14.4 KB · Views: 1,133
Originally Posted by Elvee
I cannot find reliable models for these transistors;

I've got these 2 models, I don't know how good they are; You might be able to use them.

Regards,
Ted

.MODEL 2SD669 NPN (IS=12.0986F NF=0.8124 BF=255.1933 IKF=0.7146 ISE=15.842F NE=1.1041 VAF=88.4468 RC=0.48 RB=1.9 IRB=0.1787 RBM=0.475 RE=12M VAR=24.146 IKR=2.3712 ISC=3.8496F NC=1.135 NR=0.8027 BR=9.478 CJC=53.91P VJC=0.5 MJC=0.418 CJE=59.31P VJE=0.5 MJE=0.418 TF=0.936N XTF=0.318 ITF=0.8168 VTF=3.5927 TR=53.5038N XCJC=0.618 FC=0.543 XTB=1.487 XTI=2.818 Vceo=120 Icrating=1.5 mfg=Hitachi)

.MODEL 2SB649 PNP (IS=56.5F NF=1 BF=220 VAF=197 IKF=0.6 ISE=32.4p NE=2 BR=4 NR=1 VAR=20 IKR=0.9 RE=0.479 RB=1.92 RC=0.192 XTB=1.5 CJE=202p VJE=1.1 MJE=0.5 CJC=65.1p VJC=0.3 MJC=0.3 TF=1.14N TR=789N Vceo=-120 Icrating=-1.5A mfg=Hitachi)


Thanks, the 2SD669 looks more complete, but the Cjc still looks fishy: it is too high to match with the datasheet figure, and it is practically identical to the Cje.