CFP Amplifier - Help Please!

[AndriyOL]

Cortez, I draw the schematic of IPS in LTspice. I can't set the zeners and LEDs values, there is no such option.
Could you try to simulate it and find required values?

 

[Cortez]

Yes, maybe later in the week.
But untill you can simply use a DC voltage source intead of them.
I have a LED modell at home where I can set the forward voltage to the desired value.
Zeners: right click and select modell from the list there will be a 15V one I am sure.

 

[AndriyOL]

Right click on zeners only shows options to set forward current and breakdown voltage, no any list

 

[Cortez]

Hi Andriy,

The circuit is prepared for you to any further analizations!

Changes:

• the zeners have now real models for both 4.7V and 15V
• the LED has a custom model where you can adjust the forward voltage based on your real ones
  (just right click on that long black text down under
  and change the "N" value, now its 3.2V but it depends on your actual LED type)
• I added a transient 1msec simulation I like to check even the DC conditions like so:
  run the sim with the running man and click to the node where ever you like to measure the voltage
  (by default related to GND but with a right click you can choose any node as reference)
  and also the currents clicking on the component itself
• power supply-s are tied to GND
• values given with short names

Now setting the trimmers to 60 ohms the 15V zeners are also around 5V.
I am not familiar with this design so I cant go any further right now but I hope it will be a good base for you to continue!
Good luck!

 

[Cortez]

And here is a working one with a sine source and fed back from the IPS output.
Of course it's not ideal this way because of the high feedback currents drawn from the VAS, but it's working.

To get it working I had to increase the R6/7 resistors to 400R but this
is because the default LT transistors have high Vbe arounf 800mV...

Also you can draw the OPS now.

Tip: you can check the current on a wire using the "Alt" key.
On a component it tells the dissipation which is cool.

 

[AndriyOL]

Thank you for your help Cortez!
Perhaps someone familiar with this design will be able to assist further to detect the problem.

 

[Cortez]

Yes, I came to the same conclusion, because I draw a simple OPS but couldnt achieve the cicruit to work perfectly.

 

[AndriyOL]

Btw, LC's SSA base core should be perfectly working according to SSA related threads. However this version of schematic was never built before to be publicly known. I don't think this OPS bias stage is able to work correctly.

 

[AndriyOL]

So where does the current go? AND WHERE FROM? There are still 2 possibilities and I hope you get the idea to find it by yourself.

Hi Dedegogo.
Do you know where could be the fault?

 

[de Ocampo]

Btw, LC's SSA base core should be perfectly working according to SSA related threads. However this version of schematic was never built before to be publicly known. I don't think this OPS bias stage is able to work correctly.

This is exactly what I had in mind when I suggest to contact Andrej Lakner. I was a follower of the SSA when it first started, I stopped following when the design goes into complexity. I only know of two members (Shaan & Nico Ras) who successfully built the simpler SSA.

 

[AndriyOL]

I already asked LC for the sim model of the amp. I hope he could generously share it

 

[de Ocampo]

I already asked LC for the sim model of the amp. I hope he could generously share it

He is a good guy, he helped me to properly lay out my VSSA board design, I also got his approval response when I used other jfet model for the CCS.

Regards,
Albert

 

[AndriyOL]

Just thought, if without GND voltages have doubled on components, than a bases of bottom CCS transistors were stressed 27mA which is higher than their SOA

 

[arpagon]

Hi Andriy,
Just thought that it would be better to have a schematic with measured voltages in main points of both working and defective channels.
It would be of great help to solve the mystery.

Regards
Alex

 

[Cortez]

As far as I know there is only one chanel built which is defected.
Isn't here someone how is familiar with this design and could make this circuit work ih LT Spice?
I tried it with a default OPS but couldn't make it work.
It would be very helpful to make it work at least in the simulation.

 

[AndriyOL]

Hi Andriy,
Just thought that it would be better to have a schematic with measured voltages in main points of both working and defective channels.
It would be of great help to solve the mystery.

Regards
Alex

Here is the schematic with measured potentials. Still only one channel built, sorry.

 

[AndriyOL]

Tried to change the CCS transistors - it takes longer time for IPS LEDs to glow at first start but no results.

 

[AndriyOL]

Assembled second channel - the same voltage drop on 15V zeners. Perhaps when I had 15V on those zeners the pots were set to 100Ohm - without load.
The only conclusion can be that some of my new to92 transistors are faulty or it's correct behavior for zeners with such a load. However, they are 5W types Onsemis 1N5352BRLG

 

[gannaji]

Dear andriyOL,
How did you get the PCB for this ? Is it your own design ?
--gannaji

 

[AndriyOL]

Yes, reworked from Alex MM design.

I think that you have more than 10mA or 20v across the 2k2. So what is causing the huge current draw?

Dedegogo, could you please give your thoughts on second ch. results? Schematic is through 3 posts earlier.