Redesign Crown K2/1 BCA amplifier (with schematic) pls help with the feedback

Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.
Here is the schematic of the K1/2 amp.
You can downloaded it, but you MUST delete it after downloading:) I take no responsibility for it.
The scematics are only for 1 channel!

http://onlany23.atw.hu/Crown Input Board.pdf
http://onlany23.atw.hu/Crown Main Board.pdf
http://onlany23.atw.hu/Crown Output Module.PDF
http://onlany23.atw.hu/Crown K series Service Manual.pdf

So we can start rebuild it!!!!
Its way better than ClassD, the FETs have got
separate Drivers with NO deadtime.
The high side fets have got +Vcc+12V driving source from the PSU. There is no need of Bootstrap! And half bridge drivers have dead time, so the are completely wrong for BCA!
Thanks to NO dead time the THD will be very low!!!

The whole amplifier's heart is 2 LM361 precision comparator! (you can replace with other comparator (T<=20nS differential output), and the differential output is not needed, for shutdown use SHDN pin!

I want to rebuild it (in a 100W version first:) with only using few parts:
4Mhz clock (divided to 250Khz), integrator, 2 precision comps, fet drive+fet,filter,overcurrent protection,temperature protection,and an input buffer,
----------
Please explain me how does these feedbacks work in the amp!

So there is an I monitor at the output the signals from it gos to an error amplifier (that varies the sound input), and goes to the triangle generator.

The second feedback is the "FO" and its goes to the
triangle generator and the DC servo (the dc remover).

But the feedbacks are after the coils so the signals from them are phase shifted?

What does happening when the I monitor recognises
normal currents? How its modifing the sound signal, and the triangle?

The T100 is not sensing the Over current!
It just a feedback. The over current is sensed by
a windows comparator that mesures the current flow in the mosfet line.

There are a lots of R-s and C-s on the feedback line what do they do? Do they correct the shifted phase?

The white paper of BCA is not complete, the BCA isn't
using only 2 inductors! It has 3 LC lowpass filters at the output (two 250Khz and one 500Khz, the 500Khz is for bridging), so there is 5 aircore 30A inductors!
But we cant remove any of them because its affect
phase shifting so the feedback will not work.

If DC servo circuit sees DC on the output what does it do?

What doese the response EQ circuit do?

The I monitors output is 1V/5A, so the maximum is about 2,5V.

What kind of diode do you recommend?
The service manual says only : 4x300V ultrafast TO-220
(current rating?)
I read the service manual many times but there are some things that not mentioned.

So please help me to understand these feedbacks,
I only need that to redesign the amp (ofcourse just for education purpose!).

This thread is not about:
-triangle generation
-MOSFET driving!
-the type of the MOSFETs
-PIC controlling
-SMPS
-Bridging
-Stereo BCA
-soft start

There are some unneeded:) circuits like:
-Thermal LED
-Clip
-IOC
-Output signal LED
So please forget these.

And please dont try to reinvent BCA, you CANT do better BCA amp then the professional K2/K1 series
Please concertrate on the feedback.

And before you post read carefuly the service manual!

Download the PDFs soon as possible (and delete after it), I cant hold on the server for a long time, because its not really legal! And please dont reupload, so this thread could work for a long time.
And dont forget the Crown workers could read this thread to:)
 
Gaussian filters:

From http://www.filter-solutions.com/gaussian.html
The Gaussian Filter is the filter type that results in the most gradual pass band roll-off and the lowest group delay. The step response of the Gaussian filter NEVER overshoots the steady state value. As the name states, the Gaussian Filter is derived from the same basic equations used to derive the Gaussian Distribution. The significant characteristic of the Gaussian Filter is that the step response contains no overshoot at all.

Cheers
 
Its really not that complicated!

If we want to build a mono amp!
We dont need these:
-Common mode input (XLR) /U100,U101
-Compressor /U1B, U102B
-Thermal limit control /U4B,U4C,U3A,U3B
-sense of the output power (LED) /U2A,U2B
-clip detector /U1A
-distrorsion LED /U102C,U102A,U102D
-we dont need spleep circuit:) /U201C,U5C,U5D
-bridige /U101C
-stby /U5B
Mainb:
-crowbar /U2D, U7D,U7C
- and >10 OPamps


What we need:
3 OPamps for triangle generating (from 4Mhz precision)
1 OPamp input buffer
1 OP DC servo
4 OP (or less) for low pass filter <25Khz input
4 OPamp for error calibration
2 high speed comparator (MAX963) in 1 SO!, 4.5nS!, feeded with 0-5V signals, it also has SHDN pin which is simple to use.
fet drive: MAX5054 4-15V 4A! 20nS
Max ICs are for free samples:)

4 Opamps are in one packet so 4 SO/DIP can be used(4x4=16>15):).
You can put very big current rated MOSFETs in the output stage! So you can achive 4 KW++ power, with ~90% effeince:D Maybe this inspires someone :)

We have to just work out the feedback to get a rock stable amp :)

Here is the stable working triangle generator without feedback, open with Multisim 9.
 

Attachments

  • triangle gen!.zip
    80.2 KB · Views: 689
phase_accurate said:
A filter of that high an order is IMO not necessary at the input of a switching amp.

Why this?
Are you speaking in general? Or are you refering to the Crown BCA circuit? :confused:

AFAIK many switching amplifiers take their feedback before the output filters, easily visible when loaded with very low loads. It seems to be much more demanding to include the filters in the feedback loop due to the corrosponding phase shift at higher frequencies.
Perhaps this leads to the necessity of such a complex input-filter?

Or other way round: Why should the Crown engineers include a circuit of unneeded complexity into this amp?

I'm new to this subject (switching amps), just asking to understand the design objektives. :angel:
 
Gaussian Filter not a big trade of its just a precis low pass filter,and its needed.
Tha manual says that it couse some DC offset so the DC servo circuit bypass this.

I will try to reduce the 3 filters in the output to 1 LC 250Khz lowpass) But it will modify the phase of the feedback so it has to be corrected.

How can I compare the phase shiftes of the line in Multisim 9?
 
http://onlany23.atw.hu/Triangle.PNG
Here is the triangle generator, the original schematic uses 74AC02,74HC4024N,74AC74. So please prefer for those parts (my program only supported these parts)

There is only virtual crystal in the program so I couldnt
set to 4Mhz its running about 2,24...Mhz.
But its very sharp:)
I used a virtual clock source running @ 250Mhz.

The third amp is a simple integrator that changes the bipolar triangle into a 0-5V triangle.
Its needed because the comparator can only work in this range.
The supply (+15V -13V) is not important!
The results are the same when using +-10V for powering. You can find other OPamps.
This triangle generator is very simple, and its very (crystal) stable!

What about reinventing (just) :) the negativ feedback?
I think it could work with just voltage feedback, and without current feedback.
The current limit stays with current window comparator.

The FO (voltage) feedback is very simple, it isnt using any phase correction filters. Its just correcting the triangle wave (and the DC servo).

Please watch the Mainboard F2 cella and the D9 cella.
Any idea for a simplier current feedback?
 
I'm In!

I have a more simplified PWM generator but here I can't upload, maybe administrators will do something regarding the size of attachments

EDIT: Running desktop@ 800/600 will allow me to upload...

:bawling: :smash:

Also, Crown use 74AC74, Hi speed flip-flop,bt I have some other couple questions...

1 - Have U seen on output stage of the Negative output part ,the MC34151 mosfets driver,are drived by TPL2601 wich has INVERTED signal output regarding to the MC34151 mosfet driver on Positive output part, wich is drived bu HCPL 2611 opto wich output is not inverted?

I'm telling this because on U101 and U103 (LM361) ihe signal inputs of PWM and audio, isn't defazed...

The 180 grades defazage is made by the inverted output signal of TPL2601?
 

Attachments

  • 250 khz pwm generator.jpg
    250 khz pwm generator.jpg
    28.6 KB · Views: 1,927
U have to drive the Fets like this:
when the comparators are off-> All Fets MUST be powered off!

What about leaving the current feedback?
And just using the voltage feedback?
The voltage feedback doesnt have any phase shifts,
so its no need to correct anything.

And we can leave the: I monitor+T100+ phase corretction

And what about using this voltage feedback on the input audio line?
 
Simulate

I made a simplier working version of the amplifier:
How can I feedback the red square wave to
the green input sine wave?
(the input audio line is divided into 2 lines and one is inverted)

Is it possible to feedback after the filter stage?
I mesured the phase shift is -7degree.
pls help me! thx.

An externally hosted image should be here but it was not working when we last tested it.
 
Status
This old topic is closed. If you want to reopen this topic, contact a moderator using the "Report Post" button.