Old-fashioned amplifiers with a new twist (2)

Jump to Latest
#22
Elvee said:
R29 C16 should not be necessary, but you can include them on the PCB and leave them empty.
Click to expand...
In a perfect world, every maintenance man stagehand and security guard does not carry a 5 W radio. Every car driving by doesn't have a 1000 W CB radio. BUT in my world such RF emitters exist. R29 & C16 are an external RF trap. Also should be included and not shown, 11 turns fat wire around a 1 cm coil form between speaker jack and amp output, parallel a 10 ohm 3 watt resistor. These are left out of most lab designs on here, and included in most successful commercial amps. Keep these components away from the high impedance feedback path and the input, IMHO. In my AX6/(ST120 djoffemod) build, the zobel r29 & c16 are outside the 11 turn inductor, mounted on the back of the dual banana jack for speakers in fact.
C3 the 150 pf parallel the input may be a bit heavy handed. I had to go up from 0 to 68 pf parallel the input jack on my AX6 to get rid of an AM radio station. Sports talk radio, particularly vile.
These RF traps are only fully effective in a steel or steel mesh case.
 
Last edited:
#23
OLF2 majorR2

Dear LV,
Made all modifications as suggested. Kept all options available in the scheme. If any thing is to be included further, please suggest.
.dip file is included with extra .asc extension. just remove the asc extension to open the file. Hope it gets uploaded.
--gannaji.
 

Attachments

  • OLF2 majorR2.png
    OLF2 majorR2.png
    241.9 KB · Views: 341
  • OLF2 majorR2.dch.asc
    OLF2 majorR2.dch.asc
    277 KB · Views: 91
#25
indianajo said:
R29 & C16 are an external RF trap. Also should be included and not shown, 11 turns fat wire around a 1 cm coil form between speaker jack and amp output, parallel a 10 ohm 3 watt resistor.
Click to expand...
R29 and C16's primary role would be as a Zobel, not a RF trap: their purpose is to keep the output terminated by a low impedance at high frequencies, to prevent the loop gain of the amplifier from wandering into hazardous regions.
As a side effect, it will also attenuate and dissipate some of the HF ingress, but that is not the reason why it is included in 99% of the amplifiers.
(If you need it, I can show you an example of a real RF filter)

Here, this network is split between C15/R28 and C14/R27: together they form the equivalent of a 44nF/4.1ohm Zobel network, but they also stabilize the CFP OPs.

An inductive Zobel can be useful, for capacitive loads (relatively rare) or relatively long cables behaving capacitively at some frequencies.

gannaji said:
. If any thing is to be included further, please suggest.
Click to expand...
R18 still needs to be changed to 100K
 
#28
Elvee said:
R29 and C16's primary role would be as a Zobel, not a RF trap: their purpose is to keep the output terminated by a low impedance at high frequencies, to prevent the loop gain of the amplifier from wandering into hazardous regions.
As a side effect, it will also attenuate and dissipate some of the HF ingress, but that is not the reason why it is included in 99% of the amplifiers.
(If you need it, I can show you an example of a real RF filter)
Here, this network is split between C15/R28 and C14/R27: together they form the equivalent of a 44nF/4.1ohm Zobel network, but they also stabilize the CFP OPs.
An inductive Zobel can be useful, for capacitive loads (relatively rare) or relatively long cables behaving capacitively at some frequencies.
Click to expand...
I took it as given you had managed the high freq loop gain of the base circuit already. When dynaco post ST120 shipment suggested 1000 ohms series .047 uf on the back of the speaker jack OUTSIDE the 11 turn inductor , it pretty obviously had something to do with the outside world. Like more portable FM transmitters in 1970 than in 1966. And warrenty calls. Other thousands of "zobels" may have to do with internal stability as you say.
Scored a full parts kit including exotic leaded BAT48 & HTE5L100 (schottky 5A 100 v). 44 v 2 A transformer for 50 v rail. Rains of October due in 2 weeks. Build of OldFashMax to follow.
 
#35
Dear elvee,
Please find the R3 files. The png, Diptrace and Eagle format files with .asc added to enable uploading. Please remove the ase extension and the files can be opened in Diptrace and Eagle.
Any PCB gurus to help please ?
--gannaji
 

Attachments

#38
This is the final PCB shape I would prefer depending on the Heat-sinks I have.

The height not more than 3.8 inches to fit in 4in height heatsinks and box. Distance between power BJT mounting holes to be 4 in . The actual pcb width can be more or less as suits the design, but not more than 8 inches. Thanks for any help.
 

Attachments

  • sample.png
    sample.png
    15.7 KB · Views: 318
#39
Dear all,
I do not know why, there is some problem in the Diptrace file posted. I made many changes, compared to the png file posted, and reposted as R4. Please remove the asc extension and it can be opened in Diptrace.
The eagle file posted is exported from diptrace file.
The png file, R3 is unchanged.
--gannaji
 

Attachments

#40
gannaji said:
Dear Elvee,
I started in my own way to start the lay out in Diptrace. the resultant pic is as below.
The Eagle file is only the schematic.
--gannaji.
Click to expand...
Difficult to tell without seeing the finished result, but it doesn't look promising.

It looks like the raw output of an autoplacer

gannaji said:
This is the final PCB shape I would prefer depending on the Heat-sinks I have.
Click to expand...
It could be more promising, but more is needed to be able to judge
 
You must log in or register to reply here.
Top Bottom