Fairchild 661 repair

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

This site may earn a commission from merchant affiliate links, including eBay, Amazon, and others.

Noth

Active member
Joined
Nov 19, 2021
Messages
36
Location
Fr
Hi,

I have a Fairchild 661 gate which I intend to convert into a 663 compressor with buffered I/O, balanced inputs, etc.
But first things first: I want to make this 661 work "as it".

(This is linked to the thread: Fairchild 661tl Gate... convert to Compressor? but I prefer to create a dedicated thread for this debug issue)

Here is the original schematic:

Fairchild 661 gate.png

All the capacitors were originals (and very dead) and the light bulb too.
Thus, I replaced all of them and the light bulb with a LED.

I did also convert the board to a 8V DC supply by removing the diode rectifier, C5, C6, C7, R9 and R10.

So here is the current schematic:

Fairchild 661 gate - DC.png

Now it appears that the LED is always "on", whatever the input signal.

Since all capacitors were replaced, I suspect a dead transistor.

My input test signal is (I turn it on and off to eventually "see" the gate effect, and thus the LED turn off):
DS1Z_QuickPrint7.png

At the Q1 base, with threshold set to minimum I have:
DS1Z_QuickPrint9.png
-> The amplitude of the input signal looks very small to me to be usable afterwards, ok we're just after the transformer but...

At the Q2 base, with threshold set to minimum I have:
DS1Z_QuickPrint10.png
-> Without any surprise, the test signal is now almost lost, thus if I turn it off I have almost the same output

At the Q3 base, with threshold set to minimum I have:
DS1Z_QuickPrint11.png

At the Q5 base, with threshold set to minimum I have:
DS1Z_QuickPrint12.png


And the same for Q6, thus the LED is always on, whatever the input signal is.

Does someone has any idea of what might be wrong ?


Thanks a lot for any advice!

Edit: updated the modified schematic (with DC supply), VCC and GND were inverted
 
Last edited:
Probably with the bulb, they ran a little current through it at all times to keep the filament hot. The led just goes on at very low current.
 
Hi, Thanks for your reply!

I see your point but I monitored the tension across the LED and it is fixed whatever the input is, actually just like all transistors.
I was expecting the input signal to be amplified by Q1/Q2 but it dosn't seems to be the case!
 
So, I ended up unsoldering all the transistors and testing them:

Q1 : hfe 120
Q2 : hfe 147
Q3 : hfe 87
Q4 : hfe 61
Q5 : hfe 84
Q6 : hfe 33

So, everything looks fine!
Once everything back in place, I tested again (some solderings were not super clean) and.. the LED went super bright and burned. So obviously, I got more gain! But the circuit stills has the same behaviour..
 
Ok, so I unsoldered my dead LED and decided to measure the current (I put a 1K resistor in place).
With no signal at the input: 3.3 mA
With a test signal : 7mA

So, radardoug is perfectly right, I cannot replace a bulb with an LED like this since this circuit actually delivers a fixed voltage and a variable current.

Now I'll have to source a lightbulb...!

But to me it doesn't really explain why I have a constant fixed voltage on the base of every transistor. To me, at the end of the day, the tension at the base of Q6 is supposed to control the current flowing to the bulb (and D3 adds some magic for the Release Time, I still don't really understand how). So how is it that, with no measurable voltage change on Q6 base we have a small current change through the bulb?
 
Last edited:
Looking at the circuit again, I see a major problem in your circuit. You cant just replace the bulb with a led. You need to add a current limit resistor on the led. Read up about leds! Also, as a bulb lights up, the filament resistance rises. This circuit may be relying on that. By putting the led in with no current limiting resistor, the base voltage on Q6 will hardly change.
 
So, radardoug is perfectly right, I cannot replace a bulb with an LED like this since this circuit actually delivers a fixed voltage and a variable current.

Now I'll have to source a lightbulb...!
Or also a parallel resistor to draw some current...
 
Thanks for your replies! I see, I'm not super familiar with how light bulbs behave and clearly just dropping an LED with a resistor in series won't do the trick.
I will try to find a replacement bulb!
 
Back
Top