ea*r 660

[Ptownkid]

g-mail account

 

[SUPERMAGOO]

[email protected]  please!!!!

 

[hop.sing]

I did something like a clone of that comp, it was a LOT of work, though.
I have the schematics, but do not want to share them openly, the device is still in production, but PM your mail adress and I send them to you.
I used some Haufe input transformer I had lying around and a Lundahl 1660 output configured as 4,5:1.
It is not the perfect match, I guess, something with a higher ratio would better, but with the correct termination it works reasonably well.
There is a transformer after the sidechain circuit, where I used a Pikatron 1:2 tranny I got cheap of Ebay and it works great, I was probably lucky. I modified the timing circuit to be much more flexible.
And because I built it stereo and ran out of real estate in the enclosure and did not want to stress the power supply to much I used just two pcc189 in parallel, cascoded with two ecc88 per channel. I tried four pcc189 without the cascode, that presents a lower load on the output transformer and is theoretically probably a better circuit choice (with that transformer at least), but I like the sound of the cascoded thing better, the metering is also more in line with reality.
With proper output termination, the frequency response is +-0.5 dB from 20 to 20.000 Hz.

 

[KrIVIUM2323]

you say so?

Yes something like that but you must double the number of sections for each tube (8 sections for pcc189 and 4 for pcc88).

The grid resistor is per grid not one for the whole tubes. There is an other resistor between each anode of pcc189 and each cathode of pcc88 too.

And last but not least the bias section of pcc189 is from the 660 not the 670. There is one small detail which make a BIG difference between the 2. Something like a source of negative voltage between the Rk and pot ( they're not directly connected to ground as inbalance needs to be adressed).

 

[pucho812]

g-mail account it is. use at your own risk 8)

 

[hop.sing]

And last but not least the bias section of pcc189 is from the 660 not the 670. There is one small detail which make a BIG difference between the 2. Something like a source of negative voltage between the Rk and pot ( they're not directly connected to ground as inbalance needs to be adressed). Wink

I think the major difference is that the grid of the 670 biased with a negative voltage of a few volts and the grid of the 660 is biased straight to ground.
But there is more to it...
An interesting discussion of the subject is here:
http://www.groupdiy.com/index.php?topic=31788.100
But anyway, I heard a difference in the compression characteristics between that rather complicated biasing arrangement in the EAR (Fairchild) 660 and the simple way with a trimpot and a resistor from the cathodes to ground. I tried both ways in my compressor. The 660 way sounds somewhat smoother to my ears.

 

[SUPERMAGOO]

I did something like a clone of that comp, it was a LOT of work, though.
I have the schematics, but do not want to share them openly, the device is still in production, but PM your mail adress and I send them to you.
I used some Haufe input transformer I had lying around and a Lundahl 1660 output configured as 4,5:1.
It is not the perfect match, I guess, something with a higher ratio would better, but with the correct termination it works reasonably well.
There is a transformer after the sidechain circuit, where I used a Pikatron 1:2 tranny I got cheap of Ebay and it works great, I was probably lucky. I modified the timing circuit to be much more flexible.
And because I built it stereo and ran out of real estate in the enclosure and did not want to stress the power supply to much I used just two pcc189 in parallel, cascoded with two ecc88 per channel. I tried four pcc189 without the cascode, that presents a lower load on the output transformer and is theoretically probably a better circuit choice (with that transformer at least), but I like the sound of the cascoded thing better, the metering is also more in line with reality.
With proper output termination, the frequency response is +-0.5 dB from 20 to 20.000 Hz.

you experience is very interesting.

 

[SUPERMAGOO]

you say so?

Yes something like that but you must double the number of sections for each tube (8 sections for pcc189 and 4 for pcc88).

The grid resistor is per grid not one for the whole tubes. There is an other resistor between each anode of pcc189 and each cathode of pcc88 too.

And last but not least the bias section of pcc189 is from the 660 not the 670. There is one small detail which make a BIG difference between the 2. Something like a source of negative voltage between the Rk and pot ( they're not directly connected to ground as inbalance needs to be adressed).

thanks for the schematic.
I and observed the different connections of the cathodes. do not understand why you should work so different between the 660 and 670.
I need some time looking at the schematics ...

 

[SUPERMAGOO]

And last but not least the bias section of pcc189 is from the 660 not the 670. There is one small detail which make a BIG difference between the 2. Something like a source of negative voltage between the Rk and pot ( they're not directly connected to ground as inbalance needs to be adressed). Wink

I think the major difference is that the grid of the 670 biased with a negative voltage of a few volts and the grid of the 660 is biased straight to ground.
But there is more to it...
An interesting discussion of the subject is here:
http://www.groupdiy.com/index.php?topic=31788.100
But anyway, I heard a difference in the compression characteristics between that rather complicated biasing arrangement in the EAR (Fairchild) 660 and the simple way with a trimpot and a resistor from the cathodes to ground. I tried both ways in my compressor. The 660 way sounds somewhat smoother to my ears.

thanks for the link.
rotheu to delete your schematics!
is a pity 

 

[hop.sing]

Rotheus schematic was cool for his all tube sidechain, the gain reduction section was pretty standard Imo, the 660 is more interesting in that regard. It would be nice to hear if it really makes a difference, if the sidechain is like the solid state EAR one or an all tube one.
With the 1:2 tranny after the EAR SC I get enough negative voltage on the grids of the pcc189s for 25 dBs of gain reduction even though the SC amp is running only on +-18 Volts. Output impedance of SC amp is reasonably low, too. With a 0,2 Mf timing cap and no series resistor (like the original) I get 0,2 mS attack, which is immediate for all purposes.
If you use pcc189s be prepared to try out quite a few, they are all over the map, it is not easy to find a matched set. Even though some manufacturers do better in that regard (and it is NOT telefunken in that case).

 

[SUPERMAGOO]

Rotheus schematic was cool for his all tube sidechain, the gain reduction section was pretty standard Imo, the 660 is more interesting in that regard. It would be nice to hear if it really makes a difference, if the sidechain is like the solid state EAR one or an all tube one.
With the 1:2 tranny after the EAR SC I get enough negative voltage on the grids of the pcc189s for 25 dBs of gain reduction even though the SC amp is running only on +-18 Volts. Output impedance of SC amp is reasonably low, too. With a 0,2 Mf timing cap and no series resistor (like the original) I get 0,2 mS attack, which is immediate for all purposes.
If you use pcc189s be prepared to try out quite a few, they are all over the map, it is not easy to find a matched set. Even though some manufacturers do better in that regard (and it is NOT telefunken in that case).

excuse me I have trouble understanding. I have trouble with English.
but you try to say that the solid state sidechain could be better than the tube?
to be low impedance is more comfortable to handle the time constant?.

the selection of tubes is not a problem.
I build guitar amps and I have a curve tracer.

 

[SUPERMAGOO]

and I recommend on page 90 of book morgan jones "valve amplifiers. "
there is a very good explanation about cascode connection

 

[hop.sing]

excuse me I have trouble understanding. I have trouble with English.
but you try to say that the solid state sidechain could be better than the tube?
to be low impedance is more comfortable to handle the time constant?.

Well, with a transformer to scale up the voltages, it seems to hold up pretty well. Maybe the tube sidechain has some benefits like a certain kind of saturation or the dc threshold gives a different sound in use, I don't know. But I know that the solid state design of the EAR works very well and saves a lot of powersupply resources.
Since the time constants of the fairchild seem to be very similar to those in my EAR copy I think the impedances are kind of close.

 

[SUPERMAGOO]

off topic:

I wanted to show an interview with john mayer. He says using the ear660 in their voices on the last album. But I can not find the link. someone he has seen?