Federal AM864 release modification

GroupDIY Audio Forum

Help Support GroupDIY Audio Forum:

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

ibvee

Well-known member
Joined
Mar 2, 2010
Messages
45
I did quite some searching on the forum, but couldn't really find a clear answer, so here it is.

Somewhere on the web I came by this website: http://akavalve.blogspot.com/2009/07/federal-tube-limiter-release-time.html
In order to shorten the release time, this guy swaps C1 (1uF, the release cap) with a 0.1uF. A shorter release time is something we'd welcome on our Federal. However, in order to be sure that this is the way to go, I decided to search here for some extra information. Somebody changed out R1 (2M) with a 1M resistor, but besides that I didn't really found anything. Surely this must be a common mod among the users of this thing?

So my questions, what would be the best way to go? Could changing the capacitor result in any bad side effects?

Any help is appreciated!

Regards Mic
 
Hi ibvee,

The formula for calculating the time constant to 63% (usual level) is uF x M ohms = time in seconds, hence the 1uF x 2M = 2S.  If you change it to 0.1uF you will obviously get 0.2S, if you also change the 2M to 1M you'll get 0.1S etc.  The only snag would be increasing the resistor above what the 2 tubes are allowed in their maximum grid leak spec, there is no problem in reducing the resistor value.

The same formula applies to the attack calculation thus: 1uF x 0.047M = 47mS.  But you can't go reducing that regardless because the output resistance of the drive circuit becomes the ultimate limit of the attack time constant.  The 6SN7's have only slightly more poke than 12AU7's so they will never be as fast as the 6V6 powered limiters that have 10~15W available to charge the capacitor.

The Fairchild has an output resistance of only 100 ohms so the attack time is 2uF x 0.0001 = 0.2mS.

I hope that helps!

best
DaveP
 
Thank you very much for the explanation, very clear!

So R11 (47k) together with C1, sets the attack time? This would mean that lowering C1, to 0.1uF would also change the attack time, whereas changing R1 to say 1M, would only change the release, right? And when decreasing the value of C1, it would shorten the attack time, without loading the 6SN7?
 
Yes ibvee,

Think of the cap as a bucket to fill, 0.1uF would take 1/10th the time, so to return it back to the original spec you would have to increase R11 to 470k, in effect narrowing the pipe!  Leaving it at 47k would give you a theoretical attack time of 4.7mS.

Theory is fine but it's best to check it on the bench in case some other unforseen effect takes place.  Try to put yourself in the makers position, why did they use a more expensive 1uF cap when a 0.1 would have been cheaper, it could be that working for the US army meant that cost was not an issue or there could be another issue.  It could be that they had bought a job lot of 1uF for other devices and used them because they were to hand, you won't really know until you try the various values for yourself.

best
DaveP
 
I don't believe the 6sn7 has anything to do with the sidechain, it's just a makeup/output amp. the 6sq7 is doin all the work here.

AM-864.JPG
 
DOn't worry about the attack time.  Play around with resistor and cap values on a 3 pole switch that changes the TC with caps.  use resistor and pot combo while the switch takes care of the caps.  Then you be able to have quite a few possibilities for release.  The pot will vary subtle values, while the caps will take you through more extreme CT values.
 
Yeah my mistake, Gemini86 is right, its the 6SQ7 that is the limiting factor not the 6SN7's, thats what comes of doing posts in the small hours.

best
DaveP
 
Thanks everybody for the help! Really appreciated, ordered some parts for some experiments. We're thinking about making the release variable, with a 2M potentiometer in series with a 1M fixed resistor, in the place of the 2M resistor. When lowering C1 to 0.1uF, this would result in a variable release from 300ms to 100ms right?
Only problem is, all resistors in the circuit are rated 1/2W, and from the rating on the current C1 in the circuit I expect something around 150V there, and I can't seem to find any decent potentiometer of 2M with these ratings or higher. Any ideas?

Regards, Mic
 
Back
Top