Bottlehead Forum
Bottlehead Kits => Crack => Topic started by: Ritchie on October 11, 2013, 01:27:25 AM
-
I am using my crack as a preamp, therefore I'm able to lower the output cap size depending on my amps input impedance.
If lowering the output cap should I also be increasing the output resistor?
Thanks,
Ritchie
-
Ritchie,
You have to consider the parallel combination of the amplifier's input impedance and the output resistor. Calculate the parallel equivalent and insert into the equation:
Fo=1/2*(3.141)*R*C (that is 1 divided by the product of the string on the bottom of the fraction.)
where:
Fo is the 3 dB down point
R is the calculated value
C is the intended coupling capacitance
Here are the rules:
Fo is what everybody calculates to, but it is flat a decade above that. I.E. if you calculate Fo of 3 Hz it is flat at 30 Hz.
You can swap things around, swap C and Fo to solve for the capacitor value if you have decided on the the targeted flat frequency.
Resistor and capacitor values are nominal, you need to measure them to know what is really going to happen.
I hope this helps.
-
I've been planning to do this on my Crack too by tapping a 3.3-4.7uf Clarity cap before the electrolytic cap and add RCA connectors. I'll leave the phone jack alone.
If you remove the electrolytic cap and phone jack, I'd put a 100k-500k resistors at the RCA jacks.
-
Thanks guys, my amps input impedance is 1.8k, so with the 2.5k resistor in the crack I'm looking at approx. 1k. I usually like to calculate my -3db point to 2hz so with the resultant 1k I would be looking at approx. 80uf. When going with higher quality film caps it can get very expensive.
I would like to use a much higher output resistor to keep my impedance closer to the amplifiers 1.8k so I can get away with 40-50uf.
Will going with a high value resistor on the output have any effect on the preceding circuit.
By the way the crack really surprised me as a preamp, sounds great especially driving a difficult load and long interconnects.
Thanks,
Ritchie
-
That's a difficult load for a tube preamp. You can disregard my previous post. :-[
-
Yes it is a biamped set up which brings the impedance down, the crack circuit has no problem driving this load with the 6as7 cathode follower and the right size cap.
-
What amp are you using?
-
The amps are Linn Majik 4100 which are 4 channel amps with built in active crossovers. So with four channels in parallel the input impedance ends up at 1.8k.
I know a solid state pre may be a better match but I much prefer the sound of tube pre amps. I'm using the crack now and not noticing any hint of bass roll off.
-
You could probably do 50uF (47uF) and a 3.3K resistor across the output.
So you are using 8 channels of amplification? Or is each pair of channels driven by one half of the Crack?
If the Crack is only driving one 4100, then the amp will see about a 3.7K load. In this case a 15uF cap would work, and you could go with a 10K bleeder on the jack.
The suggestion to add dedicated RCA output jacks is a good one. At some point, someone will plug their headphones into your Crack and draw some unfortunate conclusions.
-
I am using two 4100's one per speaker as I have 4 way speakers. Each channel actively drives each driver. So the crack would be looking into approx. 1.8K per channel.
I will be using it as a dedicated preamp without the headphone jack. I was hoping to use 44uf output caps as I have four Mundorf supreme 22uf caps on hand. I will eventually just try them but wanted to get close with the math first and figure out the allowable output rrsistor.
-
It can be tested easily with some jumpers. Go for it! Let us know what you hear. Then it will sound better when you solder them in.
-
Ok I will give them a try, but can someone recommend the output resistor value when using 44uf output caps driving amplifiers with a 1.8k input impedance?
Thanks
-
1M ohms will give flat response at 20 Hz. The resistor is there to eliminate the DC that bleeds through the output capacitor at startup. It would drive your amps wild if they have a good low frequency response (like to DC) but most amps just thump. The answer is to turn on the amps 30 seconds after the Crack.
-
Thanks Grainger.
-
Cap Calculator (http://www.v-cap.com/coupling-capacitor-calculator.php)
That will help with the frequency response calculation.
A 1M Ohm Resistor and a 44uF cap will give you a crap ton of startup voltage, but very good frequency response. As you go down in resistance at the output, you will reduce this startup transient. If you go too low, it will begin to combine with the input impedance of the Linn amp and start messing with the frequency response. I wouldn't go any higher than 5K with a 44uF cap.
-
Thanks guys, this was educational for me, I have always built pre amps with small output caps as I was using tube power amps with high input impedance, so I never paid much attention to the output resistor as it was usually 1M.
I now realize the difference with a very large output cap.
I have a soft start timer circuit which I can use to have the preamp output shorted for 30 seconds on start up.
Would this be an option to allow me to use a much higher resistor such as 1M?
Thanks
Ritchie
-
When calculating with a 5K resistor my -3db ends up around 2.7 hz, that should be fine?
-
Ritchie,
I couldn't remember what it was that bothered me about this thread. I have it now.
It is a good design parameter for first order 3dB down point of 0.2 Hz, flat at 20 Hz. But... You and I live in the real world. Neither my speakers nor my room could produce a flat 20 Hz. It is extraordinary for a speaker to be FLAT at 20 Hz. It would take a distance from your woofer to the "back" wall of 56.266 feet for a full wave to emit from the woofer before reflecting. This doesn't even go into the lack of 20 Hz notes in 99.99% of recorded music.
What I am saying is that the reason you will hear no ill effects from a smaller capacitor. Give one 22uF a try, then parallel in the other. You are likely not to hear a bass improvement with two in parallel. However, you might get some smearing of the image with two in parallel. (I did when doing a cap comparison in my Eros)
-
I will definitely give that a try as I would much prefer just the single cap also.
I have found paralleling caps to create a bit of smearing but usually with different values as they have different time constants.
One is better than two or more, but I have always been under the assumption to only parallel with equal value caps.
I agree, we tend to over design to ease the audio nervosa.
Thanks Grainger,
Ritchie
-
G,
I know that I am side stepping a bit here, but regardless of that, I am curious about something. Maybe I don't recall correctly, but wasn't there a time when Multi-cap made a capacitor that had an internal bypass capacitor? Maybe I just dreamed it.
-
I believe they still do. Haven't tried one in a long time though, maybe they have some manufacturing voodoo!
-
Greg,
I agree. Yet, still, I didn't like the parallel 0.22uF KK Teflon caps as much as a single 0.22uF. The design value was 0.47 for that flat 20 Hz. Still, it just didn't sound as coherent.