Hello everyone. I come here to seek help in the design of a compressor with a focus on mastering. I have two 6DJ8 valves and one of the requirements of the project is to use one or both of them.
Before I start, I want to make it clear that one of my goals is to minimize costs, even if it means worse sound quality - I'm doing it because it's a grad project and also for learning, but I don't have a decent budget at the moment, and the logistics and component production here is very costly and slow.
I developed a basic prototype in LTSpice to simulate with .wav files, and managed to compress drum sounds pretty effectively, at least in theory. However, I have no knowledge of the real, practical aspects behind it, and therefore I would like to receive help from the forum folk.
Some important points I would love to have contributions:
Essentially: what works in this circuit? What is absolutely garbage? EVERY piece of information is extremely valuable. Feel free to contribute in any way. Thanks!
Before I start, I want to make it clear that one of my goals is to minimize costs, even if it means worse sound quality - I'm doing it because it's a grad project and also for learning, but I don't have a decent budget at the moment, and the logistics and component production here is very costly and slow.
I developed a basic prototype in LTSpice to simulate with .wav files, and managed to compress drum sounds pretty effectively, at least in theory. However, I have no knowledge of the real, practical aspects behind it, and therefore I would like to receive help from the forum folk.
Some important points I would love to have contributions:
- I want to avoid using transformers as much as possible (I know you can disagree with that, but tell me if it's possible to do it without transformers, or at least convince me why is it so critical/necessary).
- The LED-LDR (in green) group represents a makeshift Vactrol I made with heat shrink, where the LED excites the LDR directly. I measured the set and developed its equation, which is under the LDR in the schematic.
- I never saw an amp op buffer directly into the a compressor's input, therefore i feel in my guts this looks absurd in some way. Shed a light on this, would you?
- The 6DJ8 uses 130V max plate voltage. As I live in a place where the electrical network is 127Vrms, I intend to rectify the signal directly to supply the valves. I heard about humming. Can I do it?
- To supply the NE5532, I would like suggestions: switching source, voltage regulator... What is the most suitable solution?
- The output represents an old radio output transformer (2400:8) and my laptop as load. How necessary is the transformer? Can I get rid of it? What are the implications of directly connecting, for example, my phone in the input and laptop in output?
- I know mastering should be done in stereo. Can I just replicate this circuit for the Left/Right channels? Is there any (transformerless) other way?
Essentially: what works in this circuit? What is absolutely garbage? EVERY piece of information is extremely valuable. Feel free to contribute in any way. Thanks!
I mean, if I were to put a 1:1 transformer for the tube supply (130V) and then rectify it to DC, what would be the difference, concerning safety and over-voltage?
130V is RMS. When you rectify, you'll be getting a little over 180V DC.
The fact that you're asking these questions concerns me.
Chris
The fact that you're asking these questions concerns me.
Chris
I am aware of that, which is why I came here to ask for suggestions. My initial idea was to make a buck or flyback converter to supply the valve, but it seems like a lot of work.
So, if possible, I am in search or ideas for a reasonable alternative approach.
You can certainly do this, and you can use an extra series resistor and bypass cap to lower the DC voltage to suit your valve. Just use Ohm's law, along with your expected anode current(s), to calculate the resistance needed to drop from roughly 170 volts down to 130 volts, or whatever your intended final B+ voltage is.
By the way, the 6DJ8 / ECC88 can handle up to 500 volts DC (not 130 V) according to the datasheet (see attached image). But there's a maximum anode dissipation limit of 1.8 watts, the reason for the recommended 130 V anode voltage is probably to keep dissipation to no more than 1.8 watts (i.e. keep cathode current below 14 mA with Vgk=130V.
The way you modelled the LDR + LED combination in LTSpice is clever. But it is also inaccurate for your purposes (compressor design), because doesn't take into account the slow (time) response of the LDR. 🙁
In reality, an LDR takes tens of milliseconds to lower its resistance in response to an increase in LED current, and can take hundreds of milliseconds, even thousands of milliseconds, to increase its resistance back to normal after the LED current has been reduced to zero. It is a very slow device, and this slowness is part of the reason why it works so well in audio compressors - the slow response of the LDR is part of the compressor attack and release time constants.
One other thing to consider: can dangerously high-voltage signal peaks (a 100-volt or more transient) get through C5 and damage U3? It might not be a bad idea to use some sort of voltage clamp between C5 and R9 to prevent op-amp damage. Low-voltage semiconductors living in the same circuit as high-voltage valves are a lot like rabbits living in the middle of a pack of hungry wolves; they are in a great deal of danger, and unless protected, will not live long. 🙂
Good luck with the project!
-Gnobuddy
Attachments
You're very welcome! I hope you'll show us your final circuit after you get it working!
-Gnobuddy
-Gnobuddy
Atack/release times needed can vary orders of magnitude, it depends what your using it for. Thats why many compressors let you adjust these. Since the LDR has its own time constants you wont need the timing cap/resistors, unless you want to slow things down more. There are some very good LDR (opto) controlled comercial compressors out there so this is a way to go. But,Since your using tubes have you looked into variable mu from one of the tubes as a gain controll?