Another 300B Push Pull (not WAD)
#286
No, I've always been straight out of the DAC chip. I did buy a better op-amp back when we were wondering if the DAC could drive the Miller cap of a valve, but that turned out to be a red herring, so the op-amp never got used. I had the AK4393 at first, ran that taking the signal straight off the DAC pins. Then got a AK4396 board, ran that the same way for a while, then installed 1K series resistors, taking the signal off after them.
#287
Talking of Miller capacitance reminds me, I realised something today.
In my application, Miller cap is about 80pF per triode. I'm in push-pull so two 80pF caps in series, so 40pF total across the DAC outputs. Other applications may have a bit more, add some long leads, lets say 300pF across the DAC output.
The Ebay board actually uses 3n3 across the output as part of the LPF - 3300pF, way more than the loading from the Miller cap and cables. So unless I'm missing something, the AK4396 can drive any likely tube output stage very happily - just remember the 1k series resistors.
I know Nick had already confirmed that this wasn't an issue when he was checking out Ali's DAC, but I thought it worth mentioning.
In my application, Miller cap is about 80pF per triode. I'm in push-pull so two 80pF caps in series, so 40pF total across the DAC outputs. Other applications may have a bit more, add some long leads, lets say 300pF across the DAC output.
The Ebay board actually uses 3n3 across the output as part of the LPF - 3300pF, way more than the loading from the Miller cap and cables. So unless I'm missing something, the AK4396 can drive any likely tube output stage very happily - just remember the 1k series resistors.
I know Nick had already confirmed that this wasn't an issue when he was checking out Ali's DAC, but I thought it worth mentioning.
#288
Yep, thats how I read the spec, if the load is capacitive, as long as the resistor is used, any value of cap can be driven.
Whenever an honest man discovers that he's mistaken, he will either cease to be mistaken or he will cease to be honest.
#290
I discovered that I was getting some electromagnetic hum pickup from the mains transformer into the inter-stage transformers. For the most part, the balanced circuit is excellent at rejecting hum and noise. But in this case, because the noise is being generated within the core of the inter-stage, its producing differential mode noise which is appearing at the output. Its barely audible even with my ear right next to the speaker, but it was bugging me. Somehow I suspect that this hum might be modulating with the signal to produce audible distortion. I tried changing the orientation of the inter-stages, but that wasn't very effective. I decided that the only solution was to move the mains transformer and the inter-stages as far apart as possible.
I wanted to retain the existing chassis, but this has forced some compromises. The ideal solution would be to keep the power supply on the far right of the chassis, and move the inter-stages to the far left. But the position of the volume control is fixed, and there isn't enough room between the volume control and the chassis side panel to mount the inter-stages.
There is just enough room for the mains transformer and first choke to squeeze in between the volume control and left panel, and the inter-stages fit comfortably on the far right. So that's how things will have to be. The downsides are:
The mains transformer and choke are now right up against the volume control - so I just have to hope that the CMR of the circuit rejects the emf-induced hum in the line-level wiring.
The ventilation slots in the lid are all on the right hand side, so I have elected to keep the EZ90 over on the right. Obviously this means long runs of AC heater supply and the HT secondary leadouts across the chassis to the rectifier, then back across to the first choke, then back across to the first cap, second choke etc.
I'm wiring this up but it just feels wrong. The way that everything was laid out before it felt right because there was a logical flow of the wiring. I just have to tell myself that there are good engineering reasons why I am laying things out this way.
Another option would be to make a new chassis. But I'll only entertain that if I can get things sounding really really good first, so I know its worth the expense.
I guess a third option would be just to make a new base panel and lid, incorporating ventilation for the rectifier over on the left. This wouldn't be too expensive, and would allow me to incorporate 1" radius corner pieces in walnut to match the output stage mono-blocs.
I wanted to retain the existing chassis, but this has forced some compromises. The ideal solution would be to keep the power supply on the far right of the chassis, and move the inter-stages to the far left. But the position of the volume control is fixed, and there isn't enough room between the volume control and the chassis side panel to mount the inter-stages.
There is just enough room for the mains transformer and first choke to squeeze in between the volume control and left panel, and the inter-stages fit comfortably on the far right. So that's how things will have to be. The downsides are:
The mains transformer and choke are now right up against the volume control - so I just have to hope that the CMR of the circuit rejects the emf-induced hum in the line-level wiring.
The ventilation slots in the lid are all on the right hand side, so I have elected to keep the EZ90 over on the right. Obviously this means long runs of AC heater supply and the HT secondary leadouts across the chassis to the rectifier, then back across to the first choke, then back across to the first cap, second choke etc.
I'm wiring this up but it just feels wrong. The way that everything was laid out before it felt right because there was a logical flow of the wiring. I just have to tell myself that there are good engineering reasons why I am laying things out this way.
Another option would be to make a new chassis. But I'll only entertain that if I can get things sounding really really good first, so I know its worth the expense.
I guess a third option would be just to make a new base panel and lid, incorporating ventilation for the rectifier over on the left. This wouldn't be too expensive, and would allow me to incorporate 1" radius corner pieces in walnut to match the output stage mono-blocs.