Hi Jason,
With the LED in place I have to raise the pot to 3K2 to get 16mA on the drivers and 120mA on the outputs. Would it be better to change the 2k2?
Thanks, Terry
I don't think so. If the values get too high then you start getting base current causing errors. The lower leg on the board is a fixed resistor and a potentiometer, so you can choose something like a 500R potentiometer and change the value of the fixed resistor to suit the build.
That resistor value will be chosen to produce about 2.4V across the VBE multiplier in the real world, the simulator will have you setting the bias very low due to how easily the Cordell models turn on.
That resistor value will be chosen to produce about 2.4V across the VBE multiplier in the real world, the simulator will have you setting the bias very low due to how easily the Cordell models turn on.
A quick sim suggests that you might get into the required range more easily with a 1N4148 diode rather than an LED. The voltage drop over the LED is too great to get the transistor to operate properly.
2K2 on top, 500R potentiometer+2K4 on bottom and a 1N4148 diode in place of the LED should be pretty close in real life but won't likely work quite right in simulation.
2K2 on top, 500R potentiometer+2K4 on bottom and a 1N4148 diode in place of the LED should be pretty close in real life but won't likely work quite right in simulation.
Last edited:
Not sure I follow. Which resistor are you talking about? You mentioned a 500R pot. I have to lower R131 to 400R to get the pot down to 500R. I'm just looking for a starting point so I don't burn something up.
Thanks, Terry
I did the layout with both a pot and fixed resistors, so I'm sorry I was referring to a physical build. In the simulation VR132 has a value that is the combination of of the potentiometer and the fixed resistor. I want to re-iterate the sim will give you values that likely won't work in real life.
Always start with the lower leg of the VBE multiplier at its highest resistance and reduce it until the target bias is achieved. The reason for putting a fixed resistor in combination with the pot is to prevent you from being able to basically shut the VBE multiplier off and allowing the bias to go totally out of control.
Always start with the lower leg of the VBE multiplier at its highest resistance and reduce it until the target bias is achieved. The reason for putting a fixed resistor in combination with the pot is to prevent you from being able to basically shut the VBE multiplier off and allowing the bias to go totally out of control.
Attachments
Hi Jason,
Thanks, I should have looked over the layout a little better. I was difficult with not part numbers or values. If you get a chance could you post the schematic you used for the board and maybe some values to start with. I think I'll stick a 2k pot in the to start, just to insure I have enough range. I can lower it later if it seems too sensitive. I have the boards ready to etch in the morning. I'll go from there.
Thanks for all your help with this stuff. You really help keep things interesting.
Blessings, Terry
Thanks, I should have looked over the layout a little better. I was difficult with not part numbers or values. If you get a chance could you post the schematic you used for the board and maybe some values to start with. I think I'll stick a 2k pot in the to start, just to insure I have enough range. I can lower it later if it seems too sensitive. I have the boards ready to etch in the morning. I'll go from there.
Thanks for all your help with this stuff. You really help keep things interesting.
Blessings, Terry
Impressive
amplifierI would love to build these with Toshiba mosfets like 2SJ201 & 2SK1530
Greetings
Ok, I didn't bother creating a new silk layer because I figured it would be pretty easy to follow. I will do a quick schematic and silk layer and post them later so there is something to use as a guide.
I'm not sure that was ever a goal, low parts count that is.
This latest exercise isn't really intended to be a 'real' implementation. Terry was interested to pursue a simplified OPS and I'm just facilitating. If anyone else wants to give it a go they are more than welcome.
As for an IC based IPS, why not? It doesn't turn my crank and I'm not too familiar with what offerings are out there in IC-land but there is nothing stopping someone from turning one out to try.
Sorry for the confusion guys. I actually laid out my plan in the Slewmaster build thread and then realized that it was probably pulling that thread off course. I have a bunch of extra IPS boards and was just looking at learning how to implement different types of OPS. My first thought was to just tag a couple of laterals on an IPS, similar to the peeceebee where they run directly off the VAS. Through the discussion Jason suggested using a vbe multiplier rather than just a couple of resistors. Then spice showed a big improvement in THD with the two dirvers. That is how we got here. He was kind enough to modify his OPS layout to give me something to try out. Good thing with this layout is I can still try the very simple design if I want. It would just be a matter of reducing parts and adding some jumpers. Anyway, I now have something to play with and learn while we wait for OS to return and give us some more of those IPS designs he spoke of before he left.
I don't know that it would be a proper use but what about using the NI3886 as a driver for the input section? From what I read elsewhere this is one of the best sounding chips if you can stay well below the clipping point as the protection circuit makes some nasty noises once you reach that point. I doubt that it would be as nice as some of the discrete designs that we have here but who knows using it optimally?
OK, I got it put together. Here is a report of the things that are happening.
First thing I did was hook it up with just 15K resistors from rail to PD+ and ND-. Not well balanced. 500mV offset. So the.n I hooked up a Symasui IPS to it and the servo was able to zero out the offset so that was good. Now for bias, with the 500R pot maxed out bias is at 80mA. With the pot turned all the way out it is at 105mA. Not much adjustment. Now for the worse part, with the output at about 3.5vac output I see some kind of oscillation on the scope. Suggestions?
First thing I did was hook it up with just 15K resistors from rail to PD+ and ND-. Not well balanced. 500mV offset. So the.n I hooked up a Symasui IPS to it and the servo was able to zero out the offset so that was good. Now for bias, with the 500R pot maxed out bias is at 80mA. With the pot turned all the way out it is at 105mA. Not much adjustment. Now for the worse part, with the output at about 3.5vac output I see some kind of oscillation on the scope. Suggestions?
Attachments
Missed INP return.
Attachments
Last edited:
