No, I haven't changed the emitter resistors in the ARC, which are 0.2 ohm.
It has occurred to me to try changing them to a higher resistance as a "fix" for the apparent runaway hazard. That would directly translate to an increase in the amp's output resistance, though, since there is no global feedback.
Note that the ARC has six pairs of output transistors vs the three pair in the ON Semi app note.
It has occurred to me to try changing them to a higher resistance as a "fix" for the apparent runaway hazard. That would directly translate to an increase in the amp's output resistance, though, since there is no global feedback.
Note that the ARC has six pairs of output transistors vs the three pair in the ON Semi app note.
My own HD220 is an AR reworked one with the current production 3281DG parts/
They dont shutdown or overheat. My unit doesnt have the trim pot that controls solid state driver current.
So was it a runaway or a tube stage cap DC leak sending DC to driver stage?
They dont shutdown or overheat. My unit doesnt have the trim pot that controls solid state driver current.
So was it a runaway or a tube stage cap DC leak sending DC to driver stage?
You quite likely weren't asking me, but after rebuilding my example in the most straightforward way possible, there was definitely a current-hogging situation among the sets of output devices, despite my efforts to match them. I believe this would have led to a damaging runaway in the "hogs" if it wasn't addressed.
DC leakage from tube stages to SS stages was not an issue in my example. (It may have been in the OP's example, but my take is that that is unknown.)
I'm not proud of it, but I addressed the current-hogging in my example by increasing the emitter resistor values for the hogs until they no longer threatened to run away. (BTW, having already compromised output impedance, and having concluded that there really was a user safety issue involved, I also added DC rail fuses.)
My example did not have a trimmer resistor for the SS stage idle current either, but it was clear that selected fixed resistors had been used.
By any chance, have you measured the per-device idle currents in your example's output devices? I would love to know what the range is.
Nothing wrong with emitter resistor fix to balance the devices
Thats exactly how I get new production Vishay mosfets to revive dead Counterpoint SA100s. I dont hear any degradation in sound at all.
Someone thinksI should go to hell for siggesting that...meh.
Good question on the queiscent current thing.... My HD220 has no trimpot and it seems that the middle combo of paralleled resistors smack at the middle that adjusts bias are at 576ohm (schematics noted 1.2ktrim R36)
As far as output devices...the 0.2 ohm buggers measure a measly 11-13mV right after turn on. The heatsinks are warm to touch (not hot) while playing though.
Maybe ARC lowered it dramatically to mask the problem when they reworked the unit with all -DG output devices.
Thats exactly how I get new production Vishay mosfets to revive dead Counterpoint SA100s. I dont hear any degradation in sound at all.
Someone thinksI should go to hell for siggesting that...meh.
Good question on the queiscent current thing.... My HD220 has no trimpot and it seems that the middle combo of paralleled resistors smack at the middle that adjusts bias are at 576ohm (schematics noted 1.2ktrim R36)
As far as output devices...the 0.2 ohm buggers measure a measly 11-13mV right after turn on. The heatsinks are warm to touch (not hot) while playing though.
Maybe ARC lowered it dramatically to mask the problem when they reworked the unit with all -DG output devices.
Last edited:
Thanks for the info. That's about where I set the the idle (for the "non-hog" devices) in my example.
Audio Research schematics show 20mV
Also I just found a mention by Bob Cordell in his amp book that it's 218mA for these exact devices.
Also I just found a mention by Bob Cordell in his amp book that it's 218mA for these exact devices.