Current draw is THE clue, I have 5 scopes, (excluding soundcard scopes) many of the forum readers have none...Current draw is only a clue.
And everyone here who doesn’t have at least a 300 MHz analog scope, and intends to repair amplifiers or build high performing ones, gets routinely chastised.
Hee, hee. I probably should not mention my 5MHz Heathkit oscilloscope.
In my experience, stability can be 90% addressed through design - understanding feedback and high-speed techniques. The remaining 10% requires testing due to unknown parasitics.
My last two amplifier designs worked with essentially no debugging. I have to go back three generations to find one that included a surprise oscillator.
Ed
In my experience, stability can be 90% addressed through design - understanding feedback and high-speed techniques. The remaining 10% requires testing due to unknown parasitics.
My last two amplifier designs worked with essentially no debugging. I have to go back three generations to find one that included a surprise oscillator.
Ed
Unfortunately AU-X1 has already been designed so the only thing left to do is error tracking/debugging 😉My last two amplifier designs worked with essentially no debugging.
My first scope was an old round green screen. Not even a MHz. My usual is a gasp DSO. 2Gsamples/sec, so it will capture most oscillations. My good analog is one of the boat anchor Teks from the old days. I think they used to call them “transistor sets”. I get it down and fire it up when necessary.Hee, hee. I probably should not mention my 5MHz Heathkit oscilloscope.
In my experience, stability can be 90% addressed through design - understanding feedback and high-speed techniques. The remaining 10% requires testing due to unknown parasitics.
My last two amplifier designs worked with essentially no debugging. I have to go back three generations to find one that included a surprise oscillator.
Ed
Designing for the parasitics is simply the way to go. But even then a circuit will sometimes simply fail to cooperate. I built two amps recently, with essentially the same amplifier circuit. Both EF3’s, using monolithic darlingtons and a separate TO-92 (pre)driver. The first one ran class H with quad rails and worked almost the first time. I had a sticking rail switch because I juxtaposed the MPSA06/56 driving the minus side rail switch. That’s all. The other with a lower voltage standard supply - but using TIP142/7 instead of the good Sankens. It fought me tooth and nail. Gave up and put in the faster outputs and it’s been an angel.
Posts from SMERSH have been removed after he was found to be a sock-puppet.
There is some unavoidable collateral.
Hugo
Moderation team
There is some unavoidable collateral.
Hugo

Moderation team
AU-X1 is a specific amplifier, the design assumptions and specifications were beyond research capabilities at the time
Most important:
Common problem is the inability to set the +/-11V voltage, then you need to change the value of the R73 and R04 resistors from 5.6k to 3.9k, the lack of voltage stability most often results from problems with the limiting resistors R41 to R44, the diagram obviously contains an error, in fact these are two resistors connected in series with a resultant value of 13.8 ohm It is worth adding an interesting fact that engineers improved the design of the amplifier practically throughout the entire production period, the last modification was a complete redesign of the main board of the phono preamplifier.
If you need any other info let me know, i restored hundreds of them.
If you will make those modifications the amplifier will be stable and will sound like it should without decrasing its performance
One i own personally, after all the mods its stable with square wave 10Vpp 8R+0.47uF up to 200kHz, rise time measured is 0.3uS, the bandwidth is up do 600kHz -1.5dB, and 1Mhz -3dB
Most important:
- Replacing Black flag capacitors
- Modification of the GND of the line preamplifier through 3 star-connected capacitors
- Modification of the GND of the main board of the turntable through 3 star-connected capacitors (Most often already introduced at the factory)
- Replacing fuse resistors less important but worth doing
- Changing the value of pre-driver decoupling capacitors from 33pF to 47pF (Often already introduced at the factory)
- Removal of ceramic capacitors from MM preamps
- Replacement of MV-12 diodes with 2 series-connected 1N4148 diodes (8 in the power supply, 2 in the power amplifiers, 2 in the preamplifier) Does not affect failure rate but improves offset stability
- Modification of the GNC of the MM preamps, cutting off the non-inverting input and connecting it with a 0.33uF capacitor, then connecting the transistor input with a 1M resistor to the output of the MM stage
Common problem is the inability to set the +/-11V voltage, then you need to change the value of the R73 and R04 resistors from 5.6k to 3.9k, the lack of voltage stability most often results from problems with the limiting resistors R41 to R44, the diagram obviously contains an error, in fact these are two resistors connected in series with a resultant value of 13.8 ohm It is worth adding an interesting fact that engineers improved the design of the amplifier practically throughout the entire production period, the last modification was a complete redesign of the main board of the phono preamplifier.
If you need any other info let me know, i restored hundreds of them.
If you will make those modifications the amplifier will be stable and will sound like it should without decrasing its performance
One i own personally, after all the mods its stable with square wave 10Vpp 8R+0.47uF up to 200kHz, rise time measured is 0.3uS, the bandwidth is up do 600kHz -1.5dB, and 1Mhz -3dB
Last edited:
FRA
10VRms 8R + 0.47uF 40kHz
Performing stress tests, its stable even with such connections
Every one is repaired with original NOS transistors
10VRms 8R + 0.47uF 40kHz
Performing stress tests, its stable even with such connections
Every one is repaired with original NOS transistors
- Home
- Amplifiers
- Solid State
- Preventing Sansui AU-X1 output stage oscillation burn out