AU-101 Bias revisited

Hi all,

I have recently done some work on a couple of 101's, one of these is the original "plated rear panel" design without bias pots, and with flying saucer drivers, October '72 vintage, and the other the last version with the black painted rear panel, updated driver transistors and no emitter resistors (links fitted instead)

The later model (which I received prior to the earlier build unit) was in immaculate condition, completely original, and I think that I was the first person to lift the lid since it came off the production line some 40 years ago. The other interesting change noted was the use of discrete rectifier diodes in place of the three legged rectifier pair fitted to the earlier models (one diode mounted on the top of the board, and he other underneath)

The originality and condition of this unit got me to thinking regarding bias settings for this amp. I have been suggesting that the bias current sould be set to 20ma, based on advice I had received for a QR-1500 that has almost identical topology and components in the output stage, so I thought that I would check the bias current in this unit before doing any cleaning, changes or adjustments. Bias current measured at the + power supply rail fuses feeding each output stage was measured, and came in at 9.4ma for one channel, and 10.7 ma for the other. A bit lower than the 20ma that has been previously advised.

I received the earlier build unit a short time after, and the thought struck me that althought I have done several of these earlier not bias pot amps, I have never done a bias current measurement on any of them. A quick check showed that bias was 8.4 ma on one channel, and 18.7 ma on the other. This amp was in for a complete restore, so I was keen to see whether the replacement of the driver "flying saucer" transistors would resolve the bias imbalance.

After completing the restoration ( new caps, and all transistors except the outputs replaced, the bias was rechecked - 11ma on each channel and rock steady!

So the moral of the story is that the 101 bias settings should probably be closer to 10ma than 20ma to get near the factory setting (for those later 101's with bias pots), and it is worthwhile checking the bias on the earlier no bias pots units, as this may show up issues with the driver transistors

And some other information re the transistor replacements

CDC 8001 - BC639
CDC 9001 - BC640
XA 495BL - 2SA970GR
2SC871 - BC550C

The gain of the CDC's were measured at 70 to 100, and the BC639/640 at 130 - 160, the XA 495BL gains 100 and 120 and the 2SA970 at 200 - 220 and the 2SC871 at 520 to 580 and the BC550c at 600 to 640

I might post up some photos of the two units, and I also did a before and after frequency response of the older unit, including plotting the loundness setting effect on the frequency response.

Apologies for the long post !!

Cheers

John

No need to apologise John - I am a big fan of your work, and the excellent information you disseminate about the repairs and restorations that you do.

Nice write up thanks for sharing the info! Much appreciated.

This is good info,thanks.

Just recapped a AU-101 early model,without trimmers fot BIAS setting. According to a topic here I placed 2 trim resistors 500 ohm(R823 and R824) to set the BIAS but right channel is aprox 70 mV,could not be lowered, measured across one emiter resistor wich is to much I think ?:scratch2: Left channel can be set at 10mV wich correspond with correct BIAS.

So I think I have to replace flying saucers. Sad , on PCB board ECB locations of transistors is not indicated. Hope I get it right in there !

Just a quick note on my first post - the "bias current" values that are noted in the post are the currents measured at the fuse holders as per a number of bias current measurement procedures for these early Cap coupled amps. If doing an emitter resistor measurement, then the voltage measured should be calculated for around 8ma to 9ma value.

Cheers

John

One thing I've noticed with untouched and stock Sansui amps of this era is that the bias is usually set to around 10-15mA even when the SM calls for 25-30mA. I get the feeling they were conservative setting bias at factory...

btw, i was playing with an AU-999 last week and measured output distortion as a function of bias.... i will post more on that later, but long story short 10mA or 30mA hardly matters from a distortion point of view (at least in the AU-999).

Hello. Here's a picture of my own restoration. I hope I helped you.