Bias Circuit Operation - explanation

[Hyperion]

Here are some tips which you might find useful when you have a bias problem.


Below is a schematic snip of a driver board for an AU-517.



TR06 (centre left) is the bias control transistor, just below it is VR03 the bias trimmer, the transistor and the trimmer are vital to the operation of the bias circuit.

How this bit of circuitry works to control the bias and make it adjustable.

TR05 and TR07 are turned on enough to provide current flow through R23 (from the positive rail) and R25 (from the negative rail). This current also flows through R20, R21 and VR03, these last two present a voltage to the base & emitter of TR06 turning it hard on, corresponding to minimum bias.

With the bias trimmer VR03 wound all the way to the bottom of the track to present all of its track resistance to the circuit TR06 will be turned fully on. With TR06 turned fully on the bases of TR08 and TR11 will be effectively almost shorted together – along the red lines – and so will have insufficient voltage to switch on so, little or no bias current should flow in the output transistors. The voltage present at ‘V’ will be 1 to 2 Volts or thereabouts with the bias trimmer set to minimum bias.

Adjusting Bias
If we then try to adjust bias we will alter the slider position on the track of VR03 in the direction of the purple arrow (upwards). As we do this TR06 will gradually turn off. As TR06 turns off, the voltage at ‘V’ is allowed to increase. As this voltage increases, the voltage across the base and emitter of the driver transistors increases, (green lines) thus gradually turning them on. When this happens the Output transistors also turn on. And with the output transistors turning on, bias current can be measured (as a voltage drop) across the output transistor emitter resistors - This is normal operation. Note that representative voltage drops can be seen across R34 & R35 and also R36 & R37 indicating current flow in the respective pairs of driver transistors.

If you can’t adjust bias:-

i.e. no bias, - OR bias stuck at wrong reading.

– things to check.

First check your power supply voltages (always).

Check that TR05 & TR07 are turning on. voltage B-E=0.6V and their schematic voltages are correct*.
*Use minigrabbers attached with the power OFF then power on to measure the voltage.

Check TR06 – adjust the trimmer and see if the voltage at ‘V’ alters** – if not suspect the trimmer or TR06 faulty. Voltage at ‘V’ will be around 1-4Volts, dependant on bias setting, and amp design.
**Use minigrabbers attached with the power OFF then power on to measure the voltage.

If good up to this point check R29 & R30 (out of circuit - with Ω Meter). Also check R38 & R43 – same procedure. If any of them are high value or ‘open’ they will either starve the drivers of current (R29 & R30), or the Output transistors (R38 & R43). .

If still no-go check the driver transistors, by removing them from circuit and using either multimeter on ‘diode test’ range, or a transistor tester.

Check the Output Transistor emitter resistors, lift one end of each and test.

If still no-go check the Output transistors (remove them) and test as above.

These are the basics, I have tried to explain the operation of the bias circuitry in as simple terms as possible – I hope this helps.

 

[Alex22]

This is bloody brilliant!!!!!!!!!!!!

I will surely be reading it a few times to try to sink it into my little brain.
I tried to adjust my 555a a couple of weeks ago but couldn't get a reading on one channel, but the amp is working apparently ok.
Once I have another project out of the way I will use this information and have a crack at it, just haven't made time for it recently.

Thanks Hyperion for some very useful information.

 

[Aud10file]

Nice writeup @Hyperion !!!

I love the note you added regarding the minigrabbers. I’ve seen a few amps come through where the owner was using test leads with power on and then a slip and they watched helplessly as the magic smoke escaped.

“*Use minigrabbers attached with the power OFF then power on to measure the voltage”

 

[Scott Muni]

Thanks for the explanation. I'm going to read this a few times to wrap my head around it.

If you could explain feedback circuits, that would really help me.

 

[ignoramustot]

Brilliant and typically generous share, @Hyperion ! Wow. Thanks!

 

[JimEGR]

Thanks so much! Super helpful, and as @ignoramustot said - generous. ESAK University. An exceptional community. Thanks @Hyperion

 

[Ray Gianelli]

@Hyperion John, very nice writeup! IMO this would be a good sticky.

 

[KeithD]

Thanks for this, very, very hellpful! If you want to comment, what are the differences between a direct coupled amp bias adjustment (such as the 517 you illustrated) and a cap-coupled amp bias adjustment?

Thanks again, wonderful post!

 

[Hyperion]

what are the differences between a direct coupled amp bias adjustment (such as the 517 you illustrated) and a cap-coupled amp bias adjustment?

Well, there really aren't any differences between to two output topologies when it comes to bias setting/adjustment, however there are significant differences when the discussion is about 'DC offset' for these two output topologies.

The procedure for doing the actual adjustment might be different, but the aim is the same. You might be asked to measure/monitor a current, rather than a voltage. performing an in-line current measurement, rather than measuring a small voltage across an emitter resistor, (representative of the current through the resistor), and so on.

Remember, the reason for having a bias setting system is to null the bipolar transistor B-E turn on voltage. In other words to get the driver and output transistors just past the 'point of conduction' so that the audio signal doesn't have to perform that task. Thus resulting in a huge reduction in distortion, compared to a system with no bias setting arrangement. In the case of the AU-517, there are three transistors in the positive half of the amplifier and three in the negative half (i.e. top and bottom). If you add their required turn on voltages together - 3 x 0.6V + 3 x 0.6V = 3.6V - sound familiar? this arrangement is called '3EF' or 'three Emitter Follower' for obvious reasons.
Notice how the Sansui engineers have thoughtfully provided a precision way (TR06 & VR03 etc), of applying up to ~4V DC to the bases of the six driver and output transistors. This requirement is no different whether the output stage is 'capacitor coupled' or 'DC coupled' to the speakers.

There are some minor differences when the bias circuit contains 'reference diodes' instead of a 'bias transistor' but the basic principle remains unchanged.

 

[JimEGR]

Well, there really aren't any differences between to two output topologies when it comes to bias setting/adjustment, however there are significant differences when the discussion is about 'DC offset' for these two output topologies.

The procedure for doing the actual adjustment might be different, but the aim is the same. You might be asked to measure/monitor a current, rather than a voltage. performing an in-line current measurement, rather than measuring a small voltage across an emitter resistor, (representative of the current through the resistor), and so on.

Remember, the reason for having a bias setting system is to null the bipolar transistor B-E turn on voltage. In other words to get the driver and output transistors just past the 'point of conduction' so that the audio signal doesn't have to perform that task. Thus resulting in a huge reduction in distortion, compared to a system with no bias setting arrangement. In the case of the AU-517, there are three transistors in the positive half of the amplifier and three in the negative half (i.e. top and bottom). If you add their required turn on voltages together - 3 x 0.6V + 3 x 0.6V = 3.6V - sound familiar?
Notice how the Sansui engineers have thoughtfully provided a precision way (TR06 & VR03 etc), of applying up to ~4V DC to the bases of the six driver and output transistors. This requirement is no different whether the output stage is 'capacitor coupled' or 'DC coupled' to the speakers.

There are some minor differences when the bias circuit contains 'reference diodes' instead of a 'bias transistor' but the basic principle remains unchanged.

This is the kind of post I like to have with my morning coffee .

Hopefully this is a valuable contribution. Here are two excellent books @Leestereo recommended to me last year. A bit like drinking water from a fire hose for someone newer to this like me, but enormously informative.

Can't express enough how much fun and how much value I've received from this community. Thanks @Hyperion and all!

 

[kevzep]

Nice write up John!! Well executed...

 

[Hyperion]

Thanks Kev, it seems to have been well received.

 

[jd3rdcoast]

Very straightforward and succinct explanation. Thanks John! Will add to my troubleshooting notes for sure. To put a bow on this post, how about a quick run through of how to calculate the Bias Voltage across emitter resistors when a service manual only offers the in-line current measurement option using Ohm's Law?

 

[Hyperion]

how about a quick run through of how to calculate the Bias Voltage across emitter resistors when a service manual only offers the in-line current measurement option using Ohm's Law?

That's a difficult one because often in the SM method the current supplying the driver transistors is measured as well as the OP transistor current. e.g. AU-9900/ AU-11000. So you have to guess at the driver current (or measure it separately) and subtract it from the total bias current - then you can calculate what the voltage drop across the OP transistor emitter resistor(s) should be. Also the above models have 2 pairs of OP transistors per channel - so the total current minus the driver current has to be divided by 2 to get the current per OP transistor pair.

However the AU-517 / '717 / '719 / '919 are the simpler ones with one measurement (per ch) - and proper bias test points too!

Just like always V=IR, so for example with 0.33Ω emitter resistors, and measuring across 2 of them, for a bias current of say 20mA

You get:-
V=IR V=0.020A x 0.66Ω = 0.0132V or 13.2mV = the 'bias current' voltage reading required across ONE pair of emitter resistors. (correct for AU-9900/11000 for example - guessing 10mA for the drivers and 20mA per OP transistor pair for a total bias current of 10+20+20mA = 50mA).

Or if you are told the voltage required = 20mV and just want to know what the current is
You get:-
I=V/R I=0.020V/0.66Ω = 0.0303A = 30.3mA (correct for AU-517/717 for example)

 

[Yused Towork]

This is most helpful, as I am in "BIAS-HELL" at the moment...

 

[Mike_S]

Thanks for your "Ted Talk" on bias, is idle current next?

Now I understand what bias is really for, and how it affects the amps distortion. Low enough to trip on the circuits, but not high enough to overpower them, causing distortion.

 

[Hyperion]

is idle current next?

That's the same thing, just another name, I prefer 'Bias'

 

[Mike_S]

That's the same thing, just another name, I prefer 'Bias'

Got a MOSFET amp that requires, idle current to be set, then input DC, lastly output DC. Always thought idle and bias were different.

 

[Hyperion]

Got a MOSFET amp that requires, idle current to be set, then input DC, lastly output DC. Always thought idle and bias were different.

Well, it's possible they are using term 'idle current' to describe the current an amplifier draws in quiescent conditions - which might be different (more) than current which is strictly for biasing the output devices.

My bias description post(s) were for biasing help for bipolar transistors, not MOSFET.

 

[Mike_S]

Well, it's possible they are using term 'idle current' to describe the current an amplifier draws in quiescent conditions - which might be different (more) than current which is strictly for biasing the output devices.

My bias description post(s) were for biasing help for bipolar transistors, not MOSFET.

You are most likely correct. It kills me that we have a dozen different terms for the same thing.

It's all the same, until it's different.