Setting bias on Sansui AU-517 puzzle

Phototone

Super Member
I have acquired a Sansui AU-517 integrated amp, and I downloaded a PDF of the service manual. It has the procedures for setting the bias and DC offset of the amp.

Well, following the procedures I am able to easily set the DC offset to near .0 mv. reading from the speaker terminals as indicated.

Now, for setting bias, there are 2 test pins (+, -) that you are to connect your voltmeter to, and another pot to adjust. Bias is supposed to be set at 20mv.

I can't get a reading from either set of test pins. (one set per output block) They read 0. Turning the bias pot does not change this, so I left the bias pots about mid range. The amp plays nicely, and the output transistor heatsinks are almost cool to the touch, one is just a wee bit warmer, but nowhere near warm. The power supply caps are cool, the dual power transformers are just a wee bit above room temperature..so no obvious "issues".

I am very puzzled about the inability to get a reading. I know my voltmeter works because I can use it to set the dc offset down in the same millivolt range.

Does anyone have an idea?
 
Last edited:
If you connect your voltmeter between the collector and the emitter of TR06, do you have a voltage of the order of 3V? Does it vary with the position of the bias trim pot?
 
If you connect your voltmeter between the collector and the emitter of TR06, do you have a voltage of the order of 3V? Does it vary with the position of the bias trim pot?

Good call IMHO.
 
Last edited:
Just a reality check for your meter... Are you sure you are set to read DC mV not AC mV?

The reason I ask is that the zero offset pots can be a bit touchy and (at least for me) it is hard to get exactly 0 mV.

Check your meter by trying to read a 9V battery also. Maybe one of the internal fuses be blown.

Russ
 
Also maybe double check your procedures.
Some bias adjustments may require selector set to Aux, speaker selector set off, volume minimum, etc. depending on the amp and wait a few minutes to let amp warm up.
 
Also maybe double check your procedures.
Some bias adjustments may require selector set to Aux, speaker selector set off, volume minimum, etc. depending on the amp and wait a few minutes to let amp warm up.

Yes, and I observed all those things. The meter is SET for DC mv. It works on setting DC balance, and yes it is hard to get "O" DC offset, but I was successful in getting "near" "O", reading from the speaker terminals, with amp set to aux, and the volume all the way down. No problems setting offset on this or other amps of mine, it is just the BIAS adjustments on the particular amp...I am not getting any kind of reading from the bias test points, no matter where I turn the bias adjust pots. Meter set on same range as testing DC offset. I know the meter works fine, as I was able to go from 150mv down to zero-ish and into negative offset with my auto-ranging meter when setting offset, so I know it functions in the correct range for setting 20mv DC bias. The procedure is the same for setting Bias as DC Offset. Amp set to aux, volume all the way down. I should also add, there is only ONE set of test terminals per power amp block, and the manual clearly shows them, I am not making a mistake there. The odd thing, is the amp operates well and sounds good, no heat, but I still wonder.
 
Last edited:
If any of the driver or output stage transistors have been substituted, then the bias pot may no longer have sufficient range. You wouldn't necessarily know this by listening, but it would stand out in distortion measurements. Do the test suggested by ecluser.
 
Sansui AU-517 zero BIAS reading problem solved!

As in the first post in this thread, I did not get any millivolt reading at the + and - TEST pins on the output cards. (The R-channel had no output, but the L-channel had a very distorted low output.) The BIAS pot was ineffective. I set up one channel output amp outside the chassis so I could safely measure voltages and they were all screwed up! I tore down the setup and started measuring resistors because locations that should have had voltages did not. What I eventually found was that both of the 150 ohm emitter load resistors, R23 AND R25, in line with transistors TR05 and TR07, respectively were open on both channel cards! I did not have the fusible resistors at the time (I have ordered from DigiKey their Vishay PPC150ACT-ND, 150 ohm, 0.33 watt, fusible metal film resistors for both channels. IF SOMEONE KNOWS IF THIS WATTAGE IS CORRECT, PLEASE POST OR EMAIL ME. Tx.), I temporarily installed a couple of 1/4 watt carbons and voila! I could adjust BIAS and DC offset just fine and the sound was great!

Now, why? I had re-capped the entire amp boards and power supply before even testing. I also had used Deoxit D5 on all the controls and switches as well as the BIAS and DC BALANCE pots on the output boards. I figure that at some time in the past the customer must have decided to "tweak things a little" or something like that and because the BIAS pots were dirty, the fusible resistors did what they were designed to do - thew blew in order to protect the transistors from overload! And boy was I glad! There are 14 direct-coupled transistors (including the outputs) on each output board! (I have seen as high as 16 taken out on a single channel in a Phase Linear, so I know it can happen!)

Thanks for all the ideas posted in this and similar threads that helped me reach my conclusions and eventually the solution! Be brave! Post 'em!
 
So, the reason you had no bias is because there was no current flowing through the bias network. These things often make a great deal more sense *after* you find out what failed. Cheers!
 
If any of the driver or output stage transistors have been substituted, then the bias pot may no longer have sufficient range. You wouldn't necessarily know this by listening, but it would stand out in distortion measurements. Do the test suggested by ecluser.

Sorry to raise this thread from the dead, but it'd be dumber to make a new thread to ask the same question.

Like the OP, I can listen to the amp and it sounds fine, but I am unable to read any voltage at all when connected to the test points on either of my driver boards. Also like the OP, I am able to perfectly dial in the speaker DC offset.

  • I have replaced all 4 power transistors with NTE 281/280.
  • I also changed all my electrolytics capacitors, trim-pots, power/fuse resistors and even the main relay.
  • I checked each and every transistors and diode in the entire amplifier and replaced a handful, including TR12 and TR13.
  • Checked all resistors on both driver boards.
  • I even removed all old solder, pad by pad, and resoldered to ensure no bad connections.

I did the test at TR06 on both drivers, both of mine are reading around 0.6V (a good ways from the quoted 3V above) and adjusting the bias trim-pot doesn't seem to change it.

So you all are saying direct replacement transistors for the stock units have altered performance characteristics enough to cause the test points to read no bias (current) voltage?

Thanks in advance for any assitance you can provide.
 
Last edited:
Okay, I found what was causing my issues. My problem is a blown output transistor. Specifically my NTE 280/281 transistor replacements on the left driver board, R36 went down with it in flames.

I believe this was because I mistakenly put insulating washers between collector and collector hardware, which is not a good idea with how they designed the power transistor assemblies. The hardware/screws literally make the collector connection with the assembly on the reverse side of the heat sink. Insulator between collector and the screws would prevent this connection from being made. How the amp worked at all in this state is a mystery to me. Lesson learned!

I've seen that some people steer clear of the NTE replacements, but I forget what they said the better choice/s were.
In short, make sure you don't add or remove anything when replacing power transistors :rockon:
 
I've seen that some people steer clear of the NTE replacements, but I forget what they said the better choice/s were.
NTE replacements are quite handy when there is nothing else that will do, understand however that while they may 'work' they are 'approximate replacements' at best. If you try to buy the same NTE part in a few months it will very likely be a different device which just happens to be similar enough to be called the same NTE P/No. At the moment there are 'certified devices' freely available from authorised suppliers (not auction sites), for any or all of the transistors in an AU-517 - that's what you should use in my humble opinion.
 
Okay, I found what was causing my issues. My problem is a blown output transistor. Specifically my NTE 280/281 transistor replacements on the left driver board, R36 went down with it in flames.

I believe this was because I mistakenly put insulating washers between collector and collector hardware, which is not a good idea with how they designed the power transistor assemblies. The hardware/screws literally make the collector connection with the assembly on the reverse side of the heat sink. Insulator between collector and the screws would prevent this connection from being made. How the amp worked at all in this state is a mystery to me. Lesson learned!

I've seen that some people steer clear of the NTE replacements, but I forget what they said the better choice/s were.
In short, make sure you don't add or remove anything when replacing power transistors :rockon:
Those are T03 style outputs. Should only need a thermal pad between it and the heat sink.
 
Well it took me a while, but I got back to the AU-517 and this time I fixed it. I had an issue on the Power Supply and Protect board, that I think was affecting all my voltages. It was the ten 10D7C silicon diodes, that had to be changed out. Then I went ahead and replaced TR602 and TR603. I made a few more minor tweaks, full details on my site: https://hallmanlabs.com/2020/05/29/more-sansui-au-517-preview-of-latest-issues-solutions/

When I originally worked on this Power & Protect board, I had no idea just how sensitive it is to component changes. You can't just change diodes all willy-nilly :)

I was able to dial in the bias of TR06 to exactly 3.000V for both of my driver boards! I think when I did this the first time, I was measuring between Base and Emitter, which is why I had about 0.6-0.7V. After sucessfully making it through Semiconductors-I for my Comp-E degree, this makes perfect sense now! :D

Then I was able to zero out the DC on the speaker terminals to about 0.05 mV.

Transformer Voltage: 35.6 Volts (AC)
I measured Rail Voltage to Drivers: +48.7V and -48.7V (+- 48.5V on schematic)
AC Ripple: 0.19V-0.24V

The best part, latching time: 1-2 seconds!! Half the time, not even my modern Yamaha and Pioneer amplifiers manage 1-2 seconds latching time.

Next, comes the audio testing! :lurk:

I still wasn't able to measure anything across those test terminals. When I checked the negative terminal for continuity to the grounding points on the underside of the amplifier, I got an Open Circuit!

Thanks for the assistance, knowing we could measure across TR06 was vital for me to setup my amplifier (due to Test Terminals not working).
 
Last edited:
Edit: I thought I figured out what had messed up the test terminals, but nevermind! It quickly zeroed back out shortly after.

I just don't see how all other ground points have continuity to ground, except this test terminal.
As far as audio quality and amp stability, everything is doing AMAZING for me now! :banana: I have really missed this AU-517 sound.
 
Last edited:
Edit: I thought I figured out what had messed up the test terminals, but nevermind! It quickly zeroed back out shortly after.

I just don't see how all other ground points have continuity to ground, except this test terminal.
As far as audio quality and amp stability, everything is doing AMAZING for me now! :banana: I have really missed this AU-517 sound.
you need to mention what messed up with test ?

do you put negative lead on negative terminal and, positive lead on positive terminal ?
 
Back
Top Bottom