Sansui AU-70 amplifier distortion in both channels from the speaker terminals.

It is a cathode feedback design, which is pretty sophisticated for the time. I do hope it's possible to salvage this one. Even if DC resistance is only half an ohm from cathode to ground, daveyh should find about 20mV across that resistance if the output tubes run at 40mA bias current. If the finding is truly zero mV, then the tubes are not conducting and we can conclude that those tubes are kaput. Well, assuming that the heaters are lit up.

All output tubes were checked by two different tube checkers. An TV-7 and a Mighty Mite TC both have served me well checking other tubes. The Mighty Mite was just rebuilt and calibrated by an older gentleman who worked for Motorola for over 25 years. I even installed another set of output tubes. Both set of output tubes have the same distortion.
 
It the distortion like a breaking up, or a cut in frequency response?

It does both. Worse when I turn up the volume. When I quickly unplug the unit from the power as the voltage fades the sound becomes cleaner until it fades.
 
You still haven't confirmed bias current flowing in the output tubes. Use one of the methods discussed above. If normal bias current is confirmed, then the logical next step is to test the OPTs for internal shorts. If there's no bias current and voltages on all tube elements are nominal, then the tubes are bad regardless of readings on any tube checker. The only other possibility is bad tube sockets, which is extremely unlikely on every one in a set of four.
 
Last edited:
Have you tried Headphones? It also could be a dirty speaker switch.(this happened to me on a HK 230 Ballad integrated amp. it's confusing because it is both channels, having 2 bad transformers, or 2 or more bad output tubes is unlikely.
 
Might be worth tacking a 10 ohm resistor between pin 3 on each tube socket and the lead on the output transformer that it normally connects to. Reading voltage across that will tell you what the tubes are actually doing.

Bit confused by the 7189 datasheet. it says with 400 on the plate and 300 on the grid, a pair of tubes will idle at 15 ma in pentode mode, but it idles at 70ma in ultralinear. Somehow that doesn't seem right to me. I'd expect a fair bit more than 15ma per pair of tubes for idle current.
 
Have you tried Headphones? It also could be a dirty speaker switch.(this happened to me on a HK 230 Ballad integrated amp. it's confusing because it is both channels, having 2 bad transformers, or 2 or more bad output tubes is unlikely.

Tried it with headphones still the same. All have cleaned all the switches more than once.
 
Last edited:
I did try tacking a cathode resistor to ground from pin 3 and see if it was getting current. I was able to only get some voltage but not much. Using my signal tracer I was actually able to get some sound but still distorated. Pin 2 is clean.
 
You still haven't confirmed bias current flowing in the output tubes. Use one of the methods discussed above. If normal bias current is confirmed, then the logical next step is to test the OPTs for internal shorts. If there's no bias current and voltages on all tube elements are nominal, then the tubes are bad regardless of readings on any tube checker. The only other possibility is bad tube sockets, which is extremely unlikely on every one in a set of four.

Test the OPTs for internal shorts? How do you suggest I do this?

I have removed and cleaned all the tube sockets with Deoxit before I started this project.
 
Test the OPTs for internal shorts? How do you suggest I do this?
That depends on what you have available for test gear. One of the simplest ways is to use an inductance meter or LCR bridge. The method I often recommend for folks without that capability is to yank the output tubes and drive the OPT 8-ohm secondary backwards from a 5VAC or 6.3VAC source such as a small filament transformer, with a one-ohm current sampling resistor in series. The voltage across that resistor tells you how much current the unloaded transformer is taking. A good OPT will consume only 50~200mA of magnetizing current in this test, while a shorted one could easily take over ten times that much. Make sure the sampling resistor can handle at least 5W. Voltage on the primary side will normally be quite high during this test. For example, an 8K:8R OPT has a turns ratio of 32:1, so primary voltage would be 6.3 * 32 = 202V end-to-end or half of that from CT to either end. If magnetizing current is suitably low, then check primary voltages as confirmation of a good OPT.

Be aware that interrupting live connections between the filament transformer and OPT secondary can have bad consequences due to unloaded inductive reactance, so make those connections secure and switch power only to the filament xfmr primary. Use a variac instead, if possible.
 
Found the problem. I had a friend of mine look at it to give me a second pair of eyes and beside that he's been doing this for a lot longer than I have! There is a resistor that someone had installed that was an 1800Ω, It is not on all the schematics for the AU-70. It is however on the schematics under the hood of the receiver. It is suppose to be a 2Ω / 2 watt. Not an 1800Ω / 5 watt that was installed. The resistor cut the voltages in half thus causing distortion in the channel outputs. A 2Ω / 2 watt resistor was installed and all clear as far as sound.

Thanks for every ones help.
 
Is this the 2R resistor in series with the 155V secondary on the power transformer? The voltages you reported previously on the power tubes were not at half nominal, they were very nearly normal. Please explain.
 
Is this the 2R resistor in series with the 155V secondary on the power transformer? The voltages you reported previously on the power tubes were not at half nominal, they were very nearly normal. Please explain.

My Fluke 87 meter I have used for over 5 years and I bought used was bad. I used a cheap meter to take the readings.
 

Attachments

  • 100_1614.jpg
    100_1614.jpg
    392.2 KB · Views: 22
Last edited:
The only problem I have now is the voltage has to be set at around -17v on the wipers. If I go to the -13.5v as the schematics says the tubes get way to hot. I am running 7189A now. Pin 7 voltage is 432v.
 
What's the voltage on the screens (pin 9)? If it's anywhere much over 300v, then I can see why you need so much negative bias voltage (by the datasheets, they need -15v bias with 400v plates and 300v screens).

432v is running the plates perilously close to the design max of the tube- so it's not at all surprising that so much negative bias is required. I'd be a bit concerned about the tubes wanting to run away (I've had a few 7189s do that, even Mullards and Amperexes), with that much voltage on them.

Regards,
Gordon.
 
Well, it's a great relief to know that your OPTs aren't blown. To keep them from being damaged and to get the best sound from your amplifier, you should now attempt to bring its operating conditions on target. One of the most common causes of excessive B+ voltage is AC line voltage above the original spec. You can check this by measuring heater voltages. They should be very near 6.30VAC if the unit is operating on the AC line voltage it was designed for. Also, it would be smart to add current sensing resistors in series with each output tube cathode. I suggest 1-ohm 1% 1/4W resistors in this case. These resistors make it convenient to measure each tube's cathode current, so you can match push-pull pairs and set the bias trimmers for optimum performance, and they can act as fuses to help prevent OPT damage if power tubes fail shorted.
 
What's the voltage on the screens (pin 9)? If it's anywhere much over 300v, then I can see why you need so much negative bias voltage (by the datasheets, they need -15v bias with 400v plates and 300v screens).

432v is running the plates perilously close to the design max of the tube- so it's not at all surprising that so much negative bias is required. I'd be a bit concerned about the tubes wanting to run away (I've had a few 7189s do that, even Mullards and Amperexes), with that much voltage on them.

Regards,
Gordon.

Pin # 9 is 325v
 
Well, it's a great relief to know that your OPTs aren't blown. To keep them from being damaged and to get the best sound from your amplifier, you should now attempt to bring its operating conditions on target. One of the most common causes of excessive B+ voltage is AC line voltage above the original spec. You can check this by measuring heater voltages. They should be very near 6.30VAC if the unit is operating on the AC line voltage it was designed for.
Also, it would be smart to add current sensing resistors in series with each output tube cathode. I suggest 1-ohm 1% 1/4W resistors in this case. These resistors make it convenient to measure each tube's cathode current, so you can match push-pull pairs and set the bias trimmers for optimum performance, and they can act as fuses to help prevent OPT damage if power tubes fail shorted.[/QUO

"I suggest 1-ohm 1% 1/4W resistors in this case".

I have tried adding the resistors from Pin 3 to ground. It only introduces distortion, I have rebuilt a Sansui 500 and two Sansui 1000a's. I did the cathode to ground mod and it worked great. I did however use 10 ohm 1/4W resistors. When I removed these 1 ohm resistors the distortion disappeared.

If I am correct Pin 4 is the heater and it reads 52.5v :yikes:
 
Back
Top Bottom