AU 717 problems

drbacklash

Active Member
This is a continuation of an earlier thread about the 717 not coming out of protection mode but I followed Leestereos instructions on improving the power and driver boards so now I have different problems.

Some background, I got this unit a few years ago and cleaned all the boards, replaced most of the caps, cleaned the pots and replaced the relay. It worked fine and I used it for a while until I stored it for a couple of years. Recently I pulled it out again, turned it on without using the dim bulb for protection (I know, not smart) and it would not come out of protection. I couldn't get the DC offset adjusted so I put it away again until I saw Leestereos post.

I pulled both driver boards and the power board and replaced the aluminum caps, diodes and resistors greater than 1/2 W with metal film resistors. I also replaced the relay. I checked all solder joints and checked the traces for continuity.

The good news is the 717 now comes out of protection mode. The bad news is it probably shouldn't. With my MM test probes on the left channel (F2721 board), he offset starts at 0 mv and increases by one to 400 mv and starts over again. That is with the speaker switch off. If I move it to A (where the probes are), the MM reads a steady 6.04 V. At B it does the recycle from 0 to 400 mv loop thing. At A+B it reads 6.05 V. At all these settings, the protect light is steady on, not flashing and I heard the relay click at the beginning.

Moving to the right channel (F 2722), the same thing happens. I get similar readings throughout the speaker select range. 0 to 400 mv recycle, 6.06 V and 6.06 V. VR 01 and VR 02 are not functional in that moving them does not adjust the readings on either channel. If I ground the negative probe and put the positive one on the test points where you adjust the Bias, then the offset can be adjusted but puts the unit in protect mode. With the unit out of protest mode, the test points show -63.9mv on the right channel and -397mv on the left.

I have a test speaker I use and when I connect it to the 717 and turn it on, it comes out of protect mode. It goes into protect mode when I switch to that speaker (A). Same for L and R. Bias current is adjustable and is on the money.

So, what next? I didn't test or replace any transistors and I'm thinking that is about all that is left to replace. Did I overuse the metal film resistors? I'm willing to keep working on this but I sure hate to have to take that power board out again. I also tested the pins and plugs at the bottom of the driver boards but I didn't know what I should be looking for.

Any suggestions will be appreciated.
 
Please post large clear pictures of one or both of the driver boards.
and resistors greater than 1/2 W with metal film resistors.
- this puzzles me a bit... you replaced all resistors over ½ watt with metal film types? why?

So, driver board resistors R38 and R43 are both ½ watt, so you will have replaced these? - this catches some people out, what resistor value did you use as replacements?
 
Last edited:
I have pictures to post of both driver boards but can't find any instructions to do so.
Why did I use so many metal film resistors? Good question. I knew that the problematic fusible resistors should be replaced with metal film so I assumed that they must be good replacements for other aging resistors too. Mistake?
Here are the values in ohms for all the 1/4 watt resistors I replaced:
R23 and R25 150
R15 10K

The 1/2 watt resistors
R16 27.4K
R09 3.9K
R31 47
R28 47
R38 4.7
R36 56
R43 4.7
R37 56

The 2 watt resistors
R10 1.5K
R04 2.2K
R46 10

I also used metal film resistors on the power board
1/4 Watt
R07 to R14 10
R33 22K
R34 22K
R604 470K
R608 39K
R622 4.7K
R619 5.6K
R613 100K
R615 8.5K
R612 100K
R614 120

1/2 watt
R25 1.8K
R26 1.8K
R617 4.7K

2 watt
R03 to R04 3.3K
R01 and R02 3.3K
R37 and R38 10
R601 390
R602 and R603 2.7K

So, how badly did I mess up by doing this?
How do I post pics?
 
F-2721(2).jpg F-2721.jpg F-2722.jpg

These are the two driver boards after upgrade/changes. Two pics of the F 2721 and one of the F 2722.
 
Which transistors should I pull and test first? I will do all of them if I have to but I would like to start with the ones that may be causing the problem I described above. What transistors, if bad, could cause the DC offset to "cycle" from 0 to 400 mv?
 
Thank you for posting the pictures, it is clear you have done a pretty good job of component replacement, I commend you for that. I think I was hoping to see something in the pictures that would explain the problem you are having but so far - I can't see anything that would produce such strange behaviour.

When the amp is doing this 0-400mV thing are all the connectors on the driver boards reconnected - including the input connectors - just at the moment I am thinking 'grounding issue' - and having all the connectors connected is important for that. (a huge clue being that the offset pots don't work)
 
Last edited:
Thank you for posting the picture, it is clear you have done a pretty good job of component replacement, I commend you for that. I think I was hoping to see something in the pictures that would explain the problem you are having but so far - I can't see anything that would produce such strange behaviour.

Thanks Hyperion. I think the problem is related directly to the cold start up without the dim bulb tester but I don't know what component was affected. I measured all the ones I removed and never found a problem. I didn't replace or test any transistors so that will be my next step. Thanks for taking your time to look at it.
 
Thank you for posting the pictures, it is clear you have done a pretty good job of component replacement, I commend you for that. I think I was hoping to see something in the pictures that would explain the problem you are having but so far - I can't see anything that would produce such strange behaviour.

When the amp is doing this 0-400mV thing are all the connectors on the driver boards reconnected - including the input connectors - just at the moment I am thinking 'grounding issue' - and having all the connectors connected is important for that. (a huge clue being that the offset pots don't work)

Yes, all the plug/pins on the driver boards and the 4 on the ps board were connected. In fact, I cleaned them to make sure they had good contact. I was wrong on one thing I posted. The offset pots are functional if I connect the mm positive to the driver board test point ( where bias is measured) and negative to ground. I can adjust output close to zero but it has no corrective effect on the cycling. The pots don't work with the leads where the speakers are normally connected.
 
Is it likely that the initial start up power surge affected another component (fusible resistor) on another board (other than driver or power) and that is causing the problem? Just a thought.

Also, is there an AK member I can send just the boards to check? For a fee, of course.
 
Try removing wires from pins 3 and 4 leaving pins 1,2,5,6 connected on one card, and remove all wires from pins 1 through 6 on the other. This removes the protection circuit from the mix. As I recall, and perhaps another tech could chime in as I do not have one on the bench at the moment,, you then will be able to measure the offset across pins 3 and 4. Try measuring with the input molex disconnected. If your offset is diving one way or the other that can point to a bad fet, but lets not go there until you get some measurements.
 
I have a test speaker I use and when I connect it to the 717 and turn it on, it comes out of protect mode. It goes into protect mode when I switch to that speaker (A). Same for L and R

I've never seen one of these units so please indulge me. Given the seemingly radical change in symptoms when you have a speaker in circuit I'd be interested in the output emitter network and R46 (which you've replaced); their values and connection integrity..
 
Ok, give me a day or two to test and get measurements. I took the drivers out (again) to test transistors so I will have to put them back in. I'm getting pretty good at doing this now.
 
I've never seen one of these units so please indulge me. Given the seemingly radical change in symptoms when you have a speaker in circuit I'd be interested in the output emitter network and R46 (which you've replaced); their values and connection integrity..

Not familiar with the term output emitter network. Are you looking for voltages at R46 or somewhere else?
 
Not familiar with the term output emitter network. Are you looking for voltages at R46 or somewhere else?

Overundr1's post wasn't visible to me when I posted so I suggest you give his response priority. What I'm suggesting is that perhaps you have poor R44/R45 emitter resistor connections/values or similar at R46
 
Try removing wires from pins 3 and 4 leaving pins 1,2,5,6 connected on one card, and remove all wires from pins 1 through 6 on the other. This removes the protection circuit from the mix. As I recall, and perhaps another tech could chime in as I do not have one on the bench at the moment,, you then will be able to measure the offset across pins 3 and 4. Try measuring with the input molex disconnected. If your offset is diving one way or the other that can point to a bad fet, but lets not go there until you get some measurements.

A point of clarification: Looking at the driver board, moving down from the Molex, the pins are numbered thusly:
F2721
02 white
03. White
08. Orange
13. Purple
12. Yellow
04. Gray

F2722
02. Black
03. Black
08. Brown
13. Green
12. Red
04. Blue

So pins 3 and 4 would match up to pins 08 and 13 per the board. Correct?
 
For example, on F2721----
yellow and grey are the power supply feeds to the board
white and white are the two ground wires from the board to the chassis/main cap ground points
orange and purple are the two signal output wires that go through the protection circuit and then to the speaker terminals.
 
For example, on F2721----
yellow and grey are the power supply feeds to the board
white and white are the two ground wires from the board to the chassis/main cap ground points
orange and purple are the two signal output wires that go through the protection circuit and then to the speaker terminals.

Understood.
 
Back
Top Bottom