protection light? papa's worried about his "baby"

My AU 717 is keeping the protection light on after the relay goes click, sometimes for a bit, and its making me worry. :sigh: It started this a couple of days ago.

It has for 27 years simply gone steady immediately after the click.

I checked for any shorts on the speaker wires and found none.

I'm going to pull the cover, and vacuum it out, and look for any internal shorts, but barring this, any suggestions for me to look at? I have the schematic here.

outlawmws said:

I'm going to pull the cover, and vacuum it out.

Click to expand...

good idea :yes:

and check the offset see this thread :thmbsp:
http://65.219.61.150/forums/showthread.php?t=5634

Dominic.

Offset was checked not long ago, .7 and.5 mv respectively. I will double check though.

OK, things are not good here, :sigh: 2 out of 4 of the big caps have leaked. Just a little, but they are obviously done. I am working at reading the schematic to find them, but thats slow going. It will be easier after I blow it up big at work tomorrow.

In the meantime, Anyone had to replace these and can you recomend best available? (Nothing is too good for baby!) Marks on top are 70902. Brand is Nichicon, and ithe sleeve has a "black negative" on it (funny, it's all black, case and sleeve). Value is 15,000mf at 63 WV. (seems low for voltage...)

There are a couple of smaller caps tied across them marked 225K, 100V a triangle symbol (with heavy corners), then a :H: Should these be good or be replaced?

I'm still cleaning, but overall it looks good otherwise.

EW, you watching this? (or anyone else with 717 experiace!)

The smaller caps are film caps...Sansui bypassed the big caps with these for low ESR at high frequencies, always a good idea.

The determining factor here is size. Measure the diameter of the ones you have. They should be either 51mm (about 2") or 63mm (about 2.5"). Length can be variable to a point, as usually the cap can be mounted deeper into the bracket than the stock one to keep it from being too tall for the case. But you need to check the length of the ones you have and determine if there is some leeway for a slightly longer one if it comes to that.

Report back...

Not even close to 2"...

Here it is:

Cap Diameter =1.60”. This includes sleeve.
Max length = 3.8 and this allows for the extra space.

Markings:

Nichicon 70902
15,000 mf
63 WV

I'm off to try to size this, but any help would be deeply appreciated!

That diameter is considered 40mm. A little odd, as I thought they were larger. The 63V rating is more than enough, as the 717 uses +/- 56V rails.

Michael Percy sells a Nichicon KG Gold 15,000µf 63V for about $24 each. It measures 40mm x 100mm (100mm = 3.93 in). I suspect that in measuring the max length, you didn't take into consideration that the caps can be mounted lower in the clamps, with some of the excess length below the base panel. The hole in the base panel looks like it will allow the cap to extend into the bottom of the chassis (I do this all the time).

Digikey sells T-HA series caps that are also fairly well regarded, and much cheaper. There is a 15,000µf 63V in a 40mm x 50mm case (Digikey PN P10638-ND) for about $11 each (squat little bugger), and a 18,000µf 63V in a 40mm x 60mm size (Digikey PN P10640) at about $13 each. However, these Panasonic are snap-in style, and the Nichicon's have solder tabs. Soldering to the Nichicon's would be a lot easier, but the price difference might be enough to convince you to go with the Panasonic caps anyway. The Panasonic's also have extra terminals designed for mounting stability, and these ought to be clipped off before mounting to get 'em the hell out of the way. You might also consider replacing the stock Sansui film caps...not because there is anything wrong with 'em, but because it might be tough to desolder them and have enough lead left to connect to the new filter caps. The Panasonic ECQ-E(B) metal polyester caps are pretty small and would work well here. A 2.2µf 250V is only 21mm long (Digikey PN P10983-ND).

I'm familiar with the wiring that Sansui uses to connect to the caps. You've got a hellova project to get these caps replaced.

Edit: Sansui used a horrid glue to hold their caps in place during assembly. It can and does turn conductive after a lot of years, but don't mistake that glue for a leaking cap.

BTW, getting the protection board out (which is also the regulator board) is a serious job on this amp. If you've got a fair bit of experiance with repair, go for it. Otherwise....

Thanks EW, how do I contact Michael Percy? If those are the best I'll get them.

But it's not going well here!

:grumpy:

I didn't get to the web, I went and looked closer and some of the caps on the boards are also leaking.

Amp boards:

470mf 63V .715 dia, 1.575 long axial, .300 lead spacing - 4 of these

33mf 25V .312 Dia, .550 long axial, .200 lead spacing - 4x

Protection board,

100mf 63 v .500 Dia, 1" long axial, .250 lead spacing - 2x

220mf, 63V .75 Dia, 1.25 long axial, .300 lead spacing -2x

All seem to be nippon chem-con? ans have black sleeves.

All the other brands look brand new...


:cry: I'm not having fun here!

It's going to be a pain to get these boards out!

Michael Percy Audio

Yep...those are the best caps that will fit without doing a buttload of mods. I don't believe the 100mm length will be a terrible issue as long as some of the cap can be hidden 'below deck'.

Don't expect a lot of correspondence from him. When I need something, I PayPal the money (price of part + shipping fees + PayPal fees), and in my PayPal comments I tell him what I want. I usually have the part in a couple of days. If you try to shuttle a lot of email back and forth regarding your purchase, you'll have your part sometime next year.

Like I said, don't mistake the 'famous' Sansui glue for leaky caps.

I might say...again...that you're attacking a serious job here. Even doing just the power supply filter caps is not going to be a cakewalk, and considering that the wiring to the new caps will be soldered directly to the cap itself, neatness counts!!. Seriously consider letting a pro tackle this for you.

Lastly, I might say that, although the original power supply caps have apparently vented and we all agree that they should be replaced, they are probably NOT the reason you are having problems with the protection circuit...

EW,

I appreciate your concern about doing the job right. I know you have seen too many wrecks come in for me to not understand your concern.

If I had any doubts about my ability to get it apart and back together again properly, I wouldn't even attempt it. While I freely admit I'm not an electron pusher, Per Se, I'm no beginner to soldering, electronics, schematics, PCB design, electronics packaging, mechanical design and general mechanics.

After over 20 years actively designing ever kind of circuit medium around, and baby sitting those designs through all levels of processing, I'm comfortable with most assembly processes. I have done single-sided, double-sided, and multi-layer PCB's to space specs, thick and thin film substrate design, and hybrids, supporting "DC to daylight" analog, mixed A/D and pure Digital circuits. I spent 3 years designing all of Ford Aerospace’s R&D microwave designs 25 years ago. About all I haven’t done is actual IC’s,

Been through ESD training several times and audited NASA soldering classes. I'm not the absolute best, but I'm good.

I can understand my earlier comments about how not fun getting some of these boards would also raise your concerns. I'm not looking forward to it, but I'm also not worried about goofing it up either. I am extremely careful to keep track of things when taking them apart, and these days Cindy photo tracks my work any time I get into anything like this. She’s incredibly good with a camera, digital or otherwise most ways, so I stay out of her way.

If I do get stumped somewhere you can bet I'll stop, post pic's and ask questions!

Thanks for the link to the power caps.

I’ll look at the other “leaks closely and try to get pics. I know that at least one is a leak as I saw some corrosion in one lead of one part. Got to get that cleaned out. Last time I had to clean up after a leaking cap, (ECM on my Suzuki SJ410) I cleaned the leak up with a solution of baking soda to neutralize any acid, then cleaned the baking soda off with alcohol.

Do you have any better suggestions for that cleanup?

Any advise for sourcing the smaller caps? Same place, or somewhere else?

I forgot to mention the protection circuit,

What it was doing was taking longer to stop blinking than was normal. It did stop, just not immediately after the relay clicked. Less than a minute, probably less than 20 seconds. I suspected either any caps that have died, or where the corroded lead is may have compromised a part. I'll try to get in for a pic, but until the board is out that may not be doable.

Any guidance here? Do I need to spring for the shop manual and start verifying voltages, or is this a job for a scope? I do have the supplied schematic, and while its small, it's readable, especially after using a Xerox to blow it up a bit. going to go even bigger tomorrow...

The manual isn't such a bad idea, but if you think it's going to steer you towards the problem, you're mistaken. If you want it in order to have all the info at your diposal that you can get, then by all means get it.

Try here.

I don't see how the LED can keep flashing after the relay kicks in, at least not for very long. On first power up, the multivibrator TR606 and 607 alternate and turn the LED on and off. Once C604 charges and TR603 turns on, the base of TR607 is pulled down close to ground potential and TR606 stays on.

What I would suspect is that TR603 is getting lazy, and its C-E voltage is rising to the point that the base of TR607 is not pulled down enough to stop the flashing. After being on for a few seconds and having a chance to warm up a bit, the gain increases (as it does in all transistors) and the C-E voltage drops to the point that TR607 is finally forced off.

In this scenerio, TR603 should be replaced with a good high-gain/low VceSat transistor.

Welcome to AudioKarma costin1998. This is a great place to get information on your amp. The AU G55X is a very different amp than the AU-717. You should start a new thread so we can keep the information for your amp separated from an AU717.

Your amp probably doesn't have fusible resistors. They were discontinued by the time it was made. Do you have a schematic? That will be helpful to find a test point. There is usually a coil between the output from the amp board and the protection relay. It usually provides a good test point for DC offset.

- Pete

EchoWars said:

That diameter is considered 40mm. A little odd, as I thought they were larger. The 63V rating is more than enough, as the 717 uses +/- 56V rails.

Michael Percy sells a Nichicon KG Gold 15,000µf 63V for about $24 each. It measures 40mm x 100mm (100mm = 3.93 in). I suspect that in measuring the max length, you didn't take into consideration that the caps can be mounted lower in the clamps, with some of the excess length below the base panel. The hole in the base panel looks like it will allow the cap to extend into the bottom of the chassis (I do this all the time).

Click to expand...

For the record
The original 15,000µf 63V caps in the AU-717 are 40mm dia, and 100mm long - the ones EW suggested from Michael Percy could not be a more perfect replacement, for size, looks, and performance - just a little pricey. I still thought is was worth it and I had them shipped to the UK !

John

Ancient thread...I've used those Nichicon caps from Michael Percy in about 20 AU-717's since that post. They are essentially the modern equivalent to the original caps, and are indeed perfect replacements for them.

Ah ! - didn't see the date till it was too late - I got fooled by that post about the AU-G55X... Grrrr

Thanks for the re-confirmation EW. (20 '717's ! thats a lot of caps !).

John

Hi,

to "abouse" this topic a bit for a Q about Percy.

I've sent him an email with parstlist and request for confirmation if everything is obtainable + what are total costs with shipping to Europe for my order - after 18 days - no answer, nothing ... In mean time I allready sent another email (he sugests us to do it if no reply in couple days) .. nothing.

So - EW suggestes simply to pay him via PP and send list of ordered parts - OK - BUT how to (in)calculate the shipping costs?




OK, another one - my AU-717 does same as the one described above - after powering up - led flashes (OK, that's fine) - BUT after relay kicks in - led still "blinks a bit" for another 10min and then it goes in permanent ilumination ...
Bias is OK, DC offset is OK, etc ...

I like this EW's idea about replacing TR603. Might be worth of trying. (btw - partslist sent to Percy is refreshment material for AU-517 (friend's amp) and mine Au-717 :-D

You're going to have to keep rattling his cage till you get his attention. Shipping costs are fixed for domestic orders, but you're going to have to get a total first.

Don't be shy...keep sending emails.

Edit: thought of another way to get his attention. PayPal him the amount, with the PayPal comments on what you need, and say in the comments that you are aware that you have not included the shipping total, and for him to get back to you with the costs for shipping and you'll send that right away (make sure he has your email address).

Hi,

yes that's the main problem - since I don't know how high would be shipping costs - I can't simply pay in advance

Thanks for updated tip - that's exactely I got in my mind after reading your first comment a bit earlier about Percy (that you simply pay him)!

Maybe start a new thread it could be a interesting fix/story
Peter