Leestereo's Restoration/Upgrade Of A Sansui AU-717

Nice work!

Would you consider posting a list of the parts you used? It doesn't have to be a Mouser BOM list, any list would do. I have someone interested in me restoring one of these amps for them. I don't mind the leg work of sourcing parts, but it's a shame to re-cover existing ground. If you ever do a NAD 3020, you'll find a whole list of parts over on my threads. :D

Lee (another one).
 
...Would you consider posting a list of the parts you used? It doesn't have to be a Mouser BOM list, any list would do. I have someone interested in me restoring one of these amps for them. I don't mind the leg work of sourcing parts, but it's a shame to re-cover existing ground...

Unfortunately, I did not compile a BOM specifically for this AU-717 restoration/upgrade since I already had many of the replacements in my own parts stock. Also, in most cases, there are several different capacitor brands/series that are appropriate; I consider the type of capacitor and the proper fit more important than the specific brand/series and hence there are different capacitor brands/series used in my restorations.

For low ESR types, I typically use capacitors from Panasonic (FM and FR series), Nichicon (PW series) and United Chemi-con (KZE, KYA, and KYB series). For signal path duties, I typically use Nichicon ES bi-polar types to replace the larger electrolytic capacitors; for smaller sizes (e.g., <4.7µF), I will use Panasonic or WIMA stacked film types. For C0G capacitors, I will typically use AVX brand, rated at 100V. All of the components used in this thread are available from Mouser.

There should be enough detail in the thread for others to replicate the work, but if you have any specific question(s) I would be more than happy to try to answer them. Please post your question(s) in the thread so others may benefit from the answer.
 
Impressive

Thanks for the rational, logical, detailed and fact-based explanation of your fine work.

xxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxxx

On the topic of component selection; on a $50,000 communications product I asked the engineers how long left to get it into production, they told me that they were cost optimizing the design which would take a couple of months. De-rating and lower-cost components. When I asked for examples, they were literally shaving pennies per component, they selected the best parts for their design-prototypes to ensure that it worked. I instructed them to skip the cost optimization from now on and use the best (within reason) components above rating.

They responded that this "went against everything that was hammered into their heads in Engineering School". Apparently perplexed by my instructions, they reluctantly went along. I had the same conversation a few minutes later with my friend the HW engineering manager, also bewildered he reluctantly agreed stating that "it made his job easier"..... at the time it sounded like a bad thing.

If I recall correctly, the gross margin on my product lines was between 75% and 90%. The performance and reliability were stellar.
 
On the F2663.
The small value (0.47µF) low leakage capacitors, C19, C20, C21 and C22, were replaced with 0.47µF stacked film types.

Could you clarify this. The film capacitors I found are of bi-polar but the capacitors named here seem to be polarized. Please excuse my ignorance of capacitor since I am new to restoring. BTW, I love your work and follow every bit of them.
 
Last edited:
On the F2663.
The small value (0.47µF) low leakage capacitors, C19, C20, C21 and C22, were replaced with 0.47µF stacked film types.

Could you clarify this. The film capacitors I found are of bi-polar but the capacitors named here seem to be polarized. Please excuse my ignorance of capacitor since I am new to restoring. BTW, I love your work and follow every bit of them.

You are correct, the stock 0.47µF capacitors are polarized electrolytic types and there is no problem using the film capacitors as replacements (all film-type capacitors are bi-polar/non-polar by design).
 
I have to say this is one of the top informative threads on this entire site for me at least. Can't say i have read them all however :) You have done a fantastic job at documenting your work so that the masses can be educated. Everyone who is going to refurbish their old stereo should read this thread first. I'd say it is sticky worthy to say the least.
As for the Au-719 upgrade, I can suggest to change the o/p bjts. Long story that I say why. Might add some controversy, but it is a discussion subject none the less.
This amp, like most sansui's were rated for 8 ohm loads, not 4 or worse 2 ohm. So it limits what speakers you can use for one thing and how many you can use at once.
The short story, is to use newer o/p bjts with larger SOA but have all the same or similar characteristics as the originals, which are IIRC 100W (2sc2581/2sa1106) devices in a TO-3P package, with ft of 20-30MHz. Onsemi,Semelab,Sanken,Fairchild are all available as improved subs. semelab make a 300W device in a to-3p package!! Quite an improvement from the original.
this is a list of what Mouser offers, digikey is selling sanken these days
http://ca.mouser.com/Semiconductors...yzhg3Z1yyziiqSGTax1shZer4gZ1z0j3y4Z1z0j3xnSGT
In EE reliability terms, it is a weak link. Weak links usually are places for failures or lessen the MTBF.
90Wrms into 8 ohms is still very loud BTW, so if you do not crank it, it might be a moog point and why the o/p devices have lasted so long.
It would have been nice if sansui used a pair of o/p's but they did not, so increasing the SOA in a new device is a real easy solution.
I and others have played around with this sansui design in simulations. i can post it for you if interested. I did up the Au-x1 design in ltspice. it is the same design toplology as used in many of the high end designs of the era.
For the owner(s) of the au-719, they can think of the o/p bjt upgrade as extra preventative maintenance, when they decide to change the big ecaps. It is not imperative if the device is not driven heavily, like for headphone use.
Cheers Rick
 
Last edited:
oops, or duh, I got au-719 mixed up with au-717, but the same analysis generally applies. au-717 is the older design. The to-3 pkg is more reliable over a to-3p. (to-3 metal can is hermetically sealed and has better theta jc)
Those o/p devices are still 100W devices(120v,10a,FT=10MHz), but in the metal can to-3 package, which is more reliable than a plastic package but limits your selection, Onsemi is your easiest sub but has lower ft, so that could affect stability. Have to look through the threads to see if people used onsemi mj21193,4,5,6 devices, to see on their success. It is possible to fit a to-3p in for a to-3, with some finesse.
 
Last edited:
The output high-pass filter consists of a pair of polarized 100µF/6.3V capacitors in series (forms a "bi-polar" capacitor) for each channel (C27/C29 and C28/C30). Each capacitor pair was replaced with a single 47µF/16V Nichicon ES bi-polar "audio grade" capacitor (jumpers were installed in the vacant capacitor positions).
I am working on a 717 tone board planning to incorporate Lee's many suggested changes. Just need clarity when he says jumpers were installed in the vacant cap position. Do I just put a jumper on the underside of the board in the now vacant cap location across the plus and negative lead locations?
 
One quick question. On the equilizer circuit board #F-2723 at the end Lee says "The local decoupling capacitors C31-C34 were originally rated at 10µF/50V and the replacements were low ESR types rated at 47µF/50V." He means C31 thru C34 including C32 & C33. Forgive me for the rookie question.IMG_20160105_211702566.jpg
 
Starting a AU 710 rebuild using these specifications - developing a parts list - with pricing . Lots of threads related to the AU 717. Any info on wiring leads - ie puling each module/section and ensuring wiring leads go back correctly?
 
Back
Top Bottom