AU-517 vs AU-719 - a post-recap comparison

pete_mac

Super Member
I've had an AU-517 for a little while now. It's in very very nice condition and has been fully recapped (apart from the PSU filter caps) and has had the usual diodes and diodes replaced.

I acquired a troublesome AU-719 about 6 months ago and skippy124 (Dad) sorted out the issues for me. The amp fought us all the way - we even went to far as buying a NOS driver board, only to find that some of the resistors were incorrectly fitted and knocked the gain down noticeably! Dad may choose to share some details of this later.

Thankfully, she's singing beautifully now, and has been recapped just like the AU-517 (no PSU filter caps yet). Both amps have had Panasonic FC or FM caps in most places, with a few Elna Silmic IIs and either Panasonic or Epcos stacked film caps in the 1uf 50v locations. So, both of the amps are very similar in terms of electrical condition.

A few thoughts:

BUILD QUALITY

The 517 clearly prevails here - black chassis, black cover panels, nicer switches, cleaner-looking transformers. Obviously, you don't see the amp with the cover off very often, but the 517 still is a thing of beauty when naked.

517:

pete_mac08.jpg



719:

AU719insidetop.jpg



Both amps were partly or wholly converted to new 5 way binding posts on the rear panel to facilitate the use of speaker cables with banana plugs. The 719 was easier to do as it came with binding posts (basic ones) as standard. However, the 517 is also very presentable.

517:

AU517andAU3174.jpg



719:

AU719newspeakertermsrear.jpg



The 719 in its natural habitat (only a TU-417 tuner, but one day we'll find a 717 or 719):

au719finishedstack2.jpg


au719finishedstack3.jpg




LISTENING THOUGHTS


Both amps were set up in my main system, comprising a Squeezebox Touch, an audio-gd digital interface as a digital signal reclocker (running a Tentlabs XO and revised hex inverter) and an audio-gd NFB-3 DAC. My speakers are Dynaudio DM 2/7s. It's a good quality source and speaker setup that shows the differences in amps quite well in my experience.

I've been listening for the 719 for the past week or so, and have become accustomed to its sound. Going back to the 517, it sounded fantastic but a little bit coloured - perhaps a bit warmer and not as accurate as the 719. Switching back to the 719, there was a noticeable improvement in clarity and definition, and it just sounds natural with no 'additives'. I'm really impressed with this amp. IMHO, the evolution of Sansui's circuitry to the Diamond Differential circuitry pays clear dividends for the listener who values sound quality and accuracy. I'd expect the difference between the 717 and 719 to be similar to this comparison, as the 517 and 717 feature the same general circuit design.

I can, however, see how the 719 (and 819/919) and later generations of Sansui amps might be 'a step too far' for some listeners, as they lose the warm vintage sound and become more 'hifi'. The 719 is right up my alley and it complements the rest of my system perfectly. I'm very very happy that I can use a 30 year old Sansui amp in my system which is not that disimilar in terms of sound when compared to modern hifi amps.

Whilst I still lust after the 517's beautiful innards and warmth, the clarity of the 719 comes up trumps! It's my new reference amp. Even though she's not as beautiful as the 517, I can appreciate her superior singing voice!

I'm now tossing up whether to use some Pansonic 15000uf 63V caps that I've got lying around to replace the filter caps in the 719, or to splurge and spend $80 on some Nichicon Gold Tunes from Partsconnexion during their 20% off sale. I think the 719 will be a keeper, so I may just end up going the Gold Tune option (or some 15000uf 80V Panasonics which are 40mm x 80mm vs the 40mm x 50mm 15000uf 63 Panasonics which will be dwarfed once installed. Time will tell.

I'm also aiming to go through both of these amps and replace a lot of the small transistors, certain diodes etc to get the amps into tip-top shape for the next 30 years.


The AU-517 and 719 ready for the listening comparison:

Vintageshootout6.jpg


Vintageshootout5.jpg



Lastly, just for giggles, I hauled out my recapped/modded Realistic Mach Ones for comparison against the Dynaudios. Unsurprisingly, the Dyns sounded more coherent and accurate (and went deeper and had more control than the Machs). However, the Machs have a BIG sound... great party speakers for sure. They weren't totally embarrased by the Dynaudios which I was very happy about.

Vintageshootout1.jpg


Vintageshootout4.jpg
 
Last edited:
Nice write up :yes: - nice pics :yes::yes: - accurate listening test conclusions IMHO. :yes::yes::yes:

I recently had an AU-719 for repair, and was able to compare it to my AU-717, all of what you say struck chords with my views on these two amplifiers.

Nicely done :thmbsp:
 
Thanks John.

I've also got a few newer generation Sansui amps to listen to in future which (circuitry-wise) bring even more to the table - super feed forward, X-balanced, 'alpha' etc. It will be interesting to see how they sound too.

The 719 seems to hit the sweet spot for me though - great performance whilst maintaining the classic looks.
 
Very good review;you nailed it pete_mac :thmbsp:

Nice write up :yes: - nice pics :yes::yes: - accurate listening test conclusions IMHO. :yes::yes::yes:

I recently had an AU-719 for repair, and was able to compare it to my AU-717, all of what you say struck chords with my views on these two amplifiers.

Nicely done :thmbsp:

and you haven't said anything about it...shame on you John :D

Thanks John.

I've also got a few newer generation Sansui amps to listen to in future which (circuitry-wise) bring even more to the table - super feed forward, X-balanced, 'alpha' etc. It will be interesting to see how they sound too.

The 719 seems to hit the sweet spot for me though - great performance whilst maintaining the classic looks.

I am eager...which one you talking about? :scratch2:
 
Nice report pete_mac. :thmbsp:

I too use an AU-719 along with a TU-719 and other sources are a SB3 into a CA DacMagic, a Pioneer D9 SACD and a Technics SL-QL1/Ortofon OM40 (the phono stage of the 719 seems really "rather good" too). I have not used any other 'sui than this, but have owned plenty of "modern" hi-fi previously and I don't find anything lacking at all with this 32 year old stunner.

I have re-capped, re-zenered and re-fusibled my amp and just have the main PSU caps to swap out. I have bought some Mundorf M-Lytic AG+ caps to fit to mine. They are 10,000uf/80v/105c but quite a lot smaller (35mm x 50mm), but to me, it's more important to fit some good quality caps to last another 30+ years, than what it looks like under the hood.

I will have to get some photos to post up here after I have fitted them.

Well done and nice write-up/photos. Enjoy !!!

Cheers,

Lee

PS: When you said you found incorrect resistors on a NOS power amp board, were they incorrect values with respect to the circuit diagram or with respect to the original board, as I have found a couple of dodgy values on the original service manual circuit diagram that are actually different on the boards in the amp? They were fusibles in the phono amp listed as 5.6k but were actually 56k in reality. How many more errors are there? I also believe John (Hyperion) found incorrect screen printing on his power board for one of the 'lytic caps. :scratch2:
 
Thanks Lee. Glad to hear that you're enjoying your 719 too.

Regarding the difference in the board, skippy124 (dad) would be better poised to explain that. There were basically a bunch of resistors that were literally swapped, and this resulted in about a 6db reduction in gain in one channel.

The board was a NOS replacement board but was a slightly revised board. The original boards in my 719 has a few of the traces cut from the factory. The later boards had this change incorporated as standard. The rest of the boards were identical, but for some unknown reason these resistors were wrong. Dad swapped them over and everything functions as we'd expect. The values were different with respect to both the original board and the circuit diagram, and they were definitely 100% NOS boards which were never molested.

There still is a VERY slight gain imbalance between the channels, so we will track this down soon. It's hardly noticeable, but it bugs me.

I found the error in the silk screening on the power supply board that John noted. There was an erroneous '+' symbol, but the line that extends beyond the circle to signify the negative side of the capacitor was shown correctly. I always align the negative to the line, so I would have been fine even if the silk screening error hadn't been pointed out to us previously by John. I checked three times anyway just to be sure ;)

re: the filter capacitors, Dad did a thread on a 717 that he fixed where he made up some little sleeves to accommodate 35mm caps in the standard cap clamps. They should work fine.

The Panasonic caps that I already have a spec'd nicely and are 40mm x 50mm so I might end up using those and see how they sound. The lure of physically large caps is strong though, and the Nichicon Gold Tunes look so OEM!
 
I am eager...which one you talking about? :scratch2:

I should be able to have a listen to the following in my system at some stage

AU-D11

AU-X301

and maybe another if everything pans out OK.

Even though the AU-X301 is only low in the model range, it still has the later generation circuit design AFAIK.
 
I have bought some Mundorf M-Lytic AG+ caps to fit to mine. They are 10,000uf/80v/105c but quite a lot smaller (35mm x 50mm), but to me, it's more important to fit some good quality caps to last another 30+ years, than what it looks like under the hood.

I just noticed your reference to the caps above. My version has the 12000uf 63V caps as standard (and the 470uf 63V caps on the driver boards) vs the 10000uf 80V filter caps and 470uf 80V caps on the driver boards. My 517 has the 12000uf caps and I always intended upgrading those to 15000uf, so I will probably do the same with the 719 methinks.
 
PS: When you said you found incorrect resistors on a NOS power amp board, were they incorrect values with respect to the circuit diagram or with respect to the original board, as I have found a couple of dodgy values on the original service manual circuit diagram that are actually different on the boards in the amp? They were fusibles in the phono amp listed as 5.6k but were actually 56k in reality. How many more errors are there? I also believe John (Hyperion) found incorrect screen printing on his power board for one of the 'lytic caps. :scratch2:

Hi,

There were some "anomalies" as Pete mentioned in the driver boards between the original ones in the amp (which agree 100% with the service manual we have) and the later NOS driver board I got. I am in the process of doing an LTspice circuit simulation of the driver circuit so I can see what the differences in values mean to the overall circuit gain. The main differences were in the diamiond differential section, plus a few different values around the differential fet input pair (for example, the original boards and circuit have at the signal input a 220K to ground, and a 1K in series to the fet input, whereas the NOS "later revision" boards have a 47K to ground with an 820 ohm in series to the fet input). Once I get the simulations done I'll start up a thread on it.

Cheers

John
 
Thanks for the info John. That is a strange one indeed. I will have to have a look at the values on my boards next time the lid is off.

Look forward to the spice results.

Regards,

Lee
 
I just noticed your reference to the caps above. My version has the 12000uf 63V caps as standard (and the 470uf 63V caps on the driver boards) vs the 10000uf 80V filter caps and 470uf 80V caps on the driver boards. My 517 has the 12000uf caps and I always intended upgrading those to 15000uf, so I will probably do the same with the 719 methinks.

Personally, I don't really recommend changing the value of the capacitors. I tried this with a pair of Marantz 2238b receivers that I endearingly call "the twins". I had thought to "upgrade" the sound, and for one unit I purchased a full compliment of Nichicon Muse and Elna Silmic II capacitors, in larger values for the power supply filter caps. The tech told me that he would recommend trying the other unit with equivalent replacements, and just good quality, as opposed to boutique, caps. He stated that the engineers in that day basically used 2-3 banks of caps to filter and smooth the sound. So, if you use really boutique caps, with different values, you might totally skew the filtering as designed by the original engineers. When I got the twins, allowed them 2 weeks to burn in, I can say that there really is a difference in the sound between them. The one that I spent gobs of money on the high-end caps, with upgraded PS filters (gold tone, with small Mundorf mcap supreme bypass cap for each ps filter cap) has a bit more extended bass, however the coherence of the tone is rather smeared compared to the unit that used exact equivalent values. The bypass caps did provide a tiny bit more detail, but the coherence is not quite as good as the stock unit. And, I spent a ton of money to "upgrade", more than I could even sell the unit for!!! I would say, all in all, stick with the stock values. Sansui engineers knew what they were doing, and if they felt larger values would be beneficial, then they would have designed the circuit thusly. You never know until you try, but I can say that I wasted a bunch of money with this experiment and wound up with a unit that has small improvements in some areas to balance against some decreased performance in others. If I wanted to sell the unit, I would say that the changes created deeper bass extension and slam, and improved some microdetail. But, I would just fail to mention that the overall coherence is skewed, it is now bass-heavy with some smearing of the midrange tones.
 
These were my thoughts exactly. Replace like for like when it comes to values. The originals in my 719 were the 10,000uF 71v, so I have replaced them with 10,000uF, 80v and also moved up from 85c to 105c so hopefully that should give a little extra headroom with the voltage and the electrolyte quality, whilst retaining the original capacitance.

Like you say - If it needed more capacitance, they would have added it. One can sometimes over egg the pudding...... :nono:
 
Last edited:
I run an AU/TU-719 in my main system. I agree with your findings.

My AU-719 has the black cased transformers. Stock. I have the inner covers.trim pieces from an AU-517 that will be fitted into my AU-719 someday.
The chassis looks the same, and the holes are there to mount them.
I may even have the heat sinks anodized black. Don't know if I will strip it down far enough to paint the chassis black, but one never knows.

Rob
 
Regarding the fitment of slightly larger filter caps, EchoWars has done exactly this on several AU-517's in the past (moving from 12000uf to 15000uf). If the master does it, I'm quite comfortable with doing it myself. After all, the 717 uses 15000uf caps and is basically the same as the 517 but with a few extra watts. If EchoWars had encountered issues with the sound after fitting larger caps to the AU-517/719, I don't think he would have advised me that it was OK to do this.

It is a decision mainly driven by necessity - there are bugger-all 12000uf 63V caps out there with decent specs, correct sizing etc. The Nichicon Gold Tunes (and Panasonics) are a direct fit which is certainly attractive.

We'll see how it pans out anyway - both the 517 and 719 sound brilliant with the original filter caps, so there is no hurry to upgrade them.
 
Last edited:
I'm not saying it wont work. I am just relaying my experience with trying the same thing (albeit on a completely different beast). I also had the benefit of having another exacat same unit that was redone to specs to make an absolutely fair a/b comparison. It's not like changing the cap value ruined the other unit, but it didn't do what I expected, which was to dramatically improve the sound. But the results could be completely different on the Sansui. And the only way to know for sure would be to try it. Then a/b it against another unit that was recapped with exact values. Otherwise, your ears might tell you it is better because that is what you expect. And the tech that consulted me that what happened to me probably would happen. Well, he is probably the best vintage gear technician in Michigan and has been doing this exclusively as his career for over 30 years, so it's not like it's the guy at the corner store that works on anything people bring in to him. Not saying he is any better or worse than EW ('cause there is no argument that he is the man), I'm just saying he knows what he is talking about.
 
I do appreciate your input mate :)

Thankfully I do have a bunch of Panasonic 15000uf 63V caps ready to go (these were $2.99ea from apexjr). I can very easily fit these and do some comparisons and testing to see how they sound. If they sound fine, I could then move to the physically larger Nichicon or Panasonic 80V caps.
 
Give it a shot! Post a review afterward and share your experience. For 2.99, you can't really go wrong. It's just a minor bit of work anyway and easily reversed if you don't like it...
Cheers!
 
The size and type of the main bypass caps will affect the sound: I really don't think increasing the value of the main filter caps has much if any affect. The Quality of the main cap would... that's why i stick with the Nichicon Gold Tunes whenever possible.

Scott T.
 
Back
Top Bottom