Czech Republic 9090DB driver amp pirate posing as ToliToms

[Skywatcher]

I don’t like his top silk legend much, but I’m nitpicking here.
Skywatcher, do you know if the schematic for the ToliTom F-2624 varies from the original Sansui F-2624?

The early and "original" F-2624 schematic in the Sansui service manual is of the early version that uses the 2SC984 BIAS transistors. The ToliTom boards are from the latter version that uses the STV-3H BIAS Diodes and therefore is different.
The original boards by Sansui are identical though for both versions. There are however differences reflected in resistor/jumper changes and BIAS pot value. I believe the "updated" F-2624 schematic was shipped with the owners manuals as a "supplement", but it seems nobody kept them around, or haven't thought to make photocopies. I was very young when I bought mine new and never even thought about it so mine was lost somewhere along the way. I do faintly remember the supplement in the back of the owners manual.

 

[tnsilver]

The ToliToms V2 F2624 is similar, but not identical to the stock driver amp. We've designed this back in 2012 and it's not fresh in my mind, but out of the top of my head there's the bias overshoot safety LED's that are not on any of the stock runs. Maybe some other minor differences too. The layout is familiar but is definitely different than stock and it's double layered, so structurally it's very different than the stock PCB.

We've spent an awful lot of time deciding on the best components for this PCB. It was 100% TOTL components, including some none traditional transistor replacements (like the better ZTX differentials and the expensive Murata MLCC's). The PCB is very high quality with a self healing scorch resilient layer to protect it from stray soldering iron tips.

For a while, these boards were assembled by senior EE's at my former company. They did it for fun in their spare time (I did a few too) and they took pride in the fact that strangers were willing to buy their work in appreciation. They were works of art with less than two dozens ever made. The lucky few that had purchased those driver amps in the custom plywood boxes got a true bargain. Each driver amp would now days cost over $1000 in parts and labor to produce, if we'd commission the same outfit to do it again. Back then they were auctioned and sold for an average of $150 - $180, which is peanuts in comparison.

I'm no longer associated with that company and the set of circumstances that allowed this to happen will never congeal again. For those who grabbed them back then - I say way to go!

An assembled F2624 V2:

For those who wonder... I was planning on a boat and a Benz with all those big bucks coming in but instead I've poured the stashes of cash back on the team. We've invested most of it with booze and snacks and big boys time on weekend parties. I've blown the rest on some new soldering stations for the guys because they've asked me to, and on paying those Mouser bills, because we don't grow quality components on trees. Thank you all.

 

[bubhardy]

Tom, I assume that the image you posted above is the real thing?
It looks good-too bad there are no more being made!
I have an 8080 and a 9090 which I intend on restoring in the near future, and I’d like them both to be as good as they can be.
And I don’t want to use these substandard clones.
Is there any documentation for your V2 board available?
Schematic, CAD files, BOM?

 

[bubhardy]

The early and "original" F-2624 schematic in the Sansui service manual is of the early version that uses the 2SC984 BIAS transistors. The ToliTom boards are from the latter version that uses the STV-3H BIAS Diodes and therefore is different.
The original boards by Sansui are identical though for both versions. There are however differences reflected in resistor/jumper changes and BIAS pot value. I believe the "updated" F-2624 schematic was shipped with the owners manuals as a "supplement", but it seems nobody kept them around, or haven't thought to make photocopies. I was very young when I bought mine new and never even thought about it so mine was lost somewhere along the way. I do faintly remember the supplement in the back of the owners manual.

Skywatcher, thanks!
But I realize that I really should have addressed Tom with my questions, so I am!
I figured that I might as well cut to the chase, rather than wade through a gaggle of messages here.

 

[tnsilver]

I assume that the image you posted above is the real thing?

That's as real as it gets. The schematics, layout and BOM are on Toli's DIY blog but the Gerber files are not public domain.

 

[Skywatcher]

Tom, I assume that the image you posted above is the real thing?
It looks good-too bad there are no more being made!
I have an 8080 and a 9090 which I intend on restoring in the near future, and I’d like them both to be as good as they can be.
And I don’t want to use these substandard clones.
Is there any documentation for your V2 board available?
Schematic, CAD files, BOM?

The 8080/9090 Non-DB are different animals altogether than the 8080DB/9090DB and use completely different boards.

 

[bubhardy]

The 8080/9090 Non-DB are different animals altogether than the 8080DB/9090DB and use completely different boards.

Skywatcher, thanks!
OK, well that answers that-I hadn’t delved into non-DB vs. DB models.
You learn something new every day!
I haven’t looked at the service manual yet.

 

[bubhardy]

That's as real as it gets. The schematics, layout and BOM are on Toli's DIY blog but the Gerber files are not public domain.

Thanks!
But Skywatcher’s info below showed me that I was barking up the wrong tree.

 

[Skywatcher]

Skywatcher, thanks!
OK, well that answers that-I hadn’t delved into non-DB vs. DB models.
You learn something new every day!
I haven’t looked at the service manual yet.

The 8080/9090 (and a few others) use the F-2436 or the F-2436-1 driver boards. The -1 rev being the later and can interchange if the correct components are installed for the particular model. There are also Service Bulletins applicable to the 8080 and 9090 so when you're ready, start your own thread and you'll learn about them too.

 

[JJCalvillo]

Unfortunately, you can’t copyright a board layout, or a circuit.
This is why I can do reverse engineering.
But I restrict myself to obsolete industrial stuff that is no longer manufactured.
Morally it’s wrong, but there are no laws against it.
Unless it’s patented, and you can’t patent a board layout, either.

As far as it being morally wrong, I think it's wronger to send these old Sansui's to a landfill because of a part that COULD be replaced. MHO.

 

[bubhardy]

As far as it being morally wrong, I think it's wronger to send these old Sansui's to a landfill because of a part that COULD be replaced. MHO.

Agree, most wholeheartedly.

 

[tnsilver]

As far as it being morally wrong, I think it's wronger to send these old Sansui's to a landfill because of a part that COULD be replaced. MHO.

Waking up this zombie thread...

I don't think it's morally wrong to reproduce an unobtanium vintage classic - I was complaining about the pirating of our old listing in order to sell some sub-standard crap. I felt it was insulting to be associated with those miserable clones, and it's far worst than just morally wrong. It's actually criminal and that's why those listing evaporated in a sec.

Which brings me to a new (to me) and interesting discovery that I was just notified about by Toli. We find it quite interesting actually. Someone from NJ by the name "Christy Electronics" makes new reproductions and sells them on the auction site. Look up "Sansui 8080DB F-2624 Driver board". They're blue and actually look quite nice! It even includes an implementation of our bias overshoot LED idea (albeit at a different layout location). User feedback seems all positive and people seem happy with it. Good for them! I'm kinda glad someone picked it up from where we left it.

No Sui's in the landfill then - Eh?

BTW: Do we still need to mention "no affiliation"? I did - just in case. We've got nothing to do with it. Any of you guys know anything about it? Tried it out? I'm real curious.

 

[bubhardy]

@tnsilver, I’ve seen the same listing on the ‘Bay. Wondered about it, if those were any good. They look like they’re nicely done. I’ve thought about doing similar with an F-6013 layout that I had fabbed for my Sansui 5000s. But all of the potential pitfalls of a board assembly that has to be hardwired in, as opposed to plug in, stop me.

 

[Skywatcher]

Waking up this zombie thread...

I don't think it's morally wrong to reproduce an unobtanium vintage classic - I was complaining about the pirating of our old listing in order to sell some sub-standard crap. I felt it was insulting to be associated with those miserable clones, and it's far worst than just morally wrong. It's actually criminal and that's why those listing evaporated in a sec.

Which brings me to a new (to me) and interesting discovery that I was just notified about by Toli. We find it quite interesting actually. Someone from NJ by the name "Christy Electronics" makes new reproductions and sells them on the auction site. Look up "Sansui 8080DB F-2624 Driver board". They're blue and actually look quite nice! It even includes an implementation of our bias overshoot LED idea (albeit at a different layout location). User feedback seems all positive and people seem happy with it. Good for them! I'm kinda glad someone picked it up from where we left it.

No Sui's in the landfill then - Eh?

BTW: Do we still need to mention "no affiliation"? I did - just in case. We've got nothing to do with it. Any of you guys know anything about it? Tried it out? I'm real curious.

I sent him a message about R03/04 input resistors. The 8080DB/9090DB use 1k resistors, and all others use 10k resistors. The difference is the Pre-out/Main-in jumpers on the DB models. The others do not have them.

 

[tnsilver]

^ Yes, we've used 1K for R03/04 too. I'm not complaining just a little concerned though. When we've developed our own reproduction (already a decade ago) we've had some clear objectives from the get go. We wanted the product to look and feel familiar and to introduce as little as possible design changes in order to avoid any notable sonic differences. We basically wanted a better, more robust driver amp with the best and quietest components available - but with as little changes as possible in the process. I've personally reiterated through Mouser's bill of origin to make sure the pedigree of each and every component meets our expectations and standards. Toli was testing parts of the design in the university labs and at some point I've shipped an entire 9090DB unit for him to test on. When we finally had a working version - I've sent samples to some prominent techs around here, fresh out of the mint. @DougBrewster had actually saved our behinds with more than one very valuable rejects (IIRC we've missed the pinout of some transistor, and some other minor issues) and there was other important feedback from techs and customers. We've swapped to an improved V2 upon receiving feedback from just the first 5 customers and implemented the bias overshoot safety LEDs. In other words, this was an iterative community development process with a genuine peer review.

So, I'm not complaining but just a little baffled... How can this new reproduction had been developed with, as far as I know (and I don't really) an under-the-radar process?

 

[Hipocrates]

under-the-radar process?

CADs, simulations, prototyping printing on JLBC (and similar)... and definitely not the same attention to detail as yours. I've seen some folks on the yamaha and marantz subforum making some power supply and amp boards, the good old xraytonyb making MC2100 boards as well, not to mention folks on diyaudio making way more complex stuff than our old behemoths (class D, AB, SMPS, composite chipamps). It is very fun.
I can see this being more common as the times goes by.

 

[Hyperion]

I'd be upset if someone else had appropriated the original intellectual property of this project for their own gains. If they made it under their own steam, (or perhaps with the assistance of Tom and Toli, which we now know isn't the case), then fine, let them get on with it.

 

[tnsilver]

We've kinda made this a transparent open source project from the get go with the intent of publishing the schematics, layout and BOM of this driver amp into the public domain. We kept no secrets, and we're beyond getting offended by anyone picking this project up for whatever personal reasons, as long as there's driver amp for anyone who needs them. On the contrary, we encourage it. We would have actually helped if asked to. For example... I'd be advising about enlarging the driver transistors heat sinks, to start with.

We're not stupid though. The gerber files are not part of that public domain deal. Toli, who spent hundreds of man hours on the design, is the rightful owner and he's the only one who has them. I don't even think he keeps them on-line. If those ever got out there, this project would have been poorly mass produced by you-know-who in no time, with matching sub-par crap components, pretty much like everything else coming from the global piratesville. Then we'd really get upset. That's not ever happening though. Not if we can help it.

 

[Hipocrates]

I'd be upset if someone else had appropriated the original intellectual property of this project for their own gains

I get what you are saying, and yeah they probably piggyback on Toli and Tom's work, but all in all the F-2426 schematic, board layout is in the service manual widely available, any CAD designer or even an enthusiast will make a copy with no problems at all, all circuitry is 50 years old technology implemented by a defunct company several decades ago.
What made the tolitoms boards unique in my point of view was the components that they used, for example: they used ZTX devices (matching those things is time consuming and expensive), STV3H and the over shoot led and overall obsessive (in a good way) attention to details. That is theirs and no one else and they should be proud of it.
Subsequent boards (specially from a ebay seller looking for profit) will have a different approach.

 

[tnsilver]

Thanks! You're probably right on most points but the most significant mod is the double layering of the PCB. This driver amp is rated at 125 wpc but some performance tests implied it was able to deliver 165. It's the double layering that's responsible for that. ...and there's the quieter Zeners, low noise resistors with carefully chosen power dissipation properties, the best transistors available at the time, the informative silk screen, the optional pinouts... Honestly (unbiased as I can get - duh!) it's a work of art.


There's no need for the service manual, which has the outdated schematics with the 2SC984 temp-co transistors in the bias circuit. Our schematics are out there for free on Toli's blog. 10 years ago, no one had them and some prominent techs around here were struggling to find the time to redraw them. Toli took care of that for posterity. We encourage anyone to use them.