F2624 Resurrection: A modern remake of a vintage classic

Tom and Anatoli, you guys have done an amazing job with all the R&D accomplished on this F2624 board. I enjoyed reading this entire thread, with the exception of the one ungracious poster (caveat emptor indeed!). It is a shame that this labor of love will barely get to see the light of day. I'm hopeful you will reconsider sales of your board. Most of us owners of 9090DB's would be extremely grateful of your efforts, not to mention being blessed with a Sansui which works better than ever! Don't let the fruits of your efforts be hidden. Best regards, Ben

Thanks for this heart warming post. We appreciate it. :thmbsp: It's no secret we've slowed down activities with this project. The logistics are tough and both Anatoli and myself had been way busier recently than in the first R&D year.

Doug Brewster turned out to be a most helpful and contributing resource and he actually trouble shot the last two issues with this project. We thank him dearly for that. Thank you so much Doug - we couldn't have done this without you!

In the aftermath of this project - none of you would even begin to believe the dollar value of Anatoli's hours in this project. I thank god I don't actually have to pay for all those hundreds of man hours and I wish to god I could get paid for just half of mine. Designing and printing boards, ordering components, shipping it all back and forth between Boston and Tel-Aviv (including a 50 pounder DB unit) and putting it all together doesn't come cheap either - but it's just a small fraction of the man hours involved. In short - it's a time gobbler financial disaster!

However - it's been a heck of a fun thing to do. Nothing like the shear plain satisfaction I get every time I turn on that unit knowing that a vital part in it's heart was actually designed and built by Anatoli and myself. There's a lot of satisfaction in it. I also feel reassured that this central and vital part of the 9090DB, will never go 'extinct' and by now is actually replaceable. I feel good knowing those units got the potential to be listened to for decades to come.

We're positively discussing the option to make most of this knowledge base (schematics, BOM, etc...) public domain at some point in the future so those wonderful receivers may outlive all of us. It's not going to happen soon and we'll hold on to those bragging rights a while longer - but we're not going to take it to the next world. We thank each and every one of you that's been involved in this thread, whether actively supporting the process or offering tips and advice. It's all very appreciated.

Thank you all.

Tom

Tom,
Thank you for you response! It is good to know that both you and Anatoli will consider passing on the fruits of this project to those who will appreciate it. And please don't let the occasional issues with your project get you down! You will always get that certain percentage of folks that won't appreciate the efforts you guys put into this. You won't even be able to console them with logical excuses, as my experiences as an aerospace avionics tech and manager have taught me.
Your willingness to accept constructive criticism from Doug, Echowars and others shows the passion you have to ensure that you only will present the best product possible to the consumer. For that you should be proud, and I for one will give standing ovation for the work that both of you have accomplished. Bravo!!

I’m afraid that Tom (in post #84) is allocating way too much credit for my modest input – in terms of effort and expertise – than I’m due, having come in at the tail end of the project when the major effort had been completed by him and Anatoli.

I recall from my days starting up Westinghouse power plants that the last 5% of the project seemed to take 50% of the effort and hopefully I’ve been able to take a bit of pressure off their schedules.

It’s very pleasing to be part of a venture that clearly gives a lot of pleasure to the many contributors as eiraved’s post illustrates rather nicely…

Your opinion please...

You all probably heard stories about how shot bias trimmers with dead spots can kill a driver amp and fry the op amps and outputs. The OEM trimmers are of semi-open design and are susceptible to oxidation and a wide array of related issues over time.

In our reproduction board, we use Bourns trimmers of a some what better design and we see no such problem now - but who knows how these are gonna be like in 20 years? Before printing a 2nd version, we thought of adding a diode in parallel with the trimmers, so if something goes wrong with them the current will pass through the diode and will be limited to ~0.7V at most. This shouldn't pass more than a few 100's of mA's in the high power transistors, which is something they can survive.

Before we actually measured the trimmer's resistance at 30mA bias (45 - 47 Ohms) we considered a few options, as Anatoli laid out for me:

1. leave it as is (e.g. drop the idea).
2. Use a 1N4148 diode, and it'll effectively make the bias adjustment resistor have almost no change in bias once its above 90ohm or so. (As I mentioned we have about half of that)
3. Use 2 1N4148 diodes in series, which will limit this voltage to the Vf of 2 diodes. This means that the maximum voltage over the bias network will be 5X diode voltage. Or about 3.5V or so. This will result in a DC current of about 5A maximum (in both pairs of devices - so that 2.5A each) if my hand calculations are good enough. This is high, but something the transistors can sustain for a few minutes with no damage. The output resistors will probably handle this for a couple of minutes as well.
4. There are other solutions like using an LED which will also give a visual notification

So I was wondering - what do think about this? How would you guys feel about a LED in that circuit shining red when your trimmers are shot? Should we just drop it and leave it be happy with the current version?

Cheers

Tom

Tom:

Many kudos for the outstanding efforts and significant accomplishments that you and Anatoli have contributed via this project. I have enjoyed reading along as I have tracked this thread. Thanks for sharing.

This trimmer pitfall seems to be pervasive to many (most?) of the Sansuis, and a decent solution would certainly be a boon to us all.

The first suggestion Kevzep made when I started in on my G-6700 was to replace the trimmer pots, and two of them literally fell apart during removal. My vote is for a collective study and discussion of this problem. Hopefully the outcome could be a set of instructions on how to evaluate the problem, followed by suggestions for possible solutions for each identified category of ‘problem’.

Education is one of the key factors here, and I sure could use some more knowledge about the trimmers and the bias circuits.

I like the idea of being alerted when the trimmers are shot, or when there is a bias problem. A LED would work fine, but only if the listener has eyes on the unit. You might consider some type of audible alert, or a cutout that could disable that channel and leave the LED on.

Keep up the good work! I hope you get lots of suggestions.

Thanks Steve. The problem domain is very restricted. The Sansui trimmers are sort of a crappy design. Well, they're half open so over the years air, humidity, pollution, take their toll and lead to oxidation of the copper wiper which leads to the trimmer not 'trimming' voltage. Basically, a variable resistor that doesn't resist at some random points. Can't blame them really, I guess no-one expected those units to last for 40 years but this may lead to a bigger problem. During the biasing procedure, where adjusting the trimmer is required, a dead spot on a shot trimmer's wiper may lead to high rise in voltage (no resistance) to the power transistors bases and a surge of current to the outputs. This generally fries the driver amp's semi-conductors, opens resistors, fuses and possibly blows the outputs.

Some prominent techs around here have developed their own methods to overcome this issue when they suspect a shot trimmer and still perform the bias adjusting routine flawlessly regardless. Some just swap the trimmers for new ones - to be on the safe side and most of us just push our luck. Keep doing that and eventually, you'll run out of it. Of-course, a DBT can save our back ends in this case - but not always and not for too long.

In our reproduction board we use way better trimmers. Yet, we want to implement what is really just a simple fail safe device for insurance. It consists of a diode in parallel with each trimmer. If something goes wrong with the trimmer the current will pass through the diode and be limited to it's forward voltage. This shouldn't pass more than a few 100's of mA's in the high power transistors, which is something they can survive. In the previous message we were exploring the possibilities.

Now, someway or another, it could also be a LED. If it's a LED, it would glow when it's doing it's rescue mission in those (hopefully) rare cases when a trimmer fails. Please note this is not an owner's eye candy or anything like an audible alarm mechanism serves. No one but the tech doing the biasing procedure would ever catch a glimpse of that LED lighting up and even then, we really wish it would never have to. This is only for a specific problematic moment during the bias setting procedure. Once the bias current is set - this feature is irrelevant any longer, until the next service. We can also avoid the LED and use a none lighting standard diode that'll do the job without the visual notification.

The question is really a matter of taste and fashion and old habits. How would you, as vintage collector/tech be willing to accept a modern LED in your circuit? My LUX uses red demonic leds. It's cool! I'm not sure it's in line with a 9090DB. Just asking your opinion.

The other question is do we really need to bother with it? I mean, those are 2013 Bourns trimmers. Seriously? Do we really need to worry about them 30 years from now? Huh?

OK. I guess that I have answered my own question. I won't need a diode (LED or otherwise) because I will always replace the ancient single turn trimmers before I do anything else.

However, I would have no problem with a 21st century LED inside my 70s vintage receiver as a diagnostic tool.

Thanks - - -

Just received a reproduction board!

Hi everyone, I just happened to be following this thread over the past year or so and low and behold I saw one of these for sale on that auction site. I just received it today, and it is a true work of art. Excellent craftsmanship. Perfect, and beautiful. I am very impressed. I am by no means any type of electronics whiz, but I love my vintage gear. Everyone did an outstanding job, and I am extremely happy with this board and very proud that I have one plugged into my 9090db. It sounds great and works flawlessly. I cant express how happy I am with it.

Where can I get one of your redesigned F-2624 driver boards ?

We only printed 10 prototype PCB's and assembled 6. One is installed in my own unit, 2 were left with techs and the rest, well... 'distributed'. 2 blank boards are kept for archive by Anatoli and myself and the other 2 were, well... 'distributed' too. The karma was spread. Nothing left. Sorry.

The good news is we're in the process of testing the new protection LED idea I was talking about earlier. If that works, we'll issue a version II and print another batch of 30 boards. We won't be assembling any of them and they'll be designated for techs to install by a BOM we'll make available. My guess is 6 - 8 weeks before any of this happens.

Cheers

Tom

This is FANTASTIC!!! Excellent work.

Hi tnsilver! I am the proud owner of the last one you sold, I just received it in the mail and am going to replace all the TO3 outputs with NTE 284/285's or OnSemi MJ21193 and MJ21194's (or the ones with a "G" after the numbers) which would be better? I am almost certain the board will need to be adjusted to the new outputs, I found this to follow (see post #153 on link) and was wondering if that would be all there was to it or if there were some little tricks to know about.

http://www.audiokarma.org/forums/showthread.php?t=144580&page=11

Congrats!!! I prefer the term 'distributed'

I'm not sure why you need to replace your outputs. If your outputs are blown, it means something blew them and you need to address that 'something'. It is common to find some of the fuses on power supply board are blown too in this case and the typical cause is some fault with the driver amp board - but not necessarily. You really need to address the cause. That been said, I posted a 9090DB recap list here some time ago.

You'll find it as a pdf attachment at the bottom of the first post. At the bottom of the document you'll find my modern replacements for the 9090DB outputs. I chose them based on a lot of reading here, not because I was feeling experimental. On-Semi MJ21194G (16A 250V 250W) for the NPN and MJ21193G (16A 250V 250W) for the PNP. Still good. You'll need to adjust the bias current on the new board and the DC offset too.

Please note - I'm not saying you need to be a NASA scientist to do it, but there's a lot that go wrong, starting with choosing the wrong trimmer for bias, going through blowing the DMM fuse and thinking there is no current, using the wrong DMM scale, shorting with the probes and a wide array of other issue that will trigger a crap storm. With such a lovely nice new driver amp... perhaps it's better to invite over a tech friend or just pay a few bucks to have it done by a pro?

Cheers

Tom

Oh, and about those tricks (Abracadabra...), well not really tricks but here are two tips and my 0.02 for this procedure:

1. Use a current limiter on any suspect unit. It may save your rear end (well not your's but the unit's - literally, it's rear end). You can look up a DIY DBT (dim bulb tester) and build your self a one for peanuts. It's a life saver.
   
2. For the current taking measurement use a spent fuse tool. This a common cheap device made of a blown glass fuse where you solder leads to the terminals of the fuse body and connect your DMM probes to sockets on the other end of those leads. You then place the device inside the fuse holder instead of keeping your probes inside (which will eventually slip - for sure). Make sure that fuse is spent or you'll meet the smoke monster :yes:

Hope this helps

Tom

The unit was my dads and I "found" it when I was about 15. I didn't know quite what I had and just tortured it with mismatched speaker loads (at that age didn't know any better). It shorted 2 of the outputs, and damaged the driver board shorting a pair of transistors, a couple caps, various resistors and one of the temp sensor cans. I took it to a shop to be repaired and while the repair got the sansui working again it didnt last, the counterfeit toshibas shorted and blew apart the driver board again. I had found your thread on here back when you had first started this project, and decided to dig out the old 9090db. The board is in salvageable condition, but the timing of the sale of your last board was way too perfect. Ill look for anything else to replace while its apart, but so far new driver board, new outputs. I'll bring it up slow and and adjust bias and offset, and all should be well. Anything else I should specifically replace while apart or should look at a little more closely?
I remember this being a really nice receiver and having a ton of power, I can't wait to get it back in action.

tnsilver said:

Oh, and about those tricks (Abracadabra...), well not really tricks but here are two tips and my 0.02 for this procedure:

1. Use a current limiter on any suspect unit. It may save your rear end (well not your's but the unit's - literally, it's rear end). You can look up a DIY DBT (dim bulb tester) and build your self a one for peanuts. It's a life saver.
   
2. For the current taking measurement use a spent fuse tool. This a common cheap device made of a blown glass fuse where you solder leads to the terminals of the fuse body and connect your DMM probes to sockets on the other end of those leads. You then place the device inside the fuse holder instead of keeping your probes inside (which will eventually slip - for sure). Make sure that fuse is spent or you'll meet the smoke monster :yes:

Hope this helps

Tom

Click to expand...

Hi Tom
It would be very helpful if you can upload couple pictures about this. Thanks.

A current limiter in the form of a dim bulb tester (DBT) is a very common DIY project. You can follow this AK thread, for example, and checkout the links. You'll get the hang of it pretty quick. Other forms of a current limiter are known as VARIACS and are based on coil induction and are difficult to build, more expensive, and less common with non-pro fans.

My spent tool is very simple and I use the fuse holder chopsticks to mount the tool in it's place between the fuse holder clips - instead of sticking my fingers down the power supply board.

my spent fuse tool



Cheers

Tom

F2624 Remake

How do I get one of these new boards ?

Thank you all for the most wonderful comments about this project.

We're moving on to version II

Cheers

Tom